LIBRARIES —:—- RETURNING MATERIALS: PIace in book drop to remove this checkout from your record. FINES wiII be charged if book is returned after the date stamped beIow. ECOLOGY'ALD LAhD-USE RELATIOLSHIES CF SLAll.LAJlLLS (II A LICIIICnJC FARM BY Joseph Paul Linduska 7- A THESIS Submitted to the School of Graduate Studies of Michigan State College of.Agriculture and Applied Science in partial fulfillment of the requirements for the degree of ' DWTOR OF 311103033 Department of Zoology 1949 THEsm TABLE CF CONTENTS Introduction . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . The area . . . . . . . . . . . . . . . . Soils and climate . . . . . . . . . Fauna of the area . . . . . . . . . Part 1. HethOdSOfSt‘ldyo...‘-.g.o... Comparative pOpulations of small mammals Annual numbers in cropland . . Prairie deermouse . . . . Other species Relative numbers in various crop types Prairie deermouse . . . . Other species . . . . . . Numbers with reference to land uses and farm practices Strip—cropping . . . . . . Field distribution of Field borders . . . . . . Harvest . . . . . . . . . Shocking corn . . . . Cultivation and plowing . $ . ‘t |~J o... .9“ fig! )1... Studies in field and cropland habitats deermice O ’43 1:3 h? 149 55 58 61 Species account and interrelationships . . . Small mammals . . . . . . . . . . . . . Prairie deermouse . . . . . . . . Seasonal numbers and biology Movements . . . . . . . . . . Homing instincts . . . . . . Food habits . . . . . . . . . Jeadow vole . . . . . . . . . . . Seasonal numbers . . . . . . Movements . . . . . . ._. . . Breeding . . . . . . . . . . Causes of winter breeding . . House mouse . . . . . . . . . . . Seasonal numbers . . . . . . Movements . . . . . . . . . . Other species . . . . . . . . . . BOg lemming . . . . . . . . . Pine vole . . . . . . . . . . Jumping mouse . . . . . . . . Striped ground squirrel . . . Houseratoooooooooo Prairie “1018 o o o o o o o o Short-tailed shrew . . . . o LTaSkEd Shrew o o o o o o o o LeaStShre'Wooooooooo 6S 65 65 65 7o 75 82 86 86 9h 100 103 107 107 107 108 108 112 112 113 116 116 120 122 J‘. O Predators of small mammals . Birds . . . . . . . . . Harsh hawk . . . . Red-shouldered hawk Red-tailed hawk . Broad—winged hawk Cooper's hawk . . Other hawks . . . Great horned owl . Long-eared owl . . Screech owl . . . Crow . . . g . . . Iimmmls . . . . . . . . Skunk . . . . . . weasel . . . . . . Opossum. . . . . . Raccoon . . . . . Red fox . . . . . Link . . . . . . . Badger . . . . . . House cats and does Summary and discussion . . . . . Part 2. Studies in farm woodland habitats History and description of study units WOOlet analYSiS o o o o o o o o o 122 126 126 131 133 133 13h 135 135 137 1&0 us 1147 1118 11:9 150 151 152 152 15h 162 161; 167 brethr-‘ds Of StUdy I O O C O O O O O O O I East crop measurement . . . . . . Comparative pOpulation levels of the Sciuridae Annual'trends...........o.. Relative numbers with reference to and grazing of woodlots . . . . Species discussion . . . . . . . . . . Fox squirrel . . . . . . . . . . . _Red squirrel . . . . . . . . . . . Populations . . . . . . . . . Habitat preferences . . . . . Effects of grazing . . . Upland and lowland types Outlying habitats . . . BiolOgical notes . . . . . . Poyulation turnover . . Longevity . Q . . . . . Trapability . . . . . . Sex ratio . . . . . . . Age ratio . . . . . . . Chipmunk . . . . . . . . . . . . . P0pulations . . . . . . . . . Habitat preferences . . . . . Effects of grazing . . . Upland and lowland types Outlying habitats . . . 169 170 173 17 3 176 180 180 180 180 188 188 189 192 196 196 197 198 200 200 202 202 201; 202; 207 209 Biological notes . . . . . . . . . . . Pepulation turnover and longevity Trapability . . . . Sex ratio . . . . . Hibernation . . . . Flying squirrel . . . . . . . Pepulations . . . . . . Habitat preferences . . . Biological notes . . . . Pepulation turnover Movements . . . . . Trapability . . . . Sex ratio . . . . . White—footed mouse . . . . . . Populations . . . . . . Habitat preferences . . . Biological notes . . . . Distribution movements . Sex ratio . . . . . Short-tailed shrew . . . . . . Other woodland inhabitants . . Acorn weevils . . . . . . Summary and discussion . . . . . . Literature cited . . . . . . . . . . . . Appendugooooooooooooooo 212 212 2111 2111 215 218 219 221 22 3 223 226 227 227 229 229 230 233 233 233 237 2111 21:11 2M: 2118 261 WCTIG Interest in ecological research has been concerned nostly with wild or native plants and uni-ole in non-agricultural areas. 1 However, this hes not been without the recognition that-new long-standing problas of the fern and range probably sill wait final solution through ecological inquiry. In 11 presidential address to the Ecologi- eel Society, Hanson (1939) upbssised the need for research in crop- lands perticnlsrly es regards the possibilities for nininising losses to noxious species. Recent trends in agriculture and the growing concept of nultiple-use of ell lands further uphesise the need for comprehensive ecological studies of farmland. According to figures node public by the Lead Planning Omittee or the National Resources Board (1935), nearly seven tillion acres of Hichigen farmland are effected by accelerated erosion. This constitutes lore then a third of the land in terns. On a notional scale, the situation is equally critical. The Reports and Records Division, Soil Conservation Service, U. S. Department of Agriculture, estineted that s billion acres, of the one billion, 1&2 million ecres of agricultural land in the United States, were in need of sue remedial treatment. is of January 1, 1919, about 93 million sores (less than 10 percent) had received full treetnent under the Soil Conservetion Service progress, and about an equal ecresge hed bene- fited by independent progress. It can be expected that the easily duonstreted values of sound fern planning eventually will encourage nest of the tenors in this country to adopt these new principle e of lend-use. the departures from eonventionsl turning nethods, in effect, sill constitute a major ecological change on my lillions of acres of land. Most species of fern wildlife will be narkedly affected. Economic biologists recognise the possibilities in this program for nanaging wildlife and sons studies dealing directly with the subject have been made. Dubach and Good (19140) and Good and Danbach (191:3) have reported on the songbird populations of farms planned for soil conservation as compared with the numbers on ferns outside of soil conservation districts. Moss (1939) has shown the relationship between upland gane bird pepulations and agricultural land-use. The effects of grazing of woodlots on songbirds have invited tn attention of several workers, all of when have reported conspicuously higher populations in protected woodlands (Saunders, 1936; Mayan, 1938; Mach, 191.0). A few basic studies, which night be exemplified by those of men (1938), Ssith (19103, 19103) and Danth (1914b), contribute to the picture of wildlife and land- use. 1'0 provide needed information for the management of all gene on farmland, the Gene Division, llichigan Deparhent of Conservation, established the Rose Lake Wildlife mperlment Station in 1939. One of the main functions of the station has been to deterrine the response of farm wildlife to the new conditions established in con- verting iron exploitive to conservation farming. The intentions for long-stern study, and the broad scope of the work, distinguish the project tron anything heretofore attempted in ecological studies of vertebrates in agricultural areas. 11') The econouic importance of farm rodents and their significance as key organises in the wildlife complex justified their inclusion in the program. It is the purpose of this paper to give an insight into the status of mall nannnals on this representative unit of Michigan farmland. Presentation of the data will include four major considerations: (1) the conparative pepulation levels of associated species of slsll manuals; (2) the effects of various land-use and farming practices on the mmbers of small annals; (3) interrelation- ships within the group and with other mam. species; and (1;) species discussions to include such biological information as night con- tribute to a better understanding of this study and to the ecology of fern wildlife in general. ACKNOWLEmMENTS Grateful acknowledgnent is nade to Mr. H. D. Ruhl, Chief of the Gene Division, Michigan Department of Conservation, for making possi- ble this study and publication of the results. Other personnel of the Gale Division, and particularly embers of the Rose Lake staff, contributed to the investigation in many ways. much assistance in the small namnal trapping in 19111 was given by llr. I. J. Thompson, and in 191.2 by llr. a. R. Rafferty. Field studies of predatory birds in 19141 were conducted by Dr. C. '1'. Black with the assistance of 111‘. P. S. Baungras, and the work was continued in 191:2 by Dr. G. J. Iallace. Accounts of these investigations in Rose Lake reports have been drawn upon freely in the preparation of this paper. Ir. H. D. HcGinley, Mr. I. I. Shapton, Ir. Iilnur Bartels, Ir. )1. E. Cooley, llr. Oscar Iarbach, and other forner embers of the Rose Lake staff, participated in the extensive progran of box- trapping. These data have provided a basis for the portion of this report dealing with mall annual populations of woodlets. A special debt of gratitude is owed Dr. D. L. Allen, fornerly biologist in charge of the Rose Lake station, for his constant aid and encouragement in the study, and for a critical reading of the manuscript. Dr. G. I. Bradt, biologist in charge of the Rose Lake station, and Dr. D. I. Douglass have been helpful in various ways. Dr. I. H. Burt of the kuseum of Zoology, University of Michigan, gave generously of his time in discussions of the work and, also, loaned a masher of traps in the first year of study. 13 For guidance, gratitude is here expressed to Professor J. I. Stack, Chairman of the Graduate Comadttee; to Dr. H. R. Hunt, Head of the Deparhent of Zoology; and to Dr. B. '1'. Ostenson, then a amber of the Department of Zoology staff of Michigan State College, who gave much assistance in the initial planning of the work. Acknowledgment also is ads of the helpful editorial suggestions by Dr. R. H. llanville, and by Dr. G. J. Iallace, whose earlier participation in the Rose Lake program provided an over-all familiarity with the work and nade of him an especially valuable critic. THE AREA 14 .hvsue tinge.“ whoa no «one on». me! one .33.“ Ed.“ Hemaonfim 23. .3325. ocean he ceeonone 9.25 0.378» Hanan: 05. .eeae oxen enem— efiv mo newton e .«o to“... Hated .H .mmh a. 1., x. .1 4.. . f .40 v F F — .\ ,C . . 4 . 4 w 1 o . .. . . _ H t . . . ‘ I'Dl 165 $0115 AND CLIHATE The a... m. Wildlife Experilent Station, principal locale .: this study, .lies mostly in the southeast corner of Bath township, Clinton County, kichigan. The land area in state ownership was 1,172 acres at the tine studies were started in 19140. Additional acqui- sitions increased the total project-sine to nearly,2,000 acres in 191.6. Topography of the area varies between comparatively level and rolling. The steepest slopes are wooded in nest cases, but early settlers cleared each of the rolling terrain that since has become badly eroded. Many level areas that might have withstood these early exploitive farm practices were left in woodland. Soil fertility is moderate to low, and the land averages second- to third-class for agricultural use. The soil pattern is complex as characteristic of heavily glaciated areas. Twenty-seven soil types have been recorded. In the well-drained uplands, Newago loan, Bellefontaine sandy loan, and Caloma.loaaw sand appear in the order named. The poorly-drained soils of kettle-holes, narshes, md lake and strean battens are predounantly Rifle and Greenwood peats, and Kerston and Garlisle nucks. Considering the project as a whole, cropland and pasture constitute 61; percent of the total area; aarsh, asap and swale comprise 21; percent; and woodlots lake up 12 percent. However, in the BOO-acre area of nest intensive study, about 71 percent of land is in pasture and crops; 16 percent is in the narsh-swanp-swale :1- ‘7 category; and the remaining 13 percent is in woodlots. The latter land-use percentages are fairly typical of nest private ferns in the mediate area, and in this respect the Rose Lake farn is representa- tive of nany southern Ilichigan ferns of comparable productivity. Parsing of the region is diversified ucept for limited develop- nent of luck lands for the growing of mint and truck crops. Corn, oats, and wheat are the principal cash crops grown. Dairying and beef production provide a considerable share of the income on nest fans, and the conversion of these lands to conservation farming usually requires that these phases be intensified. The climate in this region is favorable to diversified farming. The nean length of the growing season is 158 days between lay 5, the nean date of the last killing frost in spring, and October 10, the mean date of the first killing frost in autumn. Precipitation is fairly well distributed throughout the year. During the period of study, certain notable departures from "normal” conditions were recorded. Average snowfall for the calendar year 19140 was the greatest on record, amounting to 82.7 inches, compared with a normal of 56.14. August rainfall in that year also established a new record with a total of 9.21 inches ne'asured in the immediate area of study. This precipitation was 6. 39 inches above nonal. Conditions in 191:1 were about average, although 00- tober precipitation was h.86 inches above nornal with a total rain- fall of 7.33 inches. The year 191.2 was characterised as the wettest on record and the total precipitation of 37.02 inches at East Lansing was 5.59 inches above nomal. The year's snowfall was about a third heavier than nornal. Details of neteorological conditions during the period of investigation are given in the appendix. 18 19 oneza 2H .a .3 opens .83an cheetah“ 0.2.82: 83 20m on» no ESEd eonnonountfi en fins 42“.»; 20 FLU!“ OF THE AREA The bird and manual life of the Bose Lake area is varied. ineng gane species the cottontail rabbit (Sylvilagus floridanus nearnsii Allen) ranks first in abundance. The average hunting season kill of this species for the years 19m to 19112 was 12 indi- viduals per 100 acres. Hunting pressure in these years averaged 270 gun-hours per 100 acres for the entire open season for rabbits which runs from October 15 to December 31. During the first 22 days of the season (October 15 - lovenber 5) hinting is legalalso for pheasants and fox squirrels, and during this period a major share of the hunting effort is devoted to these species. Hunting removes nearly half the fall population of rabbits which, before the start of the season, umber about 30 to 10 individuals per 100 acres. Pheasants (Phasianus colchicus Linnaeus) occur in fair numbers. A peak population in 19111 allowed for a harvest of 17.14 cocks per 100 acres. A general decline beginning in 1910, which apparently occurred over the entire North American pheasant range, resulted in a much reduced population after that year. For the years 1910 to 191:2, the average hunting season kfll was 13.0 cocks per 100 acres. Hunting pressure for the October 15 - Novuber 5 period averaged 218 gun-hours per 100 acres during these sane years. Hunting ruoved about tines-fourths of the cooks and pro-season populations consequently averaged close to 17 cocks per 100 acres. A nearly even pro-hunting sex-ratio was usual on the project. This period included a year of Imusually high productivity for the pheasant and 21 it is believed that average hunting season productivity for the area is closer to 10 cocks per 100 acres. During the sane open season as given for pheasants and under the sane hunting pressure, the 19140 to 19142 1:111 of for squirrels (Sciurus Egg rufiventer Geoffrey) averaged 39.8 individuals per 100 acres of woodlot. Pro-hunting populations were judged to be about twice this figure or approximately 80 - 100 for squirrels per 100 acres of woods. Populations of this species varied widely with- in the period of the study and the 19152 level was well above that of either of the two preceding years. A harvest of more than a squirrel per acre was recorded in 19112 for seven woodlots totaling 102 acres. Other game species and their approximate fall densities included: bobwhite (Golinus 1° micanus Iinnaeus), 3-1; covies for ‘ the entire station; woodolmck (liar-eta g. £1355 Iinnaeus), 35 to he per section; and white-tailed deer (Odocoileus vilginianus berealis filler) , 10-12 for the entire station and inediately adjoining lands. The status of other gene and non-gene residents which inter- act nore intinately with the snall namals of the area will be dsoussed at various points in the text. The nomenclature of nemals is taken from Burt (19116) and of birds from A.0.U. check lists for the years 19hh-h8. The studies reported here include observations on the small nannals of fara woodlots and those of cropland. Small nasal associations in the two general types (woodlands and fields) are Iutually exclusive to a considerable degree. It appeared 22 profitable to employ somewhat different study procedures and in other respects to consider then as distinct problems. The data are reviewed separately here, and discussions of netheds and other information dealing with studies in these two najor habitats appear in each section. PART 1 STUDIB IN FIELD AND CIDPLAND HABITATS Na d Hermes OF STUDY One objective of this study was to neasure the comparative m-bers of snall namals in various habitats and at various tines of the year. To accomplish this, live-trapping was employed principally and according to procedures to be described. lhere other techniques were used to provide information on specific points, descriptions will be given at appropriate points in the text. A wide variety of methods have been used in population studies of seal]. manuals, and investigators have utilised some variation of the quadrat nethod of trapping in an attemt to show acinal densities. It is evident, however, that econeny in sise of quadrats and relieve]. of individuals from the plots have. invalidated many of the population figures obtained in these studies. A failure to appreciate fully - the extent of drift among small annals, the sise of individual. ranges, and the speed with which depopulated ams are invaded, have been cemnon sources of error. Recent discussions of the subject would so- to indicate that most short-cut procedures fail in their attempts to show real densities. In most cases, true censuses of stall manuals require intensive live-trapping on quadrats several acres in sise (Dice, 1938; Bole, 1939; Blair, 19m; and Stickel, 191.63). lith this procedure the sise of the quadrat and the number and spacing of traps are variable dependirg on the species being studied, population density, sise of the ecological unit under stuck, etc. it best, however, the method is laborious and probably has its greatest utility in intensive studies where the determination of actual population densities is a leading objective. For various reasons the procedure was not adaptable to much of this study. Iany farm situations (fencerows, field borders, strip-crops, etc.) are of such shapes and dimensions that quadrats cannot be laid out. Furthermore, it proved desirable at times to sanple concurrently the rodent densities in more situations than could be trapped cen- venimtly by intensive means. The types of comarisons planned in this study did not require a trapping method which would provide counts of actual densities. However, since some identical situations were to be sampled peri- odically, it appeared advisable to avoid amr possible error that night result from removal of animals. Various other considerations necessitated the use of some compromise method which would permit -a relatively quick and accurate sanpling of populations without causing undue disturbance. The most practical approach to the type of study attespted appeared to be live-trapping along lines of measured length as suggested by Dice (1938) and successfully used in some of the work reported by Blair (1938). Unless otherwise indicated, all population indices were obtained by using a series of 25 traps spaced at 22-foot intervals along a measured straight line 8 chains long. Trap-lines extended from the field border toward the field center and were located michay betwoen the field edges paralleling the line of traps. A few larger fields, 25 acres or more in sise, were trapped with two lines of traps. In this case each line was placed 2 chains in from the field border mich it paralleled. Traps were left in operation for three successive nights. ‘u- 26 The trap used was modeled after one develeped and used by William H. Burt and his co-workers at the University of Michigan. Construction is of pressed fibre board and the inside dimensions are 1-3/h x 1-3/h x 6-1/2 inches. The trap mechanian, which operates without benefit of springs, proved to be efficient and was easily maintained in good working order. In preliminary tests the trap proved to be more effective than the spring-operated metal- type trap in cemon use for mall-mammal work. A. few initial tests indicated that live-traps unbaited were about as effective in making catches as when several types and combinations of baits were used. However, a standard bait combi- nation of sunflower seed and peanut butter was used through the trapping, and a small wad of dry non-absorbent cotton was placed in traps to provide nest material for trapped animals. it the first capture, individuals were marked for future identification by ear-tagging with nunbered fingerling tags, by too- clipping, or both. Follcsring each capture, the animals were given a routine examination and the sex, age, breeding condition and other pertinent facts recorded (figs. 3, 14, and 5). Separate card files were maintained for each individual and also for each habitat trapped. The standardised spacing of traps, length of line, and period of trapping followed here allow for various expressions of mammal densities. The “catch per trap-night'l method of indicating abundance (Grinnell, 1911;) has been used frequently. HOIever, evidence presented by Dice (1938) indicates that the number of 2’? individuals taken along a given trap-line is more directly related to the length of the line than to the number of traps or the interval of operation. It appears that the records to be presented here should have the best comparative value through the use of an expression similar to that suggested by Dice (g. 15:11.), which considers the catch per unit length of trap-line. Trapping results consequently are given as individuals per trap-line. mom a computation of the average catch for two or more lines, fractional figures comonly were obtained. The conversion of the data to whole figures in some cases would introduce an appreciable error. Consequently, the one-decimal calculations have been retained without implying that level of accuracy for the data. 2M3 53 5.30 :15 a on mean. 23. Sony octane-.5. each angel: HHS—m on .uea Fig. 4. The animals were grasped by the nape of the neck and pushed through the opening of the bag for marking and examination. Fig. 6. Animals were marked by toe—clipping or with ear-tags. 30 COMPARATIVE POPULATIONS OF SMALL W5 Annual Nmbers in Crapland Practically all the small manuals listed by Burt (19h6) for the southern portion of nichigan' s lower peninsula were taken in the course of this work. "Most of these, however, occurred so sparingly that a consideration of their numbers would be of doubtful signifi- cance. In the discussion to follow, particular attention is given to the prairie deermouse (Peromyscug mamiculatus bairdli Hoy and Kennicott), the commonest small mammal in the area. The meadow vole, (morotus p_e_nn§ylvanicus pennsylvanicus 0rd) and house mouse (E13 musculus Linnaeus) occurred in nmbers sufficiently large in some situations to allow comparisons. The status of other species is dealt with at appmpriate places later in the teoct. Prairie deermouse In his work on the prairie deermouse on the Edwin 3. George Reserve, Livingston County, llichigan, Blair (191,03) made special note of "a very marked decrease in deemuse abundance from 1938 to 1939”. On one plot studied by Blair, a maximal: population in 1938 of 9.2 prairie deermice per acre was found. During a corresponding period in 1939 the density was only 2.9 per acre. On another study area, a mid-June, 1938, population of 2.0 mice per acre dwindled until none were living on the plot in late August of 1939. Blair suspected that the dry spring and failure of the bluegrass to set 3i seed may have accounted for the small prairie deermouse population in 1939. At Rose Lake the spring and sumer population level of this rodent was at a low point in 1910, and it was the Opinion of farmers and other observers in the area that it was less nunerous than it had been for several years. Intensive studies were inter- rupted during the years 1910-1916, but there is some indication that prairie deermice in the area began an upward trend in 19151 and probably approached a peak in 19h6. Since the intensity of trapping over several years varied seasonally and with different habitats, a direct comparison of all trapping results would hardly give a valid demonstration of the population level. In fig. 6, however, an index to numbers is given for a 2-month period in midsummer. This has been based on individual catches from those lines considered to have been located in accepta- ble deermouse Imbitats. These results indicate a l9hl density twice that of 19h0 and a 19h2 population that was more than five times as high as that of 1910. Limited trapping in the summer of 19146 indi- cated that the trend had continued upward, since the catch for that year was over seven times as great as that of 19140. 5... - PRAIRIE DEERMOUSE [:1 OTHER spams 4.3 4 ._. U i z i 7’ 5: Q < 3- ( C “I P 3 o: 0 U o 3 a 2 2- ( < D 5 O 3 S 2 E. .2. YEAR |94O NO OF TRAP-LINES 30 60 . $2 7 Fig. 6. An upward trend in nmbers of the prairie deeraouse was indicated by the marked increase of live-trap catches in each succeeding year of study. LY' .3 -_n \3 051) other species Species other than the prairie deemouse occurred only sparingly in the upland fields. Howover, among these "incidental species" (mostly the house mouse). a drop in nmbers was noted from 191:1 to 19h2 at the time when the deermouse was showing evidence of increase (fig. 6). At no time dining the study did the meadow vole appear in any numbers in cropland during the summer months. However, populations of this species at Rose Lake were conspicuous in some habitats during autumn and winter and showed an unorthodox behavior which will be discussed separately. Relative Numbers in Various Crop Types Although the literature dealing with small mammal populations is extensive, the reported studies invariably have dealt with natural associations. A few investigations have contributed inci- dental notes concerning the numbers of small manmals in and around farmlands.. Hanson (19%) working in Wisconsin found pepulations of the prairie deemuse to be conspicuously higher in a prairie- relic situation than in any of several farm habitats (grain stubble, alfalfa, and bluegrass-sweet clover fields). Hanson's trapping also showed a fallow field heavily grown up with beggar's tick (M), martweed (PolEonum), yellow foxtail, and ragweed to have been one of the most predictive situations studied for several species of small mammals. Johnson (1926) comonly took . ___,i.-.i _‘—-. ..— 342 the prairie deermouse in fields of gain and alfalfa in Illinois but indicated it was uncommon in "neglected areas such as the matted grass of roadsides'. Calhoun (19131), in Tennessee, found greater numbers of house mice in a Legedeaa planting than in any other of a variety of situations trapped, and Southern and Laurie (19146) in England found greater densities of house mice in wheat ricks than in either cats or barley. A program of winter trapping in Ohio by Dalmbach (1915) showed small mammals to be somewhat more plentiful in fields of shocked and machine-picked corn than in several other crop fields. Prairie deermouse In 19140, preliminary trapping was done in fields of wheat, oats, timothybclover pastures, and sedge meadow. Among these habitats, the greatest numbers of prairie deerndce were taken in wheat fields. During midsummer of the years 191d and l9h2, a number of fields of each of several common crops were live-trapped to determine the relative numbers of small mammals. Trapping was done at a time when the crops were at or near maturity. However, an attempt was made to adjust trapping dates so that all fields could be sampled as nearly concurrently as possible. In addition to the work in cropland, several units of idle ground were also trapped. In each case these idle areas were grown up principally with bluegrass (£9: compressa and g. pratensis) , although there was considerable variation with reapect to species and amounts of other herbaceous types. Fig. 7. The prairie deermouse ranked first in abundance among the sun mammals of the area. Fig. 8. 0f the common fem crops, corn fields were among those most heavily populated by prairie deemice. hiring both years (191:1 and 19112) the only small mammal taken in sufficient mmbers to permit comparisons was the prairie deer- mouse. The relative densities as found for this species are shown in table 1. With the extremely low population present in 191:1, no real difference is evident among the numbers taken in each of three types of grain fields or in alfalfa fields. The numbers in idle ground, however, were considerably fewer than in any of the adjoining croplands studied. With a considerably higher (but still low) papulatlon in 19h2, there was still no convincing evidence of any crop-type preference by the prairie deermouse. is was found to be the case the previous year, idle-land habitats supported appreciably smaller populations than did farm fields. It is of possible significance that among croplands, red clover fields had the lowest numbers of deentice in both years. In 19142 a dense, 20—acre stand of red clover which immediately Joined fields of alfalfa and wheat was trapped concurrently with the other two. No prairie deermi.ce were taken in the clover, whereas trapping of equal intensity accounted for 18 deermice in the alfalfa and 8 in the wheat fields. Hanson. (19W) in referring to the prairie deermouse commented that: ”....this species seemed to prefer situations where the inmediate ground cover varied from light to moderate densities'. It is not unlikely that this factor was important in limiting deermouse numbers in red clover fields which, in each case, were extremely dense and matted. o...” men m..." 0.0 Dem woo men men 2579.3. a e I I Hangman e 06 s so. 3.388 a ”so a esoeeenho: menu: 356.5 33 m «A m .853 com m men 5 5:34 o How NH Eco OH Com m fees? 3 men mm 38 seams: «Nagoya you ceamfisu mono mflssfiapgnu up." can sues-nos a.“ 33.33 vacate?“ was no.3 2.62.? a.“ oesoenoov 92,—8.3 one no shoe—5a chap-season 2" snag as In the late fall of 1912, most rodents in the area were considerably more plentiful than they had been at any time during the mmmer. A number-of the same situations which had been studied several months earlier were trapped again in October and November to determine possible habitat preferences. In these data, which are sumariaed in table 2, the relative ratings of various crops with reference to deermouse numbers did not show any marked changes over conditions found in the two previous summers. Trapping results suggested a greater preference for alfalfa and wheat to oat fields, and in one area of strip-cropping a rather remarkable difference in populations in these three types was observed. The three craps were in adjoining strips, each 100 feet in width. Both the oats and wheat had been harvested with a combine, and the alfalfa was about 10 inches in height following regrowth from a second cutting. All were trapped in an identical manner during the period Novaaber 3 - 5. The comparative nunbers of deemice for each was indicated by a take of 23 individuals per line in the alfalfa, 22 in the wheat, and only 11. in the cats. The data shown in table 2 were obtained following the harvest season. Certain characteristics of specific fields, such as crop residues, type and density of ground cover, etc. , remained to distinguish the particular crops grown on these fields. It would be well to point out, however, that the croplands in this late-fall period were similar in many respects to the annual weed stage in succession from plowed ground, which Allen (1938) describes; and in a large part to the marched-foxtail situation studied with 3:; Table 2. Comparative numbers of the prairie deermouse in various crOp and idle land habitats in late autumn (October - November), 1942. No.of fields Individuals CroE sampled per traprling_ Oats (stubble) 3 6.3 Wheat (stubble) 1 22.0 Rye (seeding) 2 21.0 Corn (mechanically picked) 6 11.0 Alfhlft 2 14.5 Red clover (out) 1 4.0 Pasture 2 5.0 Idle ground: Mixed herbaceous 3 6.7 Fencerows (sod) 4 0.8 40 respect to its small mammal inhabitants by Hanson (191414). At the time of trapping, most of the fields were vegetated with lesser ragweed (Ambrosia elatior), foxtails (Setaria lutescens and §.’ viridis), lanb's quarters (ghenopgdim M), snartweed (Polygonun spp.), and to a lesser extent by barnyard grass (lichinochloa omegalli), redroct pigweed (Amranthus retroflexue), tmbling pigweed (A. graecisans) , black bindweed (Polyggnum convolvulus) , and old witch- grass (Panicum capillare). In l9h6, populations of most, if not all, snall mammals at Rose Lake were conspicuously higher than they had been for a number of years. A limited amount of trapping was done through the spring and summer to determine whether relationships noted during the period of low-to-nediun population density still existed. In larch and April, two fields of corn stubble, which still contained a number of shocks, had the highest population of the several situations trapped. The average catch was seven prairie demice per trap-line. The second-highest density, averaging h.6 individuals per line, was observed on three trap-lines located in fields which had not been cultivated for several years . Vegetation on these plots was similar to the annual wed association already described. In two rye stubble fields an average of two deermice per line were taken, and in five field and roadside situations of typical grassland habitat the lowest take, of 0.6 individual per line, was recorded. In June, four of these same bluegrass associ- ations were again trapped without tmg a deemouse. 4.1 During early August of this same year, adjoining fields of corn, oats (two fields recently cut), alfalfa, red clover, and bluegrass sod (tam fields) were trapped with the following results: corn - 10 prairie deemice per trap-line, alfalfa - 7, cats - 5.5, red clever - l, and bluegrass fields - 0.5. In considering the over-all results from trapping— conducted intensively during two summers and periodically in two other years— prairie deemouse preferences for specific crops are not strongly shown. However, certain consistencies appear in the data which sees to have significance. In general, grain fields supported higher papulations than other habitat types, and in this category corn, wheat, and rye fields were more densely populated than fields of cats. Trapping in alfalfa regularly produced catches which indicated densities comparable to those in grain crops. The nmbers taken in red clover plantings showed it to be the least acceptable to the demouse of any of the annual crops. Idle-ground situations, which were mostly covered with bluegrass, maintained the lowest populations of any of the farmland habitats. In the area of this study the natural plant succession on lands permanently retired from cultivation would lead in a decade or so to brushland and later to oak-hickory woodland, the most advanced seral stage present. Much of the idle ground in proximity to crop- land is naintained as a grassland habitat, however, due to various lumen disturbances (burning, pasturing, cutting, etc.). The compara- tively low populations of prairie deermice in these areas is not surprising in view of the observations by Dice (1923:, 19232), "IN Johnson (1926), Allen (1938), Hanson (191414), and others. Johnson (1926) has indicated that the species originally belonged to some sub-cum communities in the Andropogon prairie- climax association. In Michigan the prairie climsx was present in only limited portions of the southwestern part of the state, and the distribution of the species was undoubtedly confined to those openings. As has been shown by Hooper (19142) its spread over most of the southern peninsula of Michigan was a result of deforestation and cultivation. Recommendations for the progressive operation of Michigan fame call for increased acreages of permanent pasture, and the retiremmt of many hilly areas now in cultivated crops. The effect will be that of placing a substantial percentage of untilled land in a seral stage succeesicnally advanced from the 'armml weed-cropland association," the type of idle ground observed to be most favorable for the prairie deemouse. Other species Aside from the prairie deermouse, the house mouse was the only other mall mammal to appear consistently in cropland studies. While the numbers of this animal handled were too few to justify an detailed consideration of its status, certain observations are of interest. In general, the comparative population levels of the house mouse in major farm habitats indicated the preferences of this species to be similar to those of the prairie descrmouse. Of the few individuals trapped in summer and fall studies, most were taken in grain fields and the fewest in idle ground (table 3). On the basis of frequency distribution rather than actual. density, a similar tendency is noted in the data, the species having been taken in 18 (140%) of 145 grain fields studied in summer, 8 (35%) of 23 alfalfa and clover fields, and in only 3 (16%) of 19 idle land situations. In discussing the status of the house mouse on English farms, Southern and Laurie (19146) state that farmers usually report more damage in cat ricks than in other types of grain. Their studies indicated, however, that these mice may have been more plentiful in wheat than in other kinds of ricks. On the basis of summer trapping at Rose Lake the species appeared to be of more general occurrence in cat fields than in other farm habitats. Their frequency distri- bution in various situations was as follows: oats - taken in 10 (53%) of 19 fields studied, corn - 6 (50%) of 12, alfalfa - 6 (32%) of 19, idle land - 3 (16%) of 19, and wheat - 2 (114%) of 114. Numbers with Reference to Land Uses and Farm Practices strip-creppig .On erosion-susceptible farm land, strip-cropping is being used widely as a soil- and water-conserving measure. The technique results in fields which are in marked contrast with the pattern on conventional farms. Crap types are closely interspersed, and cultivated fields are alternated with close-growing crops which in many cases, are undisturbed by cultivation and plowing for extended periods of time. The extent to which such intermixture of several 4%? .13” moo." O.“ 3 m.o .38 Esutuflu .. ._.— ufilua 3m I .8 5.3. sogéfim .. use 026: ensom I 5 each house: I 5 252536 can.“ I an neglmeua m you .n 53» need emcee: * a4 a.o m.o can w.o t. 06 m5 To m6 m...” m.~ a o.» m a.ua «a m.H m." H3 .2 in on «.o n.” n...” ma o.a o: .36..“ send” . I993 He .0: .9333 aqua-.3.“ He .25.. none- eennv 5 odes-8w dud-e we anon—3s surges-adieu .m 0.33. «.58» 03H 393 how 398 338 959% 98H .998 he: nacho sauna unseen ech 23.8 new nacho 43.8 2:3 as om? N49” sepalboz L338 ma has 23 €53 I52 .33 I} crop types might encourage rodent numbers has not been investigated. Neither has it been determined if relatively undisturbed habitats, as represented by hay strips, would maintain populations and serve as centers of infestation to the adjoining crops. An.attempt was made to compare the relative populations of small mammals under the two methods of fanning. Using identical procedures, fields of cats, wheat, and alfalfa in an area of strip- crepping‘were trapped concurrently with fields of the same crop grown under conventional methods. In.each case fields under compari- son‘were in comparable stages of growth and were located in the same immediate areap- frequently in.adjoining quarter-section.land units. Results of this study are shown in table h. Since the cover requirements of the meadow'vole would appear to be more exacting than those of other species in the area, it might be expected that populations of this mouse would show the greatest response to differences associated with the two methods of land use. It was eSpecially regrettable that significant summer populations of the species were not available for the comparison. In the case of the prairie deermouse, however, it will be seen that the variety of habitats provided by strip-cropped plantings gave no evidence of encouraging populations of this species. And, in. fact, the catch in both grain and hay fields was somewhat higher under conditions related to the conventional "block" fields. This apparent lack of response of the prairie deermouse to the greater interspersion.of herbaceous types was further shown‘by other work summarizedfibelow. 46 wow van «on Dem 2:795 «on :88 «H in on ma eeaanINeua no .on «a. «H .33...— t .8 H16 apnoea—co venue.“ 0|?“th Handed—:80 common "than“; 352. at ouegoueIndnvu one gm hand—83:028.. emcee a.“ ecu—Been outdone we anon-5n chanteuse .0 menu “and”: .hdm Cheek one 3-3 .35 :96 g r“ ._.... «to Field distribution of prairie deermice.— In the sunmer of 19140, a number of fields, selected for having dense marginal growths, were trapped, using lines which extended from the edge of the field to a point beyond the field center. In this (preliminary work the mmber of mice caught in traps located well within the field was actually higher than in traps located near the field border (Linduslca, 19146). To determine further whether or not the field distribution of prairie deerfice bore any relationship to the edges, trapping returns for 19111 and l9h2 were analyzed. It will be recalled that a standard- ized procedure of trapping was used in practically all of this study, and that one requirement concerned the placement of 25 traps in a straight line beginning at the edge of the field and extending in the direction of the center of the field. The first trap in the line was set at the field edge, and, in each case, the 25th trap was 528 feet from the edge. In many cases the crop margin was adjacent to a fencerow, roadside, ditchbank, or other uncultivated land. These supported a variable type of vegetation which included successional stages from annual weeds to mature oak-hickory. The primary function of the trapping was to establish comparative population levels in particular crops. Consequently, the first trap in each line was placed in the edge of the crop rather than in the marginal vegetation. lhile the trapping returns can be expected to show possible influ- ences of crop-border vegetation, they do not allow for a comparison of rodent numbers in fields as canpared with numbers in the field margin! 0 ’ I.1 1111‘ "I 1 ) It .330 sang we neon-sq no 332.3088 a no concern has some 3 sodas.“ ease-Seep 0393 05 no.” apnoea." A33 Ho mush—”sea :4 .m .3..." moon DJME ..na_4 >m OwUZwOSm m< MOOm 04m; O._. Fuwamwm It; ZO_FDm_N:.w_O uU_§mmmo u_m_.an mas—Eel 3!.» 2: ensue .eesoEeeu 3.393 on» you vacant—ea so a Moon he voted on» on on 683%.. hens—262 5 guest 25. #918 SE“ “WWW Nwo. C(u> duo .>02 #00 H 15.20: 73 ._-e nucvu 8-: 26. a.-. on... 93—2-: e... «in «:3 2-2 :6 Y. «-o o m>02m¢ OZ( UZEQ<¢P >0 23075 m( mPZU2u>O§ 20:303....90 UU_2¢UUO NEE”: O3ddVUJ. S'IVOOINONI JO UBBNHN ’75 The nearly continuous 'ranoval trappingfl to determine seasonal drift rates provided an opportunity for observing any "homing" tendencies in the prairie deermouse. Three release points were established along a line perpendicular to the lines of traps at distances of 200, 1:00, and 600 yards. A fourth was located in the direction of the trap-lines extended and 265 yards from the nearest trap. A total of 136 releases were made from one or another of these points. Records for all individuals which returned to the plot one ' or more times following release are given in table iv in the appendix, and the results are summarised in table 13. Host of the animals released from points 1, 2, and 3 were taken in the first period of trapping and the rmainder in the second period. Since presumably these were practically all residents in the area there would be an apparent incentive for a return to the plot. All of the individuals released from point )4, however, were animals taken on the plot after "removal trappingfl had been in progress for about 6 weeks. Probably most if not all of this group were tran- sients, and without having established home sites their return to the plot could be attributed largely to random movanent and chance finding of the traps. A total of 70 of these "non—residents" were involved in 99 released from a point 265 yards from the nearest portion of the trapped plot. A return of 38 percent of the adults and 30 percent of the imamres was obtained (table 13). Considering that most of these animals should have had no attachment for the area of their first capture, the rate of reappearance was unexpectedly high, and can be explained only on the basis of widespread and erratic move- ments. The fact that several days had elapsed before a number of them returned would indicate undirected movement. One of these individuals reappeared on the plot after 131; days, which included 39 days of trapping. Animals which were resident in the plot had little difficulty in finding their way back from the release point 200 yards away. Returns were obtained from all of eight adult releases and from three of four releases, of young animals. The fact that 9 of the 11 which returned made the trip the first night (recaptured the morning following renoval), and in many cases to the same or an adjoining trap, indicated well-directed movement. Over a distance of 100 yards their success was lower. Only slightly over half (four of seven) of the adults returned to the point of capture, and none of five young returned. In this case, however, a possible obstacle to their return was imposed. This was in the nature of two strips of marsh- land nearly convergent across a part of the route to the plot. However, the straight-line path from release point to the plot was unobstructed and at least 75 feet wide at the narrowest point where it passed between the projections of marsh. In the case of releases from the 600—yard station, no returns were obtained from six adults and seven juveniles. The 600-yard "point of liberation” was an extension of the distance of the hOO-yard station and the same conditions of travel applied. 861i .332..." have“! .3 fine.“ one." seam snow hue-pone one one»! as 0.36:5er on: one: anon- ealflne Baruch; gonad! pen an a..» .nv a. cannon .an» ». annapnna-nopoe on. .Hann_ ..»H... 3.3.... 2. anus... .2... 3.. .3» P as?» .Hsavnefl £4 \m .333 mean. an... a... $33 mesa:- 3: Ion.“ 350..» can.“ no adds a loan ens—53.533 Has no defies—en was «mama-u». 5 none». eodfieee sienna no canon» end .3" 0.3g 81 o-ouflmauv or». on». 5253 .9359. an: o» Hogfiun on: a so :3 ¢ £32" 0333 3995 no 33 2.3 3. 53:32:5qu an: a no vapooon 0.3.! n can N .H .952” 0:0qu ... on m» o o o S 2. 8H 35% :2... 39c wanna»: «820m 3 n o o 0 ¢ n a 3.03 vs ©2393!» waves 3 o N. m m u v m 3309 no: vu 3:38 7pr m H o o o H .n n 30.3 an @9399: Saw: .0093»: con mm m o o o m H n 50:38 5am: 32:8. $23.3qu 2 m o o o n .m a $on 8 ““3632 .HSuFHBd go .02 8 m p o o o n a $52" not 8t 32.3.: Sgfipgfi «o 62 .55 9mm :5 .dm .fifi Hm hum-H1 um «:58. no send .Egmhmm .Eambpm sum» Ev .Efl com 33m 5 9.3 £88“ 3 883.3 ¢ n N H thong Eon ouuoaom c.3633 ugcflsofiaua uncuwfioaom new... :32: £93.85: Eoum 3.3:.” no ragga .3 93:9 Food habits.- The food habits of many species of small mamals are imperfectly known, and in this reSpect the prairie deermouse is no exception. Since an appreciation of the economic importance and ecological status of the species presumes some knowledge of its food habits, a few observations were made and these are recorded. Attempts to establish food preferences through examination of the stomach contents were mostly unsuccessful. Materials were always finely macerated, ‘and the method would permit only rough comparisons of the relative amounts of insect and plant ranains. It appeared likely that the storage traits of the species might provide means for food-habits studies. In this connection, 914 chalk boxes, and other boxes of similar size, were distributed over the study area. These were provided with a hinged top and had a 1-1/2 inch hole near the upper margin of one end. About half of the boxes were buried to a depth of 1 foot or more and provided with a wooden tunnel leading to the surface. Most of these were set out during the summer, the remainder during early fall, of 19142. Examinations of a few boxes were made at frequent intervals, and systanatic inspections were made of the entire group during the third week of Septanber and second week of November of the same year. Although there was ample evidence of the boxes being used for protection and shelter, they received only limited use for storage of food. Among 30 which had been placed in woodlands or woody situations, eight included food and food remains which probably were the property of white-footed mice. Only 14 of the 614 boxes, which had been located in the habitat of the prairie deermouse bi‘h (crap fields, fallow ground, etc.) contained food items. Two of these four had been placed near an isolated hickory tree (21.113 galls) in the center of a strip—crOpped field. Eccept for a few insect remains, the nuts of this tree were the only itans stored. It could not be definitely established that this was the work of prairie deermice. A third box in a fencerow adjoining a corn field contained about SO c.c. of shelled corn and a few pieces of unidentified fungi. The fourth cache [was taken from a box located in a strip-cropped field. While the volume of the stored food items was not large (80 c.c.), it included the same wide variety of materials in two natural caches which had been. found previously while excavating woodchuck dens. The contents of the storage box last mentioned together with items identified from the two natural caches, are listed in table 11;. During the fall and winter of 191.0441, Baumgras (191.3) made quantitative studies of the weed-seed and crop residue on the Rose Lake area. During October the rather remarkable quantity of 211.5 pounds of weed seed per acre was present in wheat stubble, which was the most productive farmland habitat studied. Ragweed was the most abundantly available species through most of the winter. Leedy (1939) also found ragweed one of the commonest fall foods on Ohio farmland. He reported an average production of 85.9 pounds per acre of ragweed seed on 15 wheat fields sampled. It is interesting to note that of the 10 species of weed seeds found by Baumgras to be most comon on the Rose Lake area, 8 were included in the 10 species found in caches of the prairie deermouse. Table 1h. A list of food.items taken.from three caches of the prairie deermouse. Food Ite- Anbrosia elatior mafiamfinma POI onun fieriicaria R¥§mmmmflfifie ll mp 118 Meg) GryIEs app» ) .Mumhmrumfluu Amaranthus aecizans getaria“lutescens REHZHHHT' mm“ 3—133”? cmca e Rand umuqmnus hwutmm Bhfibhbud Misc. insect remains Rama mmnmpunu hflwfufiflqnu @unmnubuus fifllfiuudmaay mauwmmu mm» wunwmu) In“ mm Mbmphflrmdm Mummy @wnmm HHHHPHHHHHHN NNNu 84 Fig. 23. '; . [It Ti.“ L' {-1 I'c‘.‘L".A1!.-QI\!‘SS iHan. ~,.*.',;',1.1 ' : The contents of a prairie deermouse food cache. 86 A random collection on the part of the mice was further suggested by the fact that ragweed and lamb's quarters, the two most plentiful species on many parts of the area, were also greatest by volmne in the food caches examined. Allen (1938) found these same two species were among the most available, and by far the most frequently used by songbirds in winter. There was no evidence during any of the winters of this study that shortages developed in the supplies of wildlife foods. The observations of other workers have shown, however, that prolonged periods of deep snow may bring an acute competition among various species of wildlife (Allen, op. £11.; Errington, 1931). In these periods, the food-gathering of the prairie deermouse and other rodents might contribute to privation for more desirable game and songbird species. Meadow vole In each of four years of study the meadow vole at Rose Lake followed a definite pattern of late-autumn increase and spread into the uplands. Various circumstances associated with the winter population peaks and movements of these mice appeared to be atypical of its usual behavior as reported from other areas. Seasonal numbers .- Throughout the period of the study meadow voles rarely were taken in cropland during the sumer months. In 191:0, summer trapping for 2,1;146 trap-nights failed to take any voles, even though the work included habitats known to have had substantial populations (bring the previous winter and in early spring. As late as the first week in September, 191.0, no voles had appeared in trapping, and indication of recent activity could not be found. A note by D. L. Allen at this time stated, "Meadow voles have been conspicuously scarce at Rose Lake during the past year. It appears that the species is at a cyclic low." In late September the first evidences of voles were noted in low areas of dense bluegrass growth, and on October 17 a nest of seven young was found in the uplands by Clare File. Notes by several staff members reported a rapid increase chlring Novanber, and by Decanter these mice were common in many situations. Fields of shocked corn, which were uninhabited by voles in Septmber, were supporting an abundant population in late January (Linduska, 191.23) . Over 300 individuals were taken from 315 corn shocks on 28 acres, and over 70 percent of the mature females were breeding. The substantial winter populations of this mouse continued through the winter and early spring of 19141. By the time the last snow covering had left the fields (early April) signs of a decline were apparent. Snap-trapping in stubble fields, which a month previously were well populated with meadow mice, failed to produce catches. By the time farming activity had begun in spring, even the favored winter situations were without voles, and in the first two weeks of July only one animal was taken in over 1,200 trap- nights in fields of alfalfa, red clover, wheat, oats, and grassland. This adult fenale was taken in a retired bluegrass pasture inedi- ately adjoining the field of shocked corn which during the winter 813 had a very high population, and in which a heavy loss of corn was measured. Intensive trapping through the remainder of the summer, for a total of over 6,000 trap-nights, resulted in the taking of only three additional animals. All three of these were taken in late August on a trap-line located in dense bluegrass sod in an upland oak-hickory woodlot (woodlot number 6). In this same locality ll short-tailed shrews were trapped, the only individuals of that species taken in summer trapping. Although intensive summer trapping failed to reveal any other colonies of meadow voles, there was some evidence that the summer population in 191:1 had not dropped to the extreme low of the previous summer. Careful examination of most of the marsh and lake-edge habitats in the area disclosed a scattered population, and four individuals (three of them breeding females) were taken in a sedge- cattail marsh. This summer population exhibited the same behavior as was observed the previous year, i.e., an autumn increase and spread into permanent sod areas and meadows, and in the winter an invasion of most of the cropland on the farm. In one crepland habitat, l9 snap-traps operated for 1 night in February took eight sub-adult voles. The decline during the spring of 191.2 followed the same pattern as described for the previous year. A concerted attempt was again made during the summer of 191.2 to locate possible colonies in the area by field observation and by trapping in suitable habitats. Of the season total of over 9,000 trap—nights, nearly a third was directed at this species, but only 25:11 three individuals were caught from July through September. During the first 2 weeks in October signs of these animals were noted.in several upland sod areas and an occasional specimen was taken.in trapping. The first nest of the season, one of five young, was found by D. L. Allen on October 23. By late OctOber Microtus was present in fields of clever and cat stubble as well as in many of the blue- grass areas, and.in the period October 28-31, 56 were caught on 10 trap-lines operated for a total of 727 trap-nights. As in previous years a rapid increase followed with an annual high.being reached in late January or February. In.midsFebruary, SO snap-traps operated for h nights took 36 voles from less than half an acre of’timothy- bluegrass pasture which had been uninhabited as late as September (figs. 28 and 29). By April of 19h} croplands again were practically devoid of meadow voles. Only occasional individuals could.be located and these were in the low'marsh areas. In 1915 and 19146, meadow voles were reported to be more plentiful over many parts of Michigan than they had.been for several years (personal coummnications from Game Area Managers of the Game Division, Michigan Department of Conservation). .Although.an intensive study was not made, sufficient work was done at Rose Lake to determine samething of seasonal pepulation trends in a year when the numbers 'were appreciably higher than they had been at any time during the 19110-142 period. 9U Fig. 24. The meadow vole was never abundant in cropland during the growing season. Eig. 25. In late fall and winterllicrotus was salmon in most situations. 91 .8335 .3 .s .3 8.9a 5?; a8?» one? n30». houses .3 manna: ad .3005 one .am .maa .832...“ .3 .3 .3 325 :32 so use“? at as 38% ES 59¢ some» 82 08 has 985 no fee 4. .3 .wE 1CD,- “ V 92.: As in earlier years the population in 19145 increased rapidly _ in late fall and rose to a peak in midwinter. There were indications, however, that the spring decline occurred somewhat later than usual. As late as the last week in April, 1916, voles were still common in all cropland, and during the last 2 days in March and through April, an average catch of over five per trap-line was made on nine lines of 2S traps operated for 3 days. In one field of rye stubble, trapping of this intensity produced 17 voles in a 3-day period of trapping from March 28 through 30. The results from one line of 50 traps operated periodically in a dense roadside sod give an indication of the subsequent rate of decline. hiring the first week in April, the 50 traps took 12 animals during a 3-day period. The same intensity of trapping produced eight individuals in mid-April, and four in the last week of April. None were caught when the area was trapped during the first week in June. That this reduction in numbers was not a result of a beginning period of sexual inactivity was shown by a high rate of breeding in a sample of snap-trapped animals taken during this interval. Among 23 mature females taken in March and early April, 17 were pregnant and the other 6 either were lactating or showed other signs of sexual activity. There was some evidence that the l9h6 decline was not as pronounced as in other years. During midsummer, voles again were scarce in cropland but apparently not at the extreme low of the l9h0-h2 period. Less than 500 trap-nights of operation in cropland produced four in August, whereas trapping of from S to 12 times this intensity in the same areas yielded none in l9h0, four in l9hl, and three in 19h2. Signs of activity in wet marsh areas were noted throughout the summer of l9h6, and there is no doubt but that a greater population carried over than in any of the preceding years of study. -It has been reported that the meadow vole arrives at an annual population high about September, and cessation of breeding in winter, together with a somewhat higher mortality rate through this season, brings the annual level to a low in March (Hamilton, 19323). At Rose Lake the annual peak appeared to be reached in late January or February, and the reduction of this pOpulation in early spring, rather than being a dwindling process, occurred rather rapidly. The factors responsible for this sudden elimination were not entirely ‘ evident. The crash decline of plague populations is usually attributed.to epizootics (Piper, 1909; Howell, 19233‘Wayson, 1927; Selle, 1928; Findlay and Middleton, l93h; Elton st 21” 1935; Hamilton, 19322). Pathological causes were not given detailed consideration in this study, but certain characteristics of the population suggested that factors other than a contagion were instrur mental in the decimation of these animals. Voles, while abundant, did not approach the outbreak level usually associated with the appearance of epidemic diseases. Also, the fact that the drop occurred in several successive years would suggest that factors other than disease were involved. In each year the reduction, and in some cases nearly complete elimination, of voles from the area, was identified with the period of early spring thaws. It is conceivable that conditions resulting from repeated flooding and freezing would be critical for these animals, and the relationship is held to be of considerable importance by at least two students of mall-mmal populations. Sviridenko (1931;), working in Russia, found the temperature regulatory mechanism of mice and voles to be relatively unstable, and noted that, "....any great deviation from optimum conditions rapidly upsets the balance of its biological functions, and often leads to death." He has applied these findings to field conditions in explaining the sudden disappearance of large rodent populations by abnormal weather conditions. In considering an abrtqat decrease of Microtus in Norway in 1935, Jessen (1937) rejected epidemic diseases as the primary cause. The true reason, he thought, was a sudden thaw which directly killed many by drowning and freezing, and produced conditions leading to the loss ofmany more through a combination of loss of body heat and deficiency disease. Jessen theorized that, "....meteoropathology and deficiency diseases first and foremost, not epidemiology, ought to be focussed upon when dealing with the problems of fluctuations in the populations of mall game.” Movements.- Much of the attention given to Microtus has centered around its cyclic behavior and economic importance during epidemics. While many studies report on movements, and/or extensions of the range of the animals from what is ordinarily considered to be their normal habitat, the incidents usually are associated with cyclic peaks, a heavy cover of snow, or some prominent change or disturbance in the preferred habitat. The occasional appearance of _‘ ._. I. 31.9 these animals in uplands during winter is well-known. At these times damage to shocked grain and orchard trees or other woody vegetation is a common occurrence. The factors responsible for these winter movements usually are assumed to be a shortage of food in the summer range and a blanket of snow which provides needed cover. Hatt (1930) said, "The inhospitable winter conditions in these areas [swamps] are a major factor in keeping down the mouse populations, and in causing them to move to the uplands in the winter." This view is shared by Hamilton (19h0), who also indicates that the movements are facilitated by a covering of snow. Numerous other authors reporting on winter damage by Nicrotus have suggested a relationship between deep snows and winter movement (Lents, 1906, 1907, 1918; Bailey, 1900, 1921;; Siegler, 1937 3 Allen, 1910, 19141; Parker, 191m). The invasion of upland areas by the meadow vole at Rose Lake represented a situation somewhat different from that described by other writers and is of interest from several standpoints. It was observed in each year of study under a variety of weather conditions, and appeared to be of usual occurrence. The earliest signs of activity, were noted in uplands prior to the first killing frosts, which in 1910 and 191:1 were recorded on October 16 and 27 respectively (table iii, appendix). In 19142 a killing frost occurred on Septanber 29. However, no appreciable change in marsh vegetation was noted until after repeated frosts in late October. By this tine voles were common in cropland. It was evident also that a snow covering was not a requirement for movement of these mice, since in each year the species was established in cropland before permanent 9&5 snows. It is possible, however, that winter snow was important in promoting the substantial population increases that followed. The factors which initiated movement to uplands and also encouraged high winter populations on this area are not well-known. It has been shown, however, that Microtus has strong "migratory" tendencies. Blair (l9h0'b) found that transient animals made up about 12 percent of the total population in July and August trapping periods. These wandering habits also were noted by Harper (1929), who found a colony in a small, well-isolated area of favorable environment. Townsend (1935: p. 99) cited instances in which the animals apparently had crossed fairly wide stretches of hostile ground to colonize remote and isolated patches of their habitat. He states, "The examples cited suggest that occasionally merotus may travel relatively long distances in search of new locations..... Such a 'pioneering' tendency might have an important bearing on the successional changes in the small-mammal population of the meadows- for example.” In connection with these early fall population shifts, some observations by Blair (op. git.) are of interest. In his work with Microtus on the Edwin 5. George Reserve, he noted from trapping on dry grassland habitat that the first vole appeared in mid-June. With further reference to numbers in the dry grassland he stated, "The late September muimum is in marked contrast with the decline at that season in the moist grassland, and suggests movement from the moist to the dry habitat." These studies by Blair did not extend beyond September and the population trends through late fall and winter were not detemined. Fig. 28. Devoid of meadow voles in September, this timothy-bluegrass pasture supported an abundant population in mid-winter. Fig. 29. Close-up of a portion of the field in fig. 28. The moth-eaten appearance of the sod is indicative of the sub- nivean activities of these animals. 5N5 0.5.323.“ doe 0.3» 30.6 nova-v one: game 5.. eon-Has a." vase.“ none» #913 on» .33» no use» nose aH _ , e\'.. NR. » . . . .. 3"; VJ... ”as... .8 SE At Rose Lake the first individuals were noted in upland dry-grass areas in September and October (fig. 30). In general this movement to high ground was observed coincidentally with evidences of a popu- lation increase in adjoining marsh and swamp areas and probably was related to it. While these earliest appearing individuals probably were invaders from the wet grassland, it is very likely that the subsequent high populations of fall and winter resulted from multi- plication in the uplands and not from a mass population shift. The fact that voles did not appear in cropland until the autumn months, after increased populations were noted in marshland,. would suggest that the uplands in this area are in some way unfavorable during the summer. It is possible that the comparative dryness of even the bluegrass habitat in this morainic farmland is a factor in some smnmers. Although June to August precipitation in the area tends to be comparable in monthly amounts (table ii, appendix), the cooler temperatures of early autumn produce more humid ground conditions. It is usual at this tine for many of the herbaceous plants to undergo a short period of rapid growth, and the vegetation during early fall appears to be in a condition generally favorable to voles. Furthermore, many of the upland grasses remain succulent under the winter snow which normally persists from December to March. It is possible that the autumn movement to uplands, and the unseasonal breeding, with its resultant delayed annual population peak, is a reflection of the xerophytic conditions in this area. An observation of Hamilton's (191413) would tend, in part, to bear 101:; this out: "Drought conchtions at any period during the most active breeding season (March through October) would limit the growth of suitable foods and reduce the cover to such an extent that the mice might migrate to a more moist area and thus conceivably disrupt the normal processes of reproduction. It is well-known that field mice do leave areas when such areas become unusually dry, to take up their residence in more favored localities. While it can be demonstrated readily that severe drought conditions result in a reduced mouse papulation, the reduction appears often to be of a rather local nature." Breeding.- It was obvious throughout the study that vole populations under the xerophytic conditions of the morainic Rose Lake farmland differed from the behavior of the species elsewhere. Although a detailed investigation of the biology of the many small mammals dealt with was not within the sc0pe of this study, it was evident for the meadow vole that certain life history incidents were atypical. In the summers of 19140 and l9hl, when scarcely an animal could be found at Rose Lake, Don W..Hayne of Michigan State College was studying a substantial breeding population less than 12 miles distant. One small alfalfa field in this mesophytic situation being studied by Hayne had a population of about no voles per acre. In spite of repeated efforts, spring and sumer colonies could not be located at Rose Lake, and the extent of breecflng in the normally active period, consequently could not be determined. It would seem logical, however, that if production of litters had begun 101' in early spring and continued at a high rate, some evidences of voles would have been noted in favorable habitats by July or August. With minor exceptions this was not the case. In this connection observations of other workers offer a possible explanation. Bailey (1921.) says, "Scarcity of individuals may often tend to retard or entirely to check reproduction. The mice may be left stranded out of reach of mates, when breeding perforce ceases until stimulated by a gradual influx from other localities." Hamilton (1937a) also has suggested that the chances of females in oestrus meeting fertile males is greatly lessened in a low population, and under such conditions females might repeatedly pass through receptive periods before mating occurs. In this connection Hamiltonobserved that following the crash decline of the vole population from the winter of 1935-6 peak, '....a smaller share than usual of the surviving mice were breeding.' However, his tabulation of percentage of nursing mice that were gravid during each month in a h-year period, shows that the reduced 1936 population bred at a high rate in October (73.3 percent). In the period of cyclic increase (1933-35) the breeding rate in October was appreciably lower in all 3 years (37.5, 20.0, and 136.7 percent). Considering the extent of the late-fall to winter increase of voles noted in this study, a high rate of breeding must have been maintained over this period. While the details of this unseasonal breeding were not investigated, some observations were made. In connection with a crop damage experiment in 19140, 288 individuals were collected for study in January, 19141. Among 150 females, .102 70 percent of those judged to be mature (weighing 26 grams or more) were pregnant or lactating. Although these animals were taken in or near shocked corn, limited samples from other habitats indicated a general reproductive activity (Linduska, 19h2_a_). Sufficient trapping to establish possible differences in the breeding rate of individuals living in shocks as compared with those in other situations was not carried on. The studies of Venables and Leslie (l9h2) in England have shown, however, that winter populations of the brown rat in grain ricks breed at a significantly higher rate than those in non- riclc environments. In 1913, 80 individuals were snap—trapped at intervals from January 29 to March 5. Of 16 mature females taken, 12 (75 percent) were breeding and of these 8 were pregnant. The litter size for the series averaged 14.6. Five of the mature females were taken during the period February 1-5 and of these, four were pregnant. The relative number of young mice (5h 01' 80) also would indicate that breeding was general. There was evidence that the 19145-146 population, which was somewhat higher than usual, also bred throughout the winter. A midwinter sample was not obtained, but among 25 mice taken early in March, 7 of 12 females were pregnant and 1; others were either lactating or in oestrus. A month later 10 of 11 females trapped were pregnant. The average litter size for both groups was 5.14. This anomalous picture of seasonal population levels and breeding schedule is of interest in that it finds a reasonable parallel in the behavior of Microtus populations at the crest of the cycle. 103% Hamilton (1937a) observed the annual peak in September of most years of his New York work, although during the cyclic high he found the greatest densities in December. The intensive studies by Hamilton also showed the breeding season to be variable depending on the density. He noted that voles usually breed in central New York from mid-March until mid—November, but in years of abundance, some of the animals produce young throughout the winter, although the litter size and frequency of breeding is reduced. Causes of winter breeding.- While instances of unseasonal breeding have been reported for a number of species of small mammals, few attempts have been made to determine the factors responsible for such behavior, and no one has yet satisfactorily explained the phenomenon for wild populations. Baker and Ransom (1932a, l932_b_) attempted to correlate various dinetic factors with the normal reproductive period of the vole (Microtus agrestis) in England. Under field conditions they found no particular relationship between the period of breeding and rainfall or temperature, but a fair agreanent with total monthly hours of sunshine. The obvious seasonal factors - light, temperature, and food - when studied in the laboratory, were all found to have some affect on breeding. Animals on a diet of summer food, and kept in "summer" light (15 hours daily) bred to , only a limited extent at low (winter) temperatures. Breeding was general when the animals were held under conditions of winter food, winter temperature, and summer light. These same authors concluded that light was the limiting factor, since voles bred consistently 1! )4 when maintained under conditions of summer food, temperature, and light (15 hours daily), but had few successful matings when under the same conditions the light was reduced to 9 hours daily. Among wild populations of Microtus pennsylvanicus, Hamilton (191413) was unable to correlate winter breeding with such climatic factors as light, temperature, snowfall, and summer precipitation. He observed that wintexabreeding individuals appear in numbers only during a period of high population and seemed that dense populations are a requirement for reproduction in winter. In a 3-year study of small mammals in Bagley Wood near Oxford, England, Baker (1930), and Elton and others (1931), found considerable yearly variation in breeding rhythm for two common woodland species, the wood—mouse or long-tailed field-mouse (Apodemus gzlvaticus) and the bank-vole (Clethrionomys glareolus). During the first winter of study a complete cessation in breeding occurred from October to March. In the second year, breeding stopped during winter but continued later into the autumn and began earlier in. the spring than it had the year before. In the third year a normal summer increase was followed by a winter of nearly continuous breeding. With reference to population fluctuations, dominated by this variable breeding behavior, and occurring simul- taneously for two unrelated species, Elton (191:2: p. 166) remarked, "like all 'explanations' in science, and in ecology particularly"... this eXplanation of fluctuations leaves one with a further, in this case physiological, mystery. As a matter of fact, we still do not know what factors control the breeding season of Apodemus or lees 1( if; Clethrionomys. Baker [I930] was able to show that temperature differences were certainly not responsible.” Explanations of modified mammalian reproduction in the wild, to a large extent have been looked for in climatic or other conditions in effect at the time. There is growing evidence, however, to indicate considerable flexibility in breeding behavior, and it appears (see Errington, 19136) that populations may respond at times to adversities of an earlier period by compensatory actions in the breeding cycles. For instance, Errington (19142) found that in a portion of a marsh where the muskrat population was markedly reduced, the animals bred later into the year and produced more young than in another section of the marsh where their numbers had not been so seriously reduced. Following a heavy winter kill of small mammals a similar prolongation of the normal breeding season was observed by Osgood (1935). Another example of how a species of mammal may modify its breeding habits in response to conditions that might otherwise prove critical has been surmnarized for the fox squirrel in Hichigan (Allen, 19113). In 1910, populations of this species were high and a complete failure of the mast crop in autumn led to a winter of privation in many areas. In some habitats (fam woodlots) animals had access to an alternate food supply, were able to maintain thanselves in fair condition, and normal breeding occurred. In some scrub oak habitats, however, the starving populations failed to produce young in the first (spring) breeding period. With the passing of the food crisis in summer, it was found that practically all females in these scrub oak habitats had produced young - a remarkable situation since the breeding population included yearling spring animals which ordinarily produce young the first spring but not in the summer. The result was a fall population that in numbers approached that of the previous year (Montgomery, 19hl). There was further evidence that this compensatory reaction carried over into the following year. In the spring of 19h2, summer yearlings produced young h to 5 months earlier than usual, during the same period as the adult females (Baumgras, l9h2g). ‘Winter breeding of the meadoW'vole at Rose Lake, could not be correlated'with any single factor. Climatic conditions would seem to be ruled out, since normal seasonal activity was observed in colonies just a few'miles away. However, certain points were obvious. In each of three consecutive years, the early spring reduction in numbers'was sudden and marked. The surviving population 'was sparse during the spring and opportunities fbr'mating may not have existed even though the scattered population should have been receptive at this time. Also, the autumndwinter breeding and population increase may have been less of a supplement to summer breeding and.more in the nature of a postponement of the annual period of repopulation. The fact that Rose Lake voles showed irregup larities in breeding, and seasonal population.1eve1s, during most years of the study would.suggest certain adjustments to local conditions. It is possible that the xerophytic habitats of this immediate area imposed certain conditions not present in other localities where the reproductive cycle, and seasonal densities of Microtus were quite different. .Llll House mouse Seasonal numbers.- Densities of the house mouse were never at a level which would allow for detailed biological considerations. Nevertheless, the Species was found to be generally distributed through most farmland habitats and from this standpoint must be considered a normal component of the small—mamnal community. During most of the growing season (late spring to early fall) it ranked second in abundance to the prairie deermouse. From late fall to early spring it was outnumbered by both the prairie deermouse and the meadow vole. O Movements.- Their partial abandonment of cropland in fall and winter probably is occasioned to some extent by lowered temperatures. There was evidence, however, that cover requirements of the house mouse are exacting, and movement from fields, to a large degree, is related to harvest. They were seldom found in roadside or fencerow sod during the summer months, but appeared in these areas following harvest of adjoining fields. Two individuals which had been ear- tagged in late summer in a strip-cropped field were snap-trapped in a farm residence nearly half a mile from the field following harvest in October. The fact that only one individual appeared in the "removal plot" (table 11) would suggest, however, that seasonal movements of the house mouse may not include the widespread wandering which was observed for other species. illiéi The late fall exodus from fields is by no means complete and some individuals maintain themselves in crepland throughout the winter. Occasionally specimens were taken in densely vegetated areas where they were out of contact with habitations, and they were common in field-shocked.corn in winter. 'Under the latter circumstance they frequently occupied the same shock with meadow voles and other species, but were never found in company with the prairie deermouse. In one field of shocked corn, house mice were recovered from 26 shocks, which frequently included other species but never the prairie deer- mouse. In 39 other shocks the prairie deermouse was taken alone or with other species but never in association with the house mouse. This would seem to indicate some measure of intolerance between the species. Other species In addition to the prairie deermouse, meadow vole, and house mouse, several other species of small mammals were taken periodically in cropland trapping. Most of these appeared so infrequently that their position in the over-all economy of the area is of doubtful significance. They will be dealt with only briefly here. Bog lemming (Symptomys cooperi cooperi Baird) .- This species was taken regularly in many of the situations occupied by Microtus, and also showed similar armual and seasonal population trends. In common with the meadow vole the species appeared to be at a low in 19140 and 191d. Summer trapping failed to capture any during either 102) ouugwosg was Sun he a Q and.” dogs... megawa— .oooEoI onion.“ apnea.» nsoeocpuo: 05 «some. magnum. Warmwnowwwumw « on 5.5.» 00.3. one .8335: magenta—scape was». an mosses one: mundane." mom on» .mfm 1113 covers»: .3021: 92¢ 3h.“ vacuum ewflufisg won no nowauHsmoA “recesses ad on 60 Moon one Medan havoc:— eae :93 05. .nelmvmfl .uo ~3de 05 a.“ one. Iaoaeo no a”; .N» .mE of these years, but several specimens were taken from corn shocks during January of 19141. In the summer of 1912 they appeared in upland sod areas at about the same time as Microtus, and during the following winter were common in many places. In December of that year they were especially numerous in a 3- or h-acre opening beside a stream bottom which was grown up to largetooth aspen (Populus grandidentata) and quaking aspen (POpulus tremuloides). In this area, and during mid—day, these animals could be seen running along the snow surface as they moved between tunnels. It was not uncommon to have four or five in view at one time on an area less than 100 feet square. The small clearing in which they were particularly active was grown up to Saint John's wort (Hypericum pgrforatmn), dew’oerry (M villosus), wormwood (Artemesia canadensis), milklweed, mullein, ragweed, and bluegrass (figs. 31 and 32). Later in this same winter (19b2-h3), bog lenunings became increasingly abundant and were seen frequently in shocked corn. In February, 50 snap-traps were operated for 1; nights in an alfalfap bluegrass pasture, and 18 were taken from less than half an acre. As was found to be the case with Microtus, bog lemmings were more abundant in l9h6 than they had been at any time in the 19140—10 period. These coinciding high periods, together with the parallel upward trend in populations noted during the years l9h0—h3, are of interest in view of other reports. Elton gt 3.3:. (1931) observed parallel variations in breeding behavior and population levels for two unrelated species of English rodents - Apodemus and Clethrionomys. A similar agreement of cyclic population trends have been reported ilifio for the meadow'vole and the pine vole, Pitymys pinetorum, (Hamilton, 19383; Linduska, 19h22). No particular effort was made to obtain life-history information on bog lemmings. However, among 1h individuals snap-trapped in the spring of l9h6, four pregnant adult females had embryo counts of 3, 3, 3, and 2. These numbers are somewhat lower than would be expected from the range in litter size of two to six young given by Burt (l9h6). Pine vole (Pitymys pinetorum scalopsoides.Audubon and Bachman).- During August of l9h0, four of these mice were taken in 227 trap- nights in an ungrazed woodland. 0f several situations studied, this was the only one in which they were feund. ILater in the fall there were reports of damage to root crops in home gardens, and.the species ‘was fairly common in shocked corn during the same winter. None were taken.by trapping in the two following years, and none were seen in shocked grain during the winter months. From all indications, this species underwent a decline in numbers beginning in l9hl. Jumping mouse (Zapus hudsonius brevipes Bole and MoulthrOp).- Although only one individual was taken‘by trapping during the course of the work, jumping mice were known to be fairly abundant in some habitats on the area. In the summer of l9h2, at least a dozen of these mice were taken from post holes which had been dug for the erection of a pasture fence. The fence line bordered on a portion of a marshy area which supported a dense stand of reedpcanary'grass (Phalaris arundinacea), and it is likely that this represented one :litf of the most productive habitats on the area for this species. Blair (191402) found that most of the jumping nice on a 9-acre plot occurred along the margin of wet or marshy areas, and in a similar situation, Erickson (1938) found surprising numbers of jumping mice trapped in holes which had been dug for a telephone line. In two holes which had been cpen for only one night he found seven in one and five in the other. He remarked that: "During the day a few jumping mice were seen, but judging from the numbers that were found in the pole holes the next morning, the population of the area was dense.....In nearly all of the holes there were jumping mice." Christian (1936) also found many of these animals trapped in post holes around swampy ground. He reported that while he took only four jumping mice by trapping in two years, 26 were taken in less than a month from 200 pOSt hOIBSe Striped ground squirrel (Citellus tridecenlineatus tridecemlineatus Mitchell) .- These animals were common in some situations but they avoided cropland entirely. Although seldom taken in the small live- traps, they readily entered larger box-traps used for the trapping of game mammals. Some records are available from this trapping. From July 11 to August 9, 19M, 69 box-traps were operated in fencerows, and along swale margins that bordered on cropland. Only one ground squirrel was taken. During this same period a line of 21 traps, which included along its length a 2— to 3-acre piece of retired cropland, took five of these animals. While they appeared to be extremely intolerant of cultivated ground, it was noted that they would establish themselves rather rapidly in areas retired from cultivation. Two such units were under observation on the Rose Lake area. One of these was an experimental sanctuary, last plowed in 19140, at which time one small plot was planted with alfalfa and another with sericea lespedeza. A second unit of land which had grown a corn crop in 1939 was retired after harvest to be used as an experimental rabbit management area. It was not replanted. A system of grid placement of traps was enployed on these two units, and 58 box-traps were distributed over 1145 acres. During the July 11 to August 9 trapping period noted above, 30 ground squirrels were taken in the 58 traps. This compared with the one animal taken with a slightly heavier trapping intensity along cover areas bordering cultivated cropland. Allen (1938) noted that burrows of ground squirrels were not present on cultivated ground, and that the animals confined their activities to undisturbed portions of the Kellogg sanctuary. ’ Intensive fall trapping, which included favorable habitats, failed to take any appreciable number of ground squirrels, and it appeared that the majority had entered hibernation by mid-September. In 191.1, 6,028 trap-days took only 11‘ individuals in a September 11 to October 9 period. In 19140, trapping of half this intensity failed to take any animals over the same dates, and only one was taken in 19142 with 14, 6914 trap-days of activity in the same autumn period. House rat (Rattus norvegicus Berkenhout) .- This animal was never taken in the live-traps used for small mammal studies, but it was captured occasionale in box-traps, both in field and woodland situ- ations. A few were taken in corn shocks. At no time were numbers A corn field in 1939, this area was densely vegetated with wormwood (Artemeia ult on. canadeneia) and associated weeds in 1942. F180 33. Spernmphilee were abundant in thin a 11. \) 1.1 G encountered which would suggest that it was important in this area as a feral form in cropland. Prairie mole (ScaloEus aquaticus machrinus Rafinesque).- No attanpt at quantitative studies of the prairie mole was made although it was reasonably common in the area. On several occasions following heavy rains in spring and summer, indivichials were brought in by dogs; others were taken by hand as they traveled along the ground surface. Their surface tunnels were observed more frequently in lawns and gardens than in any of the cultivated farm crops. Short-tailed shrew (Blarina brevicauda kirtlandi Bole and MoulthrOp).- With reference to the short-tailed shrew, Townsend (1935) stated that, "This species appears to show but little habitat Ipreference'. It is equally at home in dry woods, moist woods, and wet meadows, provided food and burrowing possibilities are ample for its needs.” Dambach (1915) found them in greatest abundance along borders of various types where runways of moles were numerous, and where the soil was loose and well-covered with litter. On the Prairie du Sac area in Wisconsin Hanson (19111;) found that they were chiefly limited to areas of bluegrass and woodlands having much duff and litter. Aside from woodland populations which will be considered separately, the short-tailed shrew at Rose Lake occurred most consistently in dense bluegrass. associations, a type which Blair (191mg), also, found to be well populated by this shrew. Roadsides and fencerews commonly provided this type of habitat in the area, 1'1 '7 and the small mammals of these situations have been dealt with, in part, under the section on Field borders. It is possibly of some significance that, except in woodlots, the short-tailed shrew was almost always found in association with the meadow vole. Although a number of sod areas with abundant tunnels, and otherwise apparently suitable for shrew occupancy were trapped, Blarina was never taken in the absence of voles. In August of 1910, three such areas were trapped without taking either of these species. In August of the following year, five areas of bluegrass sod with abundant but old signs of meadow vole activity were live- trapped and neither of these animals was taken. During this same period, two lines of 25 traps in an ungrazed woods caught nine short- tailed shrews. In 19142, with an increased population of voles, Blarina appeared in summer trapping of field habitats for the first time. One of these animals and two voles were taken in August of that year from one of four areas of bluegrass sod studied. This shrew increased along with the vole, and in October-Novanber trapping in 191:2, a total of 19 individuals were taken on four lines in blue- grass habitats. The vole population was represented by a catch of 50 animals on these same lines. Studies in crepland during this same period showed that a few shrews had moved into these areas. Eight individuals of this species and 25 meadow voles were caught in four trap-lines located in crop ground. It is possible that this appearance of the short-tailed shrew on such sparsely covered ground was associated partly with late-fall distributional movements already noted for other small 111‘} .Fig. 36. The short-tailed shrew was the most abundant ineeotivore in the area. 11 manuals. In the "removal” trapping (table 11), the greatest rate of pepulation "shuffle" for Blarina occurred in November, when 11 were taken in 800 trap-nights. The area on which this trapping was conducted had little or no close ground cover, as is usually required by shrews, and as nearly as could be determined the plot did not support a resident population of this species. Periodic trapping in the simmer of 1916 gave further evidence of a shrew-vole relationship. In that year, Blarina made up part of the catch from five trap-lines which in each case included meadow voles. The phenomenal food requirements of the short-tailed shrew are well-known (Merriam, 1881i; Shull, 1907 ; Edgren, l9h8), and a number of workers have testified to the ability of this animal to subdue mice (Plummer, 18141;; Merriam, 22. £13.; Bailey, 1923; Hamilton, 1930). The efficiency with which Blarina has been observed to dispatch small mammals is more readily understood in the light of a recent discovery of a toxic substance in its saliva (Pearson, 191:2). The extent to which this ability as a mouse killer is utilized in the wild is not well-known. Hamilton (1930, 191112), following a detailed study of stunach contents, concluded that the short-tailed shrew had a negligible predatory effect on mice in nature. Shull (33. 23.3.) and Eadie (19141;), however, obtained evidence indicating that mice were a prominent staple in the diet of this shrew, and the following comments of Williams (19 36) would suggest that Blarina might exercise strong regulatory influences on papulations of white-footed mice and other small male of woodlands: ”The reduction in numbers of mice 126 i: must have reacted directly upon the numbers of shrews, as mice enter so largely into shrew diet.....The appearance of the pine mouse and the smoky shrew in the traps for the first time (19314), when the numbers of their arch competitor and probably deadly enemy the short- tailed shrew was reduced to a comparatively low figure, seems to indicate that these annals may prosper only when the mmbers of the short-tailed shrew are relatively small." The possible importance of shrews as a factor limiting vole populations was not established in this study. Shrews were not commonly taken from corn shocks, but in these restricted quarters, ranains of mice were usually to be found whenever shrews were present. The fact that these insectivores occasionally entered live-traps and consumed trapped mice suggested possible habits in the wild. In any event, the consistency with which they were associated in this area, both by habitat and season, was highly suggestive of a predator—prey relationship. Masked shrew (Sorex cinereus lesueurii Duvernoy) .- Only 13 individuals of this species were caught during the entire period of study, and of these, seven were taken from the "removal plot" during the period November, 19142 - March, 19143 (table 11). The relative frequency with which they were trapped at that time may have been indicative of a beginning increase, since none had been taken in over 12,000 trap-nights during the three previous summers. They appeared to be somewhat more abundant in 19146, when six were taken by trapping in a sphagnum bog and in a sod fencerow. Of these, one Fig. 56. flaked shrews were infrequently taken. A sphagnum bog yielded several and an occasional individual was trapped in wood- lots. Fig. 37. Close-up of the masked shrew - the smallest manual in the area. 12 ’1 adult female taken on.April 19 was lactating, and a second trapped June 1 had seven embryos. Least shrew (Cryptotis parva Say).- Only two specimens of the least shrew were taken, both by snap-trapping. The species apparently is as rare at Rose Lake as it is elsewhere over its Michigan range. Predators of Small Mammals Most of the predators common in agricultural areas have been studied widely and their food habits have been publicized so generally that documentation here seems unnecessary. FeW‘investigationg,howh ever, have provided.more than casual information on the densities of both predator and prey species. In the broad.program of research at Rose Lake attempts were made to obtain quantitative information i on most of the vertebrate life of the area. ‘While these measurements are not as detailed for some Species as might be desirable, they, nevertheless, contribute to an understanding of some irregularities in predator food habits which were observed. Before considering the food predilections of some of the preda- tory species, it lill‘be advantageous to review briefly the status of the commoner prey animals during this period. The population levels of small mammals have been dealt with at various points in the text, and it is adequate here to recall that most of these were low throughout the period of study. The meadow'vole, in particular, was characterized by summer densities so low that a measurement by trapping methods was precluded. Rapid fall and'winter increases 12;; produced maximum numbers in late winter and early spring with probable densities in the more favorable habitats of 70-80, or more, per acre. A ”crash" decline of the species was observed in March and April of each year, and from late May on through summer the animels were extremely sparse. In contrast to p0pulations of the commoner small mammals, one game species, the ring-necked‘pheasant, reached an all-time peak in numbers during the period of these investigations. Maximum annual densities of this bird were reached in mid summer, and on this pheasant range of "average" productivity, the summer population probably ranged between to - 50 birds per 100 acres for the years 19140 and 191:2 and possibly as high as 60 - 70 birds per 100 acres _ in 19141. Cottontail rabbits were abundant each year of the study, and one rabbit for each 2.5 to 3 acres appeared to be a close approximation of the pro-hunting population. In all probability the better swale habitats supported a rabbit or more, per acre, in late summer. The non-hunted quail population at Rose Lake remained unifomly low over a 3-year period. Fall populations of one bird per 20 - 30 acres were recorded, and it is doubtful that peak summer densities ever exceeded a bird per 10 acres. 124; 0 me o o a o r a u o a o o o o o o a a n a N a 33 .32 3 8.0... 0.70 o 3:333 N She-.83 N o _ o u o o m N N N H N o o o N H N N ._n on: a d c o N 0H N 39.. .33 310:3 woman hookah: eoadooa Shoe Ho. hon—52 naked oomN no Sausages 33.3% 5.5 d5 soaps-am Hho vandalmaon d5 neoeaom as BEE 23.5 as! commas-e38 a5 echoing?" neon—05 x5 833.com E c . aoaooo xi: usage-neon an! 5.3. uflusdm 4319.83 canon 3.8 no anon-.6 .5 use .Afiéfiv Bonagaoa saga :25 .m." .33 : .. a 3'5. ,' 5..» . , ”£3, m W Fig 0 38. 1940 and 1941. Heat locations of predatory birds for the years . a t . . V l \ e—a . « . LOCATION OF NESTS 0F PREDATORY BIRDS I940 NESVS 194/ #15575 ”H - MARSH HAWK 951'!- RT!!- RED smut” mow RED-TAILID M 125: 126 Birds During the winter of l9h0—hl intensive field work was conducted over the station grounds and surrounding private land to determine the avian predator populations. Slightly less than h sections (2,500 acres) were covered.and.the estimated populations were as shown in table 15. A census the following year, while not as intensive as that in l9hO-hl, indicated that the winter population ‘was about the same as that of the previous year. A detailed study of the breeding papulation of predatory birds also was made in l9hl, and the results of these are given in the same table and.fig. 38. 'With a few exceptions this breeding population remained about the same in 19142. It should be noted that, although the shooting of avian predators was not permitted on this eXperimental area, the density of these birds appeared to be entirely comparable to that on adjoining privately owned lands. Furthermore, the numbers remained essentially constant over the period l9h0—h3. Marsh hawk (£35222 cyaneus).- In 19h0, two nests of this species 'were located, one with four eggs on.May 18, the second with five eggs on.uay 25. On June 1h, when collection of food items began, the first nest had four 1- to Zdweek-old.birds, and the other contained three newlybhatched young and one pipped egg. The nests and young ‘were confined within pens 10 feet in diameter and constructed of inohameSh poultry netting 2 feet in.height. Gullet contents (see Errington, 19322), prey remains, and a few pellets collected until July 1, provided a sample of the principal food items. The following tabulation of 23 individual items shows that, by number, immature pheasants and rabbits made up 30 percent of the total during this period: Pheasants (young) 3 Cottontails (young) 1; Songbirds 13 adults 14 fledglings 9 Meadow vole l LeOpard frog 2 In 191.1, four marsh hawk nests were located. One was deserted during the laying period, but the other three had a high rate of hatching success. One of the successful nests contained an’unusual clutch of 10 eggs of which seven hatched (Baumgras, 19).;23). Unfortu- nately it was not possible to follow up on the food habits of these nestlings. Only two collections were made from each of the three nests and these during the period June 23 - 30. Materials identified from these collections included one water snake (Est—r135 sipedon), five songbirds, two cottontails, and mnerous insect remains (mostly of the family Carabidae). In l9h2, three of the 19141 territories were re-occupied. Two of these pairs failed for one reason or another to nest successfully, one after a re-nesting effort on a new site. The third pair, which located in the Rose Lake marsh, produced a nest of six eggs. These hatched intermittently over a lO-day period, indicating that oviposition and incubation were concurrent and that the eggs were laid at widely spaced intervals. Fig. 39. Two nestling long-eared owls located in Chandler's nan-sh, west of the Rose lake station. The remains of a cottontail can be seen at the extreme left. ' ‘ ' "v‘ r' - v. , , _ , . - a ’8‘ {Nil-21579;" A; ' - \ "/1 ' ‘ \ ./ ,\ y/ vr-w . ' " ‘-'"“w ,1! fies Fig. 40. A shortage of meadmv voles was reflected in the primarily avian diet of these young marsh hawks. Photo by P. S. Baumgras. The above nest was reasonably satisfactory from a food.habits standpoint, and Sh identifiable items were obtained in 30 collections of gullet contents and prey remains. Confining the young did not seem to interfere with their being fed.by the adults (mainly, if not entirely, by the female), and the young continued to return to the enclosure from short excursions for 2 weeks after the first flight. A list of the prey recovered from this nest is given in table 16. It will be noted that songbirds constituted to percent of the total, pheasants - 10 percent, and birds of all species over 60 percent. Cottontail rabbits made up nearly 25 percent of the total. The songbirds appeared principally during the early stages of growth of the nestling hawks, and up to the normal nest-leaving age of the brood a predominantly avian diet (nearly 100 percent) was observed. A majority of the pheasant remains were recovered in.this period, also, and since they were all young birds they probably constituted an ecological equivalent of songbirds. At about the period of growth when the young were making excursions from the nest, cotton! tails became conspicuous in the diet and it was during this late fledgling period that most of the rabbit remains were recovered. Food habits studies of this species have indicated a ready acceptance of murine items. Fisher (1893) found mouse remains in nearly half of 12h stomachs examined, and.McAtee (1935) recovered mice (mostly meadow'voles) from 211 of 601 stomachs. In.a summarizing statement of food habits of marsh hawks Bent (1937) remarked that, "Meadow'mice seem to constitute the bulk of the food, according to nearly all observers.‘l Errington (1933) found 130 3.3% t 2 saunas osmmmmofl Enos Hgossoo m hawking—as .m new?» Ha I each house: A §Mo>HE new? 3.58m I dogma Mom .n 35.35.3033 .» 3.33.8 I Hagen eases» eeaaapm 3 .oolmoooaa 833.835 3563 .n «Won—Hoe mawanoHo: I consume ween 355» .3 n mag meow sum oHo: I 5.8% E H some nEmOHo: I shaken mecca-egg gflasm< I wen-loom €553 N sauce accustom I oranges gunman H matchmshMm NchoflHon I gaonem a 38%. 33a .. dos snags @ 30.233 msgmmmmm I squamous poxoonlwafim a Bananas? 3:38 .. Baa-Isa m .n 833533 . €585 5.8m 3332m and .. use Housed guinea Iolomuoflm mg mafia mzfimHg .33.. J. ensued 3. MN 056 weapon .5 warns vac: E enema d Honk 53.30500 “soon On and vOdnp—nav.“ EH oOH OHDdH. considerable variation in the frequency with which certain prey species appeared in food collections from marsh hawks. From June to August, 1929, during a period of peak abundance of meadow voles, he found that this species represented 89 percent of the total of 71; prey individuals collected. Only Ll percent of this total was birds. In 1931, vole populations were much reduced and 1119 July collections by Errington included only 11 percent of voles and other mice. Birds (mostly passerines) represented over 21 percent of the total that year. Similar relationships between food habits of marsh hawks and the numbers (or availability with reference to condition of habitat) of prey species, were noted by Breckenridge (1935) and Errington and Breckenridge (1936). The June-August food habits of marsh hawks in this study are of interest in that they reflect the extreme low in mouse numbers found by trapping during the same period of the year. Meadow voles made up only 8 percent of the total prey individuals recovered (191:0 and 19h2) , whereas an abundant cottontail population was utilized with nearly three times this frequency (22 percent). Pheasants, which were at a fairly high density, appeared to have contributed disproportionately to the needs of marsh hawks, and this also was a probable consequence of the near absence of small mammals. Red-shouldered hawk (Buteo lineatus) .- In 1910, three pairs of these hawks were summer residents on or near the study area. The only nest located contained one egg on May 8 and a week-old bird on May 25. No gullet material was collected in four visits to the nest, Iii?» but four pellets contained.mouse hair, insect fragments, and snake scales. As found by Errington (1933), bones of prey were so completely digested by these hawks that neither qualitative nor quantitative determinations could be made reliably from pellet remains. Between.April 28 and May 8, l9hl, four nests of this species 'were located. One of these was found destroyed on May 16, and the others hatched clutches of two, three, and flour eggs. Two of these three failed shortly after hatching and the fate of the remaining nest was not determined. Only two pellet collections were made befbre the nests were destroyed. Numerous insect remains (mainly Carabidae) and two snake skeletons were the only identifiable items. or three nests feund in l9h2, one failed in its final stages, one fledged one young and lost one, and the other fledged three young from a three-egg nest. When the nestlings from the latter nest were about three-fourths grown (June 17) they were transferred to an observation cage which had been attached to the base of the nest tree, and daily collection of food items was begun. Seven days later, at about the normal nest-leaving age, two of the young escaped through the adjustable cage opening which had been left wide open when it was found that partial closure prevented proper feeding by the adults. The third young, whichthad been tethered, escaped several days later. Collections of prey remains from this nest, together with pellets from the other two nest sites, revealed the following items taken from June 3—25: snakes a 5, cottontails - h, songbirds - b, snapping turtle - l, prairie mole - l, fox squirrel - 1, red squirrel — l, meadow vole - 1 (plus fur in a Zlffii few pellets), crayfish - several (?), and numerous insect remains. Fisher (1893) found small rodents in 65 percent of the 220 stomachs he examined, and Errington (1933) reported "....considerable quantities of pellet mouse fun..." from Wisconsin birds. The practical absence of mouse remains from the red-shouldered hawks under observation here was to be eXpected in view of the known low levels of these animals at the time food collections were made. This hawk'é ready acceptance of snakes and crayfish also was noted by Fisher (22, 233.) and Errington.(gp, 313.). Red-tailed hawk (Buteo jamaicensis).- One nest of this hawk was found in a dense lowland woods in 1910. Great horned owls preempted the nest in 19b1, and it is possible that the above pair nested off the study area in l9hl and l9h2, since they were not seen in the vicinity of the 19ho site in either of the subsequent years. Irregular collections of materials at the 191:0 nests, and gullet contents from two young yielded 12 prey items from May 5-- June 11;. These were as follows: pheasants - 5, fox squirrels - 2, cottontail - l, bobwhite - l, meadow voles - 2, and snake - l. Broad-winged hawk (Buteo platypterus).- One nest was found in 1910, none in 191:1, and in 19142 one bird was seen periodically at the 19140 location but a nest could not be found. It is possible that it was an unmated bird. The 191:0 nest hatched three young, only two of which survived the first week of nest life. Gullet contents of the young plus a few pellets revealed the following identifiable items during the last week in May: wood frogs - 3, garter snakes — 2, jumping mouse - 1, mouse (unidentified) - 1, crayfish - 2, and insect remains - several. Cooper's hawk (Accipiter cooperii).- A nest of this Species, possibly belonging to the same pair of birds, was found in the same portion of a woodlot in each of )4 years (1939-IQ). In 19140 four young were fledged from a five-egg nest; a two-egg nest failed in 1911.1; and in 19h2, three young hatched and two were fledged from a four-egg nest. In an effort to obtain food records the young fiom the 191:0 nest were tethered, and those from the 191:2 nest were confined in a cage at the base of the nest-tree. In both cases adult birds fed the young satisfactorily, but due to the habit of parent birds of tearing up prey away from the nests and feeding mainly soft fleshy parts to the young, few identifiable prey items were obtained. Gullet contents most often yielded strips of meat, and much of this appeared to be avian in origin as judged by bone fragments. Identifiable items near the nests and from gullet contents, included the following prey taken during the period May 12-June 25: pheasants - 5, songbirds - S, fox squirrel - 1, ground squirrel (Citellus) - 1, and chipmunk - 1. The preference for an avian diet, claimed for the COOper' s hawk, appeared in several field observations in which this species was seen carrying songbirds and other avian prey. On one occasion, a Cooper's hawk was known to have made off with the entire nest, and included young, of a goldfinch (Linduska, 19h3). Other hawks.- The American rough-legged hawk, although principally a transient, was present on the area in most years from October to May. Not over two individuals were observed at any one time. Sparrow hasz and sharp-shinned hawks were present occasionally in migration, and ospreys were seen over Rose Lake at several times in Spring and fall. Great horned owl (Bubo virginianus) .- In 191:0 two nests were found, one with two young, the other with one. In an effort to obtain information on food habits, the young from the first nest were moved to a cage placed 5 feet above the ground on the nest tree. Faulty construction of the nest box allowed the young to fall out repeatedly, with the result that they were not fed by the parent birds and died. The single bird from the second nest was over half grown when found on May S. It fledged successfully a short time later. One pair raised two young in 191d, the nest having been built over one used by a red-tailed hawk the year previously. No nests could be located in 191:2, but one pair was known to have reared two young again in 191:3. ljir Numerous pellets collected in and around the above nests in late winter and up to early April, gave a total of the following 226 items: 1% p 5 m '3 5 2‘3 '8 '° '3. 2. afiegggés o g -a -a a h .5 5: s .5 'o :1 >553 ‘9 igoohod 5°a8$33$2$e “a WMHMWME: Species: § c: an 0 3.1.5.. US E. a: m :3 an: 64 Number: 88 85 16 ll 6 1 l 1 1 9 S 1 l 226 zor. . C . total: . Mice-8h . Other mammals-9 . Birds-7 . 100 In 19146 the young of a great horned owl were held well beyond the nest-leaving age by confining them in a caged artificial nest. A total of 92 prey individuals were recognized in pellets and food remains from this nest. A marked change in food habits was observed during the period of confinement. Among 57 individual items used by the birds prior to April 9, mice and voles (Microtus, Synaptomys, and Peromyscus) made up 89 percent of the total. 0f 35 prey indi- viduals identified from collections made during the period April 11- June 11, only 20 percent were mice. While this change in food habits probably resulted from a change in prey availability (decrease in mice, and increase in rabbits and others), it is possible also that growing food requirements of the young encouraged a selection of the larger prey common in later collections. Following is a list 113?; of items taken during the active period of the nest: meadow voles - 32, deermice - 2S, cottontails - 8, bog lemming -‘l, long-tailed weasel - l, songbirds - 12, pheasants - 7, chickens - 2, mallard - 1, bluedwinged teal - l, crow — 1, and great horned owl - 1. Long-eared owl (Asio otus).- Nests of this bird were not located on the area prior to l9h2, although two individuals were taken in pole traps in February and March of l9hl. 0f two nests found in l9h2, one with five eggs in an advanced stage of’incubation was broken up by crows; the other found on April 27 fledged four young. Several pellet collections were made near the nest until May 7, when the young were placed in a cage fastened near the base of the nest tree. The adults regularly disemboweled the prey and left the entrails on the cage platfbrm, but as the young matured, the col- lections of entrails ceased, indicating that the young were being offered the prey entire. Nearly daily collections of pellets and prey remains were made until May 29, when the young were released, since it appeared that the adults had deserted.- The following items ‘were recognized in the collections: meadow voles - 17, deermice - 1h, bog lemmings - 9, short-tailed shrews - 7, house mouse - 1, jumping mouse - l, and songbirds - 3. In.March and.April, l9h3, a collection of over 150 pellets from the vicinity of two long-eared owl nests, gave 1hl prey individuals of which 86 percent were Microtus and b percent were Peromyscus. A comparison of these percentages with those of the previous year is of interest. At that time (late May, l9b2) Microtus made up only 13E 1: 33 percent of the total items and Peromyscus 27 percent. As has been noted (see section on meadow vole), voles were abundant through March of both of these years, and a rapid decline which occurred in April left the Species at a very low level from May on through the summer. The marked difference in incidence of Microtus in food col- lections from long-eared owls, undoubtedly is a reflection of this marked seasonal change in the density of this prey. Considering the scarcity of voles in May of 19h2 it is remarka~ ble that the species appeared in the diet of these owls even to the comparatively reduced extent shown. What appears here as a per- sistent demand on Microtus by long-cared owls has beennoted also by Errington (19323): "Attention might be drawn to the surprising similarity of food habits for the winters of 1929-30 and 1930—31, despite the drastic reduction of meadow mice early in 1930. Although the deermice supplanted in part the meadow mice, the steady occur- rence of the latter in the long-earis diet leads one to suspect that the food habits of this owl are ineXplicable in simple terms of availability of prey." 131522 Eig. 41. Screech owls frequently were heard and occasionally were seen on the area. 14(3) Listed below are 193 items identified from long-cared owl collections from all sources during the dates March S-May 29 for years 1910-142: 8 go 0 00 E?” {3 I g on) g 31.01 mm Ho: .45 c '31 as so 88 as E‘s an: as: a *3 Species: Eb car-1 me me; he 03303 no H a... Number: 138 22 19 1 1 9 ’ 3 many 193 ch total: 100 Voles and mice - 91; Screech owl (Otus asio).- No nests were found in 19h0, but in 19111 we were located, one with five eggs on April 7, and the other on May 7 with.four pipped eggs and one newly-hatched young. The first nest, located in an artificial squirrel nest box, was deserted by April 114, .when only one egg remained. Four of the five young from the second nest fledged successfully. The squirrel nest box used by one of the 19111 pairs again was occupied in 19142. Four young were raised from this five-egg nest. These young were confined in a cage at the base of the nest tree and held until about 6 days past the normal nest-leaving age. Daily collections were made of food remains and pellets. A study of food residues and pellets from the caged youxg indicated an almost entirely insectivorous diet during the period of their confinanent in late May. Ground beetles (Carabidae) and June beetles (Phyllophaga 3p.) were common in all pellets, and craya fish remains appeared regularly. The only sign of vertebrate prey from this nest was a decapitated woodmouse found at the base of the nest tree. ‘A number of miscellaneous screech owl pellets collected near other nests and roosts, and representing winter (January. to March) feeding included the following: meadow voles - 17, deermice - 7, bog lemmings - S, short-tailed shrews - 2, and songbirds - 2. Crow (Corvus brachyrhynchos).- In 191:0, 17 crow nests were located in woodlots and wooded creek bottoms on approximately 2500 acres, indicating a minimum breeding population of one pair per 11:7 acres or about 1:.1: pairs per square mile. The fate of all the nests was not definitely determined, but five are known to have failed before hatching of the young. One brood was shot from a nest, and seven others were successful in raising an average of 2.1.: young per nest. Over the same area a total of 15 nests were found in 191:1, a nesting density of one pair per 266 acres or 2.5 pairs per square mile. The clutches averaged slightly less than five eggs, and 3.6 young was fledged per successful nest. Eight of 11 nests examined regularly were successful. In 191:2 the number of nesting crows was well below that of the two previous years. Only six nests .were found. The numbers of crows appeared to be comparable to that observed in the spring of 191:0 and 191.1, end it is possible that the reduced number of nests may have been explained, in part, by large numbers of non-breeding individuals. Of the six nests in 14122 191:2, two were broken up so early that renesting undoubtedly occurred. The remaining four had average clutches of five eggs and hatched an average of 1:.5 per nest. One group of nestlings was lost subsequently in an experiment, and the remaining three produced an average of 3.3 fledglings per nest. . Aside from routine observations, no attempt was made in 191:0 to follow the food habits of crows on the area. However, in 191:1, Philip Baumgras develOped a method which would provide for quanti- tative studies of the food of these birds. Cages were made from nail kegs which had been sawed in half, the top covered with 2-inch- mesh chicken wire, and the bottom with l-inch fox netting. These 4 were attached to nest trees and the young were confined when they reached the age of 3 to 1: weeks (Baumgras, 191:22). Adults fed the young through the coarse top screen, and droppings and food remains which passed through the finer-mesh bottom screen accumulated in collecting trays or on a sheet of paper spread beneath the cage. Adult birds continued to feed the young throughout the period of confinement, which was 1: to 5 weeks in most cases. After this period, however, the young (about 2 months of age) became active and crowded, and hence were released. Four cages of young handled by this procedure in 191:1 provided 52 collections, totaling 3,225 grams of dried fecal material. Addition- al collections were made during the same period (late-May through June) in 191:2. Due to the thoroughness of crow digestion, the analysis of these residues presents real difficulties, and a detailed study of the collections has not been completed. Cursory inspection s \P 111: 5138.38 .m .n 390.3003 one he one a.“ Kongo: an cream .093. enema tone: :5 ea mecca dosed—we c a.“ venues ease Access. no v8.3. 4 .3 swam no sees“ a o» £9.55 9:» vac Home.“ d 2% .mfim 14 4 of the material has revealed small-mammal remnants (mostly teeth and small bone fragments) in about 50 percent of the individual collections. By volume, vegetative remains were greatest and service berry seeds (Amelanchier canadensis) made up nearly 50 percent of the total. Insect remains were common in all collections. That crows commonly prey upon farm rodents has been shown by the work of Barrows and Schwarz (1895) and Kalmbach (1918). The frequency with which mouse remains appeared in crow drOppings would indicate that on the Rose Lake area, also, this bird utilizes appreciable numbers of such prey. Numerous sight records of crows . carrying mice suggested that small-mammal predation by this species might be greatest in the fall and winter months. Mammals A primary aim of the Rose Lake station has been a study of wildlife reactions to various conditions of land use and in response to habitat changes. Large-scale collecting of predatory species for stomach analyses obviously would not be consistent with this objective. For this reason the limited studies on predation by mammals were confined mainly to observational methods. Present consideration of mammalian predators will concern estimates of their density and population trends over the period of the study. A main basis for population estimates results from an . intensive program of box-trapping which for the years 191:0-1:2 totaled 63, 780 trap-days. Differences in yearly trapping intensity and the fact that the total area involved was not the same each 145. year (some additional units were trapped as new land parcels came into station ownership) invalidates a direct comparison of the total annual trapping returns. However, selected portions of the over—all trapping effort offer a comparison of average annual densities for several species during the years l9hO-h2. In the discussion which follows, reference will be made to the total individual catch from 15,989 boxptrap days operation during the year July, 19h0 to July, 19hl, and from 17,531 boxptrap days for the same Julybto-July'period 'which followed. The "box-trap day" is a 2h-hour period of operation. In each year, a standardized program of trapping was conducted just prior to the opening of the small-game hunting season on October 15. Traps were distributed with reference to the better habitats over about 1,100 acres, of which 12 percent was woodland; 2h percent marsh, swamp, and swale; and the remainder crOpland and pasture. ‘With a few exceptions trap operation extended over the period September 11 to October 10. These data, summarized in table 17, are offered as an index to year to year population trends for the common mammal predators. Skunk (11. lephitus nigra Peale and Beauvois).- The status of the eastern skunk on this area was described in some detail by Allen and Shapton (191:2). It will suffice here to point out that a state- ‘wide decline of the animal which began in 1938 was apparent from fur dealer reports. Cellular inclusions in tissues of the brain cortex was found in sick animals, and an encephalitis epidemic was believed to have been responsible for the decline. At Rose Lake, populations 14E; b H @H m ¢ was ”ON NEH a HH mm fl m ”chm QON .mde N t NH 0 c G86 mNN 3m." locco mediums-um m a a much an 8 :Setaefl no tong: one...“ no non-s: £132 to.» 5 means-.5 supersede... 3 sunscreens? warns 0.393ch 53838. no moves nfivuwom .5 canon 111? ‘were very low at the time this study was initiated in l9h0 and continued low at least through 19h2. The number of individuals taken in similar periods of trapping (table 17) indicated that the fall population in l9h2 might have been even lower than in the two preceding years. However, comparable trapping over the lZ-month period July, 19h0 to July, l9hl, and in the year which followed, indicated similar average yearly densities. A total of 15 individu- als were taken during each of these year intervals. During the winter of 1936—37 prior to the die-off of skunks, Allen (1939) found a population in excess of nine per 100 acres in Kalamazoo County, Michigan. 'While the Rose Lake area would appear to be of about equal productivity for the species, it is doubtful if the density during the years 19hO-h2 ever exceeded one or two animals per 100 acres. It iS'wellaknown that these animals are persistent hunters of mice and other small mammals. At the low density maintained during this study, however, it is doubtful that their net effect was an appreciable factor in limiting rodent numbers. Long—tailed weasel (mustela frenata noveboracensis Emmons).- Live-trapping data of the type obtained here have limitations in indicating actual densities of some of the wide-ranging predatory mammals. In the case of the long-tailed weasel, certain sizable acreages appeared to be uninhabited, whereas, other units occasionally were found to have what appeared to be local concentrations of six to eight animals per hundred acres. From all indications a fall 14h population of six to eight per section was close to the actual average for the 2 sections of most intensive study. Results from autumn censuses suggested a somewhat higher population in 19h2 than in previous years (table 17). However, trapping totals for the July, 19h0 to July, 19h1, period showed a total of 21 animals taken in this period compared with only 12 taken from similar trapping in the following year. In each year of study the sexes of trapped weasels ran heavily to males. From 191:0 to 191:2, a total of 50 individuals was taken 8h times and in the ratio of over five males to one female (h2 males and 8 females). It is very probable that the high excess of males results from a differential activity of the sexes. The prey of this carnivore is known to include considerable numbers of small mammals (Dearborn, 1932: Hamilton, 1933; Errington, 1936; Allen, 1938; Polderboer gt_al., l9hl), and there is some evidence that the numbers of small mammals, to some extent, may regulate weasel populations (Middleton, 19303). ‘An instance of heavy winter predation on meadow voles by weasels has been reported for this area (LinduSka, 19h2a), and their is no question but that the weasel constitutes one of the most important mammalian predators of this group. Opossum (Didelphis If virginiana Kerr).- This animal was common on the area and trapping returns indicated that the population ‘was increasing during the period of investigations. A total of 29 ‘were trapped in a Julyhto-July'period of l9hO-h1, and double this 1425 number were caught in the following year. Among the 58 taken in l9hl-h2 a high percentage of males was observed, the actual count being 38 males and 20 females. .An autumn density of 8 to 12 per section is believed to be a fair estimate of the average population for the area. Dearborn (1932) and Taube (19h?) both reported more than 70 percent animal remains in stomachs of Michigan opossums taken in the fall of the year. Taube's study of 133 stomach contents showed that about 6 percent of the total was mice, and Allen (1938) found a similar percentage of mouse remains in 30 stomachs taken in Michigan during autumn. The well-known scavenging habits of the opossum make it difficult to apply such infbrmation to an evaluation of the species as a predator. The leisurely activity of the animal in the field, and a consideration of its habits in general, would not suggest an important predator effect on small mammals, except possibly on nestling young. Raccoon (Procyonnl. lotor Linnaeus).- Creek-bottom woodlands on the area supported what were probably average numbers of raccoons, but this habitat type was not extensive enough to provide for a generally high population. Trapping returns and observations ' indicated a marked increase in their numbers from 19h0 to 19h2. The best available information indicated a density of about 12 to 15 individuals per section during the fall of 19h2. This population is about one-third that found.by Stuewer (l9h3) on good raccoon range in.Allegan County, Michigan. Various investigators have reported that raccoons occasionally include mammalian prey in their diet, and Stuewer (3p. git.) found evidence of heavy feeding on Microtus during the Spring (April- June). Considering the numbers and distribution of raccoons at Rose Lake, it is doubtful that they figure very prominently in determining small-mammal numbers on the area. Red fox (Vulpes fulva Desmarest).- In common with conditions over most of its United States? range, fox numbers in this area exhibited an upward trend during the years of this study. Tracking studies in the winter of 19110-141 indicated that three individuals were using the area regularly. Bring the winter of 19111-142, four individuals ranged over the two sections of land included in the study unit. These animals invariably entered the tract from an extensive marsh adjoining the north boundary. One group of hunters killed six foxes in one week within a 10- or 12-mile radius of the station, and two other animals are known tohave been taken within a mile of the study unit during the same winter of 19111-112. This degree of hunting success was considerably higher than what had been reported for the vicinity in previous years. During January and February, 19141 and 19142, 37 miles of fox tracks were followed by various station personnel and evidences of 16 kills were observed. While numerous evidences of mouse-hunting were noted on the trails, the success of these attempts usually could not be determined. However, three short-tailed shrews were found uneaten, and considering the density of these in comparison (Q 1131 with.mice, it is probable that the toll of the latter was high. Other prey remains indicative of kills by foxes included seven pheasants, two rabbits, one wing-clipped Canada goose, one bobwhite, one leopard frog (uneaten), and one prairie deermouse (uneaten). Detailed fecdphabits studies of.foxes have been reported for Michigan by Dearborn (1932) and.Murie (1936), and for other regions in the mid-west by Errington (1935) and Scott (19117). These authors agree that food predilections of the fox are explainable in a large measure in terms of prey availability. Cottontail rabbits and meadow voles are reported as stock items wherever they occur in numbers, but other species have been preyed upon heavily where unusual popu- lation densities or adverse circumstances increased their vulnera— bility. In this study, limited evidences of a disprOportionate take of pheasants by foxes might be a reflection of higher-than-usual pOpulations of the bird and a comparative insecurity in the habitat. More complete observations on the feeding of several species of predatory birds suggested a similar relationship. Mink (Mustela vison Schreber).- Fall densities of these animals probably were at a level of two or three pairs per'section. 'Winter trapping accounted for two males in 1939, none in 19h0, and three males and one female in l9hl. The 19h2 pepulation‘was believed to be equal to if not greater than that of l9hl, although due to unfavorable trapping conditions none were taken in steel- traps . Ridge: (Taxidea t. .ith Schreber).- A 12-pound female of this species was taken in a box-trap in August, 1910, and a male weighing over 18 pounds was taken in September, 191:2. Although badgers are known to be persistent hunters of small rodents, they probably were of little importance on this area die to their low numbers. House cats and dogs.- The standardized period of fall trapping (table 17) indicated wide variations in the number of house cats present on the area at this time of the year. However, the total number of animals taken in July-to-July trapping was essentially the same in 19110-111 (23 individuals) and in 19h1-h2 (25 individuals). These totals probably represented a reasonable census of the number using the area at any one time. A majority of these were farm animals, but some were known to be living in a feral state, and at least one case was known of a litter having been reared to an independent age in the wild. Quantitative measurements of the food habits of cats were not obtained in this study, but numerous sight records testified to their success at mouse‘hunting in the fields. Any possible utility that might have been claimed for the animels in. this connection was placed in question, however, by frequent observations of a like aptitude for taking game species. One well-tended pet that frequently brought prey victims home, was known to have caught no less than 17 rabbits during the Spring and early summer of 19112. McMurray and Sperry (191:1) found that mammals composed from half to four-fifths of the food of cats in several habitats in Oklahoma, l‘ and in one area cottontail rabbits made up one-fourth of the mammal diet. Couch (1928) reports the failure of introduced cats to suppress mouse populations in orchard areas, of Washington, even though the animals were released in such numbers that "....it was not unusual to see ten or Welve cats on three or four acres." About 11; dogs were resident on the area and on adjoining private lands. Most of these were allowed to roam at will and occasional losses of pheasants, rabbits, or other game species were known to have resulted. It is doubtful that their activities were reflected appreciably in the dynamics of small mammals. 0‘: L1 0:1 v- I 154 SUMMARY AND DISCUSSION 1. Scope and methods of study. During the years 19hO-h2 and l9h6, intensive live—trapping was employed to measure comparative populations of small mammals on an area of southern MiChigan farm- land. A total of 1,6h9 small mammals were handled, of which 1,0h2 were taken in 18,329 trap-nights of live-trapping, the remainder by snap-trapping or other means. A concurrent program of box-trapping provided information on the status of the larger associated mammals, and intensive observation was employed to indicate the numbers of predatory and game birds. 2. Annual population levels of small mammals. The prairie deermouse and meadow vole were at low levels in l9h0 but increased markedly in both of the fellowing years. Trapping results indicated that the l9hl prairie deermouse population was double that of 19h0, and the l9h2 catch records suggested a five-fold increase over the 19h0 low. Studies in l9h6 showed that both of these species were more abundant than they had been in the 19h0—h2 period, and limited records for the intervening years indicated that pepulations were maintained at a fairly high level during l9h3-h5. 3. Seasonal population trends. Early summer pepulations of the prairie deermouse were comparable in l9hl and l9h2. In l9hl, the summer pepulation doubled by autmmn, but in l9h2 the numbers of this mouse had increased five-fold by autumn. The differences in rate of population increase were reflected in age ratios for the two years. In l9h2, significantly more young were trapped through- out the summer-fall period. In l9h6, the prairie deermouse began breeding early and survival of young apparently was much higher than in either of the two earlier years. A high population in August,.l9h6, was in the ratio of 1 adult: 1h young compared.with.August ratios of 1 adult : 0.5 young in 19hl and 1 adult : 1.7 young in l9h2. Although live-trapping took a high percentage of males each year, the unbalance in sex ratio was greatest in the year of lowest popu- lation and nearest normal in the year of highest population. During each of four summers, meadow voles on this area were absent from croplands and only an occasional individual could be located.in other habitats on the area. In each year it was mid! autumn before they were observed or trapped with any regularity, and a rapid.fall increase was accompanied.by a spread of the species into craplands. Peak annual numbers appeared to be reached in nddwinter, and during each of three winters it was found that a considerable portion of the population was breeding. It is suggested that xerophytic conditions might eXplain the local departure from the usual breeding behavior reported for this Species, since a "normal" trend of seasonal population levels was known for at least one area in the near vicinity. h. M0vements. A program of continuous trapping and removal of small mammals from a plot of ground was conducted over a period of 8 months. The rate of re-invasion of the plot was taken as an indication of distributional movements by small mammals. The greatest movement into the plot occurred in November and it is likely that this month also represented the period of annual "population high" for most of the species. At least four species were taken from the plot, which was of a habitat type not considered acceptable for permanent occupancy by these same species. There was practically no movement through the area during January and February. Evidence of a population shuffle in March,.which may have been related to increased.breeding activity, was shown by additional catches of the prairie deermouse. 5. Small mammal interrelationships. During the summer and early fall months the prairie deermouse was the dominant small mammal in all farm habitats. At these seasons it was associated with the house mouse, the only other small mammal to be taken regularly. Evidence of possible intolerance between these two species was obtained, since both were found to be common inhabitants of corn shocks, but they seldom appeared together within the same shock. Similarly, house mice were most consistently taken in cat fields, whereas, the deermouse was less numerous in cats than in any other crop. From.December to March the meadow vole was the commonest associate of the prairie deenmouse in cropland, and in those situations providing an abundant ground cover it greatly outnmmbered the deermouse. The bog lenming and short-tailed shrew were important secondary species in these latter habitats. we 6. Small mammal predators. No control efforts were directed at any of the predators, but mammalian species of Sporting or fur value were hunted or trapped'with an intensity approximating that on private farmland in the area. During the interval of study, skunks were extremely low in numbers as the result of a widespread die-off. Careful estimates placed their density at 20 to 30 percent, or less, of normal for the area. Another "mouser'l of reputation, the red fox, was beginning an upward trend in populations that'was to reach a peak level three to four years later. Other common mammalian predators occurred in numbers comparable to that reported for similar farmland habitats in the region. Avian predators were studied for their population levels, nesting success, and food habits. Two species, the marsh hawk and long-cared owl, were of special interest for their food habits under conditions of reduced.mouse pepulations. Food records for marsh hawks in particular appeared to reflect a known shortage of small mammals and higher-than—average populations of the pheasant and cottontail rabbit. 7. Farm crap preferences of small mammals. The prairie deer. mouse failed to show an outstanding preference for any given crop type. In general, however, populations in grain fields and alfalfa ‘were comparable, and higher than in other farm habitats. Among grain fields, corn, wheat, and rye were more densely populated than fields of cats. Red clover plantings appeared to be the least acceptable of any of the annual farm crops. Idle ground areas supported low populations generally, although variations were noted which appeared to be explainable on the basis of density of vege- tation. Dense bluegrass sod'was practically uninhabited by the prairie deermouse, whereas annualdweed associations of moderate density supported numbers approaching those in cropland. The house mouse was taken in to percent of the grain fields studied, 35 percent of the hay fields, and 16 percent of the idle- ground situations. It was trapped more consistently in cat fields than in any other habitat. In this respect it differed from the deermouse, which was less abundant in cat fields than in other grain crops. 8. Strip-cropping and rodent numbers. Population studies of small mammals were conducted concurrently in crops grown on fields of conventional plan and in others laid out in crop strips. No significant differences were observed. Furthermore, an analysis of all trapping returns indicated that prairie deermice were distributed uniformly through fields and were not concentrated along margins, crop junctures, etc. The species did not appear to be noticeably responsive to the diversity in food and cover types 'which strip-cropping provides. 9. Populations in field.borders. During the summer, field borders of sod supported higher populations of small mammals than did wooded borders. However, both of these types of border vege— tation were less densely populated than the adjoining croplands. During the autumn months, fencerows and rcadsides of sod held higher 15:2: populations of meadow voles than did any of the farm habitats. From all indications the high winter numbers of this species in crOpland were traceable to a large extent to the scattered popu- lations maintained in sod border vegetation. 10. Effects of harvest on small mammals. Live-trapping was carried on in 22 grain fields and 8 hayfields, before and following harvest. In the case of the prairie deermouse a somewhat higher catch was made in the trapping period following harvest and there was no indication of a reduction in population as a result of the operation. House mice were adversely affected by harvest as indicated by marked reductions in their numbers following combining and mowing. This reaponse appeared to be accounted for on the basis of reduction in ground cover, since fields of mechanically picked corn which retained most of the ground cover showed no significant decline in the numbers of house mice. 11. Effects of cultivation and plowing. Corn fields, which were subjected to periodic cultivation, maintained some of the highest deermouse pepulations measured in farm habitats. Similarly, they were taken in relatively high numbers from fields recently plowed and in the sparse growth of fall-sown crops. While there is little doubt but some individuals were killed in the course of tillage, the marked modifications in habitat resulting from such practices were not reflected in any decrease in numbers as measured by live-trapping. 1613+ 12. Economic considerations. Of the several species of small mammals dealt with in this study, the meadow vole generally is considered to be of greatest importance to agriculture. On this area certain departures from the usually reported population behavior of the animal were observed. This principally involved a projection of the period of annual population high to the midwinter months. Their numbers during the season of crop production were observed to be extremely low; It has not been established for the area that this represents the "normal" behavior of the species in all years. Nevertheless, it was observed in each of four years, and this anomalous trend of seasonal population levels undoubtedly minimized the importance of the species in crop and forage areas. Irrespective of its status in the present study, meadow voles are recognized as an actual or potential liability wherever they are present in agricultural land. On this area the maintenance of breeding stock during periods of adversity, and their periodic invasions of crop- lands, was facilitated by sod areas adjoining farm fields. The encouragement of woody vegetation in roadsides and fencerows should effectively minimize this objectionable function of such idle areas and, at the same time increase their utility for desirable forms cf‘wildlife. As has been noted in other areas, fieldpshocked corn in winter supported considerable numbers of meadow mice and other species. Aside from the appreciable economic losses which directly result, the practice would appear to contribute to the over-all prosperity of farmland rodents. While the pest status of the vole in relation to agriculture is a demonstrated.fact, the position of many other small.mammals in the economy of the farm is not well-known. For the most part, shrews are credited with having a neutral effect if not actually being beneficial. Excepting voles, some species of field inhabiting rodents, also, are known to utilize insects readily (Aldous, 1931; Erickson, 1938; Linduska, l9h2d; Telford, 19h313 and in the case of forest rodents some are credited with exercising a measure of control on insect pests of the forest (Graham, 1929; Balch, 1936, 1937; Hamilton, 19382; Hamilton and Cook, l9h0; Bess 22 31., 19h?). A further value has been assigned to small mammals for their favorable effects on the soil, a relationship that has been reviewed in some detail by Van Dersal (1937), and Jacot (l9h0). In this area, the prairie deermouse alone was of general occurrence during the period of crop production. However, it never appeared abundantly in any of the variety of habitats studied, and it is questionable that any of the new trends in land utilization, such as were covered in this work, will encourage its numbers appreciably if at all. The status of the species throughout this study does not justify conclusions on its economic status. . x ‘ 11 ”-1.. PA HF II STUDIES IN FARM WOODLAND RABITATS 1625 The farm woodlot represents a real, if not always appreciated, asset to the farmer. In addition to being important in providing fence posts, firewood, and rough lumber, these woodlands under proper management can contribute to the income of farms. Progressive land owners have recognized the advantages of applying forestry practices to their woodlots, and the proper management and utilization of the timber crop will establish new ecological conditions in this farm habitat. Principally, these changes will result from exclusion of livestock and a systematic program of harvest. Most farm woodlots range in size from a few to 30 acres, and on this basis the area involved is not impressive. However, in terms of total farm acreage in southern.Michigan, these units comprise about 10 percent of the land area and so constitute an.important ecological type. In addition to producing a major portion of the state's half million or more fox squirrels crepped annually by hunters, farm woody lots also contribute to the needs of other game animals. For this reason, detailed studies on the ecology of this habitat were included in the program of research at Rose Lake. A phase of these studies reported here concerns the relative pepulation densities and habitat responses of several non-game Species. By numbers and habits these forms are important as competitors of the fox squirrel, and otherwise are important for being the most numerous year-round inhabitants. 1154- HISTORY AND DESCRIPTION OF STUDY UNITS A century ago this region was covered with a relatively continuous stand of oak-hickory forest broken by lakes, marshes, and kettleholes. The land was cleared for farming beginning in 1837, when one of the earliest settlers homesteaded a tract on Rose Lake. By interviewing former land owners and older residents in the community, D. F. SNitzen- berg obtained information on some ecological events of the last 50 years or so, and notes on the past management of the remaining wood- lands. According to these reports, gray and black squirrels were common up to 50 or 60 years ago and began to disappear when the big timber was removed. Fox squirrels were seen occasionally, but did not increase noticeably until farming opened up most of the extensive forest stands. With continued clearing of the land, gray and black squirrels disappeared entirely and were replaced by increased numbers of the fox squirrel. Porcupines disappeared, and cottontail rabbits, which had been comparatively scarce, increased under the new conditions which also proved optimum for the fox squirrel. Red squirrels, chipmunks, and flying squirrels were reported to have been common then as now. On the Rose Lake area, a central unit of one and one-quarter sections has been used for the most intensive studies of farm wildlife (fig. 1). Seven woodlots, 10 to 23 acres in size, are present in this unit. One additional woodlot lying outside the station limits, selected as a typical lowland type, was added to the woodlot study Tlthb program in December, l9h0. Limited selective cutting was done in this woods in 191:1. These eight units constituted the main study areas for this work (fig. hb). The past history of these, as established by Switzenberg, showed that all were cut over for large timber at one time or another, and within the past 50 years all but one have been grazed with varying intensity. Notes on the use of the woodlands prior to 1938 are included.in appendix table v. Since 1938 no additional cutting has been done in the central area woodlots and all have been protected from.pasturing. Of the eight woodlots used for study, six, totaling 86 acres, ‘were of the typical upland, oak-hickory type. In these units the dominant Species of trees were black, jack, and white oaks (Quercus velutina, g, ellipsoidalis, and.Q, alba), and hickories (Carya ovalis and.2: agate). The other two units, totaling 26 acres, were of a lowland type in which red maple (5225 552332) was the dominant species. These two fundamental differences in vegetative composition furnished one basis for comparison of the small mammal inhabitants. Among the six upland woodlots, three, totaling hl acres, had been pastured intensively, and until fairly recently, so that they were representa- tive of a grazed type of woodland. The remaining three, totaling b5 acres, had recovered from past effects of mostly light grazing so that, by appearance at least, they had again approached an "ungrazed" state. These differences in intensity and recency of grazing also 'were considered in population comparisons of the woodland fauna. 166$ EXPERIMENTAL ‘- "' i : WOODLOTS Q18 ask) fl I‘L/ J x” L Z UPLAND wooos i I e; .351. \V ,. -_... e. - .. x\\\ LOWLAND WOODS E15. 44.‘ Location of the Rose lake woodlots. Unite 1-7 were the areas of most intensive study. If?“ 'Wbodlot Analysis / 0“. For purposes of future comparison, an anylysis was made of all woodlots in l9h0. A 5 percent sample of all woody vegetation ‘was obtained by making a detailed.examination of 0.5 square-chain plots at permanent boxptrap locations. (The box-traps were spaced one to the acre on a grid pattern.) .All samples were taken arbi- trarily on the west side of each permanent trap location, with the trap marking the center of the long side of the quadrat. A detailed count of woody vegetation was made and recorded under the following size classes: 3 feet high to 1 inch in diameter (breast height); 1 to 3 inches d.b.h.; 3 to 10 inches; and 10 inches and up. The details of this analysis are not included here, but the results have provided the basis forfthe desiription of woodlots given in table v in the appendix. 11389 ouonmsoooshosv hommoolnuvnapunoo dovcoOH havnonufiuom «no hopes oavflovom maunpuozuuonH a no needs an cenaavap new .2008 Ezvfiomaov So- no newton d .3. .wE. chosen on». a.“ vow: “canines one .3 .mE Vivaahu‘ i M s 16:} METHODS OF STUDY The broad sc0pe of this study imposed limitations on the number of specialized methods that could be employed effectively. Had the objective been a species study, some modifications in the procedure undoubtedly would have been profitable. Considering the inclusive nature of the work, however, the approach used was believed to be near optimum. Since the principal aim was to establish comparative pepulation levels of woodland mammals, intensive live-trapping and marking was the main technique used. The box-traps (fig. h5) measured 2 feet in length and 1 foot in height and width, and were covered with 16-gauge, galvanized hardware cloth of half-inchemesh. The traps 'were Spaced one to the acre on a grid pattern at permanently marked locations. Other traps located with reference to rabbits in swales, fencerows, etc., furnished additional information on.woodlot species. Ear-corn was used as bait throughout the study. Dried herring also 'was included to attract the carnivores on the area. Trapped mammals were marked with numbered tags fastened in the ears. For fox squirrels a tag of about half an inch length was used, and for red squirrels and chipmunks a somewhat smaller fingerling size was more satisfactory. To insure long-time recognition of fox squirrels one of the smaller tags also was attached to the outer toe of one hind foot (Linduska, l9h22). Flying squirreIS'were identified in subsequent handlings by means of a tag fastened.to the patagiume These methods of marking proved satisfactory for all Species except chipmunks. In some periods of trapping as many as 10 percent of the chipmunks which had.been marked in some earlier trapping period were noted to have shed their ear-tags. However, duplication of animals in any given period of trapping as a result of such loss of identity was believed to be insignificant. Field records were maintained for individual traps which were numbered within separate trap-lines corresponding to the woodlot designation. In the laboratory, information 'was transposed to individual cards bearing the record of each animal caught. In l9h2 an indeX'was obtained to the comparative numbers of mice and shrews in woodlots. The procedure was much the same as that which has been described for studies in cropland. {An attempt was made, also, to obtain an actual census of small mammals (mice and shrews) in one upland'woodlot. The intensive methods employed in this study ‘will be described under the discussion dealing with these Species. Mast Crop Measurement With the beginning of these studies in l9hO an attempt was made to develop a system for measuring mast yields which would provide an index for yearuto-year comparisons. Refinements of method were made in.l9hl and further changes in l9h2 provided what appeared.to be a reasonably workable system (Allen and McGinley, l9h7). This evolution of method.precludes any critical comparison of statistics on annual mast measurement. However, numerous observatiOns are available to show in a general way the success of the mast crop during the period of the investigation: 1939: l9h0: 19hl: l9h2: 137} Excellent. Both oaks and hickories produced heavily, and the crop in this year probably represented near maximum for the area. Very'poor. Oak and.beech trees failed completely to bear mast. Isolated hickory trees produced a fair crop, but those in woodlots had very few nuts. The crop was probably minimum for the area. Good. Hickories, especially small—fruited hickory, yielded heavily. Oaks in open situations bore well and red oaks in woodlots had exceptionally heavy crops. 2333. Red and black oaks produced fairly good numbers. White oaks failed to bear. Hickory nut production was good and isolated trees had much heavier crops than 'woodlot trees. file: .333 Massage no vesoooo you axon 003 one???» caption no oneness 253 .«o oneness Hess—nu op wooed gang» Sam .3 .3“ Itasca on. News.“ omepeoonom :33 .3384 .5 .w?» a memo _ l — SE _ oooo _ 34.5. #93 W . Rho nfiskthu ox who afixkkhu 2590 44553. OO-CUA— _ - _ «on. v .3. owe. 02.35.... H _ F «3. . .3. . 03. o , 'e-'"'" w ON G No”” x 3 l v , u u N M .v M a m a v V 3 1 N s w. n o _ . m .6558 02:4: nun: .o 3 v 3 xzszcio ii. «to n.. ._mmmSom cum I W a _ v '00. M .Q n m... mumu< No. 02333... 95.5003 l 2u>um 2. 35235.. :85 4«een Mo .3852 on on.o~ o~.oH seuafl oaup m e n m H .35 common» needy no nopssz evOQQE—P OLD? ufiahgdn 60h own nodal npdsihonosvpum .mH canoe 184'- HJH are ass as: 2% 13.3 as peyote «H on m 2 3 fiestas: eo tag #8 as: 3.8 3038: 3088: pesos. on: one 0.3: .8 pea one BMW 838: 8.2a pgmseo 82.? come .3 £53 0.3.3 0.82 no a vogue: engages pea won No sense?!" Mo hone—agem .3 03am. 1' 184'“ Fit-1|!) 0 no» I on .5293 as... 25 5 pews: :Sevsfi ~38 .. In K A I Is. 333a any 633533 soared—mom 33134 3 .952...» no so... e88. as 5 88:! :39. ~38 3 .3395. no Meek 93.038 05 5 .3332 cacti—we of. comma» spans 23 He seems: .— ounamasavlebfi we floor one 5" assumes)": catego- vz no heals: Asw HH 0 an H _. n ON 0 O a On H 0H 33%» on 82. 3 "fine Simon 5&3 n ma. nudge; an,“ 83. ._.fl ”new -539» its: mt oak New." H3." 30.3.. «on no evomvoot asbe- on nodes." casualwsflweve no cases .3 303133." seepage so." no 835.570 .8 capes no «0 N» O” an >38 mm Os. at an on nomad a» me. an an 6H 6 N96 .82 some...» H83 .. 83333 assassin use! on": :38 .2809 Meet unuuvnwssoea .03 uses and-sew?» ooh 1813.5 These movements were reflected in population trends for individual woodlots in a way which makes it virtually impossible to establish accurately the red squirrel p0pulation density for any single woodlot. An example of the error involved can be shown by results obtained in three consecutive trapping periods in June and September, 191:1, and in January, l9h2. The duration of trapping was 18 days in the June and January periods and 28 days in September. While the total munber of red squirrels taken in the seven study woods was similar in each instance (35 individuals in the June interval, 30 in September, and 25 in January), the catch records for individual woodlots showed marked variations between periods. In woodlot 2, for instance, the numbers taken were as follows: June - one red squirrel, September - four, and January - seven. In the same trapping, woodlot 3, which was adjacent to number 2 and comparable to it (both were grazed. upland woods), the trend in nmbers was the opposite, with six indi- viduals taken in June, three in September, and one in January. Woodlot 6 (ungrazed) also showed a marked decline in numbers, and the catch in the June, September, and January periods was respectively nine, six, 'and one. Woodlot 7, an ungrazed lowland woods, yielded catches in the three periods of 6, lb, and 6 individuals. In view of the behavior discussed above, it is obvious that a census of these animals will have the best accuracy when it concerns the largest possible area and is accomplished in a minimum of time. In table 20, trapping data for peak population periods are utilized in a way which should minimize the complications posed by a shifting population. The data apply to the seven study woodlots which iii? constituted all of the true woodlands on more than a section of land. Trapping returns used in calculations are limited to two weeks of each of the trapping periods. By'a simple algebraic ratio explained in the table, the apparent papulation was calculated whenever returns were adequate to allow for such computations. It is believed that the l9h0 figures are reasonably accurate and the late-summer population shown.for that year undoubtedly was the highest dealt with in this study. The calculated density for 102 acres was 70 individuals or about one squirrel to 1.1; acres of 'woods. As will be shown later, however, red squirrels showed rather ‘well-defined habitat preferences, and this summer density would hardly' represent the average for individual woodlots. As a matter of fact, 'within this period of peak population, the density for single wood» lots ranged from a.maximum of two per acre in one lZ-acre woods to a minimum of one per 12 acres in another unit of the same size. Most recorded observations on reproduction in the red squirrel indicate that the time of birth is April to June and fail to mention the possibilities of a fall litter (Merriam, 188h; Mearns, 1898; Preble, 1908; Dice, 1921; Hatt, 1929). Klugh (1927) noted. half- grown young in September and.0ctober, and suggested that while these young might indicate second broods such occurrences are the exception. and not the rule. Hamilton (1939) found in.New York, however, that most adult females have both a spring and fall litter, and Lynn (1923: p. 198) believed that two litters prdbably were produced in Indiana. 1823. _Observations on marked animals in this study indicated that two broods probably are produced, in which case peak annual populations would be expected in autumn. It will be noted.from.catch records, and from calculated populations of red squirrels in l9h0, that the fall density actually was slightly below that of the summer months. Allen (l9h3) reports reduced breeding in the fox squirrel in times of food shortage, and the failure of the mast crop in.l9h0 may have helped to account for the waning population of red squirrels in the fall of that year. The fact that the decrease continued through the following years of fair to good.mast production would suggest, however, that any effect of a food shortage in 19h0 probably was superimposed upon more fundamental limiting factors. Habitat preferences For reasons already discussed, attempts to correlate population responses of the red squirrel with ecological conditions in a specific 'woodlot did not prove profitable. Nevertheless, certain predilections 'were shown which can be expressed.more or less quantitatively. Effects of grazing.- In appendix table vi, the catch records for red squirrels are shown for three grazed upland woodlots totaling hl acres and three ungrazed upland'woodlots of us acres. Since trapping in the two types was conducted concurrently, comparisons of the catches for any given.period.should.give an index to the effects of grazing. It should be pointed out, however, that trapping periods were of different length, and the data do not provide for valid 18".? estimates of actual densities or population trends from one period to the next. For purposes of ready comparison, the data have been converted to a catch-per-unit-area basis, and the results shown graphically in fig. 52. While the relative numbers taken in the two types of woodlots do not vary widely, a fairly consistent preference for the ungrazed condition is shown. This was most evident in 191.0, and it is possible that the mast failure in that year may have been a further influence. Several authors have reported on the great variety of foods accepted by red squirrels (Klugh, 1927; Murie, 1927; Hatt, 1929), and the annals are known to be adaptable in the face of food shortages (Hosley, 1928; Williams, 1936). Most ungrazed woodlots with their abundant variety of shrubs provide greater quantities of supplementary food than do the grazed type. In years of mast failure it might be expected that ungrazed units would maintain a higher carrying capacity for species of diverse food habits. Upland and lowland types.- The red ”squirrel catch for 86 acres of upland and 26 acres of lowland woodlots is given in appendix table vi. This information provides the basis for fig. 53 in which the comparative numbers of individuals using two lowland and six upland woods are shown. As in the comparison of grazed and ungrazed woodlands the data have been canputed to a standard area of 100 acres for each type. 19( 5 .ueouoooe. oeeatoa one pecans 5 common». .epoaeook oeufiwg one possum .ofiotfisou pea no oneness ehBEemsoo .3 .m: 5.. peanut» 30.2.35 can no once—5e ebapenemfioo .mm .mah P NV! . 0*... x3 . A , ace .3 «8. .3. 3.2 a . cue as»... t! o _ (a a m a. e. a t m . - — . . , w. z m e. o m m . m U H , e. m .oe U s 3 .e _l n : co m d O 3 w .3 8 m m 8 o .. w. m a. .8. m on C ,_ w... L O _ . 38.. 0.33 D T nooos and}... U M .833;- noooi 3:52: . 62 m I. O m i m m 191 These grass nests ,‘._:~_.\. 1:.“ a p ’ ‘ “gig" Not all of the red squirrels were wood1ot dwellers. were constructed in isolated fencerow trees. Hg. 54. 19.1: The red squirrel's preference for a lowland woods habitat is clearly shown. With the exception of the trapping period, for September—October, 19141, when the catch was about equal in the two types, the numbers taken in lowlands were 2 to 8 times that of the upland situations. In this same fall period, an individual catch of 80 red squirrels per 100 acres in mgrazed upland woodlots indicated a higher density for this type than was found in lowland areas (62 per 100 acres). This can be attributed in a large measure to a temporary concentration of 2!; individuals in one lZ-acre ungrazed woodland (woodlot number 6). The conditions in lowland woodlots which contributed to higher populations of this species are not entirely known. In these study plots, the greater number of species of trees and shrubs, and the denser ground cover ' in the lowland types undoubtedly were important factors. An abundance of red maple (Acer rubrum) in both lowland woodlots, and sugar maple (£931; saccharum) in one, may have con- tributed to the greater acceptability of this type. Various parts of both of these trees are known to be eaten readily (Hatt, 1929; Klugh, 1927) , and referring to the latter, Hatt states, "This tree is among the most important food sources for the red squirrel where their ranges coincide.” Outlying habitats.- Red squirrels in this farming area were not restricted to woodlots. Practically every swale area had several inhabitants, and in many cases a few isolated trees, or a strip of brushy fencerow or roadside cover, maintained one or more of these h . ‘ 1-171“) animals.- While it is difficult to express quantitatively the extent to which such habitats were used, some indication of their importance can be obtained by reference to the red squirrel catch in rabbit lines. It is very likely that many of the animsls taken in rabbit traps were using outlying cover as travel lanes between woodlands. However, a considerable percentage of red squirrels for which five or more capture records were obtained were known to utilize such areas regularly. Over one-fourth of the individuals handled five or more times appeared in rabbit trap-lines as well as in the woodlot trap— lines. An additional 11 percent were taken only in rabbit traps and can be presumed to have spent the greater part of their time in these habitats away from woodlots (table 19). Some evidence was obtained to indicate that red squirrels in marginal habitats had a greater tendency towards stability in numbers than did those in woodlands. The papulation decline of red squirrels during the three years of this study has been discussed. On 102 acres of woodland, comparable periods of fall trapping took 62 individuals in l9h0, 30 in l9hl, and 13 in l9h2. Over these same trapping periods, six rabbit lines which.included.12h traps in.l9hO, and.th in l9hl and l9h2, were in Operation in fencerow, roadside, and swale habitats. The autumn red squirrel catch.on these lines'was 22, 16, and.lh in successive years. A further indication of the red squirrel population trends in the two environments can be shown by a comparison of the total annual trapping success. From fig. 55 it can.be seen that the catch percentage for woodlot inhabitants declined from a level of 6.h percent RED SQUIRREL ANNUAL CATCH PERCENTAGE IN WOODLOTS COMPARED WITH FIELD AND SWALE HABITATS 81 e4 — WOODLOT TRAPLINES ' --- FIELD a SWALE TRAPLINES 6- u 2 '— 5 u 4" (I U a. I ,9 5 2‘ o 1940 ' 194: ' n942 Fig. 55. Annual catch percentage as an index to the papulstiom trends of red squirrels in woodlots com- - pared with those inhabiting the scattered woody situ- ations in fencerows, road sides, etc. $1. 195. in 19140 to 1.1 percent in 1912. In outlying habitats the trapping success, while lower, was maintained at a fairly uniform level during this same period and.varied in extremes only from 0.8 to 1.3 percent. Trapping pressure which formed the basis for these latter catch percentages amounted to over 8,000 trap-days in l9h0 and l9h2, and 17,000 trap-days in l9hl. ‘Without knowing what forces operated to bring about the decline in the red squirrel population, it is difficult to estimate the full significance of the relationship discussed above. However, certain features of the nonawoodlot habitats were in contrast with‘woodlot conditions. For instance, isolated trees were found to produce mast more abundantly and consistently than did those in.woodlots. Also, outlying units of brush included a variety of fruiting shrubs that undoubtedly provided important supplementary feed in times of mast shortage. The fact that many of these marginal areas bordered cropfields gave additional opportunity for providing against food shortages. Except for the year 19h0, there was no good evidence that a food shortage actually existed in any of the woodlots, and it is doubtful that this factor figured.prominently in the population decline. Nevertheless, the greater consistency in food.production ‘which characterized the outlying habitats must be considered.a stabilizing influence in the lives of animals occupying such areas. There is no good.basis for a quantitative comparison of red squirrel numbers in the two types of habitats. ‘With generally high populations the density of animals undoubtedly was far greater in the woodlands. ‘Under such circumstances they might be more vulnerable 19s; to epidemic diseases, and possibly other decimating factors, than would the inhabitants of discontinuous units of cover. In any event these evidences of wide variations in population trends between habitats are such as to suggest a survival value for low capacity habitats. In this connection, a conclusion reached by Evans (191:2) following an intensive study of several woodland mammals in England is of interest: "The distribution of Clethrionomys appeared to be associated.with its relative density; The population 'survival' from a period of high density to one of low density appeared to be greatest in those habitats which had normally'mainp tained a low population density. It is suggested that habitats which will maintain only low densities may be essential to the ultimate survival of a species." Biological notes Population turnover.- The widespread.movements'which red squirrels underwent in this study make it virtually impossible to determine the rate of die—off and replenishment in the population. However, the combined effect of losses due to mortality, and.losses due to movement from the area, were observed to result in a marked annual turnover in individuals composing the population. In l9h0, trapping was confined largely to an area of 800 acres. Following land acquisitions in l9hl and l9h2, additional habitats around.the central BOO-acre area also were trapped, and the total area under study was nearly 2 sections. For purposes of the present comparison the data probably are most significant if the sum total 19‘? of all trapping in each year is considered. On this basis it was found that 21 percent of 169 red squirrels marked in the calendar year 1910 were retaken in 191:1, and only 3 percent of the 1910 popu- lation were still present in 191:2. Of 197 animals marked in 191:1, only 9 percent were included among the 1.12 individuals trapped during the following year. Recoveries of marked animals away from the stucb' area were too few to provide a basis for estimating the extent to which population turnover can be attributed to this factor of movement. Considering the extent to which it occurred between individual woodlots, however, it appeared to be appreciable. A single recovery of a Juvenile female 5.5 miles from the study area indicated that such movements may be extensive. This animal (no. 5630), last handled on July 5, 19140 at Rose Lake, was found dead in the road on February 26, 19111 in Victor township, Clinton County. In the case of game species, hunting returns often are helpful in revealing the extent of such movements. From this source of information it has been established that fox squirrels may move fer distances of 30 or he miles Longevity.- Several red squirrels were known to have lived for over two years, but only a single record was obtained to show a longer age. An adult male red squirrel (no's. 5163—11587), first trapped on March 18, l9h0, was last handled in the same woodlot on June 29, l9h2, at which time it was at least 3 years old. During this interval it was trapped 1;? times and changed its woodlot 19a» residence twice. On two other occasions it was trapped in a "field line". Klugh (1927) reports a record of a pet red squirrel which apparently reached the age of 9 years, and Walton, 1903 (in Hatt 1929) refers to a tame though free animal whose record extended over 10 years. A very few fox squirrels reach the age of 6 or more years in the wild (Linduska, l9h72). By analogy it is to be expected that red squirrels, with a higher biotic potential, would even less frequently attain a similar life span. Trapability.- ‘When confined in boxptraps, red squirrels appeared to be far more excitable than any of the other sciurids. However, there was no indication that such an experience discouraged subsequent entry of traps. The same animal frequentlwaas recaptured on several successive days, and on a few occasions the animals re—entered the same or an adjoining trap a few minutes after release. Nine indi- viduals were handled 30 or more times (table 18), and several animals had trap records showing 50 or more recatches. In one 3aweek period of'December trapping, six individuals in one woodlot were trapped a total of 39 times; and in an extended period of summer trapping, seven individuals in another woodlot were handled a total of 75 times. Due to differences in trapping pressure it is not possible to show accurately any yearly differences in their susceptibility to trapping. However, the animals were trapped more readily in l9h0 than in either of the later years. This was explainable on the basis of a critical shortage of natural fbods, a situation that has been discussed elsewhere. From 1,883 capture records for red squirrels {IV 19.) cease» nee-93 cassava—cud no 838396 ego. 3152:”... 32—. 3a soon :3 «no sown 38.09 a.» a: We... n6 3.. mm. «3" ace 3» 5 a.» 3m an 2.3 .6 one 3. man one 2. 33 73.335 .333 canoe—Suva” 7:353 .333 .725)ch you so co.— .uea «e .o: . uefiouwm .538 :75: seal— owe. nose .3.“ nominees was 3:55. e... .5 333 :38 .3 .3...“ 20‘}: no significant difference is noted in recatch records for the sexes (table 21). In the 3-year period, males were trapped an average of 1;.11; times and females 1:.61 times. Sex ratio.§- The over-all total of animals handled during the study showed that males were taken in slightly greater numbers than females. Although the annual differences were not great, the sex ratio was most nearly even during the year (191:0) of highest popu- lation density. In the following two years of population decline, an increasingly larger percentage of males was taken (table 22). Age ratio.- Considerable difficulty was experienced in distinguishing ages of red squirrels in late summer and fall, and the results are offered only for their possible value in indicating trends between years. This information, obtained mostly from exami- nation of females, is summarized in table 23. The most that can be said for these data is that the indicated production and survival of young is in agreement with population levels for the same periods. Year 1940 1941 1942 W 20.1 Table 22. Sex ratio of red squirrels taken in live-traps Year W 2.92.2: Number examined: 169 197 112 478 Isles: 89 105 61 255 Females: 80 92 51 223 Mala/loo fmles: 111 114 120 115 tInoludes some duplicatim of snimls between years. Table 23. Age ratios of red squirrels taken in sumer and fall trapping. Pe rcgnt yo Total Age .3/ examined Adults m uncertain hows Betimst 135 39 61 35 61 71 62 29 21 12 42 53 24 15 5 4 23 38 hug aniuls of known age. Assuming questiomble group included all young. 21125 Eastern Chipmunk (Tamias striatus rufescens 3615 and Moulthrop) ngulations Relatively few population figures are available for this species. Over a heyear period (1932-35), Williams (1936) estimated autumn densities of h to 15 individuals per acre in.a beechhmaple woodland in Ohio. Bole (1939) reported what he thought was an exceptionally high.population of 8.7 per acre for upland forests in Ohio in 1935. He remarked that "....similarly favorable conditions in 1937 resulted in but a very slightly higher figure (l.lh per acre) for 1938 than the lowpoint of the previous year (0.8 per acre).” On two small plots totaling 5.52 acres, Burt (l9h0) found populations in 1935-37 of 2.h to 3.6 per acre following appearance of the first litters in.May and June. He suspected that September populations, when second litters ‘were out of the nest, would average somewhat higher. Observations on the breeding behavior of the eastern chipmunk are not in full agreement. Condrin (1936) andwillen, E. G. (1938) report that twotnoods may be produced. Burt (l9h0) indicates that old females produce both a spring and summer litter, and that young females of the spring broods, also, may give birth to a fall litter. Schooley (l93h) states that ”....it is clear that it is the old females that breed in the spring and that the younger ones, even those born the preceding Spring and summer, are the ones that breed during the July cycle. 01d females that have not become pregnant ' from the spring mating may also breed at this time." Whatever the . 20;; .& details of the summer breeding period may be, it seems obvious that peak annual populations would be reached in early fall after the summer litters become active. 0n the seven study woodlots totaling 102 acres, summer trapping (June and July) for the years l9h0—h2 took 87, 76, and 33 animals respectively. September trapping during the same years accounted for only 65, 20, and hO individuals. This summer-to-fall reduction in numbers during two years is not what would be expected from.the animal's breeding history. However, it is not surprising in view of other marked variations in numbers noted from one trapping period to another. In l9hl, for instance, only five animals were taken with the same intensity of trapping in August which a little more than a month earlier had produced 76 animals. It is doubtful that all such variations in catch can be accepted as representing true quantitative changes, since the chipmunk is known to enter into periods of inactivity during the summer months (Seton, 1909; Schooley, l93h; Williams, 1936). ‘With reference to such disappearances of these animals near Ithaca, New York, Allen, E. G. (1938) observed that, '....there seems to be a definite tendency for chipmunks to disappear during warm, dry spells, especially after the midsummer litter is raised, but it has been noted also in July." In this study, the trap spacing of one per acre may have been too sparse to effect a catch of all resident chipmunks. However, it is questionable that this was the case, since individual animals frequently were taken in more than one trap, indicating that their range extended beyond limits represented by the sua— distance between traps. Furthermore, studies by Burt (19140) and Blair (19142) indicate a home range of about 2 acres or more for these animals. In any case, it is doubtful that the average density for all woodlots at any time during the study was more than one animal per acre. In basing this figure on the highest of the over-all trapping returns, it should be mentioned that the average population for the majority of the woodlots was well below this density. There was nothing approaching an equitable distribution of the total catch through the seven woodlots, and during both the high periods of 19140 and 191:1, a majority of the total was concentrated in a single woods. In 19140, 57 of the total of 87 animals caught in summer were from one woodlot (no. 5), and in 19141, 1d; of 76 were taken in the same 23—acre woods. This distribution of the animals would indicate a pOpulation of 2.0 to 2.5 individuals per acre for this one woodlot, while other woodlots had as few as one animal per 8 to 15 acres. Habitat preferences As has been indicated above, chipmunks showed definite prefer- ences in regard to type and condition of woodlots. A comparison of the total catch in woodlots trapped concurrently and with the same intensity will show the nature of some of these preferences. Effects of grazing.- The catch records for chipmunks, during several periods when significant numbers were handled, are given in appendix table vii. These returns were obtained from three grazed and three ungrazed, upland woodlots which totaled 141 and 115 acres 2115 w,l.\e_|\ - dent’s-mt:- - ..-{....~- j:f;; ,1.“ at» a ‘ ' W ,Ww'faa‘é "as a; -1- t _ / fi nap-'3,“ a '{gfifl “ ,7- After 5 years of protection, new seedlings are slowly Over 20 years of grazing eliminated most of the reproduction in this up- land, oak-hickory woodlot. becoming established in the bluegrass sod. Fig. 56. 2116 oegfinfino he agavensmom AME esteem: notes pcguw mo umfisdo @2396: use roam—soon: ounce 333305 a warps: 3.0.2.00: .5 5.3 211’} reapectively. In fig. 58, the results have been converted to a unit area basis of 100 acres. It should be borne in mind that these data are offered only for the purpose of comparing the relative catch of chipmunks within each given trapping period. Since trapping periods were of variable length, comparisons between periods would not be entirely valid, and for the same reason, it is doubtful that actual populations are shown with any real accuracy. Catch records at intervals over three years showed that these animals consistently were using ungrazed woodlands to a far greater extent than they were the grazed units. The numbers taken in each period were from three to six times higher in ungrazed woodlots, and with minor exceptions the trap returns for individual areas showed that each of the ungrazed units was supporting greater numbers than were any of those which had been grazed. The chipmunks' greater use of ungrazed woodlands indicates a preference also shared by the red squirrel. In view of the habits of the two species, it is not surprising that they would reach their greatest abundance in areas offering abundant ground cover. Upland and lowland types.- Trapping returns for six upland woodlots of 86 acres and two lowland woodlots totaling 26 acres are given in appendix table vii. These data have been projected to a "catch per unit area" basis and the results shown graphically in fig. 59. During the summer and autumn of 19110, chipmunks were 9 to 12 times as abundant in upland woodlots as they were in lowlands. In the summer of 19111, a similar marked preference for the upland 13112: type was shown with a catch about 11 times as high as that in lowland woods. In 19142, however, about equal numbers were taken in the two. types in summer trapping, and a slightly higher catch was made in autumn in the lowland type. Howell (1929) and Quimby (l9hh) have reported that this species shows a definite preference for upland woodlands. As has been pointed out, the difference in length of trapping periods does not allow for close comparisons of populations from period to period, or from year to year. However, the longest trapping intervals occurred in 19140, and the fact that the lowland catch of chipmunks was highest in 19142 is some evidence that an increase may have occurred in these habitats. If the higher catches in 19142 in lowland areas did reflect an increase, or even a mainte- nance of numbers in face of a decline in upland areas, no satisfactory explanation can be offered. Food should not have been an incentive for a movement, since supplies appeared to be adequate in both types. Neither is there any reason to suppose that production or survival of young was greater in lowland situations. Weather conditions, also, were about normal during the surmner and fall of 19112. and during the preceding winter. While it may be of no Special signifi- cance, it is, nevertheless, of interest that Quimby (191414) observed somewhat the same situation for this species, and during the same years in Minnesota. He reported that, "Eutamias, which I have taken most often in swamps....was more abundant in 19142 than in 19141, while '1_‘a_.r_ni_a_s_, which according to my records is more of an upland Species....was less abundant in 19142. In general, the species and 20:;— individuals preferring swamps were able to maintain approximately the same populations for the two years whereas those preferring the more open uplands decreased considerably." In explanation of this situation, Quimby suggested that weather conditions were such as to cause a differential overwintering success in the two habitats. Outlying habitats.- In contrast to red squirrels which inhabited, and.frequently visited, small patches of woody cover and fencerows lying some distance from woodlands, the chipmunk seldom ‘was taken away from woodlots. Only 3b of 296 individuals were taken in "field line" traps, and among 56 animals handled five or more times, only two were credited with residence outside of woodlands. Chipmunks also were observed to confine their activities to a given locality to a far greater degree than did red squirrels. In the case of the latter species, less than half the individuals handled five or more times limited their travels to one woodlot. Among chipmunks, 91 percent of those with a record of five or more catches were taken in only a single'woodlot (table 2h). In this farmland area, the failure of chipmunks to colonize outlying cover areas to the extent noticed.for red squirrels prdbably can be attributed.to their reluctance to cross open ground. Burt (19h0) states, "Chipmunks in this area rarely are seen at any distance from wooded.or brush- covered areas..... They always seem to desire a protective cover and certainly are not at home in the cpen." Howell (1929) observes that, "....they are.partial also to stone walls and rail fences but rarely leave their protecting shelter for any distance to enter adjacent fields." 211'» .npbupooi encased use 383: a.“ @2395 .3093? consumes use pougw 5 common». mans-mace no anon—a5: 05953500 .3 .wwm cage mo among: ozvfioaaoo .mm .3..." _ . m «we. .ve otm. . _ _ use has 1.33 to as... , «to. 3% java. game. g 3.35 Kama. ix: 0 ' Kbo_ \335 $33 K90 ub‘ Haw 3,33 $33 Kama xx: .0 - A on w V .om 9 w _ m w _ d n i m N , n y _ N S _ vow W. 00 mm M U m nlu m 0 v .00 m on W s W F m _ o J _ .8 m _ M w r. m ON_ w I o _ m . .oo. 1 r on. 300; 9.243 D nooo; 3526.. I 3003 9333 D .00. v 8003 03532: I _ 1i. N F" a; CHIPMUNK ANNUAL TURNOVER IN POPULATION 100 WHITE). so [ S 100 as (ii Q D. g so 1 ,. - TRAPPED IN 4940 2’ II II I94| ’— 2 40 111111 .. .. .942 Ll. o '— 2 LL] 3 20 1.1.1 0.. 14 / e / ' | o l940 I94! 1942 Fig. 60. The numbers of chipmunks remaining in study woodlots from one calendar year to the next. 2:1 2. Biological notes Population turnover and longevity.- Allen, E. G. (1938) gives accounts Of two captive chipmunks, one Of‘WhiCh reached the age of 5 years and the other nearly 8 years. It is doubtful that very many experience such a life-Span in the wild. In a study extending over 3 years, Burt (19h0) marked 106 Of these animals and had one indi- vidual which he knew to be 3 years Old, and three others which‘were at least 2 years Old. In this study) lhl individuals were marked in the first year and of these only one was accounted for 2 years later. This adult female (no. 58h5) was first trapped On August 26, 19h0, and was last handled on April 15, 19h2, at which time it was at least 3 years Old. Between these dates it was trapped a total of nine times, six of these captures being in the same trap, and on three other occasions the animal was taken in an adjacent trap. There was little evidence in this work to indicate extensive distributional movements for chipmunks. Consequently, the marked year-tO-year turnover in pOpulations must be credited.main1y to a high mortality. Limited information indicated that as many as 90 percent Of those reaching an independent age in.any one year succumb before the following year. In.l9hl, only 1h animals were recovered Of 1h1 marked the previous year. In 19h2, nine were still present on the area out of a total of 97 handled the previous year (fig. 60). While no particular effbrt was given tO evaluating mortality factors for woodland.mammals, one predator relationship involving the chipmunk and weasel seems worthy Of note. 'While evidence for 21:“: the importance of this relationship is mostly circumstantial, it is probably no coincidence that woodlots most densely populated with chipmunks also supported the greatest numbers Of this small predator. In suxmner trapping, 19140, 57 Of a total Of 87 chipmunks handled in the seven study units were taken in one 23-acre woodlot (no. 5). In this same period, five Of eight weasels trapped in woodlands were taken from this same woods. In 19141, woodlot 5 again produced a majority of the chipmunks trapped in summer (1414 of a total of 76). During this period two Of a total Of five weasels caught were taken from this woods, and “0 others. were trapped in the woodlot (no. 1) supporting the second highest number of chipmunks. The importance Of the weasel as a predator on chipmunks apparently has not been investigated. Allen, (E. G. (1938) snapected that domestic cats take the highest toll, and added that, "The enemy next most dreaded is, perhaps the weasel ...." Williams (1936) observed cats stalking chipmunks and indicated also that foxes and weasels might be of some importance as predators. Allen, D. L. (1938) reported, fl....weasels are most effectivelyadapted to preying upon small fossorial animals. They are persistent enemies Of such mammals as chipmunks, apermophiles, and all species of mice...“ All appearances indicate that the weasel may be the most effective check on the numbers Of the spermophile on this area." McClelland (19148) on two occasions witnessed a weasel pursuing a chipmunk and commented further that, I'The number Of chipmunks varied greatly from year to year, and one factor may have been the weasel (Mustela)." Aside from a concentration of weasels in those woodlots most densely populated by chipmunks, no good evidence was obtained as to the importance Of the relationship. ‘When released from traps weasels frequently entered.burrows known to be used by chipmunks. On a number Of occasions weasels burrowed under boxptraps containing chipmunks, and in some unobserved manner were able to chew the feet of these trapped animals. Two or three trapped chipmunks were killed by weasels which apparently had been able to grasp them through the floor Of the trap. Also, one instance Of a bold attack of a weasel upon a chipmunk was Observed. This weasel (no. 537k),'when released from a trap, remained within a few yards Of the Operator ‘while a chipmunk from an adjacent trap was examined. ‘When the chipmunk (no. 5630) was released, the weasel caught and.ki11ed the animal before it had progressed more than a few yards from the trap. Trapability.- In each year Of study the repeat catches averaged slightly higher for females than for males (table 25). 'While the difference is not appreciable, it is hardly what would be expected considering that females probably are more restricted in their movements due to their preoccupation with family duties. In studies on the long-eared chipmunk (Eutamias quadrimaculatus), HOldenried (l9h0) also recaptured females somewhat more frequently than the males. Sex ratio.- Blair (19h2) in a study of this Species in northern Michigan found an even sex ratio (51.3% males and h8.7% females) among 15h live-trapped individuals. In his studies on the George Reserve (Livingston County, Michigan) Burt (19h0) took 106 chipmunks in the .214} ratio of 60 percent males to ho percent females. Among 333 individuals handled at Rose Lake, males were found to constitute only ho percent of the total and females 60 percent. A similar unbalance of the sexes is noted when the data are considered separately by years (table 26). ‘Whether or not such differences in sex ratio are significant is hard to say. Allen, E. G. (1938), Condrin (l936),and Schooley'(l93h) have observed that males appear first from hibernation in spring, and Schooley'(gp, git.) comments further as fellows; "Early in the spring the entire catch of animals is likely to be males. This condition changes rapidly so that after about a week of favorable weather the proportion may be three males to one female. The ratio continues to shift until, after a large percentage of the females have ovulated, the balance may have shifted to one male to three females." It is obvious that such differential activity of the sexes would be difficult to evaluate in instances where unbalanced ratios are observed. Hibernation.- Trapping in early spring and late fall was not of a continuous nature which would.allow for determining accurately the dates of entry into, and emergence from, hibernation. Hewever, sight and trapping records testify that this animal is not markedly regular in the matter of hibernation. Individuals occasionally were seen in late November and a few became active on warm days in midwinter. In late December of 19b0, two were taken in traps, and another was trapped in mid-January of 19142. In the latter instance the animal emerged during a week when the average temperature was 2115 27 degrees F. with a maxiJmnn of )43 degrees. Other investigators have observed that chipmunks may emerge from winter hibernation when weather conditions are favorable (Howell, 1929; Test, 1932 3 Williams, 1936; Allen, B. G., 1938). N2 9" \T Tabl. 24. Pennenency of residence of 56 chipmunks handled 5 or more times Si tun tione Where alight _ Three l’oodlet Field One Two or more end line woodlot woodlote woodlots field line only No. of individuals: 51 1 O 2 2 Percent of total: 91.0 1.8 O 3.6 3.6 Teble 25. Total catchee of ohipmmke and averages for each sex. lhlee Fem lee Cetofiee Catches Indi- per Indi- per vidue 1e Cetohee individual vidue 1e Catches indiv‘igg 1 1940 51 180 2.55 83 267 3.22 1941 42 98 2.33 49 133 2.71 1942 32 66 1.88 _5_8_ 114 1.97 Totele 1284- 294 2.30 190‘ 514 2.71 tIncludee some duplicetion of individuals between yeere. Table 26. Sex ratio of chipmunks taken in live-trepe. Year ._.4. ._.a. .2. :22. Total enmined: 141 97 95 333 lhlee: 53 45 36 134 leee: 88 52 59 199 lhlee/loo ton-1... so 87 61 67 idléi Southern Flying Squirrel (Glaucomys volans volans Linnaeus) While most authorities agree that flying squirrels are among the most arboreal of the sciurids, there is little agreement on the extent to which they may move around at ground level. Little ground travel is indicated for the species by Sollberger (1940), and by Howell (1918) who reported that they seldom.descend to the ground, ”.... apparently never running for any distance on its surface”. Other workers are inclined to credit these animals with considerable activity on the ground, and Burt (1940) reported that they may move between woodlots by traveling under the cover of brushy areas. The trapping experiences of Jordan (1948) would also indicate frequent ground travel: "Of all traps which were success- ful in making captures, only six were located above ground level, a fact that conflicted with the previously accepted techniques of attaching them to the holes of trees.” Since flying squirrels are frequently captured by house cats, and are a common source of aggra- vation to northern fur trappers (Nelson, 1918), an exclusively ar- boreal existence is hardly indicated. :51}: Populations While observations in this study suggested frequent terrestrial travel by flying squirrels, it is doubtful if ground trapping was effective in obtaining accurate population figures. The animals entered ground traps readily, but a few comparisons with tree- 1ocated traps showed that the former were less effective. During the last three weeks in December, 1941, 10 nest-box traps were paired in location with 10 box-traps and operated concurrently. No flying squirrels were taken in box-traps, whereas the tree-placed traps caught two animals, three times. The same combination was tried in a different woodlot for 14 days in mid-March. Nest-box traps in this case took two animals and box-traps failed to make a catch. Considering these limitations of method, the data are recorded here with a realization that population figures may be low. However, this should not interfere with one of the main uses of the data, that of showing comparative densities and annual trends. In his studies on the George Reserve, Burt (1940) found that the flying squirrel population on a 3.72-acre plot ranged from 1.6 to 1.34 animals per acre from.late June to miquugust. Near.Ann Arbor, Nflchigan, Jordan (1948) took slightly more than one animal per acre on a 28-acre oak-hickory woodlot during July and August. Bole (1959) Observed marked variations in the numbers of these squirrels in Ohio woodlands. The largest population noted by him was six per acre in a beech-maple forest during late summer. He estimated that peak populations in 1935 ”....were generally in the vicinity of 3 per acre....", and in 1939 animals were so few that C 00-. 2531: scarcely an individual could be collected. Sollberger (1943) re- ported taking nearly five per acre from a small Pennsylvania wood- land and 2.2 per acre from a 10-acre woods near Ithaca, New‘York. While it is doubtful that box-trapping effected a complete catch in this study, July-August trapping in 1940 took so individuals on 102 acres. In September-October of the same year 36 were taken, and in December only eight. In 194l,trapping produced 11 individuals in June and July but none in September and October. During 1942, five periods of trapping between January and.November never resulted in a catch of more than six of these animals. Obviously flying squirrels underwent a marked decline in numbers from.l940 to 1942. There is no way of knowing whether the 1940 pOpulation represented a high I density for this area, or if even then it may have been reduced from an earlier "high". In any case, the minimum average population for 102 acres during that summer and fall period was close to one squirrel per two acres. For several years after 1942, activities of the station were so reduced that intensive box-trapping could not be maintained as a means of following population trends of many species. However, a 10-day interval of trapping from April 21 to Lay'l, 1944 took 16 flying squirrels from.one 15-acre woodlot, which would indicate some recov- ery of numbers by that year. As has already been reported for the other sciurids, the flying squirrel population in the seven woodlots was by no means evenly dis- tributed. With the highest population studied in the summer of 1940, one squirrel per acre represented the maximun.catch for any of the 2531 woodlots. This density was measured in woodlot 6, an ungrazed unit of 12 acres. In the same period, no animals were taken in woodlot 1, a lO-acre woods of similar condition and composition. This woodlot was the only one of seven which evidently was uninhabited by flying Squirrels in the summer of 1940. In the fall of the same year, this unit again was the only one in whiCh catches were not made, and it was not until late December that there was evidence of flying squirrels using the area. Two were taken in nest-box traps at that time. No further catches were made in this woodland during the following two years, a rather surprising situation since it contained many mature trees with dens, and appeared well-suited to the needs of this species. It was, however, the most isolated of all the woodlands, and after a total loss of animals, re-establishment of a population in the area conceivably might be a long—delayed event. Habitat_preferences So few critical population studies have been conducted on this species that it is difficult to estimate what constitutes their op- timum.habitat. Sollberger (1940) reported that they occur most commonly in climax forests of the beechemaple type. He attributed this to the numerous nesting cavities and the usual abundance of nuts. In his studies of such a woodland (Sollberger, 1943), he found a population of 2.2 Squirrels per acre, although the time of year is not indicated. Beech-maple woodlands studied by Bole (1939) in Ohio averaged about three per acre (presumably at the peak annual density) during a "high" period. Oakéhickory types apparently are less productive of this species, since the figures obtained by Burt (1940), Jordan (1948), and in the present study, show populations of not much over one per acre during late summer. This study failed to reveal significant differences in the numbers inhabiting upland oak-hickory, and lowland elm-maple types. With reference to the density of ground cover, a slight preference was shown for grazed woodlands (appendix table viii). This would seem to agree with the observations of Jordan (1948), who obtained the greatest number of catches in traps located where ground cover was of sparse to medium density. Both Jordan (1948) and Hanson (1944) observed that severely overgrazed woodlands were avoided by flying Squirrels. It is well—known that flying squirrels will construct outside nests in which they may bring forth young (Landwer, 1935; Cowan, 1936; Burt, 1940; Sollberger, 1943). Whether or not they require the more protective cover of tree-dens in winter apparently is not known. Cowan (pp. 313:) indicates that while dens are used in winter in the Tran- sition Zone, the outside type is preferred in summer. Hanson (1944) and Sollberger (1943) have indicated that the frequency of den trees may be important in regulating the numbers of flying squirrels in some areas. This study afforded an opwortunity to observe the use of arti- ficial dens by flying squirrels, and to determine the possible effects of such cover on their populations. In July, 1940, 24 nest boxes were installed, two to the acre, in woodlot 6. These Were subsequently ~examined for signs of occupancy in October and December, 1940, and in April and.iay3 1941. Practically all boxes were used, either by arboreal mammals, several species of birds, or bees. Nests of shredded bark, and other signs of use by flying squirrels, were found in five boxes or three examinations and in four boxes on the fourth examination. Three flying Squirrels were present in separate nest boxes at the time of the April, 1941, examination. while there were these repeated indications that flying Squirrels were making good use of this abundant den cover, the population trend for this woodlot was no different from that noted in other woodlots under study. Just prior to installation of the boxes, 12 animals were taken by trapping. In autunm.on1y two were taken, and in the following two years, trap- ping never accounted for any more than this number in this particular woodlot. Biological notes Population turnover.- Flying Squirrels were undergoing a marked decline in numbers during the course of the study, and it is possible that the turnover of individuals was higher than would be the case with a thrifty population. In any event a low year-to-year survival is indicated, just as was found for the other sciurids studied, and as is probably the case for small mammals in general. 6f 78 indi- viduals marked in 1940, only four were retaken in the following year when they represented 21 percent of the much-reduced population. None of 14 animals trapped in 1941 was recovered in 1942 (fig. 61). The Statement by Sollberger (1943) to the effect that, "Probably ‘five years represents the normal life Span of the flying squirrel," could hardly apply to wild populations. wwm mm” U5 LNm. m lam H} W 1‘ 1 0 O 0 O 9 7 4 mmmmmmmmmmmmmmmmmmmmm |||| m M. Table 27. Pemnenoy of residence of 82 flying equiuele. Situations where tn (1 Three Woodla- FIeH One he or more end line woodlot woodlote woodlote field line 0111: No. of individmle: 66 14 O 1 1 PONODt of takld 80.5 17.1 0 1.2 1.2 ‘reble 28. Sex retio of flying equirrele taken in live-tape. Year as. 1242 £933. 2.22.1. lumber examined: 78 11 16 105 Male” 39 3 6 48 Fan 10! x 39 8 10 57 nun/100 run... 100 as so 84 226 Kovements.- Considering the apparent ungainliness of these animals on the ground, a rather surprising number were found to have moved between woodlots (table 27). Fourteen of 82 aninals were trapped in two woodlots, and in all cases the transfer would appear to have re- uired travel over ground. Under the stimulus of a "homing urge", LbCabe (1947) found that one of two marked females returned from a release point at a treeless distance of 1/4 mile, and over a woodland distance of 1-5/4 miles. For several Species of rodents, Evans and HOldenried (1945) have reported the frequent double capture of individuals of a species in single-catch traps. Similar evidence of flying squirrels traveling together has been reported by LCCabe (1947). In this study, more than one-fifth (17 of 78) of the flyina Squirrels taken in 1940 were in- volved in double captures. This concomitancy in travel may have been related in part to breeding behavior, since 6 of 10 double captures involved opposite sexes.) A few of these records are of in- terest for other behavior notes which JEUv provide: On Ju1y 4, 1940, adult male no. 5551 was taken in a trap in woodlot 4 with adult female no. 5552. On July 20, 194 , the same male was retrapped in woodlot 3 with a second adult female no. 5685. The record of this male might suggest a promiscuous breeding behavior. Adult male no. 5424, captured in woodlot 5 on July 16, 1940, was retaken 2 days later at a point nearly a quarter mile away in woodlot 6. At the second capture it was accompanied by adult female 110. 54250 i323? Innature male no. 5723 was trapped with immature female no. 5724 on July 31, 1940. One week later, on August 8, this male was recap- tured in the same woodlot with an adult male no. 5663. Trapability.- Recaptures of flying Squirrels were considerably lower than for other sciurids, and among 104 individuals, nearly half were captured only once. AS was found to be the case with chipmunks and red squirrels, females of this species were retrapped slightly more frequently than were the males. Sex ratio.- Among 78 individuals caught in 1940, the sexes were equally divided. In the snall numbers handled during the two follow- ing years, females outnumbered males two to one (table 28). Burt (1940) took 50 percent more males than females in a catch of 45 in- dividuals, and in trapping 32 animals Jordan (1948) found an essen- tially even distribution of the sexes (15 males and 17 females). -Fig. 62. The white-footed mouee wee the comcneet enll manual in farm woodlote. Fig. 63. An agile climber, the F15. 64. White-footed nice often white-footed mouse usually took used nest eitee of other speciee. refuge in the neereet tree when The greee nest occupied by thie released from a live-trap. animal was nude by e red equirrel. (C p; White-footed heuse (Peromyscus leucopus noveboracensis Fischer) Trapping for this species was not done with an intensity which would permit an accurate appraisal of its year-to-year status. It is known, however, that these mice are subject to marked changes in numbers (Williams, 1936; Bole, 1939; Lindeborg, 1941). Limited catches in box-traps, together with information from.other sources, would indicate that annual levels of the Species remained fairly con— stant during the period of study. Certainly, they did not undergo the pronounced population changes noted for some sciurids. Populations Several early studies on this mouse resulted in population es- tinates as high as 200 or more individuals per acre. In the light of recent work, however, it is obvious that such counts were gross- ly exaggerated by fallacies in census methods and in calculation of densities. More recent investigations Show close agreement in fall counts of 5 to 10 animals per acre for most deciduous woodlands (Burt, 1940; Hanson, 1944; Stickel, 1948). .An average density of 29 per acre for upland forests in a known peak year, and a low of 5.2 per acre, were measured by Bole (1939). In this study, intensive methods were employed to determine the white-footed mouse population of a representative unit of woodland. For this purpose the most isolated of the study woodlots (no. 1, fig. 44) was selected. This ungrazed, lO-acre unit of upland oak- hickory was surrounded by pasture and cropland. The nearest of the 2:50 ' other woodlots was at a distance of more than a quarter mile, and the opportunity for any appreciable interchange of individuals appeared negligible. On October 6, 1942, the entire unit was sat- urated with 204 live-traps set out on a 10-foot grid. These were continued in operation for 7 nights, through October 12, by which date the rate of new captures was so low as to indicate that prac- tically all of the individuals had been captured. The total catch was 76 white-footed mice and 27 individuals of other species. This near-peak annual density of 7.6 white-footed mice per acre is prac- tically identical with that found by Burt (1940) in a similar wood- land during early October trapping. Habitatppreferences Townsend (1935) suspected that this species was "slightly more numerous in dry woods than in wet woods". Stickel (1948) took five to eight times as many in bottomlands as in upland woods. Limited trapping in this study indicated a preference for the lowland type. During the period November 1-3, 1942, traps spaced at 22-foot in- tervals, and located on single lines running through the center of two lowland and five upland woodlots, were operated for three nights. This involved 50 traps in lowland and 229 in upland units. The catch per 100 trap—nights was 4.2 individuals in the uplands and 11.8 in the lowland woods. The response of these animals to the effects of grazing of wood- lots was not clearly evident from the cursory studies made. Samples of each type of woodland, when.trapped in 1940 and 1941, produced 2552 nearly equal numbers. In the 1942 trapping described above, 126 traps were located in grazed upland woods and 103 in ungrazed up- lands. In this work a catch per 100 trap—nights of 2.9 individuals in grazed units compared with a catch of 5.8 animals in the ungrazed type. Hanson (1944) found these mice were more plentiful in a mod- erately grazed upland woodlot than in a severely grazed unit of similar type. A possible further suggestion of habitat preferences of the white—footed mouse was shown by the intensive trapping of woodlot 1. This woodland and the rows of traps areshown diagrammatically in fig. 65. The point at which each of 158 captures of 76 white-footed ndce was made is shown by an "X”. lit would appear from the distri- bution of these catches that activity over some parts of this woods was considerably greater than in other parts. While several factors might contribute to this apparent irregularity in distribution of the mice, there was a fairly good correlation between density of ground cover and number of catches. Although the woods as a whole is'a typical ungrazed type, limited grazing a number of years ago resulted in a reduction of cover along the north and south edges (top and bottom in fig. 65). This permitted the establishment of bluegrass sod that a fairly cpen canOpy had allowed to persist to the time of the study. Burt (1940) has reported that this species avoids such grass areas in woodlots, and the catch records for this study substantiate this observation. Elsewhere in the woodlot, an apparent concentration of individuals was found along the intersec- tion of vertical rows of traps 2 to 5 and horizontal rows D to F. This general vicinity marked one of the low areas in the woods. The shrub and sapling growth was particularly dense and the ground was covered with duff and litter. The greater than average take of mice in this situation is in accord with the observations of Hatfield (1938) who found this and another woodland species of Peromyscus to be most numerous on plots having an understory of shrubs which pro- duced fruits of value as food. Indications of a prOportionately large number of animals along the east (right in fig. 65) side of the woods is to be noted, also. The outer row of traps (no. 12) along this margin accounted for 36 captures of white-footed mice, compared, for instance, with only 6 catches in the two center rows (no.'s 6 and 7). Along the entire east edge of this woodland, the canopy was dense, the understory thrifty, and the immediate ground cover consequently sparse (fig. 66). While it may have been this combination which proved attractive to these mice, it is possible that a field of standing corn which bordered the edge that year may have encouraged activity in that area. Nicholson (1941) reported for this species that, "....nearly twice as many mice were captured in boxes near the edge of the woods as were taken in the center....." This tendency also may be reflected in the increased catch of animals on the east and, to a lesser extent, west borders of this woodlot. While primarily a woodland dweller, the white-footed mouse is known to utilize small patches of brushland and groups of a few trees (Burt, 1940; Hanson, 1944). In this study the animals were taken ngydhrly'in shrub fencerows, and on several occasions all their needs appeared to have been provided by a single isolated tree 0 Biological notes Distribution movements.- It is not uncommon for individuals of this species to maintain themselves in shocked corn (Johnson, 1926; Linduska, 194235 Hanson, 1944), and at times to appear in grassland several hundred yards from woodland cover (Blair, 194625 Burt, 1940). Probably such appearances in fields can be attributed to population pressure in their preferred habitat. In the "trap- ping-removal" study (see table 11) a total of 17 of these animals were taken, and the greatest movement through this idle ground hab- itat coincided with the period (November) of peak annual density for the species. Sex ratio.- Nest sex ratios calculated for this species from trapping returns show a slight preponderance of males (wansend, 1935: p. 42; Burt, 1940; Nicholson, 1941), a fact which is attri- buted to more widespread movements of this sex. The numbers handled in the present study were too low for real significance. It is of interest, however, that in the total population trapped in woodlot 1, there was a marked preponderance of females among the adult age class and a slight excess of males in the case of immature animals. In the same autumn period, a small sample ob-. tained from eight woodlots showed similar relationships. For 122 individuals handled at this time, 25 adults included 6 males and 19 females (32 males : 100 females), and 97 young included 54 males and 43 females (126 males : 100 females). .3:— .2... .1. .44.. .5... .53... .7... .9. .3... .19... $1.23.. 4] xx xx xxx xx x ml 1 x x 0| xx 0 x xx xx xx 9| xx 0 xxx xx x x xxx xxx x o x ml xx xx x xxx xxx x x xxx 0 o o x x o 0 ml x xxx x xx xx x x xx xx xx (5' xx 0 0 O x xxx xx m) x x xx xxx x x x xxx H), x x x x xxx ‘ oo '7) :oc xx xx xx xx 0 ml x x o x 00 a! x xx xxx x 0 0 :| x x o x x 00 ml 0 xx x x xx 0 0| o x x x xx ml x x xxx x x x xxx x o or| xxx m xx x Fig. 65. Diagram of woodlot 1 showing location of lino: of traps, and points at which 76 white-footed 11100 (x) won trapped 168 tins and 25 short-tailed ohm. (o) 26 time. F1‘. 66. Thin lO-ecre unit of upland oak-hickory offered excellent possibilities for etudiee of en ungrazed wood- lot. From October 8-12, 1942, 204 live-traps on a lO-foot grid took '76 white-footed mice and 27 individuele of other species. 236 25i5’ Short-tailed Shrew (Blarina brevicauda kirtlandi Bole and Mbulthrop) Probably most of these animals in woodlot l were taken in the intensive trapping of that area. The total of 25 indicated a pop- ulation (2.5 per acre) considerably lower than that estimated by Bole (1939) and Burt (1940) for similar habitats. Too few of these shrews were caught in woodlot tr’pping gen- erally to permit a detailed consideration of habitat preferences. While they were common at times in bluegrass sod bordering crop- lands, they appeared to avoid such a habitat where it occurred in woodlot 1. Instead, the greatest percentage of the population in this woods was trapped along the east and west margins (right and left in fig. 65), where the layer of duff and litter was heaviest. The two outer rows of traps along these two edges accounted for twice as many catches (17) as was made in the eight intervening rows (9). In other woodlots limited trapping indicated densities in low- lands similar to those in ungrazed uplands. Grazing appeared to limit numbers strongly. In August, 1941, nine shrews were taken in an ungrazed woods and the same intensity of trapping (170 trap- nights) took only one in a grazed unit. In 1942, trapping for 309 trap-nights in two ungrazed woodlots produced eight of these ani- mals, and concurrent trapping of similar intensity (378 trap—nights) in four grazed units resulted in the capture of only one. It is understandable that burrowing animals like the short- tailed shrew would find conditions more apprOpriate to their needs in the mulch layer of ungrazed woodlands than in the exposed hard- packed mineral soils common in many grazed areas. Other results of grazing were mentioned by Dambach (1944) in explanation of pop- ulation differences similar to those noted here. He observed that invertebrates in an ungrazed tract were two and one-half times as plentiful as in.a grazed area and that freezing of the soil is much less severe in ungrazed woodlands. With reference to the latter point a series of temperatures was taken at various soil depths in two grazed and two ungrazed woodlots in the winter of 1942. This work was accomplished by means of a Leeds-Northrup potentiometer and a number of Constantan- cOpper thermocouples (fig. 46) loaned through the courtesy of W.‘U. Garstka of the Soil Conservation Service, U. S. Department of.Agriculture. Under the conditions of the study, no significant differences in soil temperatures were measured in the two types of woodlands. This was to be expected, since a cover of snow of comparable depth was present in all units during the period when records were taken. At other times it was observed that the comparative lack of ground cover in grazed woodlands allowed wind action to remove much or all of the snow. Under such conditions low air temperatures prob- ably would produce lower soil temperatures in grazed than in ungrazed areas. A sample of records obtained for one of the ungrazed wood- lands is shown in fig. 68. 23234 This .(l, u—e-n .o.vv--’w-' I ‘- ~' fln" , a «Mr .- ; 53,. i;-w.e~._u§fi ‘\ The chart-tailed shrew wee abundant here. Trilliume ere e common early epring flower in the lowland woodlots. etudy woods, heavily cut-over in perte but ungrazed, had an excellent duff and litter layer. Fig. 67. am #33: earn» .fifififias .5. sweet so: Sausage .3»: 8a. $2 .... an...» 9202!. analogue». :0. no not: obese can 308.3! hen». n.“ evanoopoafi gonads handgun—8." e and.» eunaavook panama: no ...”!!! wad-chasm .3 on: .ifi--.-...M.H. H . . I'IIIOIIO)+‘|¢')3\U ...... i l _wmurglnScaNnuta 4 b :9 ()‘wt ' tttttttt . i 1(‘T$|O|Q\OIO -I'Ol+(|6lue c :3... 333.3% 9.3 333333. game: ”a «£33 3 keno tum . 33.3.).ka £33 . okafiwuelfimx. gawk. Sun.” S_ . . I, ._I:.. ....xfiexmu§.§due§§fiu: : . _ H . w H . - . 4 13.. watts ac mutate 223.256» 363.5 no ”cough _..mr . . -IE; ....... ......... fi\ 53%. Q\ 2%) a «3 V) ()I‘II 9.23.... .Niwprgxn . 3.413.... ...... .E a. was... San... . 9' + n 14-. .. * t4 4 4 4 ¢ ....... 4 a e n ..... 4 . o e e . a - . a . ...... o ...... o, ...... . 4 I)(§ O. e+)§|«1e O ) 10 .+. . . .......... .“ ...... . g ...... w ..... . o a ...... ........ . ,0 ¥ ..... . . 31th .. . .fi. . . . . . '17:-) .L. t 0‘1)+(|6(6Ib O . .......... . e I e t . . . c A t o x . u : a O O t u .. . . . . . . ....... .,. . . . . . . . ..... . . . ¢ . . . . . . .. . o . ...... ..n.. .. . . ,. . _... 4 -.-!(oi-..(f (ft .....(11.. - - . -. 4 L ...... £. .1 . . ). ,. \. .K.... ....... ......... ... . . .. . . .. . .. .. .. ...... . . . 4 . .. . . . . , .. .. . .. ...... I ll JQIOIOIflOI.‘ 11]» e on 0 e 10 e (O: . . A t I e y . . . t . t e. 0 t u t e . I II ) o n I iii 4 t IJIJI.’ 9 o ..... . . H . .. . a ...4.... Q ....... .* _ P o,. ...H... ’ ...... .. . . . ........ . . . ..... .. . . (4F. .+ ......... .. ..H.......W---,q-:-lwq.. w . ....o ..w ...an . .mn ...... .n. . ....1. o. . . . ........... .. . .. M ..... .. o . .u .+. e . uqfioe ov . . . . . . .. . . . . . . ... no. ......... .t .+. Ofi40.e . . . ... e g H . . .H. . .“v . H . . o . . .‘ . . . o. .0. . ... ... qvo Q .+. u1 :EgaNQQQ2NfQW \‘N HWQE‘Q“ “43.0.. ., 5.7.1:” than- :vt.......,..th$.g Whit“? {pl Other Woodland Inhabitants Several species of small mammals were taken too infrequently in woodlands to provide information on their habits. These were the meadow vole, house mouse, bog lemming, and pine vole. Several Species of larger mammals including the cottontail rabbit, opossunu raccoon, skunk, red fox, woodchuck, and long-tailed weasel, also were present in wmfioms numbers. The last mentioned is the only one of the group suspected of having an important direct effect on any of the small mammals. This relationship has been discussed under the section on chipmunks. Acorn weevils Most woodland rodents depend partly or almost entirely on the annual mast crop for food. A consideration of interrelationships of the woodland fauna would hardly be complete without some mention of what appears to be one of the most serious competitors for this food supply. Korstian (1927) has shown a high loss of acorns to insects generally, and in this study, species of Balaninus, the acorn weevil I were found to be especially important in this regard. In the fall of 1941, acorns of several species of oaks were ex- amined to determine the incidence of infestation by larvae of Balaninus. The results are summarized in table 29. It will be noted that from 67 to 81 percent of the acorns of white, jack, and black oak were infested. Only seven percent of the red oak acorns had been entered by this insect, a fact of some interest since Baumgras N v }'\ ‘ (1944) found that red oak acorns were poorly accepted by fox squirrels. Apparently both of these mast feeders find the high tannin content (fiainio and Forbes, 1941) and resultant bitterness of the red oak acorn distasteful. It is doubtful that weevil-infested acorns should be consid- ered entirely a debit in the economy of woodland vertebrates, since many small mammals undoubtedly would utilize the insects (see point 12 in summary of part 1). Gray squirrels are known to relish the larvae of Balaninus (Hatt, 1929), and Petrides (1944) has noted . this same squirrel feeding on the larvae of gall insects. However, Dennis (1930) has written of the rejection of "wormy nuts" by the gray squirrel, and Sollberger (1940) probably had reference to insect-infested nuts when, regarding food storage by the flying squirrel, he remarked that, "Torthless nuts are not buried." 24:3; s 8 8» Alan! 3283 .4: a8 com 8 «so as 937 ..mv “.8 BE: 2. 3» m2." $238.52. ..3 use use 3 3.. 8m “53:: 388.3 “.8 :83 vote: cot £5 3:238 322% £393.” 3.8: .93: go as so 6: Aim. 35.3345 :50: Eco: 2.» hp .93: «o guafioounu .mm ode-a . 24 4: SUIL‘RY ill-ID DISCUSSION 1. Scepe and methods of study. During the years 1940-42, inten- sive box—trapping was employed to determine numbers and pOpulation dynamics of the Sciuridae of farm.woodlots. Studies were conducted principally in seven units totaling 102 acres. Box-trapping during the S-year period amounted to 29,748 trap-days. Among the principal species discussed (red squirrel, flying squirrel, and chipmunk),7 a total of 748 individuals were trapped 2,176 times in woodlots. Other records for animals living away from.woodlands were obtained from 34,032 trap-days of operation in swales, fencerows, and similar hab- itats. Limited work with small live—traps resulted in the capture of 194 mice and shrews (mostly white-footed mice) in 3,210 trap- nights. 2. .Annual population trends of the Sciuridae. Fox squirrels increased markedly in numbers during the 3 years of study, and the 1942 population probably was nearly double that of 1940. Other sciurids decreased: flying squirrels to the greatest extent (by 75 percent or more); red squirrels to a somewhat lesser degree; and chipmunks the least. Red squirrels in fencerows, swales, and other outlying habitats appeared to maintain fairly uniform.population levels in the face of the marked decline among some of the other woodlot dwellers. In the same period there was no evidence among white-footed mice of any marked pOpulation changes. 245 3. Habitat preferences and interrelationships of woodland mammals. Each of the eight woodlands involved in this study was pOpulated by the white-footed mouse and all four species of sciurids mentioned above. The consistency with which these animals were associated would indicate that no marked incompatibility existed. Nevertheless, all were in direct competition for the same foods and, with the exception of the chipmunk, for the same nest sites. The intensity of such com- petition may have been mitigated to some extent by differences in daily and seasonal activity rythyms, and further by species prefer- ences for woodlots of a certain type or condition. (a) Fox squirrel. This animal showed a marked preference for upland oak-hickory woodlots, and it was taken about three times as frequently in this type as in lowland elmemaple. .A sparse ground cover appeared favorable to populations of this species, and in up- lands they were most abundant in grazed woodlots. (b) Red squirrel. In contrast with the fox squirrel, this species reached its greatest numbers in lowland woodlots. In this type, and over various seasons, the numbers trapped were 2 to 8 times greater than in uplands. The response of the red squirrel to effects of grazing was not clearly evident. However, a consistent preference was shown for ungrazed situations. In this respect it differed from the fox squirrel. The red squirrel appeared to be the most adaptable of the arbor- eal Squirrels in the utilization of habitats other than woodlots. Most swales and brushy fencerows were occupied by a few individuals, and others were taken frequently where the only wood cover was one or two isolated trees. 246 (c) hipmunk. with high populations in 1940 and 1941, chipmunks were 9 to 12 times more abundant in uplands than they were in lowland woodlots. With a reduced total pOpulation in 1942, about equal numbers were taken in the two types. Reasons for this apparent differential rate of decline were not clear. However, a relatively high population of weasels was concentrated in upland woodlots and this may offer a partial explanation of the disprOportionate reduc- tion of chipmunks in this situation. During 1940 five weasels in- habited a 23—acre woodlot which at that time was the area most densely populated by chipmunks. Requirements of this species for a dense ground cover probably account for the greater number taken in ungrazed woodlands. The chipmunk catch in this type was 3 to 6 times greater than in grazed units. (d) Flyingksquirrel. No clearbcut preference for a particu- lar kind of woodland habitat was observed among flying Squirrels. The numbers taken in upland and lowland units were comparable, but slightly higher catches were made in grazed woodlots where ground cover was sparse. 2L17’ R (e) White-footed mouse and short-tailed shrew. The white- footed mouse was the commonest mammal of these farm woodlots, where it competed directly with the species mentioned above for food and nesting sites. It was taken more commonly in lowland than in up- land woods, and in the uplands it appeared to be more plentiful in ungrazed units. An October population of 7.6 individuals per acre was found in an isolated lO-acre woodlot of an ungrazed, upland, oak-hickory type. The local distribution of this population was correlated with the density of ground cover. Sod areas were avoided, and low or edge areas, with a dense understory of shrubs but little immediate ground cover, supported the greatest numbers. In common with the red squirrel, individuals of this species frequently in- habited single trees or small units of wood and brush cover which were entirely isolated from.woodlands. In the herbaceous habitats of farmland, short-tailed shrews were most abundant and widespread in areas of bluegrass sod. This same habitat in woodlots supported few of these animals compared to the much higher numbers in situations where a dense canopy and heavy understory had all but eliminated herbaceous plants, and where a thick mulch layer was present. These requirements probably accounted for its scarcity under conditions resulting from.grazing. .A lO-acre woodlot, which by appearances offered optimum.con- ditions for this species, contained a fall population of only 2.5 individuals per acre. This density is well below that reported for similar habitats at other times. 24E LIT ERAT HIE CITED Aldous, Shaler E. 1931. A flybeating mouse. Jour. Mamm. 12: 316. Allen, Durward.L. 1938. Ecological studies on the vertebrate fauna of a SOD-acre farm in Kalamazoo county, Michigan. Ecol. Monogr. 8: 3&7-h36o -—--—l939. 'Winter habits of’Michigan skunks. Jour. Wildlife Mgt. 3: 212-228. ---l9h0. Millions of little teeth. Michigan Conservation. September. ---19hl. Relationships of winter weather to farmland'wildlife in the midwest. Proc. Central Snow Conf. 1: 95-102. ---—l933. Michigan fox squirrel management. Michigan Department of Conservation, Game Division publ. 100, pp. l-hOh. Allen, Durward L., and H. D. McGinley. 19h7. A method for the year- to-year measurement of mast yields. Jour.'Wildlife Mgt. 11: lab-185. . Allen, Durward L., and Warren‘w. Shapton. 1982. An.ecological study of winter dens, with special reference to the eastern skunk. Ecology 23: 59—68. Allen, Elsa G. 1938. The habits and life history of the eastern chipmunk, Tamias striatus lysteri. Bull. New York State Mus. n0. 31h, pp. 1-122. Bailey, Vernon. 1900. Revision of American voles of the genus Iicrotus. U. 8. Dept. Agr., Div. Biol. Survey, No. Amer. Fauna ---—1923. Mammals of the DiStrict of Columbia. Proc. Biol. Soc. ‘Washington 36: 103-138. -—-192h. Breeding, feeding, and other life habits of meadow mice (Microtus). Jour. Agr. Research 28: 523-535. Baker, J. R. 1930. The breeding season in British wild mice. Proc. 2001. Soc. London. (Part 1): 113-126. Baker, J. R., and R. M. Ransom. 19323. Factors affecting the breeding of the field mouse (Microtus agrestis). Part 1. Light. Proc. Royal Soc. London (BS 115: 313-322. -—--1932§. Factors affecting the breeding of the field.mouse (Microtus agrestis). Part 2. Temperature and food. Proc. Royal Soc. London 13) 112: 39-h6. Balch, R. E. 1936. European spruce sawfly outbreak in 1935. Canadian Entomologist 68: 23-31. ---1937. The spruce sawfly outbreak in 1936. Pulp and Paper Mag. of Canada. February. Barrows,‘W. B., and E. A. Schwarz. 1895. The common crow of the 'United States. U. 5. Dept. Agr., Div. Ornith. and Mann. Bull. 68 1‘98. Baumgras, Philip S. 19b23. An unusual clutch of marsh hawk eggs. 'Wilson Bull. 5h: 50. -_.—19M». A technique for confining nestling crows in food-habit studies. ‘Wilson Bull. 5h: 251-252. --—-l9h2c. Population studies and habitat improvement for southern Michigan fox squirrels. Michigan Dept. Cons., Game Div., Unpub. Report :1-112. -—--19h3. ‘Winter food productivity of agricultural land for seed- eating birds and mammals. Jour.‘Wildlife Mgt. 7: 13-18. --—--19hh. EXperimental feeding of captive fox squirrels. Jour. Wildlife Mgb. 8: 296—300. Bent, Arthur C. 1937. Life histories of North American birds of prey, Order Falconifonmes (Part I). U. S. Nat. Mus. Bull. 167: V111 + h09o Bess, Henry A., Stephen H. Spurr, and E. W. Littlefield. l9h7. Forest site conditions and the gypsy moth. Harvard Forest, Bull. 22: 1.56. Blair,'w. Frank. 1938. Ecological relationships of the mammals of the Bird Creek region, northeastern Oklahoma. Amer. Midland Naturalist 20: h73—526. -—-l9h0a. A study of prairie deer-mouse populations in southern Michigan. Amer. Midland.Naturalist 2h: 273-305. —--——l9h0b. Home ranges and populations of the meadow'vole in southern Michigan. Jour. Wildlife Mgt. 14: 119-161. ---l9h02. Home ranges and populations of the jumping mouse. Amer. Midland Naturalist 23: 2hh-250. -—-19h0d. Notes on home ranges and populations of the short-tailed shréi. Ecology 21: 28h—288. Blair,‘W. Frank. l9hl. Techniques for the study of mammal populations. Jour. Mamm. 22: 1h8-157. ----l9h2. Size of home range and notes on the life history of the woodland deer-mouse and eastern chipmunk in northern Michigan. Jour. Mamm. 23: 27-36. ---19h3. Populations of the deermouse and associated small mammals in the mesquite association of southern New Mexico. Univ. Michigan, Contr. Lab. Vert. Biol. 21: l-hO. Bole, B. P. 1939. The quadrat method of studying small mammal pepu- lations. Sci. Publ. Cleveland.Mus. Nat. Hist. 5: 15-77. Breckenridge,‘W. J. 1935. An ecological study of some Minnesota marsh hawks. Condor 37: 268-276. Brown, H. Leo. 19h5. Evidence of winter breeding of Peromyscus. Ecology 26: 308-309. Burt, William H. 19h0. Territorial behavior and populations of some small mammals in southern Michigan. ‘Univ. Michigan.Press, Misc. Publ. Museum 2001. hS: 1-58. ---19h3. Territoriality and home range concepts as applied to mammals. Jour. Mamm. 2h: 3h6-352. 19h6. The mammals of Michigan. The University'of Michigan Press, pp. 1-288. Calhoun, John B. 19h1. Distribution and food habits of small mammals in the vicinity of the Reelfoot Lake Biological Station. Jour. Tennessee Acad. Sci. 6: 177-185, 207-225. ChristianLlJohn C. 1936. Mammals caught in post holes. Jour. Mamm. 173 6. Condrin, J. M. 1936. Observations on the seasonal and.reproductive activities of the eastern chipmunk. JOur. Mamm. 17: 231-23h. Couch, Lee K. 1928. Relationship of predatory mammals and.birds to rodent life. Jour. Mamm. 9: 73-7h. Cowan, Ian MoT. 1936. Nesting habits of the flying squirrel. Jouro mam. 17: 58‘60. Dambech, Charles.A._ 19h0. The effect of land-use adjustments on wildlife populations in the Ohio Valley region. Trans. Fifth No. Amer.‘Wi1dlife Conf. : 331-337. Dambach, Charles A. l9hb. A ten-year ecological study of adjoining grazed and ungrazed woodlands in northeastern Ohio. Ecol. Monogr. 1h: 255-2700 -—--19h5. A study of the insect ecology and economic value of crop field borders in relation to conservation practices. Ohio State University Press. Abstracts of Doctoral Dissertations, no. h8. Dambach, Charles 3., and E. E. Good. 19h0. The effect of certain land use practices on pOpulations of breeding birds in south- 'western Ohio. Jour. Wildlife Mgt. h: 63-76. Dearborn, Ned. 1932. Foods of some predatory fur-bearing animals in Michigan. 'Univ. Michigan School Forestry and Cons. Bull. 1: 1-52. Dennis, Rayne. 1930. Rejection of wormy nuts by squirrels. Jour. Mamm. 11: 195-201. Dice, Lee R. 1921. Notes on the mammals of Interior Alaska. Jour. Hmmm. 2: 20—28. ----l9232. Notes on some mammals of Riley County, Kansas. Jour. Mamm. h: 107-112. ---1923§, Notes on the communities of vertebrates of Riley County, Kansas, with especial reference to amphibians, reptiles, and mammals. Ecology'h: hO-SB. ---l938. Some census methods for mammals. Jour. Wildlife Mgt. 2: 119-130. ---19h1. Methods for estimating populations of mammals. Jour. 'Wildlife Mgt. S: 398-h07o Eadie, W. Robert. 19hh. The short-tailed shrew and field mouse predation. Jour. Mamm. 25: 359-36h. Edgren, Richard A. 19h8. Notes on a northern.short-tailed shrew. Chicago Acad. Sci., Natural History Miscellanea (25): 1-2. Elton, Charles. 19h2. Voles, mice and lemmings. Oxford Univ. Press. pp. 1-h960 Elton, Charles, E. B. Ford, J. R. Baker, and A. D. Gardner. 1931. The health and.parasites of a wild.mouse population. Proc. 2001. Soc. London : 657-721. 22‘ Q)‘ 4‘ Elton, Charles, D.li. S. Davis, and G. M. Findlay. 1935. An epidemic among voles (Microtus agrestis) on the Scottish border in the spring of 193h. Jour. Animal Ecol. h: 277-288. Erickson, Arnold B. 1938. Parasites of some Minnesota Cricetidae and.Zapodidae, and a host catalogue of Helminth parasites of North American mice. .Amer. Midland Naturalist 20: 575-589. Errington, Paul L. 1931. Second winter. American Game, Jul.- Aug. : 56’ 60-61. ---193a§. Food habits of southern‘Wisconsin raptors. Part 1. 0W1s. Condor 3h: 176-186. ----l932§, Technique of raptor food habits study. Condor 3h: 75-86. ---l933. Food habits of southern'Wisconsin raptors. Part 2. Hawks. Condor 35: 19-29. ---1935. Food habits of mid-west foxes. Jour. Mamm. 16: 192—200. ---l936. Food habits of a weasel family. Jour. Mamm. l7: h06-h07. ----l9h2. On the analysis of productivity in populations of higher vertebrates. Jour. Wildlife Mgt. 6: 165-181. ----l9h6. Predation and.vertebrate populations. Qtly. Review Biol. 21: 1hh-177, 221-2&5. Errington, Paul L., and'w. J. Breckenridge. 1936. Food habits of marsh hawks in the glaciated.prairie region of north-central ‘United.states. Amer. Midland Naturalist l7: 831-8h8. Evans, F. C. l9h2. Studies of a small mammal population in Bagley ‘Wood, Berkshire. Jour. Animal Ecol. 11: 182-197. Evans, F. C., and Robert HOldenried. 19h3. Double captures of small rodents in California. Jour. Mamm. 2h: h01. Fenyuk, B. K. 1937. The influence of agriculture on the numbers of mouse-like rodents and the biological foundation of rodent control. Rev. Microbiol., Saratov l6: h78-h92. Findlay, G. M., and A. D. Middleton. 193h. Epidemic disease among voles (Microtus) with.special reference to ToxOplasma. Jour. Animal Ecol. 3: 150-160. Fisher, A. K. 1893. The hawks and owls of the United States in their relation to agriculture. U. S. Dept. Agr., Div. Ornith. and Mann. Bull. 3: 1‘2010 Good, E. E., and Charles A. Dambach. l9h3. Effect of land use practices on breeding bird populations in Ohio. Jour.'Wild1ife Hgto 73 291-2970 Graham, S. A. 1929. The larch sawfly as an indicator of’mouse abundance. Jour. Mamm. 10: 189-1960 Grinell, Joseph. l9lh. .An account of the mammals and birds of the lower Colorado Valley; Univ. California Publ. 2001. 12: 51-29h. Hamilton, William J. Jr. 1930. The food of the Soricidae. Jour. Mamm. 11: 26-39. ---1933. The weasels of New Yerk. Their natural history and economic status. Amer. Midland.Naturalist 1h: 289-3hh. ---1937a. The biology of microtine cycles. Jour. Agr. Res. 5h: 779-790. ""‘1937EP Activity and home range of the field mouse, (Microtus pennsylvanicus pennsylvanicus (0rd). Ecology 18: 255-253. ---l938§. Life history notes on the northern pine mouse. Jour. Mamm. 19: 163-170. ----19382. Small animals as a factor in the control of forest insects. Ecol. Soc. Amer. Bull. 19: 3h. ---l939. Observations on the life history of the red squirrel in New York. Amer. Midland.Naturalist 22: 732-7h5. -.....19uo‘. Life and habits of field mice. Scientific Monthly 50: h25.h3he ----19h1a. Reproduction of the field mouse, Microtus pennsylvanicus (0rd7. Cornell Univ. Agr. Exp. Sta. Memoir 237: 1-23}. ---l9hlb. The food.of small forest mammals in eastern United StatES. Jour. Mamm. 22: 250-263. Hamilton, William J. Jr., and David B. Cook. l9h0. Small mammals and the forest. Jour. Forestry 38: h68-h73. Hanson, Harold C. 1939. Ecology in agriculture. Ecology 20: 111-117. 19hh. Small mammal censuses near Prairie du Sac, Wisconsin. Trans. lisconsin.Acad. Sci.,.Arts and.Letters 36: 105-129. Harper, Francis. 1929. Notes on mammals of the Adirondacks. New Ybrk State Museum Handbook 8: 51-118. Hatfield, D. M. 1938. Studies on rodent populations in a forested area-o Jour. Mamm. 19: 207-211. Hatt, Robert T. 1929. The red squirrel: its life history and habits, ‘with special reference to the Adirondacks of New York and the Harvard Forest. Roosevelt‘Wild Life Annals 2: l-lh6. ---l930. The biology of the voles of New York. Roosevelt‘Wild Life Bull. s= 509-623. HOldenried, Robert. 19h0. A pOpulation study of the long-cared chipmunk (Eutamias quadrimaculatus) in the central Sierra Nevada. Jour. Mamm. 21: 1105:1711. Heeper, Emmet T. l9h2. An effect on the Peromyscus maniculatus Rassenkreis of land utilization in Michigan. JourJEMamm. 23: 193-196. Mosley, N.'W. 1928. Red squirrel damage of coniferous plantations and its relation to changing food habits. Ecology 9: h3-h8. Howell, A. B. 1923. Periodic fluctuations in the numbers of small mammals. Jour. Mamm. h: 1&9. Howell, A. H. 1918. Revision of the American flying squirrels. U. 3. Dept. Agr., Bur. Biol. Survey, No. Amer. Fauna bu: 1-6h. ---1929. Revision of the American chipmunks (Genera Tamias and Eutamias). U. S. Dept. Agr., Bur. Biol. Survey, No. Amer. Fauna 52: 1-157. Jacot, Arthur P. 19h0. The fauna of the soil. Qtly. Rev. Biol. 15: 28-58 0 Jessen, J. Baashuus. 1937. Periodiske vekslinger i smaviltbestanden. Norsk Jaeger— og Fisker-Forenings TidsSkrift 66: 12-15. Johnson, M. S. 1926. Activities and distribution of certain wild mice in relation to biotic communities. Jour. Mamm. 7:2h5-275. Jordon, James S. 19h8. A midsummer study of the southern flying squirrel. Jour. Mamm. 29: 1414-148. Kalabukhov, N. I. 1935. ("On the causes of fluctuations in numbers of mouse-like rodents"). Zool. Zhurnal. 1h: 209-2142. Kalmbach, E. R. 1918. The crow in its relation to man. U. 8. Dept. Agr. 3111].. 6218 1.930 Kendeigh, S. Charles. 19hh. Homing of Peromyscus maniculatus gracilis. Joura Mamm. 25: h05-h07. be 99 Klugh, A. Brooker. 1927. Ecology of the red squirrel. Jour. Mamm. 83 1'32. Korstian, C. F. 1927. Factors controlling germination and early survival in oaks. Yale Univ. School Forestry, Bull. 19: 1-115. Landwer, M. F. 1935. An outside nest of a flying squirrel. Jour. Mamm. 16: 67c Lantz, D. E. 1906. Meadow mice in relation to agriculture and horticulture. U. 3. Dept. Agr. Yearbook 1905, pp. 363-376. ---1907. .An economic study of field mice (Genus Microtus). U. S. Dept. Agr., Biol. Survey Bull. 31: 1-6h. ---1918. Field mice as farm and orchard pests. U. S. Dept. Agr., Farmer's B1111. 670: 1-10. Leedy, Daniel L. 1939. Some land use factors related to pheasant productions. Trans. Fourth No. Amer.‘Wild1ife Conf. :525—533. Lindeborg, R. G. 19h1. Fluctuations in the abundance of small mammals in east-central Illinois, 1936-1939. Ecology 22: 96-99. Linduska, J. P. l9h2a. Winter rodent populations in field-shocked corn. Jour.‘Wildlife Mgt. 6: 353-363. --._—19b22, Cycles of Virginia voles. Jour. Mamm. 23: 210. ---19h2c. A new technique for marking fox squirrels. Jour. WildIife Mgt. 6: 92-9h. --..-19b2d. Insect feeding by the house mouse. Jour. Mamm. 23: 212-213. ---l9h3. Cooper hawk carrying a nest of young goldfinches. Auk 60: S970 ---19h6. Edge effect as it applies to small mammals on southern Michigan farmland. Trans. Eleventh No. Amer. Wildlife Conf. : zoo-zoh. ---—-19h13, 'Winter den studies of the cottontail in southern.Michigan. Ecology 28: hh8-h5h. ---l9h7b. Longevity of some Michigan farm game mammals. Jour. Memo? 28: 126-129. Lyon, Marcus W. 1936. Mammals of Indiana. Amer. Midland.Naturalist 17: 1'38ho McAtee, W} L. 1935. Food habits of common hawks. U. S. Dept. Agr. Circular no. 370. Mccabe, R. A. 19h7. Homing of flying squirrels. Jour. Mamm. 28: hOh. McClelland, E. H. 19h8. Notes on the red squirrel in Pittsburgh. Jour. Mamm. 29: h09-h12. McMurray, Frank B., and Charles C. Sperry. 19h1. Food of feral house cats in Oklahoma, a progress report. Jour. Mamm. 22: 185-190. Mayall, K. M. 1938. The natural resources of King Township, Ontario. Toronto, Canada. June 20, 1938. Mearns, E. A. 1898. Vertebrate fauna of the Hudson Highlands. Bull. Amer. Mus. Nat. Hist. 10: 303-352. Merriam, C. H. 188h. The mammals of the.Adirondack region. Henry Holt and Co., New Yerk, pp. 1-316. Middleton, A. D. 1930a. Cycles in the numbers of British voles (Microtus). Jour: Ecology 18: 156-165. -—--l930b. The ecology of the American grey squirrel (Sciurus caroIinensis Gmelin) in the British Isles. Proc. ZooI. Soc. London (Part 3): 809-8h3. ‘ Montgomery, R. D. l9hl. Population studies and habitat improvement for southern fox squirrels. Unpub. prog. report, Pittman- Robertson project 8-R, Michigan Dept. Conservation. Moss, A. E. 1939. Relation between take of upland game and agri- cultural land use in Connecticut. Jour.'Wildlife Mgt. 3: 269-278. Murie,.Adolph. 1936. Following fox trails. Univ. Michigan, Mus. 2001., Misc. Publ. 32: l-hS. ; Murie, Olaus J. 1927. The Alaska red squirrel providing for‘winter. Jour. Mann. 8: 37-140. Murie, Olaus J., and.Adolph.Murie. 1931. Travels of Peromyscus. Jour. Mamm. 12: 200-209. ----1932. Further notes on travels of Peromyscus. Jour. Mamm. 133 78-790 Nelson, E.'W. 1918. The flying squirrel. National Geographic Mag. 323 h66-h680 2fi36v Nicholson, Arnold J. 19h1. The homes and social habits of the wood-mouse (Peromyscus leuCOpus noveboracensis) in southern Michigan. Amer. Midland Naturalist 25: 196-223. Osgood, F. L. 1935. Fluctuation in small mammal population. Jour. Mm. 163 156. Parker, Lansing A. 19h1. Factors causing rodent damage to tree plantations in southeastern Minnesota. Jour. Wildlife Mgt. 5: 297-3030 Pearson, Oliver P. l9h2. On the cause and nature of a poisonous action produced by the bite of a shrew (Blarina brevicauda). JOuI‘o mam. 23: 159-166. \‘Petrides, George A. l9hh. A gall insect food of the gray squirrel. Jour. Mamm. 25: h10. Piper, S. E. 1909. The Nevada mouse plague of 1907-8. U. S. Dept. Agr. Farmer's B1111. 352: 1.23. Plummer, John T. 18hh. Scraps in natural history (Quadrupeds). Amer. Jour. Sci. Arts h6: 236—2h9. Polderboer, Emmett B., Lee‘W3 Kuhn, and George 0. Hendrickson. 19h1. Winter and spring habits of weasels in central Iowa. Jour. Wildlife Mgt. S: 115-119. Preble, Edward.A. 1908. A biological investigation of the Athabaska- Mackenzie region. ‘U. S. Dept. Agr., Bur. Biol. Survey, No. Amer. Fauna 27: 1-57h. Quimby, Don C. l9hh. A comparison of overwintering populations of small mammals in a northern coniferous forest for two consecutive years. Jour. Mamm. 25: 86-87. Saunders, Aretas A. 1936. Ecology of the birds of Quaker Run Valley, Allegany State Park, New Yerk. N. Y. State Mus. Handbook 16: 1p17h. Scheffer, T. H. l92h. Notes on the breeding of Peromyscus. Jour. Mamm. 5: 258-260. Schooley, J. P. 193h.. A summer breeding season in the eastern chipmunk, Tamias striatus. Jour. Mamm. 15: 19h-l96. Scott, T. G. 19h7. Comparative analysis of red fox.feeding trends on two central Iowa areas. Iowa State College, Agr. Exp. Sta. Res. Bull. 353: h27-h87. Selle, R. M. 1928. Microtus californicus in captivity. Jour. Mamm. 9: 93-98. Seton, E. T. 1909. Life-histories of northern animals. Vol. 1, _ New York. Shull, A. Franklin. 1907. Habits of the short-tailed shrew, Blarina brevicauda (Say). Amer. Naturalist 141: 1195-522. Siegler, Herbert. 1937. Winter rodent damage to game cover. Jour. Mm. 18: 57-61. Smith, Charles C. 191.03.. The effect of overgrazing and erosion upon the biota of the mixed grass prairie of Oklahoma. Ecology 21: 381-397. ----19l:0b. Biotic and physiographic succession on abandoned eroded farmland. Ecol. Monogr. 10: h21-h8h. Sollberger, D. E. 191:0. Notes on the life history of the small eastern flying squirrel. Jour. Mam. 21: 282-293. --—-l9h3. Notes on the breeding habits of the eastern flying squirrel (Glaucomys volans volans). Jour. Mm. 214: 163-173. Southern, H. N.’ and E. H. 00 Laurie. l9h60 The msmouse (Lin-8- musculus) in corn ricks. Jour. Animal Ecol. 15: 13h-1h9. Spencer, D.A. 19111. A small mammal community in the upper Sonoran Desert. Ecology 22: 1:21-1:25. Stickel, Lucille F. 19b6a. Experimental analysis of methods for measuring small-manna populations. Jour. Wildlife Mgt. 10: 350-159. ---l9h6b. The source of animals moving into a depopulated area. Jour: Mamm. 27: 301-307. ---l9h8. The trap line as a measure of small mammal pepulations. Jour. Wildlife Mgt. 12: 153-161. Stuewer, F. W. 191:3. Raccons: their habits and management in MiChiga-no Ecclo Monogre 13: 203-258. Sviridenko, P. A. 1931:. Increase and decrease of mouse-like rodents. Bull. Plant Protection (Russia) Ser. M3): 1-59. Taube, Clarence M. 19147. Food habits of Michigan Opossums. Jour. Wildlife Mgt. 11: 97-103. Quit Telford, Horace S. 19h3. ‘White-footed mice feeding upon grasshopper eggs. Jour. Mamm. 2h: hOO-h01. Test, Frederick H. 1932. Winter activities of the eastern chipmunk. Jouro M31113“. 1.3: 278. Townsend, Myron T. 1935. Studies on some of the small mammals of central New York. Roosevelt Wildlife Annals h(1): 1-120. Van Dersal, William R. 1937. The dependence of soils on animal life. Trans. Second No. Amer.‘Wi1d1ife Conf. : h58-h67. Venables, L. S. V., and P. H. Leslie. 19h2. The rat and mouse popu- lations of corn ricks. Jour. Animal Ecol. 11: hh-68. Weinio,‘Walter'W.,and E. 8. Forbes. 19h1. The chemical composition of forest fruits and nuts from Pennsylvania. Jour. Agr. Research 62: 627-635. Walton, m. A. _1903. A hermit's wild friends. pp. l-30h. Boston. 'Wayson, N. E. 1927. ‘An epizootic among meadow mice in California, caused by the bacillus of mouse septicemia or of swine erysipelas. U. S. Pub. Health Reports h2: 1h89-1h93. Whelan, R. V. 1939. Vole plague at Smoky Falls, Ontario. Canad. Field Naturalist 53: 53-55. Williams, A. B. 1936. Composition and dynamics of a beech-maple climax community. Ecol. MonOgr. 6: 318-h08. 'Wood, Frank E. 1910. Mammals of Champaign County. Bull. Illinois State Lab. Nat. Hist. 8: 501-613. 261 APPENDDC 26:2: .- l 0.00 «.50 0.50 0.00 ¢.~0 0.00 0.05 0.00. 0.00 0.00 «.00 0.00 0.00 0.00 «.00 0.~0 5.00 0.00 0.00 0.05 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.50 0.00 5.50 0.50 0.00 5.00 0.00 0.00 0.0» «.5H 0.50 moan 0.00 5.00 H.00 0.H5 0.05 0.00 0.00 0.~v 0.00 0.00 «.00 0.00 0.00 «.00 0.00 0.50 0.00 0.05 0.00 0.00 «.H0 0.00 0.00 «.0H 0.50 0.00 0.00 0.00 0.00 0.50 «.05 0.50 0.00 0on0 0.00 0.0m «.00 Home 0.00 «.H0 0.00 0.00 0.00 5.H5 H.00 0.00 0.H0 0.00 0.00 0.00 0.00 «.00 0.00 0.00 0.00 0.00 «.H5 0.00 0.00 a.u¢ ~.5« 0.0a 0.0a ”.85 .25 .8: .e8 .emom dealings as... 1%.: ...da ...... 3..— .5... .00 I 000“ epoch one new on¢0dnoux owed-neg poem as nouspnuoaaov noon Hosted 0n¢.hflmvnoa .« capes fining! 000a 000a ¢¢0H ”00H «vow H00" o¢0H i ‘ 1 136 CI. 8 E saws BBB Emu the _.an ....e or: 32 $2 83 «.2 $2 33 ......» .5033: .0533 aid as 3 use not. 3.20 0323 to: no 3a .2: one: an.H» so.” e..~ s..s an.n no.» on.» He.» no.» ee.n o».s on.“ so.” on.»~ no.» so.” o~.~ on.” ee.o u».o n..~ no.» no.6 »..s as.» ee.a can» 3: a.” SJ and H»; 8.~ 2..» a... 3.» cud e»; 3.6 8.9." no; 34 8.6 on.” ens no." SJ o»... 3:... 23 as s»; a..e» o..o oH.~ on.” he.» ee.~ so.» on.» on.» as.~. so.» »e.~ ne.s 8.2.» 8.» 0..» 3.» 2.» o».s 3.. 3.. 3.0 8.6 2... 3.6 as e~.s» »e.H sa.n »».s on.“ oe.~ oe.o es.» on.» ...n so.“ no.6 e».fl 3.1.» to; S.» B.» we.” 56 on.» one no... no.“ so 86 o»... ”884 .25 .6: .s8 .38 one: was .8; am. can ...! 3...» .ea. doused. Gonna tea a. 3.82 ....an .5 no.“ soaeflaaaofia H85: on. beans. .3 .36 750m 000.." 000." $0.." 30H «as “can ova." 4.4;... S H AdHQPOA fiv¢v d 00» v mmN w on» OON OON MN 6609 OON llmolpm «up ov anom o-aoHou I-vuah nH nonaqun 3.» 950m coaOH H I my.” :8... Inn on 995 unavou uncoom 0N NN 0H ON ”N #N ON In» 002 monk ¢ mHn H I mH fl mmN n I NN ¢ 00» o H 0N ¢ NN v 05» ¢ H HH d 0N ¢ an H I wH ¢ mH d Ob» H I HH 0 bH « ow» » » o a. a. can: 0H» ¢nH an mH ¢ ”N 2 at» 2.» n H" o» 3 on ¢ moN H I mN ¢ 0N w mHn m N mH ¢ NH n 00¢ ¢ H ¢N N mN » 00¢ a N DH N NH n and HH 0 b» N . mH N 00¢ N N mH H OON N N NN H nN H CON H I NN H NH anon mmN H I mm H on anon OON H I pH H mm H OON H I NH H mH anon mmN H I on .. H mH 00 I H H . “than may“... NW NM 4mm . 9.36m 4mm . cucoH op vnHom gopao mags » anuoH many » Iom cud-Hp» Ian on . Iom » In?!» 5.. than _. \HI . ooaavnHa . . . dunno» puuHh E32 5» M3983 .. 638 B» 3983 .. 3-5 .5» a .3»: an: .5» » 5»: mu: :5 a 8»: on: .3 » 8»: 3.: .3 a mm»: 3.2 as» » »»»3 «ago» .55 » $8» 8:3 .3 » $8» in :5 m 53» T» .3 a 3:.» »..m .3 » 3H: 8-» .3 » 32.» E32 2» $388 . 3..» .3 » 3:.» 3..» .3 a 83» 3.» .3 » m2: 3..» ..3 » «3: 8...» .3 a 8:.» 3..» .3» o 3:.» IE. hm. L E: won 60H: 0 4 N8 hon—Ema 3. 95:2.» vngao 38 9.903 «833030.. on» on. 35932: .3 one @0530.» A033 68» 3253 3.3.“ no 9369 «8.5 gonnavav usOHuab voboaou onpk_nngk.00Hauoou oHuuaum no nvuooon HusvaanH .bH oHnua IFI‘I. ‘I‘l 20;» 23. op p.509 0» «0H...» fiouu 9.»! oHpup on». a... =33.qu 60932.4 32% 00¢ I N 33% 8N I H “than .93.: von on». a.“ no.5 30.32» « 3» H. H A9250.» .3.qu w on» a N .v mmN N N “v 8N H I pfiom N93. Vm Vm OuaoH 3. p52“ :33 Ian coaoHou Io» 09 In?!» 5 Eng consang— EB?» 3825 mmN OH» mm N On» mmN 8N 00» 93 oHn mwN mbN mm 3 HN mN mugs» ouaOH a». £509 . Iom ouaOHo.» I'VE» a.“ 0083an dflfiflfififl'i'i'd'i' mm»... + ouhuv 3.. 08.3 woumaHo Haven. \m .mfimafiv «o 939 \m 583». £023 5 may». .3...» 8m .. 3 3a 3.3.» 8» .. » o». 3.56.“ ouuoHE aouu :23“?an \m. H I NN w on on!» ...-5 a an»: H I 2 3 3 o»..» .35 » 032 H I mN d NN owl» .EH 9 HanH m N on ¢ r.» cNIHH .EHH a NanH 0 n HN w t» bHIHH J: a mman b « 3 a. «N sHIHH .EH a 5an » » 8 E38 2» 3.988 .. IN N N d .. ”N pHIHH g 0 N55 » n 3 H. NN NHIHH g o 9.an m. n mN d HN bHIHH anH a «was H I «N d ON hHIHH .55 a HHan . I 1mm IINIm lob IIEI .BnH IN LEE NW NNI .oz . 956 .oz vomaoHo.» new“ Mom .3852 £098 998.. . 0» SH man. can mapqum Io.» on . Iom “Ewaoo EH5 . \m Ban .. 3. 593.0% ”.mhflm idiiij occur—nu «8580 €3.20 9:. E2382 22 .b: .32 ates? thongs.“ you .Hono: Iunoooo Mao 0923 o». 28.3.30 I33 vdaHmDI 093. .0930. £033 .buonoOHEom vac MoaHp Sauna. 930.» um no.“ .613 .383 L253 .3003 .0020on hp. «8303. I800 beg—ow: M93 000:. hp .3931 onHuvnomoHHom was Mo 29: hobo can I202 28» 50550 I920. .23 Nsumwm .vongw hHH>aom 63H cHucuHHHmI Hmon .uugw 23H. hpaom , 020398 59300 owdgoovo 0998002 .33 hopOIpfio 3H I 9000.34 oN 9009003 ..txosvoum vcokwov haw vac ..nguosooxogo .303022 93 930.29... 0:50am mo umfisHo 0230 36.352 $362303 khano ivonfimfis ouoouHa .nossoo 3.2332 “3ch can .roxong .mmmH ..fiSMOHn 3 £003: 09» hhopoofifiou Mumnwfim can flame op .820 932" I33 vanmDI 093. HHHpu you 3on .onaa 900.. £80 £8 3.35 $055500 0 no.“ 33%: .3900an 08:. £9.53 no 13860 .hHHEmo: Magoo» 3:8 39%. ES HouHm coaspcam .0HmH 83H owhgoZI HHom «83.33509. Sum Fiona ”,3ch .0380 550393 3.2000: 950 .3053 033 «OH I amazed .H 8040003 yahoo 08:20 3303:: 0:3» HRHHquum 000H 0». 33580 vnosowucfl oquHvooK .«o «8302300 huggm or 03.68 apt... ‘ Hui; . r ‘ 26 ocuuamo uopwna 30. an Em 28H. svwl_vopo>oo vanam: no puoz one “0.0.2.5006 hop». “H ‘ pnmfifim .co. uuuuw aqua mavuoa ho>oo undone onplbuw huavuooou no vnoa nuanhsm homanan unnouw no nocogum vac .QOfip tosvoumon huoxoug can naumumoam .Ounumu knob ovoozwov than» can huuonaocao QEOu saw? nofivOzvogmou nomad spoovomgma vna mamas can on: Eaton undated 22m .couoppuoo nuogpkan can hauopmouu ghhuon Ixoafin .huuopoowbuom and wnwvoon huoxoas Muupwmnm oeuhamm «mauve unomkbu Hanan 0:» nu aqua Iaeou can EH0 :coa Ihoad_cno onca vmom .uooup uowuua wnosd uoaoomu econoasoo gun unapoxoas uucp chaos am you Imus» can vsnmum oHpvuo and moon» and .de0 Mega can hp con-uwAhaabdom Madam .o-uamn hup> .Awpuohv hobo «so .mnma op hogan oooaounsooo .mwma op usedbpum n“ cacao. ammua chuck pm you any can Mao opficz. mean» an havuqa «anunfiecv goo Madam , vacuum haubucm .wfiumoopm 3303: .213th nobobbo hmovuuovpp vamp» Hammonaumv mama o» u:o«>onm anosowcnmz $383 .> .33. cucuuuo sounds can was :30 unadtofi no man?» Hauau a»? boxofléao uquHm: haomunq1 0939 .auoa hvnun onauvaomoHHom anpnanfiautmn :8» .2 .. .382 3. 8.83. vouuuw enamuaa one macaw eumafi con un« gaflo 33:3 .ioxofléa undans haw-02: omha saaoa human unauvacuoaaomt Haom «in. $8.34 .n 8188: lll‘ Ill... '3 ”Q A 2132 .uoppfifi no unflpgfiaasooc pgwflam odoun no $.95 30 can out» 086 ouuopl mo nous ouopoo Hanan woo macaw 0:56 uhovvun no nowvuuomov hash hoboo v:50p¢ chomdnfin undouw and each cflu: .huuonxoafip mo ugasfio cohopvaom .vuoswoum ago» was anyone uodap and acuu 1828 50.302» mo uvnunpnokv vsopu hobo annov hauuam ohaa>aon hfluacu ma wnwvoou Erasmus» .8530 much huposo Moo~£ was nuogprnm .hpgonxoaflp no nonopam cocoa hhob qkuufl Hohokow .ounon huopupovnp ovuououm on: oqoc nachossn Ava: nook» Mao avast ovounum vnovaoov do nonfian d ha»:wan uovqwusbu ohpacnooOu hhoxoun guousuw huuboo: psnwam can Mao hauaum avoot_uo Madam .vnunuaov unanvuona 92094 #8 3.23 .0933 .2533 .35 £6 .uovuo .oan was» a“ ucfigusooo uonH noozpop Mao ovwgtccno nymph 5H .huoxOH: apanwun. you hapgwafl .350 Mosflm .vnund hunk connuc taco auodeg asnwfim .onH vac . 93300» a oat ovox hoe—13. ow." 3 .Imqap. H¢QWonaum mam” op .:o«>oum anosowcnnz 3.203 .> .35. .uoaapmuu .huoxofls uxuo cauanpu oaks ofidoa human onwcpnom noHHom hfivuoa1 ado» .NH u omaouo4 .o aqqaooz. ivouuuwnp .huoxoua .58 2:37 .93 efiuoa human ouuuvnouoanom and Edna oggoz... om .»N u .mapua¢..m Bounces 26$? .coavavcoo vouauwn: :4 venououmma naawu oban on on voduom a nose Lem wqwnagw scum copoopoum noon opus nogg3.u»aq= op eunuch ccouauwnba noavaowmwunaao unoaoum use .oaav can. an magnupw o» wovoonpsn not: opa: o» nabs» ouck_uvoacooh.aaa b nonvooa.uo uoapmeoxo esp snag: o-owoomm Pamphom2« uospo onoaphom Loam“: cud hguoso xoaan was :« uopvppaom use «cockwou hahw .hpuon ovazg.oflauavnoam .ouuaau cuppa vow Ixoaap .mpponmuam kahuna vague chum ovenaum hflvuooon opou. mo oozoaowmmcoo a“ gag» opnu:«eov.aao cued voz oH¢mH wcavvso woos hobo pound Ioswoumou nance cur: ahead can onmde com tbwpooaou wovaeaq uovwdfl coca vac com .ounov hfiuwcm ownaxoovu ovapcvo: .Amucohv hobo v50 ounausu pcavuomsw aha huponoowbpou qaao can .Hona: gov“: nuoauoE<.nnc .nomuu «hpuopxoaap .vanOuun apoopowpafl .hho anopap gumzb .nooswov lac“: psnmam .axao hahw can hoauo vom #omfip was do; .opfig: .pnaswoum oua omauw haaamaonwpm ooaoomu vaF mo noawaav canon no poafiflq oflnwuo ohflwbao: a“ mnfiooon Ivaoqoo a nonSHonw ohm napuamncm can hovcfiufiom gunman mo opccmoum mocha coo hpuono soaflm .gvkouw accouom om hauao: .vouagw noon oz ovoH ouapno specs: :« oowoomm oamwo nanomoumou can van: o>c£ ow nsonx hobo hovvfin mo aawpm no one nu onmE Iflaov aw cause wow #02 .Awumomv godpofisEdoou a>bom com oouaov.huo>.owcfixoopn abaon_hngw¢WThHucoopu hobo v50 goboo unsoua ‘Nmopuuounp ozuvn fiamwonuum mama op ozowbokm anoaowanna 31:88 .> 33.9 _..voufiwnp ceamas won an: _an .vncakbqt oaks cauoa uo>onoo can oHaooHHHMI “How .2: $8.84 .2 Sacco; .vouauwnp nuance v0» una_BHo havcandeoc Iopm .vnuflzogl oaks .udouu gonna: do aaoa oaaonflafim .:0fivpom ucuakoa 5 flog .22. Sam .3 u owufiod .5 8,883 . Ilfl. Ill} .11 0 fid7is ...—8.20m 32$ 5 @233» no.8: 3 hag o o H o p.» «.H N a gonzopoz m.m ¢.¢ ¢ N o.¢m 0.5 «a o .poou.pmom m.m ».a « » «.ao a.» a” k hfisauoasn p.e ».> n n ».om o.» am e ”dumq m.bH o.m~ m H“ n.»» H.«« a" a” huuanau .~¢ma ~.- H.5H o" » m.non «.ma 5» a" .poou.»mun p.o« n.m~ «a «a a.poa a.pu mu «m hflsanoquu ~.~¢ ».a« a” «H m.oo ”.mn am a» .gm4r.g.a .flvau o.o¢ o.¢~ ma 0 m.op m.b~ om ¢~ nongoooa «.mm ”.mm ow «H m.~o o.om {ed «a .pooa.»mom m.§m «.no ¢¢ om m.umu «.Hm com o» .wa¢rhn: p.o~ ”.5 «H » ».Ho «.pa . .ma ma .unaw.puu .owmn 03¢."qu @395 Age»: 3v T93: 3v 65. Nm “nosed 03 Anon?» 03 can: con Au Waning _.an .935 8a you £5 madam-l .ausuapflvuu mmmaao .flaauapfiuna .Hanefi>acnm a; so ufiascfipaunn 3003003 233: .639"? no.3». and 639% 0923. 5 end .3333 undated 25 van 2335 Ha- n... gaggle»: .3 no? 38.353 com Jb 039a 1." f. II. Ilsl HI 2 .ouoiom 333 an @2530 canon on has _.. 3b a... on ... can 5.»... 3 3 £8.25.» ...3 NJ“ 3 u c6» «.3 a on 3.7.93 .32 «£2 33 no 3 5. a...» m a. 59?.5» .22 as: 33 «a «H a.» «.2. _.H 3 $8.3m... «$3 9.2.. 8 S 0.2 n62 .a 3 63.»... .32 vaufiunp conga “no.3: 3v .793: d3 “HE vammm Anon: 2.3 Noun: 03 .931 cc“ a” oulwnb woufiw ufiou 00.... a mus-mg Edam :83ch 39.8 .73?8& 5.5?“an 2 8 £33.35 .3030?! no wfiuflw and 2K» 9... gaging a.“ 8325.30 no :38 .3» 030a .38.“qu 2.05 5 comm-.3 no.3. on has ... a.» «.N N H . 3. N: N « 5.?an o.o o6 o o o6 N.H o H 38.3%.» 06 NJ 0 N o6 N.N o N .HHaaaoEH. N.N ....N H H 3: ».N _ N» N Hrs? N.N ¢.N H H N.» N.N H N $3.» .NaH 3. 0.0 o o 3. o6 o o £8.33» N; N; N « u.HH ».N ... N #5723. 3. N.NH N N 06 H6 o H. ..afai .HEH N6 ...; c N N.» H; H N. 2.385 N.NH N.Nm N «N o.NN N.Nn .o N» .aoou.pmom fine t3 NN 3 NJ» QB 1: 3 .miLEH N.» 0.0 « o a.» NJ ._H N .3753 .32 coutum Nana Anon: 3V Ancho- gv v5 HH a 70.3.. m3 A273. o3 no.3. con 3:95 boa-.5 :9: con .Hu :1." an: .HSNHpHHiH fined .HSufiHHEH .HSNHpHBH 2 8 .HSNHpHHaH .3038! no gut» v5 093. on 5.3.30» an 30.933- 33: no #330 9:: 0.35 Assioncd qur’in‘y “rm.- .ng _Abbls‘uleLl 1!. man "f”:fi‘flz’z HOE $193-»: . ”C! 15 “[3 I I . Jl " 2' "I SLIM Ls . .& R:- xx a. a an" x" f ‘ I g 'r 1; _.f'z Y ‘ w'