RACCOON. MOVEMENTS IN A SOUTHERN. MICHIGAN AGRICULTURAL UPLANDI Thesis for the Degree of M. S. MICHIGAN STATE UNIVERSITY KENNETH LEE STROMBORG 1970; BINEING av " H0“ & SUNS' 300K BIND!“ INC. LIBRARY BINDERS IPIIIEPOIT. HIGH“! . ‘ 3 .fl ABSTRACT RACCOON MOVEMENTS IN A SOUTHERN MICHIGAN AGRICULTURAL UPLAND BY Kenneth Lee Stromborg This study was undertaken to evaluate raccoon movements and cover type use in an agricultural environ- ment containing small hardwood woodlots. Live-trapping was carried on for 16 continuous months. Seventeen per cent of the retrapped raccoons were found to have shifted their centers of activity be— tween woodlots. The number of raccoons captured per unit of effort was highest during the breeding season and dur— ing the summer and fall when the animals were presumably most active. A low-cost Citizen's Band telemetric transmitter was developed to allow a detailed study of raccoon move- ments. Using these transmitters and a portable hand-held receiving system, 4 animals were tracked for a total of 21 nights in the late summer and fall. Movements of these animals were found to be regular and predictable. Periods Kenneth Lee Stromborg of relatively rapid movement were found to be interspersed with periods when the animals remained in a localized area. Woodlots and areas containing corn, either as a crop or as animal feed, were found to be most frequently used. The major, although infrequent, use of pastures and grassy areas was for travel lanes. Little evidence was found of raccoons utilizing permanent standing water. Raccoons moved from 1,000 to 7,000 feet per night (mean, 3,194) at rates from 160 to 1,320 feet per hour (mean, 733). The home ranges of the four animals, based on radio-locations, ranged from 19.8 to 83.2 acres. I conclude that trapping is an efficient tech- nique for evaluating inter-woodlot movements and seasonal activity. Cover type use can be evaluated by telemetric monitoring. Some portions of home ranges were unused and I conclude that home range descriptions should take this into account. RACCOON MOVEMENTS IN A SOUTHERN MICHIGAN AGRICULTURAL UPLAND BY Kenneth Lee Stromborg A THESIS Submitted to Michigan State University in partial fulfillment of the requirements MASTERS OF SCIENCE Department of Fisheries and Wildlife 1970 ACKNOWLEDGEMENTS I wish to thank Dr. Leslie Gysel for the guidance and aid he gave me during this study and for the equip- ment he made available. My thanks also go to Dr. George Petrides and Dr. Rollin Baker for their suggestions and for editing the. manuscript. Special thanks go to Mr. Richard Thomas for his assistance in the development of the telemetric system and to Mr. David Purol for his help in gathering tele- metric and trapping data. Finally, to my wife Janet, I extend heartfelt gratitude for her constant understanding and encouragement. ii TABLE OF CONTENTS Page INTRODUCTION . . . . . . . . . . . . . 1 STUDY AREA . . . . . . . . . . . . . 3 Location . . . . . . . . . . . . . 3 Physiography . . . . . . . . . . . . 3 Vegetation . . . . . . . . . . . . 6 METHODS O O O O O O O O O O O O O 0 lo Trapping . . . . . . . . . . . . . lO Telemetry . . . . . . . . . . . . 11 RESULTS AND DISCUSSION . . . . . . . . . 21 Movements Revealed By Trapping . . . . . . 21 Telemetric Findings . . . . . . . . . 26 SUMMARY AND CONCLUSIONS . . . . . . . . . 48 LITERATURE CITED . . . . . . . . . . . 51 iii LIST OF TABLES Table Page 1. Raccoons That Were Captured in or Near More Than One Woodlot . . . . . . . . . 22 2. Animals Caught at Least Three Times Inside Hudson Woodlot and Once Outside It . . . 24 3. Average Duration and Nightly Frequency of Periods of Movement and Localized Activity . . . . . . . . . . . 33 4. Summarized Nightly Use of Cover Types by Raccoons . . . . . . . . . . . 36 5. Distances and Rates of Movement . . . . . 39 6. Number of Nights With a Given Average Rate of Movement . . . . . . . . . . 41 iv LIST OF FIGURES Figure Page 1. The General Study Area on the University Farms 0 O O O O O O O O O O O O 5 2. Hudson Woodlot and the Surrounding Agricul- tural Areas . . . . . . . . . . . 9 3. Circuit Diagram of the Transmitter . . . . l4 4. Circuit Diagram of the Loop Antenna . . . . l4 5. A Field Sheet Showing the Movements of Raccoon 1 on the Night of August 8, 1969 O O O O O O O O O O O O O 20 6. The Number of Raccoons Captured per Trap- Night by Month . . . . . . . . . . 27 7. The Movements of Raccoon 4 on the Night of October 24, 1969 . . . . . . . . 31 8. Home Ranges of the Four Radio-Tracked Animals Based on Radio-Locations . . . . 44 INTRODUCTION Knowledge of the movements of an animal species pro- vides perhaps the best basis for inferences regarding its ecology aside from direct observation. With the advent of radio-location telemetry in the early 1960's, detailed study of many species' movements was made possible. Nocturnal species are particularly difficult to study by visual means. The raccoon (Procyon lotor) is one such species which has been studied mainly by indirect methods. Prior to the use of telemetric devices, its movements were determined primarily from tracks (Stuewer, 1943 and Stains, 1956). Berner (1965) used an automatic event recorder set along raccoon runways to evaluate movements between cover types. These methods show the presence of animals, however, they fail to indicate the time spent in various cover types or the movements within types. Telemetric studies offer this potential, but have been largely limited to studies of resting sites (Mech, Tester, and Warner, 1966) and the size, shape, and shifts of home range (Ellis, 1964 and Tester and Siniff, 1965). One study which dealt with use of cover types (Turkowski and Mech, 1968) was limited to one animal from February to August. This study was carried on from July, 1968 until No- vember, 1969 with the objectives of determining raccoon movements and their relationship to habitat and home range. To meet these objectives, it became clear that a low-cost telemetric system would have to be developed and used in conjunction with live-trapping. STUDY AREA Location The study area, on Michigan State University property, included all or parts of sections 30 and 31, le, T4N and sections 25 and 36, R2W, T4N, Ingham County, Michigan. Physiography Since one objective of this study was to evaluate raccoon movements, precise restrictions were not placed on the size of the area. However, Berner's (1965) limits were generally followed giving an area of 2.5 square miles with a width of 1.7 miles and a length of 1.5 miles (Fig. l). The primary use of this study area is intensive agricultural research. The soil type of the fields north of Bennett Road is predominantly well-drained Hillsdale fine sandy loam of moderate fertility. South of Bennett Road, various well- drained loams are interspersed with small pockets of imper- fectly-drained loam soils. The predominant soil type in the woodlots is Hillsdale sandy loam. Grazed Woodlot, how- ever, is in an area of Miami loam and both Hudson and Tou- mey Woodlots contain extensive areas of Spinks loamy fine sand. All of these areas are well-drained and low to Fig. 1.--The general study area on the University farms. a, Maple Woodlot; b, Bee Woods; c, Grazed Woodlot; d, Hudson Woodlot; e, Toumey Woodlot; f, Minnis Woodlot. moderate in fertility. The extreme southern portion of Hudson Woodlot contains an area of imperfectly-drained Locke sandy loam. Vegetation The area of most intense study is centered around Hudson Woodlot. This woodlot is 19 acres in extent and consists of two almost equal-sized communities. The nor— thern half is a mature hardwood stand of predominantly sugar maple (Acer saccharum) and beech (Fagus grandifolia) with some black cherry (Prunus serotina) and basswood (Tilia americana) irregularly interspersed. The dominants are roughly 14 to 25 inches in diameter and form a dense canopy with few openings. The southern half consists of an even— aged stand which developed after clear-cutting about 35 years ago. The principal species are sugar maple, beech, and white ash (Fraxinus americana) in the 6 to 8 inch dia- meter size class. This stand is closed and contains only a few scattered larger trees left during the cutting. Herbaceous and shrubby foods available to raccoons within this woodlot were quite diffuse. It was noted, however, that clumps of red-berried elder (Sambucus pubens) bore fruit early in the summer, and this fruit disappeared quickly upon ripening. Also, gooseberry (Ribes cynosbati) was scattered throughout the woodlot. May apples (Podo- phyllum peltatum) were quite abundant early in the summer, but the fruit disappeared within a short time after ripening. On the northern border of the woods, near the center, a large clump of pokeweed (Phytolacca decandra) bore fruit, and these berries did not disappear until near the end of the winter, and then only slowly. Surrounding Hudson Woodlot are agricultural lands (Fig. 2). To the north, an alfalfa field was cropped both years. Approximately 200 feet north of the woods, in the middle of this extensive hay field, and also to the west of the woodlot, corn was planted both years. Cattle pas- tures occupied the land to the east and south of the wood- lot. Barns and feeding troughs were southeast and pig pens northwest of the woodlot. Beyond the adjacent corn fields on the west were goat pastures. On the north of the alfalfa field, corn and soy- beans were planted. On the far northeast, wheat was grown, and southeast of this planting, there was a nursery which provided ornamental plants for the campus and contained some fruit-producing trees, particularly crabapples (M3133 sp.). Beyond the area described were scattered woodlots, pastures, and fields (Fig. 1). Stock tanks at the cattle barns, pig pens, and in the pastures directly east of the woods, and a small 0.7 acre permanent pond in the pasture east of Beaumont Road were the only permanent sources of water near the woodlot. Tem— porary vernal pools in the woods and fields were the only other sources of water near the woodlot. Fig. 2.-—Hudson Woodlot and the surrounding agricultural areas. ILL ]L_ W) P STUIE PIG GIDLEY woomor 1 OYIEANS ranss ALFAlFA 'HUDSON woootor conu PASTUIE I ALFALFA stfiflb ROAD 1 ' k gm D PASTUIE CATTlE LL NURSE'Y AT nuns I I J I I'll [It \\ nonunv wooos fl ——> I 400 fee! I METHODS Tra in Live-trapping and tagging of animals were carried out from July, 1968 through October, 1969. The traps used were National Wire Company live-traps and live-traps of the same design and dimensions as those used by Stuewer (1943). In Hudson Woodlot, traps were initially set on a grid of four lines, seven traps in length. The distance between traps was paced off and marked in a uniform pattern 170 feet in each direction. During August, 1968, 16 traps were removed from Hudson Woodlot when recaptures indicated that the majority of the raccoons there had been captured at least once. This removal left three lines of four traps each with 340 feet between each trap. The remainder of the larger study area was also trapped. During the summer of 1968, four traps were set in each of the six outlying woodlots on the area. One trap was placed as near as possible to the center of each of the four borders of each woodlot to get some indications of movements between woodlots. As telemetric data later verified, raccoons used the borders of woodlots for travel, 10 11 and hence, placing traps on the borders seemed to be a more efficient trapping strategy than placing traps in the interior of woodlots. The number of traps in outlying woodlots varied from a high of 21 during the first 3 months of the study to a low of 4 at the end. Traps were also set along the borders of the fields north and west of Hudson Woodlot when the corn was ripe. During the summer of 1968, additional traps were set in the pastures surrounding Hudson Woodlot. Traps were set a minimum of two nights per week during the 16 months of the study. This procedure resulted in a total of 4,133 trap nights. When captured, an animal was weighed, sexed, and aged as to adult or young of the year. Then it was tagged in each ear for later identification and released. A chicken wire cone as described by Stuewer (1943) was used for handling animals. Telemetry The development of a suitable telemetric system con- stituted a major portion of this study. The critical com- ponent of this system was the transmitter. When this study was begun, I felt that with slight modification, a design which Dr. Rollin Sparrowe (pers. comm.) had attempted to use with hawks would serve my purposes even though it had not proved adequate for his. The history of this design was obscure although it appeared to have been evolved from 12 circuits developed by Southern (1963). With technical ad- vice from Mr. Richard Thomas of the university radio sta— tion, WKAR, the circuit shown in Fig. 3 was developed. The components shown were assembled and sealed in silicon rubber (Dow Corning Silastic), then coated with cold-cure dental acrylic (Perm by Hygenic Dental Manufac- turing Company). It was possible to reopen the component package by dissolving the potting materials in chloroform. This permitted repair or tuning of the circuit. During the summer of 1968, four transmitters were built and attached to different raccoons. Of the four transmitters tested, only the first operated for more than two days. All were subsequently recovered and it was determined that they were neither waterproof nor resistant to damage by raccoons. On the basis of this evidence, I decided that the initial design was inadequate in several respects. First, although the use of silicon rubber permitted repair of a faulty transmitter, it did not seem to form a waterproof seal around the copper antenna wire. Second, the antenna seemed to be subject to considerable movement in the in- terior of the component package, and was also easily de- formed by the animal. This deformation changed its elec- trical properties and resulted in a loss of the emitted signal. To overcome these deficiencies, I decided to sacri- fice the repair capability inherent in the original design. 13 Fig. 3.--Circuit diagram of the transmitter. T, transistor (ZN-708); R1, resistor (2,2009); R2, resistor (180 KQ); B, battery (Eveready E133); X, crystal (channel 8); C1, capacitor (300 pF); C2, capacitor (5 mF); L, antenna (25 cm aluminum strip). Fig. 4--Circuit diagram of the loop antenna. C, capacitor (7-45 pF); L, antenna loop (2 10" dia- meter turns of 3/8" c0pper tubing). 15 Instead of using silicon rubber as a potting compound, epoxy resin was used on all transmitters built during 1969. There were several advantages to using this compound: epoxy completely immobilized and waterproofed the compo- nents of the package and was easy to shape using plaster of paris molds. Molding the epoxy resulted in a much smaller package than was possible with the silicon rubber. Since the epoxy precluded Opening the component pack- age after potting, it was necessary to have some adjustment in antenna length to fine tune the transmitter to the fre- quency in use. I also considered it important to be able to remove the transmitter easily in the field. To meet these conditions, the copper antenna was replaced by a thin one inch wide strip of aluminum. Two short pieces of aluminum were potted in the epoxy to serve as leads for the connection of another strip to complete and tune the circuit. This strip was attached with two bolts on each end after determining the optimum length to use. As Ellis (1964) pointed out, epoxy is not resistant to damage by raccoons, so the completed component package and battery pack were coated with dental acrylic. The battery was coated only with dental acrylic, so when it lost power, I was able to expose and change it by dissolv- ing the acrylic in chloroform. This design was found to be waterproof and resistant to damage by the animals. It was also extremely easy to 16 attach to an anesthetized animal using only the antenna as a collar. In the field, the transmitter could be removed by cutting the replacable strip of aluminum. One trans- mitter was recovered and reused after changing the battery. This transmitter functioned well despite having been par~ tially opened to replace the battery and then resealed. I did not feel that the transmitter caused any significant abnormalities in the animals' movements. The receiver used initially was a Lafayette model HA450. It was a crystal controlled Citizen's Band unit. The only modification of this receiver was the inclusion of a beat frequency oscillator which imparted a tone to the signal making it easier to distinguish weak trans- mitter signals from static. This receiver was stolen on September 19, 1969, necessitating the use of another unit. The only alterna- tive receiver readily available was a Heathkit Mohican. This receiver was inferior to the first in several respects. First, it covered a wide range of frequencies. This ne- cessitated the inclusion of more parts and consequently was a bulkier and heavier unit. Second, although portable, it was not designed for hand-held operation and had to be set down for operation. Its biggest disadvantage was that it was not crystal controlled. Tuning was not precise enough to allow setting a frequency with assurance that it was close enough to the transmitter's frequency to 17 detect a signal. This caused considerable wasted effort in scanning the band trying to pick up a signal from a transmitter which might or might not have been in range. These disadvantages severely hampered locating the animals at the beginning of tracking periods. However, it was quite easy to maintain contact once the animal had been located and the operating frequency of the receiver fixed. The antenna used for location was built from the cir- cuit design shown in Fig. 4. It was a portable hand-held directional loop with an accuracy of approximately five degrees when line of sight fixes were used. The original antenna was lost with the first receiver and I built an- other to the same specifications which seemed to function as well as the original. The tracking system just described was adequate for this study. Its range was between one-quarter and one-ha1f mile depending on weather conditions, intervening cover types, other transmissions on the frequency in use, and transmitter power. The Lafayette receiver had slightly better range, but the Heathkit allowed better tuning to eliminate extraneous noise. In the final analysis, I felt that the two systems had comparable useful ranges. The biggest advantage of using the crystal controlled Lafayette unit was that, unlike the Heathkit, it allowed me to con- clude with some degree of assurance that the animal was not within range if no signal was received. 18 Animals were usually tracked for two nights per week, although at times it was possible to include more tracking nights in a week. On several occasions, rain precluded use of the equipment so a total of 21 tracking nights was ob- tained over the 12 weeks of the telemetric study. The animals were usually followed from shortly after dark until it was felt that they had become inactive for the day. The average length of nightly contact after the initial loca- tion, excluding three nights on which rain or equipment failure cut short tracking, was 6 hours 45 minutes. Field records were kept on mimeographed maps of the area. A sample field sheet with bearings and fixes from one night is shown in Fig. 5. Fixes were taken approxi- mately 15 minutes apart from convenient landmarks. When the animal was on the boundary of cover types, a fix coin— cident with the boundary was taken to determine which type it was in. This feature of a portable system was a major advantage over a large fixed antenna system in this study because locations with respect to environmental variables were of more interest than geometric locations. From the field sheets, the locations and times were transferred to clean maps of the same type used for field records. This permitted easier evaluation of results since extraneous marks and lines were not present, and the times were recorded in a legible ordered manner. 19 Fig.5--A field sheet showing the move- ments of raccoon 1 on the night of August 8, 1969. k IIIIII II RESULTS AND DISCUSSION Movements Revealed ByiTrapping The trapping program was intended to measure large— scale movements by the raccoon population. Of 36 animals caught more than once during the 16 months of trapping, 9 were caught in or near more than one woodlot (Table 1). Of these nine, three were caught in traps set in the fields around Hudson Woodlot. During the telemetric studies, animals were found to approach woodlots other than the one in which they resided, although there were no instances in which they actually entered a second woodlot. Therefore, the capture of an animal in more than one woodlot probably indicated a shift in its center of activity from one woodlot to another, while a capture near, but not within a second woodlot would not indicate such a shift. By this criterion, ani- mals caught in Hudson WOodlot and later in adjacent fields or by the pond were not considered as having shifted cen- ters of activity. Likewise, animals caught in fields ad- jacent to Hudson Woodlot and later caught in another wood- lot were discounted as not actually having entered Hudson Woodlot. 21 22 TABLE l.--Raccoons that were captured in or near more than one woodlot. Location of Successive Captures Ta Numbgr Original Second Third Fourth Fifth 544 Hudson Bee Woods 513 SW Hudson* Grazed Grazed Grazed 503 Hudson Grazed 518 Hudson ' Grazed Grazed 512 Hudson Maple 534 NW Hudson Minnis Minnis ll Grazed SW Hudson 85 Grazed Maple 170 Toumey Hudson Hudson Hudson NE Hudson *Abbreviated directions signify direction from the center of the woodlot to traps located in adjacent fields. 23 Using these criteria, 3 animals were deleted from the 9 which might have shifted their centers of activity from one woodlot to another during the course of this study, leaving 6 out of 36 (17 per cent) which might have shifted activity centers. Of these six, four moved from Hudson to other woodlots in the general study area, one moved from Toumey to Hudson, and one moved from Grazed to Maple. The net outward movement from Hudson might indicate that more animals were produced there than the woodlot could support. The number of traps set and the frequency of recap- ture were not sufficient to delineate home ranges of trap- ped animals. The movements of the two animals which were most frequently caught, numbers 514 and 559, could be in- ferred, within the boundaries of the woodlot, to be mostly in the eastern half and the western half, respectively (Table 2). The recaptures of these animals outside the woodlot also followed this pattern with number 514 caught east of Hudson at the pond and number 559 west of Hudson in the corn. The eight captures of number 514 were con- fined to July and August of 1968 while number 559 was captured six times from July, 1968 to February, 1969, and finally in July, 1969. The recapture records of other animals were more difficult to evaluate. Of the six other animals which were caught at least three times within Hud- son Woodlot and once outside of it, all captures were made from late June until early October of 1968 or 1969. In addition, all except number 119 followed a pattern in 24 TABLE 2.--Animals caught at least three times inside Hud- son Woodlot and once outside it during the course of the study. Location of Successive Captures Tag Number 1 2 3 4 5 6 7 8 514 531/4* NE1/4 NW1/4 531/4 NE1/4 NE1/4 E** E 521 sw1/4 551/4 sw sw1/4 559 NW1/4 w sw1/4 NWl/4 NW1/4 sw1/4 sw1/4 119 NEl/4 NWl/4 nw1/4 sw 114 sw1/4 nw1/4 sw1/4 NW 136 sw1/4 531/4 sw sw1/4 142 sw1/4 531/4 w 551/4 132 NEl/4 NWl/4 NEl/4 NW1/4 NW *Directions followed by 1/4 indicate quarters of the woodlot. **Abbreviated directions signify directions from the center of Hudson Woodlot to traps located in adjacent fields. 25 which their captures were confined to one-half of the woodlot and the fields adjacent to that half. Number 119 was exceptional in that it was caught in the northern half of the woodlot, but in the southern half of the cornfield west of Hudson Woodlot when recaptured outside of the wood- lot. During the telemetric portion of the study, raccoons were found to use the borders of Hudson Woodlot more than the interior. To ascertain whether this use was reflected by the trapping data, captures of raccoons from August 1, 1968 to July 30, 1969 were analyzed. During this one year period, 12 traps were utilized with 10 set in peripheral locations and 2 in central locations. Assuming each trap to be equally efficient, the number of captures in a trap should have reflected the relative use of that trap's seg- ment of the woodlot. Therefore, if the interior of the woodlot was used proportionally as frequently as the bor- ders, the ratio of captures in the central traps to those in the peripheral traps should have been 1:5. Of the 32 captures in Hudson Woodlot, 2 were in central traps and 30 in peripheral traps: a ratio of 1:15. A chi-square test gave a value of 2.497 with one degree of freedom. This value had a probability of being exceeded by chance of slightly more than 10%. Although the results were not significant at the 5% level, the small sample size would require a ratio of at least 1:31 to attain significance 26 and I feel that there is a strong possibility that a larger number of captures would have showed a significant differ— ence. From these results, it is apparent that both fre- quency of capture and the limited distribution of traps mitigate against obtaining an accurate evaluation of the movement patterns of raccoons. If capture in more than one woodlot is truly indicative of a shift in the center of an animal's activity between woodlots, then this gross aspect of raccoon movements can be determined by trapping. Raccoon activity by season might also be indicated by trapping data. Fig. 6 shows the number of raccoons caught per trap-night by months. The rates of capture in January, March, April, October, November, and December were roughly the same while the capture rates were con- siderably higher during the summer months of May through September. The sharp peak in February corresponded roughly to the breeding season of raccoons in this area (Stuewer, 1943) when increased activity would be expected. Telemetric Findings The most intensive aspect of this study was the tel- emetric monitoring of raccoon movements. One behavioral characteristic of the animals followed was the regularity of their nightly pattern of movements. While following animal number 1 before detailed data collection began, I was impressed by the extremely small variation in the 27 .IO‘L .094- .08-» .07dr pamHCImmuu .06-- .054- .04«» .o3«- .osz Mom omusumwo mcoooomm .01 L L L l .omo r.>oz rfioo u.ummm . .msm Twash Teach jam: 1HHHQ¢ rcoumz 1.3mm r.c0b Months Fig.6--The number of raccoons captured per trap- night by month. 28 animal's itinerary during the several hours after sunset. When I began to collect data for longer periods, this pattern was found to continue throughout the night. With two exceptions, animal number 1 followed the same general pattern, moving along the eastern and southern borders of Hudson Woodlot, into the cornfield on the west side, and finally returning along the northern or southern and eas- tern edges of the woods (Fig. 5). The exceptions were two nights when this animal moved east to the nursery and pas- ture east of Beaumont Road. While not as many data were collected on the other three animals, they also followed a somewhat regular pat- tern in their movements. As a result of this regularity, I was able to begin tracking several hours after dark merely by using prior knowledge of the animal's habits for the initial location. Animal number 2 exhibited some variability over the three nights it was followed, but generally, its main ac- tivity centered on the pasture east of Beaumont Road. It visited this pasture on all three nights. On one of the nights, it was initially located in the southern half of the cornfield west of Hudson rather than the pasture. From the cornfield, it moved along the southern border of the woodlot, stopped for one hour in the brushy area on the very edge of the woodlot, then moved relatively di- rectly to the pasture east of Beaumont Road. 29 Animal number 3 usually moved counterclockwise around the pig pens immediately adjacent to Gidley Woodlot, in which it denned, and then either returned to the vicinity of the den for all or part of the remainder of the night or moved southeast into the adjacent cornfield. On two nights, the animal moved only through the pig pen area, not into adjacent fields. The movements of animal number 4 were less regular and more difficult to follow than the others. A large part of the difficulty was due to the animal's remaining in the woods and the problem of my negotiating passage with fragile, bulky equipment through the heavy underbrush after dark. On three nights, this animal moved north along the eastern border of Hudson WOodlot. On one of those nights, the animal moved further north and into the corn- field just north of the woodlot. On another night, the same pattern was probably followed although contact was lost and the cornfield position not verified. On the third night, it remained stationary and presumably denned along the eastern border of the woods (Fig. 7). On one of the nights when this pattern was not followed, the animal moved slowly, with frequent stops, from the center of the northern half of the woods south through the middle of the woodlot stopping just inside the southern border of the woods. From there, this animal made a brief excursion into the pasture on the south of the woodlot, then back, pre- sumably denning at the place where it had previously 30 Fig. 7.--The movements of raccoon 4 on the night of October 24, 1969. ”é: A I IllIllIIlj & 32 stopped. On the last night this animal was located, it moved from east to west through the woodlot and into the cornfield on the west of the woods where the signal was lost. This transmitter had by then been operating for three weeks and probably went dead at that time. Another striking feature of the nighttime movements of these animals was that periods of movement alternated with periods during which the animal remained in essen- tially the same location during the tracking period. In summarizing the data, periods of movements were defined to include series of fixes which occurred singly (Table 3). Localized activity included fixes occurring in a series at a single location. The first fix of such a series was, however, included in the preceeding period of movement as I felt that the time from the last single fix to the first of a series of stationary fixes should be included as movement to give a conservative estimate of the amount of time the animals remained stationary. Also excluded from localized activity were series at the beginning or end of a tracking night since these probably represented den site activity and were not indicative of activity patterns away from the den site. The average number of consecutive fixes of each type of activity was calculated for each animal and for the four animals considered to- gether (Table 3). These average numbers were then con- verted to time by multiplying by 15 minutes. On the 33 TABLE 3.--Average duration and nightly frequency of periods of movement and stationary activity. Movement Activity Localized Activity Animal Duration Number Duration Number (minutes) per night (minutes) per night 1 54 3.6 39 3.3 2 51 3.7 36 3.3 3 45 3.2 29 2.2 4 36 3.2 42 2.6 Weighted Average 48 3.4 38 2.9 34 average, the duration of these patterns was about the same: 48 minutes of movement and 38 minutes of localized activity. The pattern of alternating periods of movement and localized activity is shown by the average number of per— iods of each type of behavior per night. An underestimate of the frequency of localized activity periods is shown by the consistently higher number of movement periods per night caused by the exclusion of localized periods at the beginning and end of tracking nights. If the animals had always been followed only from the onset of movement until after movement ceased in the morning, the number of move- ment and localized periods would be equal due to the defi- nition of the two types of activity. The number of periods of each type per night is an indication that these animals were not continuously moving, rather their behavior con- sisted of movements and localized activity distributed in blocks of time greater than the 15 minute tracking inter- vals. In the field, during tracking operations, this two- phase pattern was even more noticeable since the receiver was not turned off between fixes. To facilitate evaluation of raccoon use of cover types, the area was subdivided on the basis of available resources rather than on geographical criteria (Fig. 2). Therefore, the two areas of woods were included in the same subdivision although they were not geographically continuous. Likewise, the two cornfields were considered together as were the pig pens, cattle barns, and the 35 pasture east of Beaumont Road. This pasture contained loose corn dumped with manure from the cattle barns and thus, animal feed was available there in much the same manner as at the livestock feeding areas. The remaining pastures formed the fourth subdivision. Alfalfa fields and grassy areas were grouped into the fifth subdivision and the nursery comprised the sixth. The southeast corner of the cornfield west of Hudson Woodlot was included in the alfalfa-grass subdivision because it was wet and con- tained high grass rather than corn. Several general conclusions were evident from the summarized data (Table 4). As might be expected, these animals spent a considerable proportion of their time in wooded areas. Three out of the four animals spent more than 50 per cent of the time they were tracked in wooded areas. The fourth spent only 11 per cent there; however, this animal was only followed for roughly one-half as long as any of the others and there may have been insufficient data to detect the animal's true behavior pattern. Another general observation was evident from the summary. On most nights, these animals visited areas where corn, either in the form of crops or feed, was present. On 4 of the 21 total nights this visitation did not occur, but of these 4 exceptions, 2 were nights when fewer than 12 fixes were obtained. The remaining two nights were ap- parently nights when the animal did not visit an area which contained corn. 36 m N m mm mm .2309 OO OO O ON\OH OOH OH ON\OH ON O O OO ON OHxOH O OOH HH OH\OH O NO NO HH\OH O NH ON NO OOH Hence ON HO ON O\OH OOH OH ONxO O OO OO OH OH\O O O OH NO HO ON OH\O ON ON ON OH\O O ON NO OH HH NO Hmuoe O mm mm ON ON OxO O OO HN OxO N O OO NH O OH N\O O H OH OH NH NO OOH Hmuoa OO OH NO NN NH\O ON HO HN HH\O O ON OH OO ON OxO O OO OO NN OxO H OOH OO «\O NO OO ON H\O O ON OH OO ON OO\O mmMHU WMUHAN pumnouo cam an: musummm xooumm>flq mcamwmcuoo mtOOS mmxfim mo Hwnfisz wumo Hmecd mama Hm>ou Umuqu ms» cw m0me m0 wmmucwoumm .mcoooomu an momma Ho>oo no on: xaunmflc cmNOHmEESmII.v mamas 37 This observation points up the importance of corn to the ecology of the animals in and around Hudson Woodlot, particularly during the late summer and fall. Berner (1965) noted a great number of movements between woodlots and cornfields, and a large percentage of corn in the raccoon scat he examined. During the present study, I qualita- tively examined 31 scats located during trapping operations and the contents of 2 stomachs from road-kills. Only one scat and one stomach were not predominantly corn. Both of these exceptions were composed entirely of black cherry and were found in August, one in each year. One scat was found which contained fairly direct evidence of feed corn. This scat was found in April and contained corn and pieces of the black plastic used to cover various materials at the livestock areas. The last general conclusion about cover types which was drawn from the summarized data was that alfalfa and dense, tall grassy areas were relatively infrequently used by raccoons. The highest percentage of use of this type of area by an animal on a night was 26 per cent. These fixes were all obtained at the end of the tracking period when the animal moved to the north of the northern corn- field, entered a culvert there and remained until tracking was discontinued after dawn. The other fixes of animals in this type of cover were contributed by animals moving through these areas rather than spending extended periods of time there. 38 Another aspect of raccoon movements which was ascer- tained from these data was the distance moved and area covered by an animal (Table 5). The total distance covered by the animal during the tracking period, as shown in the third column of Table 5, must be regarded as a minimum figure since short movements were not detectable, and often, either the initial or final movements of a night were not recorded. The fourth column shows the hours the animal was tracked on the night in question, and the fifth column is the quotient of the third divided by the fourth, i. e., the average distance covered per hour of tracking. The sixth column shows the time during which the animal was active either after leaving the den, before returning to it, or both. This column was included because on many occasions, the animal was inactive for a period at the be- ginning or end of the tracking period and I did not feel that this time should be included in a measure of rate of travel since the animal was presumably at the den site and inactive. The seventh column is similar to the fifth in that it is a rate of movement calculated from the distance traveled and the time of tracking less those periods which were excluded from the beginning or end of the tracking period. From this table, it can be seen that the mean dis— tance covered varied from 1,000 to 7,000 feet with a mean of 3,194 feet and a standard deviation of 1,800 feet. The average rate of travel was 564 feet per hour or 733 feet 39 .musoc Hmuou cmnp uwnumu pmumsncm :0 .xmc on» How 0>OuOMCO wEoomb on: HwEOcm DH Hmumm mason cam >HH>Huom hHunoOc mo uwmco mg» mnemmb mason mmmH mnsos Ummmmac on» uHmO Owe Hmuoee NOO OON OOO.H OcoHumH>mo UHMGCMpm MOO vmm va.m mane: OOO OO.H OOO OO.H OOO.H ON\OH OOH 0.0 ONH O.O OOO.H ONxOH OOH ON.OH OOH OO.HH OOO.H OH\OH O OOO O.N OOO OO.N OOO.H OH\OH OOO OO.O OON 0.0 OOO.N HH\OH OOO ON.O OOO OO.O OOO.N O\OH OOO ON.N OOO ON.N OOO.H ON\O OOO O.N OOO 0.0 OOO.N OH\O O OOO.H O.O OOO.H OO.O OOO.O OH\O ONO.H OO.H ONO O.O OOO.N OHxO ONO O.O ONO O.O OOO.O O\O OOO ON.O OOO ON.O OON.N O\O N OOO O.O OOO 0.0 ONH.O N\O ONO OO.O OOO OO.O OO0.0 OHxO OOO.H O.O OOO O.O OOO.O NHxO ONO.H OO.O OOO ON.O OOO.O HH\O ONO.H ON.O OOO OO.O OON.O O\O OON.H 0.0 ONO 0.0 OOO.O O\O H OON O.O OOH ON.O OOO.H OxO ONO O.O ONO ON.O ONH.O H\O OOO O.O OOO O.O OOO.O OO\O Huso:\umwmv oumm muso: Husoc\uwmmv muse: Huwmwv mama HMEOC¢ «.OOOOOOOO .OOOOOOOO mama Hmpoa OOOOOOHO .ucme>OE mo mmuou 0cm moonmumOQII.m mqmoE mo oumu mmmnm>m cm>Hm m suH3 muanc mo HmbEdzII.m mqm