“VESTQGATW OF THE W WOODCOCK N MW Thai: for H» Down a! N1. D. W STA?! W Lyfie H. Blankenship 1957 $ WU unnumuunww WWI Ml ~ 3 1293 13 0669 This is to certify that the thesis entitled Woodcock Investigations in Michigan presented by Lytle Houston Blankenship has been accepted towards fulfillment of the requirements for Ph. D. dqpeein Fisheries and Wildlife an”; 41‘ ' C—\ ,1". / ¢§Tykk ;>-Jfl41(»ax7?F Major professor ' I 71 M'( Date May 22; 1957 K l 0-169 L LIBRA R Y Michigan State University INVESTIGATIONS OF THE AMERICAN WOODCOCK IN MICHIGAN by 1(n Lytle Hf'k‘Blankenship A THESIS Submitted to the School of Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Fisheries and Wildlife 195 7 ABSTRACT Investigations of the American woodcock in Michigan were made for the purpose of developing and improving management techniques involv- ing the bird. Phases especially investigated were (1) migratory habits, annual production, and population levels, (2) extent of pre- sent utilization of the resource, and (3) factors influencing the distribution, abundance, and activities of the bird. Woodcock arrived in Michigan during late winter. By the last of April, spring migration was completed. Movements thereafter were re- stricted to local activites until late summer when full migration began. Most woodcock moved through Regions I and II by October 20 while in Region III some birds had moved out and others moved in by that date. V The major part “of the breeding season occurred between April 15 and Hay 15 in the Lower Peninsula and between April 25 and May 25 in the Upper Peninsula. Singing-ground counts were made during these periods. Data obtained on these counts were used to determine factors affecting observations and woodcock activity, papulation trends, num- ber of- times a. route should be run, and number of routes to be run in the state. ‘ Woodcock breeding activity was affected by light intensities, rain, temperature, wind, and other animals, particularly man. Major factors affecting observations were wind and noise interference. Singingaground routes usually were run two or three times each but a statistical. analysis of data from one route indicated that the route would need to be run at least 27 times to obtain confidence limits 25 per cent of the observed mean. Economically, a route should be run twice the first year and once each year thereafter. A comparison of bill lengths and the total width of the outer three primaries was considered a reliable criterion of sex. A com- parison of the total width and total length of the same primaries was used also for sexing birds. Leg color was used tentatively as a criterion of age. Sex and age data in conjunction with nesting success, brood sizes, and kill figures were used to determine survival and mortality and to analyze the extent of harvest. Data indicated woodcock are not being over-harvested in Michigan. Major factors influencing the distribution, abundance, and ac- tivities of woodcock were vegetation, food, weather, and soil. Minor factors included competition, predation, diseases and parasites, and accidents. Vegetation was important in restricting distribution as well. as serving as an indicator of the habitat type. The availability of earthworms was a critical factor determining whether or not woodcock used certain areas as feeding or loafing sites. Weather, including temperature, wind, and precipitation, influenced local activity and migration. The remaining factors were of little importance except in local situations. For managing the woodcock population two methods seemed promising. First, improvement of habitat by vegetation controls should be used on singing grounds and feeding areas. Second, a change in hunting season dates to correspond more closely to migratory movemnts might be helpful for increasing the harvest of woodcock. Future studies should emphasize the relationship between woodcock populations and habitat changes, practical methods for determining seasonal occurrence of birds, increased vegetation control by various means, and better techniques for determining population trends. In the meantime, banding operations and singingcground counts should be continued and increased when possible. TABLE OF CONTENTS Introduction............. ....... ................ ....... ....... Status of Neodcock Research.......... ..... ............... Acknowledgments........................... ..... ..... ..... Distribution and Movements.................................... Dietnbution in M10hlganooo............................oo Movements................................................ Spring......................... ...... . ........ ....... Path of migration.................................. Time of migration.................................. Summer............................................... Fall................................................. Pre-season movements............................... Hunting season movements........................... Conclusions regarding fall movements............... winterOOODOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Poplfl-ation StudiGSOOOOOOO.9000000000000000.0000000000000~°OOOOO Singing-ground Activities................................ Factors affecting woodcock activity.................. Factors affecting observations....................... Marked‘woodcock...................................... Singing-ground counts................................ Neatmgoooooeeoooeeeoeoeoeeoeeeooeooooeeoeooeeeeeeeoeeeee BrOOdSOOOOCOOOOOOOOOOOOO00000.00.000000000000000000000000 Sex.and Age Statistics................................... Sex and age determination............................ Sex and age ratios................................... Production........................................... hunt Of mlooeoeooeooeoooeoooeooooooooooooooeoeoeoeooo Ec010gical Factors Affecting Distribution, Abundance, and ~A°tifiti°8°oeeeoeooeeoeoeooeeeoooeeeoeoeeeoeecoooeoeeoeooooeoe vegetation............................................... Breeding areas. Nesting cover........................................ Brood habitat........................................ Summer............................................... Fall................................................. Land-use..............o..............o........oo.o..o Soil..................................................... FOOdOOO000000009000OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO weather.................................................. Temperature.......................................... 'Wind................................................. Precipitation........................................ Nebulosity and light intensities..................... «ovi 10 10 ll 11 12 21 21 21 23 55 56 56 62 6h 78 79 82 98 102 108 108 Barometric pressure.................................. OtherJFactors............................................ Competition.......................................... Predation............................................ Diseases and parasites............................... Accidents............................................ Management.................................................... vegetatim contrOIOOOOOOOOOOOOOOO00°00OOOOOOOOOOOOOOOOOOO Hunting RegulationBOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Recommendations for further studies...................... SWW.‘.000OOOOOOOOOOOOOOOOOOOOOOOO00006.0000000000000000000 Iii-“rat”. CimeOOOOOOOOOOOOOOOO00.0000000000000000000000000° Appendi-XOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 161 161 161 16h 165 172 172 173 178 181 185 190 19h IN TRODUC TION The American woodcock (Philohela minor) is a favorite bird of many people. To some bird watchers the male's courtship flights dure ing the spring are a sight to behold. To photographers the well- camouflaged nesting female provides a perfect model as she site so closely on the nest that she may be removed by hand. And.yet, to others the seemingly quick and darting shadow of a bird is a target for the.gun. To these latter people the woodcock, though small in size, is a.favorite game bird, even though most hunters consider the timberdoodle only a welcome addition to a mixed bag of game. Regardw less of its appeal to the human, many hours have been spent by inter- ested persons searching the woodcock's domain for sight or sound of the bird. In addition, many rounds of ammunition regularly are expanded on the diminutive game bird, causing a considerable income for arms manufacturers and dealers. Thus, the woodcock is a bird of economic as well as esthetic value.l Although a member of the Order Charadriiformes, the woodcock is primarily an inhabitant of wooded swales and stream.bottoms. More specifically, it is a member of the Family'ScolOpacidae which includes such relatives as the snipe, sandpipers, curlews, and willets. The family is divided.further into three subfamilies one of which is the Scolopacinae to which the woodcock belongs. Generally this group is recognized by the three following characters: (l) eyes located in the upper part of the head, (2) a long and membranous bill, and (3) ear far below the eye and possibly a little forward. There are ten genera in the North American Scolopacinae. The American woodcock is dis- tinctive in having the three outer primaries narrowly attenuated and shorter than the next two primaries which are the longest of the primaries (Pettingill, 1936). A close relative, and Old World cmmt~ erpart, the European woodcock, ScolOpax rusticola, is found in the British Isles, Europe, Asia, and northern Africa. The American woodcock is confined primarily to the deciduous forests of eastern United States and southern Canada. Although the bird breeds generally throughout this region, the primary breeding range includes southern Ontario and Quebec, the Maritime Provinces, and the area south to the hOth parallel. The main wintering range includes portions of the southern Atlantic and Gulf Coastal states from Maryland to Texas with the major concentration being in Louisiana. Since the major wintering and breeding grounds are widely sepa~ rated, the life cycle of the woodcock includes a biannual migration. In late winter and early spring most birds leave their wintering range for the northern breeding areas. A movement in the Opposite direction occurs during the fall. When woodcock return to the breeding grounds the males locate relatively Open sites, referred to as "singing grounds ," on which they perform their courtship display (Plate'l). Throughout the day birds remain in diurnal coverts, occasionally moving and sometimes feeding if the coverts are located suitably. Then a few minutes after sundown (15* minutes, depending on cloud coverage) the male seems to become sexually excited and Often starts to sound off with a call of nasal-like quality described as a "peent." After a few minutes of rather'consistent ”peenting; the bird leaves the covert and proceeds to the singing ground which usually is within 300 feet. It may'fly directly to the site, make a short, circling flight. first, or even walk to the singing ground.‘ After reaching the area.the male usually spends several minutes peenting before starting the aerial display. This display normally commencestvith the bird taking off into the wind. It ascends in spirals to a height of at least 200 feet and then descends rapidly in a series Of zig~zags and j=shaped loops until abOut 30 feet from the ground. From.this level it glides in and lands at or near the spot of its departure. During the flight, which lasts about 60 seconds, an aerial song may be heard. This song begins with a whistle of the wings, and the wing trilling may be heard intermittently through most.of the flight. The vocal twitter- ing or chipping is not added until near the apex of the flight, prObably as the bird begins to descend. Then the vocal and wing noises seem to occur alternately. All sounds and as the male glides in for a landing. After reaching the singing ground the bird peents for a minute or so before taking off on another aerial display. This alternate ground and aerial activity continues usually between 30 and hS minutes each evening. Occasionally a female will enter the male's domain, at which time copulation may take place. When the-evening activity has ended,the male may remain on the singing ground awhile; eventually it leaves to feed. It may return to the I'ground't several times during the night, especially on a bright moonlight night, but another major period of display does not take place until around daybreak. After this morning period of activity PLATE 1 A. A typical singing ground located on the Rose Lake Wildlife Experiment Station. The site most frequently used is indicated by the arrow. B. A singing ground situated in Indian grass and small shrubs near Vermillion Creek on the Rose Lake Wildlife Experiment Station. The singing ground is indicated by the arrow. on the singing ground, the male retires to the diurnal coverts again. Status of Woodcock Research One of the first detailed accounts of the Americanwoodcock was published in the eighteenth century by Lawson (17114). This was some years prior to the formal description by Pennant (1785) who gave the bird the name of ”Little Woodcock." The first Latin name was S_g_o___lo~ E73 mi__no_r, given by J. F. Gmelin in the thirteenth editionof the §zstema Natures of 1788. George Robert Gray, in his 18131 edition of A List of the Genera of Birds, changed the generic name to Philohela. Other than the taxonomic studies and general observations little scientific work was done on the species until recent years when Pettingill‘ (1936) publist on an intensive study of the woodcock's life history. In addition to data gathered through his own studies in Maine and New York, Pettingill compiled many findings from ques- tionnaires and public atiOns . A later book, written by Mendel]. and Aldous (19).;3), was based on an investigation of the ecology and management of the woodcock in Maine. The present widelydused methods for spring breeding-ground counts were introduced by them. A large part of the most recent and intensive studies on wood- cock, including those of William Sheldon in Massachusetts, Stephen Liscinsh in Pennsylvania, and Leslie Glasgow in Louisiana, have not been published except in state and federal (Pittman-Robertson Projects) progress reports. In Michigan, investigations of woodcock have been confined primarily to observations regarding distribution and generallife history notes. The writings of Barrows (1912) and Wood (1951), on the birds of Michigan, included most of the published references to woodcock in the state and information from a few observations of their own. For many years the American woodcock has been a major game bird in the northeastern states. In the Great Lakes region, especially Michigan, the imaortance of this bird has not been emphasized until recently. There have been many woodcock hunters but the major harm vest of the bird has been by the ruffed grouse hunter. This incim dental harvest has been sufficient to place Michigan as the state with the largest annual kill. The present study was instigated dura ing the spring of 1953 to learn more concerning the "timberdoodle" in Michigan. Field work continued until the fall of 1955. The primary . objectives of the investigation were to determine: (1) Annual production, population levels, and migratory habits. (2) The extent of the present utilization of the resource. ( 3) Factors influencing the distribution, abundance, and activities of the woodcock. ()4) Management practices necessary to improve conditions for the woodcock. During the first year most of the time was spent getting ace quainted with the woodcock habitat in the state and doing basic work in preparation for future activities. The major part of the work was accomplished in the following two years with most of the spring data being gathered on the Rose Lake Wildlife Experiment Station and the Gratiot-Saginaw Game Area, while summer and fall research was carried on in the vicinity of the Pigeon River Trout Research Station near Vanderbilt, Michigan (Plate 2). For clarity, certain terms are defined at this time. A ”route" or "singing ground route" constituted the area of woodcock habitat traversed for the purpose of counting the number of breeding males which could be heard. "COOperators" were the people who assisted in running singing ground routes, keeping records on hunting success and migration, or collecting woodcock parts for sex and age data. Regions I, II, and III are state administrative divisions used by the Michigan Department of Conservation. Region I is the Upper Peninsula; Region II includes the northern half of the Lower Peninsula; and Region III includes the southern half of the Lower Peninsula (Plate 2). Other terms are considered as being self-=exp1anatory or are defined in the section in which they are used. Acknowledgments The investigation was undertaken under a contract between the Michigan Department of Conservation and Michigan State University. This study was designated as Pittman~Robertson Research Project 90=»R of the Game Division, Michigan Department of Conservation. The writer wishes to express his appreciation to Dr. Donald W. Douglass, Game Division, Michigan Department of Conservation, and PLATE 2 A map of Michigan showing the regional divisions, stucv areas, and many other locations mentioned in the text. Wltourc O M I W‘s" su. , l sanctum CHISYAL V " Hus e \ m | I . 4 BL AKV O PARK m nun um exam an m ‘ ~ -:- ‘ V" “but “In“ atom “NA ” .W ‘ l . Am ' "Ran ‘ . an: caucu- ~ un- ‘ m all-loo: *mvm- - can: xltA ,- , t am e GRAND Mm RADIOS one I Dr. George A. Petrides, Department of Fisheries and Wildlife, Michi= gan State University, for their supervision of the project and edit- ing of the manuscript. Thanks also go to Dr. Eugene A. Roelofs and.Dr. Robert C. Ball, Department of Fisheries and'Wildlife, Michigan State University, for their critical reading of the dissertation, and to Dr. Don W. Mayne, Zoology Department, Michigan State University, for his guidance in statistical matters. .Many employees of the Michigan Department of Conservation were helpful during the study: ‘William.H. Goudy assisted in field investigations;'William.Youatt did the photo developing and.printing; William.Laycock and Merrill Petoskbyb 88 Co- ordinators of PittmanGRObertson Research Projects, assisted with tech~ nical details concerned with the project; Cash Wonser directed the soils study; and Walt Palmer, Lawrence Ryel, and Lloyd Schemenauer spent much time locating woodcock and woodcock habitat for the writer as well as helping on other phases. Acknowledgments are due the many c00perators, both department and non-department personnel, who made singing-ground counts, maintained hunting and migration records, and sent in parts of woodcock used for sex.and age data. Space is not available to name each.individual but the names are on file in the Game Division records. Ed.Iaw and John VanNoy, students at.Michigan State University, helped gather data on singing ground activities. The writer is grateful to Dr. William H. Marshall, Professor of Wildlife Management, University of Minnesota, for the use of woodcock specimens taken in Minnesota and to Dr. William Murchie of Thiel College, Pennsylvania, for his identification of earthworms taken 10 from study areas. Allen Duvall, U.S. Fish and Wildlife Service, Patuxent, Maryland, furnished banding records. Facilities for summer and fall studies were made available at the Pigeon River Trout Research station by the Institute for Fisheries Research of the Michigan Department of Conservation. Thanks go to Edward Bacon and Gerald Myers for their cooperation at the station. Gratitude must be expressed to the writer's wife, Margaret, who spent man hours assisting in field studies and in typing reports. Last but not least, grateful acknowledgment is given two top- notch woodcock dogs, Lady and Pixie, both Brittany Spaniels. Without the aid of these fine dogs in locating woodcock,much of the infor- mation would never have been gathered. Appreciation is expressed too for the marw hours of company they provided in the solitude of ex- tensive untenanted woods as well as the enjoyment received from watch- ing them work. DISTRIBUTION AND MO VEI‘ENTS Distribution in Michigan During this study woodcock were observed by the writer, other Department of Conservation personnel, and singing ground cooperators in every county except Alcona, Bennie, Branch, Cass, Hillsdale, Ionia, Missaukee, Osceola, Ottawa, and Wexford. Reports of breeding birdsvvere received from hO counties as a result of singing-ground counts between 1953 and 1955. Records from the other counties were of one or more active singing grounds. In all prdbability woodcock nest in every county of Michigan. In 28 counties where breeding woodcock had been observed by the ‘writer, birds were found in the summer. In 15 other counties where breeding woodcock had been reported, the writer saw summer residents. Hunting records since 1938 revealed that woodcock, whether local or migrant, had been shot in every county. Because of the large acreage of farmland and the scarcity of wooded streams and swales, the woodcock is not as abundant in the southern half of the Lower Peninsula as in the northern half of the Lower Peninsula and the Upper Peninsula. Movements Spring. Path of migration. The best information available concerning the relationship between the breeding and wintering grounds of Michigan woodcock is from banding records (Table 1). Since some of the banded woodcock were killed near their singing grounds, it is likely that many woodcock remain in the vicinity of the breeding ground from.the time of arrival in the spring until departure in the fall. If this is the case then fall band.returns should be useful in determining the origin and destination of woodcock. 'With these data and the 12 supposition that birds travel somewhat in a direct path from the wintering grounds to the breeding range, the general direction of migration between Louisiana and Michigan is shown in Plate 3. Other paths of migration must involve other states as indicated by a band recovery in Georgia from a woodcock banded in Antrim County, Michigan. This bird lived almost two years; so it is probable that it had made one return trip to Michigan between the banding and recovery dates. That woodcock frequently return to the same area each year was demonstrated by banding Operations. Several male birds were re-trapped on the same singing ground three years in a row. V Other workers found that male woodcock frequently returned to the same singing ground or to that vicinity (Mendall and Aldous, 1913, and Sheldon, l953) . Time of Migration. The time of the major movement of woodcock from their wintering grounds is probably during February although records in the literature are scarce. Knight (191:6) , in an article which was based on information gathered by others, mentioned that before two-thirds of February is gone most of the woodcock have de- parted from the wintering grounds. Glasgow, of Louisiana State University, is planning to publish a manuscript soon regarding the migration of woodcock (personal letter). Woodcock arrive in Michigan at various times depending on the part of the state under consideration. The Lower Peninsula is ap- proximately 290 miles long. Usually much of the northern half of this area is still under snow when woodcock first reach the southern half. Most of the Upper Peninsula is still farther north and has a 13 .nosz ..oo cemsonoeo .ocshe>aozv.m .as we mn\m \oa .oo csmsonoeo mm\o~\n z - mammm mam .nosz ..oo omooeo ..sem noses doomed mm\am\oa .oo omoeeo mm\mm\s z 4 Noomm new .noez ..oo censonono .omesnm one mm\m \OH .oo omomeo mm\mm\m : a msmmm mam .noaz ..oo omoneo .Haam ram nm\m \oa .oo cheese .nm\m \m a . cannm man .nosz ..oo someonono .nsso coo nn\m \HH .oo cemeonono mags \m z - manna mmn .eosz ..oo omomeo .Hasm new mm\ma\oa .oo cemaonono nn\a \a z - aommm mmn .noaz ..oo commesmsnm ..sem .axm oxen omom mn\a \HH .oo cosmemsenm nn\om\n z 4 cases mam msmnooo ..oo emcee .oaas>mocnsm mm\ma\~ .oo senes4 .ena4 mn\oa\m . a Homes man .nosz ..oo someone .cosxoee Hn\e \oa commons .Hm\om\m . a momma mas. scsmsnoszs . .noss ..oo omoeso mm\na\oa once creche nm\mm\a a 4 semen mum .nosz ..oo cones nocnoo m2 sm\ms\oa essence shone an\m \H - - «ammo mun .nosz ..oo cswoaa4 .soacshm .m .e5 m sm\om\oa ones mesons en\oa\a - 4 ammme mmm .noez...oo_ssm sm\om\oa nested hostess .em mm\m~\ma - - cones mam .nesz ..oo monsoon .smsm mn\em\m ssmnea4 caress mm\mm\a a - chasm mmm .nos: ..oo essences .odassuseoo sm\ma\e once onsets mm\m~\ma : 4 momma mmm .noaz ..oo cannon em\m~\0H once onsets mn\mm\ma - 4 oonmo mmm. .nosz ..oo cosmmcsssa .Haosom .m sm\m~\m once mesons mm\mm\ma. s 4 enema mmn .noszfl..oo maooomo .coenmsn .3,.ss a sm\e \QH samshm mm\aa\ma . 4 Homes «mm .nosz ..oo cmmsonono .oxme sesame mm\ma\oa oxen oeeaesm Nm\am\a . 4 mmaoo mmm .eosz ..oo seahorses mn\o~\oa cannon ~m\e~\H a 4 mmeme mes .nosz ..oo smsomo .nosxaom mm\s \oa once , . . choose e.cshose, ~m\m \NH . - «seem mmn .nosz ..oo smcoensm3_.hono4 cc4 .z .s2 a em\aw\oe cannon ~m\mH\NH a - mmomm mmm .nosz ..oo oomoH .noesm none s4 mm\s \oa ecosmcoesm ~m\m \H z 4 mamam mmm .nosx ..oo oscoh4 .noma4 .m .s2 s Nm\m \oa ochsntonm Nm\m \H a 4. mmmam.mmm .nosz ..oo recesses: .eonh4 cc4 ~m\m \oa eoosmcosmm ~m\ma\a. a 4 mmemo mmn .nos: ..oo sumac .ooeao .m .s5 a mm\sm\oa cannon Hn\om\~H a - sedan mmm .nosz ..oo oceans: ~m\~ \HH emote com Hm\mm\ma z. 4 , oaoam mum .noez ..oo escape ..anese seems scene mm\m \HH once mesons Hn\ma\ma a 4 omens non .eoaz ..ou oocssoco: .chonesmm mm\-\oa senses Hm\om\ma z. 4 oasee.mom .nosz ..oo maestro .oaasscoeno mn\nm\oa cannon Hn\o~\~H . 4 cases men .nos: ..oo omamnmz .onsn room Nn\ \OH ochsnhonm am\o~\~H a .4 nmsee.mon .nos: .ocsnsz mm .s2 s Hm\ma\oa ecosmcosmm Hn\e \H a 4 anmao mom .nosz ..ou oomoH ~n\aa\oa coosmcosmm on\ma\a . 4 manna no: .noaz ..oo eaoeomo ss\a~\oa season mm\~H\H - - nmmos mam‘ .35.“de?? . .nosz .eeeasm mam oe\e~\~H .oes3_.ocseem am\aa\s z - msemn ewe no moon oemm1 coaemooa oemo somwxmmsv .oz seem even unseen 33 393 sewage“: ca motocross.” no pounce xooopoo: mcheocoo mesons.“ meancmm .H 3an PLATE 3 Relationship between banding and recovery sites of 29 woodcock, with postulated paths of migration. lS correspondingly longer winter. Woodcock may not reach parts of that area until mid-spring. _ The earliest arrival date in Michigan recorded recently was for Kalamazoo. Here a cooperator (Mrs. Gladys Hall) observed one on February 28, 1955. For the Lansing area, March 9, 1955, has been the earliest date of arrival observed recently (Victor J anson, Department of Conservation). Eugene Kenaga observed woodcock on March 21;, 19148, and March 27, 1953, near Midland. Lawrence Ryel, game biologist at the Ogemaw State Game Refuge, reported woodcock in Roscommon County during the latter part of March, 1955. In l95h and 195's woodcock were observed during the first week of April at the Pigeon River area. No recent records are available for the Upper Peninsula. Older records show that first arrival dates have continued to be about the same through the years. Gibbs (1885) in discussing birds of the Kalamazoo area mentioned that woodcock usually appeared in early March but often arrived in late February. According to Cook (1893), the birds were common. from March, rarely February, to October or November. The earliest date of arrival at Petersburg was March 2, 1887, and the average was March 17 (Cooke, 1912). In reviewing past works, Bent (1927) found that woodcock had been recorded as early as March 9 at Battle Creek and March 17 at Ann Arbor. Pettingill (1936) ,- through literature review and questionnaires, found that woodcock had arrived at Ann Arbor by March 2, 1918, and Blaney‘Park, Schoolcraft County, by April 5, 1931. Although the average arrival of wo odcock at the Rose Lake Experiment Station during the past three years has been March 12, the 16 last evident flights have been three or four weeks later. In 1953, the last flight passed before singing ground routes were started on April 20 (Figure 1). The data for 195).; (Figure 2) did not show any definite migration peaks but the greatest number of occupied singing grounds occurred between April lb and 22. The route was not run prior to April 3, but one or more flights may have passed through earlier inasmuch as observations at two sites in the Rose Lake area on March 22, 1951;, showed three or more birds present at each place. In 1955 the Rose Lake route was run earlier in the spring and more frequently than in 1951;. The fluctuations in number of occupied singing grounds (Figure 3) indicated migrants were passing through the area up to mid-April. The last peak) Imparently occurred about April 13., which would correspond to the postulated peak for 1953 and 1951;. The Gratiot-Saginaw route was run fewer times and many migrants could have been missed. In 1953 one flight of spring migrants seemed to be present when the singing-ground counts were started on April 11 (Figure 1). Since another count was not made until April 19, a pos- sible flight may have passed unnoticed. About the same may be said for the 19514 route (Figure 2). Since only three counts were made during the migration period, the true migration pattern is probably not shown by the available data. Migrants may have passed through prior to April 9 when the route was first run or before April 23 when the next count was taken. Data for 1955 revealed a clearer picture of tin possible migra- tion routine. The Gratiot-Saginaw route was run only a few times but 17 the observer happened'te be in the right place at the right time. on April 16 and My 1, 19 active woodcock were counted while only 11 were heard on April 23 (Figure 3). A study of these counts might leave the impression that two waves of migrants passed through the area.. How- ever, Kay 1 is past the normal (migration period for Region III and the low count on April 23 probably resulted from high wind conditions which prevented the observer from hearing all active woodcock. 0n the basis of the above data and explanation the birds counted on April 16 could not be considered as migrants. And yet, on the evening of the 16th it was evident that migrants were present in large num- bers. While the route was being traversed woodcock were seen flying across and sitting in the road. Sme of these birds may have been local females, but at no other time did such a large number of wood- cock appear as on that evening. At least 31 woodcock were seen or heard along the route. As a result of. the above observations it is believed that for the two areas discussed the regularly-used singing grounds were oc- omied throughout much of the migratory period. As this period of migration progressed, some singing grounds were established as per- manent breeding areas by residents while other sites were used as temporary grounds by migrants. Consequently, some woodcock moved in and stayed while others moved on, thus giving the impression of successive waves of spring migrants with the last wave usually being the greatest. By the time the last flight came through, most of the regularlydused grounds were taken so that the count made on April 16, 1955, included most of the residents plus some migrants. 18 H.294 e a. m m Hem am mm mm Headed? 9d , ._A _ _ _._.|_.I_.l_|_.|_|_|_|_.lo 3.3396... gene S .mnma .. needed higmueoflfio use , 03 owed on» do medal :33 mo none—us on» 5 udoapmmposfim A 0.335 qunoo .red pres}; xooopoon JO .tequnu 19 made he: . .Hahm4 L maommmmmemmmommafiifioa m m em ommmomemmmommaflimaoam _ d é-\O _.___~___4_____:d_44fi.___ axed omom / amdammmteogmno 2:an ... nova?” gfimsmnvogmuu use 33A them one. :0 nodes 233 no hobs—Ba one. a.“ whoapmsaosfim .N enema —-e -—N .3 ON 3. aunoo .rsd preeH swoopoon 3:0 .10qu 20 m H on mm mm 3N NN 0N ma 0H dd NH 0H w o n N on mm mm mm mm om ma 0H dd NH 0H m w a- N Hm mm _/____4___fi#._4_....d____l__l__l_.ao IIM1 . /// [II ill/II mama and: . . .I.n / nlloH smdaummnuo«umuu 1? [.om .mnma I season :adamsmavoavano use mama snow on» no means o>avos no bones: one dd escapesuosah .m enemah aunoo Jed paeeH accepoon_;o JeqmnN 21 sun-ls. In late spring and early summer, woodcock apparently began to move from the typical breeding sites into moister situations. Dry, open sites frequently used for singing grounds were no longer occupied by the birds and the adjoining edges comonly used for diurnal coverts often were abandoned. . Also, since many moist areas became restricted, woodcock had to concentrate nearer to more permanent wet sites such as stream bottoms. Often the hillsides overlooking such areas were used during the day. Woodcock seem to remain in or near these more permanent sites of moist soil throughout most of the sumer. If the summer has plenti- ful precipitation, the woodcock are more dispersed.. In late summer there was some indication of a complete popu- lation novemnt from certain locations. On several occasions during August, 1955, in the West HcMaster Creek area of Cheboygan County, as new as 15 woodcock were flushed in 30 minutes. In the same month and area, 12 woodcock were banded. But after September 1, no woodcock could be found in the area. The birds had moved but tl'eir new loca- tion could not be determined. Reports were received of an increase in numbers of birds mar Atlanta during the week ending September 3. At the same time an increase in woodcock was reported in southern Otsego County near Crapo Lake. Similar instances of woodcock move- ments will be more fully discussed in the following section. Fall. Pre-seasan movements. This period included the 30 days preceding 22 the October 1 opening of the woodcock hunting season in Regions I and II. Most of the information gathered was based on daily field obser- vations. The movements observed in parts of Cheboygan, Otsego, and Mont- morency counties were mentioned above. Another definite record of a large number of woodcock, seemingly on the move in Septerhber, was turned in by a member of the Department of Conservation who saw 100 woodcock one evening on a road in Ontonagon County (western part of Region I). On September 28, David Arnold, District Game Supervisor at Escanaba, with an observer, flushed 15 woodcock in 1% hours on a special study area. This was the peak for that location during the fall season. The greatest number of birds seen thereafter was 7 in 1 3/1; hours. It is not known whether these movements were definite southward migrations or local wanderings. The population fluctuations noted around September 1 in Region II could have been a southward migration since the number of woodcock seemed to decrease in southern Cheboygan County but increased southward in the vicinity of Atlanta and Crapo Lake. If this viewpoint is taken, it is difficult to account for one woodcock which was banded at West McHaster Creek in August, 1955, and shot Novenber- 2, 2} miles due south. Soil moisture never became a critical factor for the presence of woodcock at West McMaster Creek since a beaver pond maintained a high soil water table. Food could have been the important factor at that site since the earthworm population was mall. 23 Another cause for movement, especially in the Region II area being discussed, could have been the weather. August had been gener- ally hot and dry, but a frost occurred on September 1, followed by a few cool days. The sudden change could have caused the population fluctuations in the West McMaster Creek area; similar weather condi- tions may have been effective at other sites. However, woodcock were present in many other areas checked regularly in the Pigeon River area of Otsego and Cheboygan counties. ‘ There is a theory that woodcock may move north before finally heading south. This could be, but if they do it seems likely that they return via the same route, since seven banded woodcock were recovered in 1955 in the vicinity of banding. Mendall and Aldous (19h3) had banding records which showed a preseasonal movement of woodcock north- ward as far as [:0 miles. Regardless of why woodcock move and where they go when they do move, some areas become empty of woodcock during September while other sites have an increase in pOpulation. Huntianeason movements. The movements discussed in this sec- tion refer to those which took place during the hunting season. This season lasted from Octobertl to November 1 in Region I, from October 1 to November 9 in Region II, and from October 20 to November 9 in Region III. The major emphasis was placed upon data from Regions I and II. Methods of studying these movements included: (1) Splashing counts (2) Area observations at Pigeon River ( 3) Daily observations by woodcock observers 2h ()4) Records frcn game bird hunters (5) Banding Operations Splashing counts were undertaken to determine whether such counts were feasible as an indication of migration (splashings are chalky- colored fecal droppings of woodcock). The plan was to set up several plots in different areas and to check these plots regularly for evi- dence of splashings. It was hoped, if splashings did indicate a population fluctuation, a method for determining dates of migration could be devised for use by field personnel. This method was started during the fall of 1953 when plots were - set up in two different localities in the Rifle River and Pigeon River areas. The localities were chosen on the basis of known previous use by woodcock. The plots were one-hundredth acre in size. Five were set up in each of two areas located along Rifle River, six plots were located at Tin Bridge (Pigeon River Area), and eight plots were set up near the Pigeon River Experiment Station. Upon each visit to the plots, splashings were counted and erased. In addition, notes were taken on the number of woodcock seen on and off the plots, the num- ber of splashings near plots, and weather conditions. These plots were checked only on weekends in 1953. In 1953;, plots at the Rifle River Area were drapped and efforts were concentrated in the vicinity of the Pigeon River Experiment Station. Only two areas (Tin Bridge and Experiment Station Flats) used in 1953 were used again, and the number of quadrats was reduced to four for each area because of time limitations. However, since fall headquarters for the project was at the Pigeon River Station, the 25 plots were visited more frequently. The same type of data as was recorded in 1953 was gathered. ' The same procedure was followed in 1955 except that three areas were used. Besides the Tin Bridge site which was the only old local- ity retained, two new areas were set up, at West HcHaster Creek and the Old Grant School. All. three sites were located in southeastern Cheboygan County (Plate 1;). Ten plots were set up in each location. These plots were checked frequently throughout the fall, and again the same type (of information was taken. In addition, another method of observing splashings was under- taken in 1955. This method was a line transect projected through each of the three sites discussed above. Splashings were counted on a strip covering five feet on each side of the line. The data for 1953 were not used as much as later information in an analysis of population fluctuations since the areas were visited only on weekends and consequently often represented a week's accumu- . lation of splashes. Therefore, data from the 195).; and 1955 seasons (Table 2) were used to determine the population fluctuations. In 1951: the areas were visited more frequently than in 1955 and the amount of time spenton the areas was less in 1955 because more plots and transects were involved and other phases of study were started. However, data from the fall of 1955 were as reliable as those of 1951; since at each check both old and new splashings were counted and the age of splashings considered. Moisture was considered since it hastens the deterioration of splashings. However, moisture did not eliminate anylsplashings before they were counted. 26 The 1951:. data regarding splashings correlated more closely with other information regarding times of migration than the 1955 data. A large number of splashings often occurred on the Tin Bridge plots at the same time a peak in actual number of birds occurred in the Pigeon River Area. More often, however, when a peak in the number of wood- cock was observed, splashings were not found on an plots. Little correlation was seen” between the amber of woodcock flushed and the nmber of splashings recorded from plots and transects in 1955. Fewer woodcock were seen in the Tin Bridge area than in 1951:. The decrease in woodcock in that area may have accounted for the splash- ings being less indicative of population fluctuations. Data gathered incidental to regular plots emphasized the necessity of a great number of sample areas and plots and thus the probable impracticality of the method. Specific instances illustrate this conclusion. On October 1, 19511, at Tin Bridge no splashings were found on the four plots, but eight woodcock were flushed on an acre enclosing two of the same plots. Four dqs later in the same area no splashings were seen on plots, but 15 were found a few feet from one plot. In 1955, on October 114, only one splashing was found on plots at Grant School while a great number of splashings were found near the plots. At the same time, seven woodcock were flushed in and adjacent to the plot area. During the three years only eight woodcock were flushed from the quadrats. Bio of these birds were flushed in 1953 and the other sir were flushed in 1955. Four of the latter woodcock were moved from 27 PLATE h Detailed map. of the Pigeon River Area showing the location of study areas and distinguishing landmarks. 1.- use“. 7"‘ v “as. u nae-- LEGEND SECTOU D SECTIOC 0 OLD CCC CA" mecca MI SYAYION FLATS TIN BRIO“ TIN emcee TIN MDGE GRANT “M “”511! CEEK YN SHANTV canoes: 114 MY save: “$7 *MASW. C‘CK WEST muesm cm gut HILL mans"! wALL seam run mm M! farmer sum SCALE (2... “ILE'O 28 N o o o o N 0 HH 0 o o o o o o o e o o o o o o o o o o o a on H HH o o o o o o o o o o e o o o o o o o o o s o m m o a o o n o o o o N o m o o o N N H o n o H o n o o o m H o o o H o o o o N. o o 0 mm H H o o 0 0H s mm H n o o o m o o v o .o o o o o o o o o o NH 0 o o o n H m o a m o o 0 NH S H e HN m . conch eeeJIlesxez chem indecent eehlnnsseh 93.5 833m chain .828 cohesem e the scene see: :2. scene use: :3. ..m secure a: cream .m secure a: _ stomach. 3on am J32." fl n H . smmH it 3me no.3 non venom nmfinnaaam xoooeoos mo sonasz .m oases 29 along the transects. The splash-plot method has not proven worth while as a means of determining migration. (The major difficulties involved are: (l) the great number of sample areas and quadrats needed which raises the time, effort, and cost, (2) the variation in use of an area by wood- cock from year to year as well as between seasons, and (3) the varia- tion of use within an area. Another method employed in, the Pigeon River Area for determining population fluctuations was the use of bird dogs‘on special study areas. In l95h, two such areas were set up, one of which was near the Tin Bridge and the other at Kingfisher Wall on Cornwall Creek. The areas were checked primarily in October, the critical period for observations. Record was kept of the number of hours spent in the area and the number of woodcock flushed (not counting reflushes when such could be determined). Ten study areas were located for the 1955 fall observations. All ten were in southeastern Cheboygan County and northeastern Otsego County. Observations were terminated on November 3 because of a heavy snowfall. All sites used were known woodcock areas, being used either during preceding falls or the sumer Just prior to the fall observations. For the Tin Bridge and Kingfisher Wall areas there was a definite time during October when birds were moving. In both years the great- est number of woodcock were found during the first three days of October. Some of these woodcock could have been migrants but it is 3o likely most of them were residents since both areas harbored high summer populations. The decrease after the first few days of hunting probably was not a result of migration but a result of birds being hunted and consequently killed or dispersed. In both.years another increase, though smaller, occurred around October 7. Another peak was on October 12 although the number of woodcock seen was greater in 195h. The next peak for 195k came on October 19 which was two days later than the next high for 1955. The last prominent increase took place on October 22 in 195k and October 2h in 1955. In 1955 the Pigeon River Area was separated into east and west sections. This division was made to determine if the population fluctuated differently between the two sections. The cast section was . composed of those five areas (Grant School, Grant Farm, Duby Lake, Nest McMaster Creek, and McMaster Creek) located east of the Osmun Lake Truck Trail. The west section included those sites (Tin Bridge, Kingfisher Wall, Elk Hill, and Pigeon River Station Flats) located west of the truck trail and primarily along the Pigeon River. Population peaks never occurred at the same time in the two sec- tions although occasionally they nearly coincided (Figure h). These data indicated woodcock came into one area.one time and the other area the next time. An analysis of variance was made to see if any consis- tent difference existed in the population fluctuations between the two areas. The resulting variance ratio (F) of 0.11 was not significant at the 1 per cent level, indicating no evidence of such consistent differences between times or places. 31 .895262 e N 3,: eenoeoo an N N N HN ma NH ma 9” .3. lam/ii» v3 .mnma I no.3 nozm comma..." 23 no mmoamaban .503 can mean 23 £95 meaoog madman hp eopeoflxud me «moavmnposam moaned—flea 3: .ouswdm NH .moH .red pouch-[g nooopoon JO .reqmnn 32 Population peaks were evident when observation records for all of the Pigeon River Area were considered (Figures 5 and 6). The 1951; curve contained records primarily from the two study areas while the 1955 curve was constructed with data from nine study areas mentioned above, plus records from Tin Shanty Bridge. This study area was not included in the east-west division since it would not fit into the separation. For both years population peaks occurred about the same date, in can cases on the exact date. Such a fluctuation could mean that migra- tory flights, even though small in numbers, did occur in the Pigeon River Area and that they came about the same date each year. In addi- tion woodcock were present most of the hunting season with the greater part of the migratory population passing through by mid-season. A total of 169 woodcock were seen in the Pigeon River study areas in 1951;, and of these, 113 were seen prior to October 20. In 1955, 219 woodcock were seen and 163 of these were seen before October 20. Most of the pOpulation in the latter half of the hunting season was repre- sented by one last wave of migrants near the season's end. Three major difficulties were involved in analyzing data from the Pigeon River Areas. First, low counts on certain dates could have been a result of less time in the field rather than actual pOpulation fluctuations. For instance, on October 16, 1951;, only l/h hour was spent in the field. No woodcock were seen during that brief period but several could have been on the stuck area visited or at some other Sites 33 monsoeoz nonopoo e s N. Hm N N N N HN H H ....H H HH m N m H _ _ _ _ . _ _ H .1 l - _ _ _ ...an .. cheeses 3355 .3 eeeseHssH no so: .338 doe-mam one. 5 ngavmsvosfiw coavmgaom .n eeBMHn anon .rad pouch-[g xeocpoon go .19qu In 3 noaegoz nonovoo e N an N N N N N 9.. NH ma ma 5 m N. m m _l q a ‘l _ J _ q _ _ _ ~ _ u 6an .. encodes geese an 85.53% mm no.2 sober comma..." one 5.. escapmnuocam coincide—mom .0 otdmwh .moH red peusum accepoon JO requmN 35 Another factor was that the number of areas checked varied from day to day. Thus the Pigeon River pOpulation may not have fluctuated so consistently if a more systematic method of checking each area had been used. The third point is that marv of the areas often were hunted by other persons. Therefore, observations often were made on a study area after some hunter had passed through. The third method used for studying population fluctuations in- volved the assistance of observers located at various points in the Upper Peninsula and at Mackinaw City. These observers were instructed to check special areas several days a week. Some observers spent six days a week on study sites while others averaged every other day. Records were kept of the number of hours spent in the area looking for woodcock, number of woodcock seen, and weather conditions. The purpose of this method was not only to determine migration times and routes but also to determine if woodcock were moving across Lake Superior from Canada. In 1951:, observations were made at sites near Dunbar, Marquette, Eagle River, Crystal Falls, Escanaba, and Moran in the Upper Peninsula and Mackinaw City in Region II (Plate 2). One area near Marquette and three near Eagle River were chosen with the idea that if any woodcock crossed Lake Superior the population in those areas would show increases. Two sites near Dunbar were used to observe papulation fluctuations caused by birds moving across the St. Marys River from Canada. The remaining three areas, located in the southern part of the Upper Peninsula, were used to observe fluctuations caused by woodcock moving 36 out and others moving in from.more northern parts of the region. The single Region II area at Mackinaw City was selected to observe movements in that area in relation to the other areas. Since Mackinaw City is located on the south side of the Mackinac Straits where two land masses are separated by a narrow stretch of water (five miles} the area was considered a good place to check woodcock:movements. In 1955, observation areas were set up in the general locations used the previous fall. However, the study areas near Dunbar were moved closer to Sault Ste. Marie and a new location was chosen near Paradise. Two observation areas each were set up near Sault Ste. Marie, Marquette, Crystal Falls, Escanaba, and Mackinaw City while three such areas each were set up at Paradise, Eagle River, and.Moran. Only a few birds were seen at any one time in the northern Upper Peninsula observation areas. Small increases in number of woodcock seen from one day to another could have been a result of birds moving across Lake Superior but more likely the increases represented wood- cock moving in from.adjacent areas. Mac Frimodig, Manager of Fort ‘Wilkins State Park in Keweenaw County, reported that he normally flushed five or six woodcock near the park during the course of a hunting season, but that in 19Sh.no woodcock were flushed by him.and five other hunters hunting in the area. Definite records exist of woodcock.crossing Lake Superior from the direction of Isle Royals. Davis J. Smith, a worker on a Great Lakes ore boat, found two dead woodcock on the boat during the latter part of April, 1950. The boat was about halfdway between Isle Royale and the Keweenaw Peninsula. Two woodcock landed on the boat in 37 mid-October, 1951, near the same location. These woodcock could have been residents on the island since breeding birds have been found there. Other woodcock possibly moved into the Upper Peninsula from Canada, but it seems doubtful that the numbers were great. Data gathered on study areas near Sault Ste. Marie did not indi- cate that birds were moving in. from Canada. But since there was a large breeding population in the Algoma District of Ontario in 1955: many woodcock probably entered Michigan from that area.‘ These wood- cock probably accounted for the large papulations often recorded near the Straits of Mackinac. Population peaks were evident at Moran and Mackinaw City probably because this area served as a funnel and received migrants from parts of the Upper Peninsula farther north and from Canada. Increases in woodcock numbers occurred about the same time for both places, indi- cating that woodcock were coming into the Straits area in large num- bers and dispersing rather generally. Data gathered from observation areas located in the southern part of Region I could not be correlated with the northern part, except that occasionally a decrease in the woodcock pOpulation on the Marquette areas was followed by an increase in the papulation on study areas near Escanaba. The length of time when woodcock were present was greater in 1955 than in 1951;. Except for one large peak in the Mackinac Straits area between October 26 and 29 most birds deserted the observation areas by October 20 in 1951;. Birds were observed on the last check- ing date on many areas in 1955, but most woodcock were seen before OCtOber 200 38 Records from woodcock hunters were collected and analyzed to determine population fluctuations. Two sets of records are repre- sented. One set includes information received from hunters who at times hunted only woodcock, while the second set came from hunters who were requested to check a favorite hunting spot at least three times a week. Hunters who gathered the‘first set of records were part of a group of sportsmen chosen to keep records of all their game- bird hunting activities, but were requested to keep separate the hours which were spent in woodcock cover. Few hunters looked for woodcock or kept. separate records; hence, the area covered and the number of hours represented were not very large. Figures 7 and 8 show results obtained for the three regions. There is little correlation between the different regions in regard to number of birds and time of flights. In 19511 some woodcock were still present in Region II at the end of hunting season. There was little day-to-day fluctuation for the region even though extreme local population fluctuations occurred. The 1955 data for Region II showed much the same variation except that the papulation definitely decreased after October 21;. The fluctuations could have resulted from a lack of data on certain dates, or from local variations. In Region I, the data indicated that woodcock population fluctu- ations were prominent; however, the peak on October 15, 1955, was the result of one hunter's observations. Only 145 minutes were spent in one location near. the Straits. Data from two hunters accounted for the peak on October 21, 1955. Thus the records indicated local more 39 amnss>oz honopoo mm: mammmmwmmmmammanaSandwmumma _J T. . .11. _ _ . . fl 4 _ _ 0 Q H :oawom 0H .0 HHnoamom .m HHH cosmos 0H :3an I shoemaaooo nouns: Moooeooz 89C menace.“ he confluence no 3033905..“ cognac..." .N. 0.33m .moH .red peqsntg aooopoom JO .raqumn hO honsoboz noDOpoo e e m an mm mm. mm mm am as 3 ma. 2 d a n a ‘11.. _ q _ _ _ . _ _ _ a _ _ _ _ \ 2 . \ Eugene \L n / /\~ I4. 1 H cosmos / \ .man I mnovmhoaooo amazon xeooeoosysohm meaoooa ho_eoodsuovoe as mnoavsoposam ecoapanmom .m anomah O H N H anon Jed peqsntg xeoopoon JO Jeqmnn than regional population fluctuations. There was a high woodcock pOpulation in many locations on opening day of hunting season in Region III. This was noticeable on the southern game areas where much of the hunting took place. The popu- lation curve (Figure 7) for l95h showed the highest peak at the begin- ning of the season. On opening day, one hunter reported 20 woodcock flushed in li'hours in the Gratiot—Saginaw Game Area. This area had a high breeding papulation in 19Sh and possibly the birds had concen- trated in the site which was hunted; however, only a few splashings were Observed, indicating that most birds had just moved into the area. Hunter reports in 1955 indicated a high population near the open- ing day. In the Deford Game Area, 36 woodcock were killed for 719.5 hours in the field. This population was not large for the amount of hours recorded; however, hunters were after all types of small game. At no other time during the season was a comparable ratio obtained. 'Wbodcock were present in Region III throughout the season. On October 25, one hunter flushed 11. woodcock in 3 1/1; hours in the Stan- ton Game Area. On November 9, 16 woodcock were flushed in 2 hours from approximately three acres in the Waterloo Game Area. When records from.the hunters who were checking favorite spots were analyzed, definite trends in the woodcock population were noted (Figure 9). The 195h reports showed.major fluctuations occurring dur- ing the first half of the season. Also, the trend of the population size was downward from.the first part of the season. In 1955. the reports indicated the main fluctuations occurred about mid-season. 142 In a further attempt to determine if hunting records could be used as an indication of woodcock migration, reports received from game bird hunters were analysed. The report forms, which included places to record counties, dates, different species seen, and species bagged, had been mailed to approximately 100 hunters prior to the 1951; and 1955 fall seasons. Population fluctuations were evident for both years in Regions I and II. These records showed that the number of woodcock flushed per hour remained conqaaratively high throughout the season in both regions, particularly Region II where most of the data were collected. Two major peaks occurred in Region I and four in Region II (Figure 10). Population peaks were more evident in data gathered for 1955. Three major peaks were noticed in the Upper Peninsula, while several papulation highs occurred in Region II (Figure 11). The number of woodcock flushed per hour varied between one and two birds most of the season indicating a rather constant woodcock pepulation for the region. There was an alternating population fluctuation between the two regions in 1955; that is, a pOpulation peak in Region I usually pre- ceded a peak in Region II. The similarity was less pronounced for the 195k data. A final criterion for studying fall migration concerned band re- turns. Different methods were involved in capturing woodcock (Table 3). Most woodcock caught during the spring were taken by a singing- ground decoy trap patterned after one used in Massachusetts by Sheldon (1955). This trap was set for both evening and morning flights. Normally, two traps were set on the same singing ground 3 h monsouwoz nonovoo m w e m an mm mm mm mm am ma 3 ma ma Q m N. .m m a 1 _ 4 _ _ _ _: 4 a a _ _ _ _ moss _ a _ a _ a __ _ _ _ _ a _ n e _ some .3.on modems: 03.85.“ wcaxoono 3355 xoooeoos Ho manage." Bonn congruence mm weedpmsposam coapmddaom . m ousmah .mon .red peqsnm xooopoon 3:0 .19qu noeeouroz nonopoo w a N Am mm mm mm mm Hm ma ma ma mm .3 m N. m H ql_.__iqi_____.7o~ 7 \o > \I 7/ \ \ \ / o . .. H coamom \ I L HH coamom .emma n advantage nouns: than seam mo apnoea." no.8 ponds—Asset mm uncapssveoam gavegaom .oa madman .moH .red peqsnIg xooopoom JO .1qu}; 1:5 Table 3 . ho Number of woodcock caught by various methods used in banding Operations. Methods Year Decoy Trap Hand Net Chicks Wist Net Clover-leaf ITrap Totals 1953 195h 1955 Totals Table )4. Band No. 3 2 h In 9 3 h 16 h6 65 20' 2h Banding and recovery records showing the extent of travel and the tendency of some resident woodcock to remain late in the fall. Band Data Records . B ' cation ~ Date Returns Location 513-96901 553-33002 513-96910 523-3350? '523-33515 523-33519 523-335h7 523-335h9 5/10/ 53 u/28/55 5/20/55 7/ 2/55 7/29/55 8/ u/ss 9/ 5/55 9/28/55 5 mi. N. Alba, 2/19/55 Antrm CO. Pigeon River Exp. 10/21/55 Sta., Otsego Co. Rose Lake Wild. 11/ 7/55 Exp.Sta., Clinton Co. Tin Bridge (Pigeon 10/18/55 R.), Cheboygan Co. Old.Grant School, 10/ 2/55 Cheboygan Co. W. McMaster Creek, 11/ 2/55 Cheboygan Co. Elk Hill, 10/ 3/55 Otsego Co. Elk Hill, 10/ 5/55 Otsego Co. Barnesville , Ga. Pigeon River Exp. Station Rose Lake Exp. Station Elk Hill, Otsego CO. Old Grant School, Cheboygan Co. Old 000 Camp, Cheboygan Co. Elk Hill Tin Bridge, Cheboygan Co. h? (Plate 5). In the evenings, a painted plastic decoy (Plate 6) was set on one trap while the other remained blind (without a decoy). For morning flights both traps were left blind. The blind traps set for morning flights proved most successful, especially those traps which had held a decoy on the preceding evening. Perhaps woodcock were caught more often on these blind traps because the performing male "remembered" the decoy being there. It can not be said that the traps were successful because they were set on the main performing spots; on the contrary, they usually were set to one side. Apparently more woodcock were caught in the morning since the traps seemed better hid- den when activity was beginning at a period of decreasing darkness. Another method of catching breeding birds was the mist not. At first, regular Japanese mist nets of one-inch square mesh were used but these were unsuccessful. Eventually, a not was constructed which was successful for the few times used on singing grounds. This not was nylon gill netting of zit-to three -inch stretched mesh constructed as a mist net and dyed black. To capture a woodcock the net was stretched vertically along the edge of a singing ground on the wind- ward side. A long-handled net (lZ-foot bamboo pole) and a six-volt headlight were used several times but their success was limited. Difficulties involved were too much dry vegetation on the "grounds" and the excit- ability and nervousness of birds during their breeding season. Several chicks were caught and handed. Most of these were found with the aid of bird dogs, but some were found by systematic searches near singing grounds. 1:8 PLATE 5 A. ‘Itto decoy traps set on one singing ground at the Rose Lake Experiment Station. Both traps were set blind for the morn-- ing flight and each trap caught a woodcock. B. A close-up of one of the above decoy traps showing the male woodcock which was caught. 119 PLATE 6 A. A singing ground in the Gratiot-Saginaw Game Area on which a woodcock (513-96916) was active three successive years. The bird was attracted once by the plastic decoy shown in the lower picture. B. A plastic decoy set on a camouflaged trap in the Gratiot- Saginaw Game Area. 50 During the summer and early fall woodcock nets were used again. The nets were stretched across back trails and streams where birds had been seen flying at dusk (Plate 7). Since the woodcock only flew about 10 to 15 minutes in the evening, nets were erected prior to that time and removed immediately afterwards. Another method used for taking woodcock in the summer was a clover-leaf trap (Plate 8). Trap specifications were received from Stephen Liscinsky of the Pennsylvania Game Commission. To use this trap feeding or diurnal concentration areas had to be located. Once an area was found, the trap was set and either two or three leads were erected (Plate 9). Traps were checked each morning and occasionally in late afternoon. 1 Band returns indicated that some woodcock stayed most of the fall season (Table h). Since information presented in previous graphs showed that woodcock p0pulations were fluctuating most of the seasonJ it is possible that some or most of the resident birds in the Pigeon River and Rose Lake areas stayed in the vicinity throughout the sea- son while migrants dr0pped in for a short-stay bef0re moving south- ward. Mendall and Aldous (l9h3) observed many resident birds present during the latter part of the migration period in Maine even though migrants were still passing through. Of the eight banded woodcock which were recovered in the past three years, seven were shot near the banding site. The other wood- cock was taken in Georgia. There have been a number of woodcock banded in Louisiana and shot in Michigan as described previously (Plate 3). Assuming that woodcock return to the same wintering 51 PLATE 7 A. The Tin Bridge Road crossing the Tin Bridge of the Pigeon River. Evening flights of woodcock were seen at this site almost every evening during the summer and fall. Nets were stretched across the road and the river. B. Two woodcock caught in a net stretched across the Pigeon ‘ River at the Tin Bridge. 52 PLATE 8 A. A clover-leaf trap showing the relation of the leads to the trap. Notice the woodcock to the left of the far entrance. This trap was set at West MCMaster Creek. B. A close-up of the clover-leaf trap showing an entrance, the gathering cage on the right, and a woodcock. 2. Nu. It :“:x 3‘ 53 PLATE 9 A. A fifty-foot wire lead going to a clover-leaf trap located near Elk Hill in the Pigeon River Area. B. A close-up of a clover-leaf trap showing the gathering cage connected to the trap. This trap was located in the mMaster Creek study area of the Pigeon River Area. grounds each year, the fall path of migration between Michigan and Louisiana would be similar to the one described for the spring migration. Conclusions reggding fall movements. Most of the fall movement of woodcock in Regions I and II took place prior to October 20, al- though large numbers frequently were observed through the first week of November. A few woodcock remained even longer. This fall movement appeared to start as early as September 1 in some localities, and by October 1 many woodcock had left their owner range. By October 20, migrants were present in Region III. However, band returns indicated that some woodcock in Regions II and III stayed in the same area much of the fall. season. Therefore, the papulation fluctuations occurring in many areas probably were a result of migrants coming in from more northern locales, staying awhile, and then moving on south. Birds in the northern part of the breeding range possibly were. stimulated to move earlier because of weather and decreasing light conditions. The characteristics of the movements were quite varied. Some woodcock were observed in what appeared to be southward flights in the twilight hours of morning and evening, and at night prior to 11:00 p.m. Although an occasional large "flight" of twentyor more woodcock was seen in an area, the birds generally moved through in small nunbers. In some locations woodcock could be found every day during the hunting season while in other spots woodcock were found only once or twice. The papulation of birds in some coverts remained at a peak for as long as a week while other coverts "held" birds for only a day. SS liner Even though woodcock depend primarily on areas of soft, moist soil where they can probe for earthworms, some birds may be found in Michigan long after most of these areas are frozen. These woodcock are usually found in sites where spring waters keep the ground soft. Records indicated a number of woodcock have been seen in Michigan during the winter. Gibbs (1885) reported that woodcock were taken in all months in Kalamazoo County. According to Cook (1893), R. H. 'Wblcott found a woodcock on January 1, 1888, at Grand Rapids when there was a.foot of snow present. F. F. Tubbs, of the Department of Conservation, observed a woodcocleecember l, l9h0, along a spring- fed creek off Harrison Road near Lansing. On January 10, 1952, Bert Stoll, an outdoor writer, took a picture of two woodcock feeding in the bottom.of a drained trout pond at the Cedarbrook Trout Farm.near Harrisville (Alcona County). This area contained running spring water. The most recent record of wintering woodcock was reported.by Martin Pollok (Michigan Department of Conservation), who observed two wood- cock several times during December, 1955, along a stream in southern Ingham.County. Since most, if not all, of these woodcock were in good condition it is doubtful that the birds remained.because of an inability to travel. No reason can be given at this time for their apparent lack of the migratory urge so strongly characteristic of woodcock. 56 POPULATION STUDIES Singing-Ground Activities The singing ground serves as the center of a woodcock's breed- ing activity. Because such a definite site exists and since the male bird visits this site rather regularly most of the spring, the sing- ing ground has become the basis for the spring breeding counts. These counts are used purely as an index to p0pulation fluctuations from year to year. Factors Affectianoodcock Activity. - Activity on the singing grounds begins shortly after woodcock arrive in the spring. Some birds may even begin courtship displays before leaving the wintering grounds or while they are on the north- ward trip. In 19514, however, the first woodcock were reported for the Rose Lake Area on March 9, but the first sign of display on a singing ground did not occur until March 13. The main portion of the breeding activity occurred between April 15 and May 15 in the Lower Peninsula and between April 25 and May 25 in the Upper Peninsula. The time when most of the woodcock activity on singing grounds was terminated indifferent parts of Michigan in 195).; and 1955 is shown in Tables 5 and 6. Some records were obtained regarding woodcock courtship displays in the fall, but no evidence was presented to indicate actual attempts S7 at breeding. Two factors which are difficult to separate and discuss individ- ually are the time of day and light conditions; therefore, the two are combined. The evening performance usually began about 15 minutes after sundown, but varied.from.5 to 30 minutes depending on light conditions. When a clear sky prevailed birds began courtship displays about 20 to 30 minutes after sundown, but on a dark, heavily clouded day the ac- tivity started as early as 5 minutes after sunset. Variable starting times existed.for different percentages of cloud cover between the two extremes. On a hazy day when light conditions were affected, birds commenced activity slightly earlier than on a clear day. There was a correlation between the time Of day, light conditions, and commencement of woodcock.activity. As the season progressed the sun set later and thus the woodcock activity began later. A regular photographic exposure meter was not sensitive enough to record accu- rate readings Of conditions at that time of the day. Some readings taken by an exposure meter showed.from.two to five foot-candles of light apparent when evening activity started on the singing grounds. A much more sensitive meter was tried in Michigan, but details for Operation of the instrument had not been completely mastered and further tests are desirable. Morning activity began shortly before daybreak in almost com- plete darkness and continued for a similar length of time as the evening activity. The performance, however, seemed to be more vari- able in intensity and length. 58 Table 5. Date of last singingéground activity observed in various locations in Michigan in 195h. ’ Direction and Distance Route Locations from.Lansing (miles) Date* Kellogg Bird Sanctuary SW - h5 May 30 Rose Lake Exp. Station ENE - 10 June 3 GratiotsSaginaw Game Area N - hO June 1 Houghton Lake N - 105 June 5 Cusino Ekp. Station NNW - 260 June 20 Table 6. Date of last singing-ground activity observed in various locations in Michigan in 1955. Direction and Distance Route Locations from Lansing (miles) Date* Kellogg Bird Sanctuary SN - h5 May 30 Rose Lake Exp. Station ENE - 10 June 3 Gratiot-Saginaw Game Area N'- hO June 6 Greenwood (Ogemaw County) NNE --100 June 10 Pigeon River Area N - 165 June 16 Cusino Exp. Station NNU - 260 June 9-28** * The peak of activity passed long before this time. ** One woodcock was heard on June 9 but another count was not made until June 28. 59 Light produced by the moon Often affected the activity of perform- ing males. This effect was most pronounced on a bright moonlight night. When such conditions existed the courtship display became more. irregu- lar. The male bird started later, continued longer, displayed erratic periods of peenting and flying, or performed at intervals most of the night. The erratic display may have resulted because the bright light made birds feel less protected. Woodcock approached on such nights were more wary. When approached the bird Often ceased all sounds and flushed more quickly than on a dark night. Probably the bird could see the intruder more clearly, but even when the intruder remained hidden the woodcock was more wary of noises on a bright night. Such erratic display did not take place in the full moon period when a heavy cloud coverage existed. In Michigan, especially the Lansing area, the average cloud cover was 50 per cent or more for the spring months. Thus the effect of bright moonlight often was eliminated. Weather was one of the most, if not the most, important factor controlling activity on the singing ground. This does not mean all activity ceased under adverse weather conditions but it became more erratic and sometimes lessened depending on the situation. All wood- cock were not affected similarly. Weather data were collected from several sources and by several methods. General climatic conditions were reported by singing-ground cooperators, while more detailed information was taken from daily reports accumulated by the East Lansing Weather Bureau. In addition records were kept on time of day, per cent of cloud coverage, phase 60 of moon, wind velocity, amount and type of precipitation, and temper- ature, for the Gratiot-Saginaw Game Area and Rose Lake Experiment Station routes each time a count was made. Thermometers were placed near several singing grounds for the 1955 spring season. Readings were" taken at the time the cOunt was made at that stop. on the route. A recording thermograph and a mgrograph were set up on one singing- ground at Rose Lake. Temperatures were taken at ground level and at 10 feet with these two instruments and the thermometers. Temperature was one of the critical weather factors. Woodcock activity was more consistent and persistent on warm, clear nights. On cold nights the performance of singing males decreased consider- ably. However, this does not mean that the activity was directly proportional to all. temperature changes. Once the minimum temperature had been exceeded, subsequent increases did not insure increased num- bers of active males. This minimum temperature occurred near 35 degrees Fahrenheit, although the Fish and Wildlife Service recommended that no singing ground counts should be made when the temperature fell below 25 degrees. On three different routes run in 1953 and one in 19514, the number of birds heard per count was below the mean when- ever the temperature was below 35 degrees. In no instance was the count above the mean when the temperature was colder. Wind was another important factor affecting woodcock activity in the spring. The most consistent performance occurred on calm nights or when the wind was below 10 miles per hour, assuming other factors were suitable. When the wind velocity reached 13-18 miles per hour, activity became erratic. The male bird made fewer flights than normal 61 and spent more time on the ground peenting between flights. Alain and snow definitely influenced the activity Of singing males. On rainy nights the courtship display started earlier than on a dry, cloucb' night since darkness came earlier. Also, the activity became erratic and shortened, even when a light drizzle occurred. As the intensity Of the precipitation increased, activity seemed to decrease. Observations were made on some rainy nights when almost no singing grounds were occupied by performing males. If a rain occurred during the day but stopped before sunset woodcock seemed more active. Snow caused the same trend toward an erratic and shortened ac- tivity; however, the erratic tendency was less pronounced. On April 19, 1953, a singing ground count was made in the Gratiot-Saginaw Area during a snow storm (about three inches fell). The night was calm and the temmrature was 35 degrees. Seven woodcock were heard on the first half of the route but none on the last half where three to five birds usually were heard -- the mean count for the route was 11 birds. The seven woodcock were as active as on arw other night, but their enthusiasm ended quickly. Another incident of woodcock activity in a snowstorm was Observed in 1956. Again activity was quite normal but shortened. Relative humidity and barometric pressure seemed to show little correlation with the amount of breeding activity displayed by a male woodcock. Since these two factors were studied only slightly, further investigations might be warranted. Performing males were influenced occasionally by activities of other animals. Man and his activities exerted the greatest influence. 62 On one instance a singing ground was burned over and the male was forced to vacate the area for the evening display. Often, farming activity, continuing late into night and passing near or through a singing ground, caused the male to move. Occasionally, the movement of observers near a "ground" caused an erratic display. At other times and on the same or other singing grounds birds were not bothered by fanning practices or movements of observers near the "ground." Nearby car traffic, low-flying planes, and building construction did not cause the male to stop performing. Other animals were not observed to affect the male woodcock's display to any extent. Stray dogs Often could be heard barking near- by, but the noise did not influence the bird's actions. If a dog chanced to run near the singing ground while the woodcock was peenting, the bird would either pause momentarily or take off into a flight dis- play. One night a fox was seen within 50 feet of a singing ground.but the animal's presence either was not noticed or was ignored by the bird. Often deer and elk walked through or near singing grounds in the Pigeon River Area; again the bird's reaction was either to ignore the animal if it was not too close, take off in a flight display, or sit quietly until the creature passed. Other birds, particularly song and game birds, did not produce any effect on a male woodcock and his courtship display. Factors Affecting Observations. There are several groups of factors which.may prevent the gather- ing of suitable data on a route. Of these, weather is very important. 63 However, only the wind is critical to the observer. Fewer woodcock usually are heard as the velocity increases beyond a certain point. The direction of the wind becomes important when the wind is blowing about 10 miles per hour; woodcock on the leeward side are difficult to hear. If the wind exceeds 15 miles per hour, it becomes increas- ingly difficult to hear woodcock frmm any direction. Low temperatures may be uncomfortable but cold does not affect the hearing, except psychologically. Also, it is uncomfortable to make a count in rain or snow but it would be unusual to make a count under such conditions. Snow does not affect the hearing ability of the Observer, but rain makes so much noise that an observer would not be able to hear all active woodcock. I Another group of factors would include noises of other animals, again including man. When a route is selected it should be in an area free of such interference, but it is seldom possible to escape every cause of distraction. Even on trails, another car often is encountered. Occasionally, if a route is set up as far as two miles from.a black-top road, the noise of vehicles, especially trucks, still may cause a certain amount of interference to the observer. If part or all of a route is along a paved road, passing cars make it impossi- ble to get an accurate count. / Other human or man-made disturbances include airplanes, tractors, construction, trains, motor boats, and playing children. Other*animals'which affected observations at varying intensities, depending on the number, distance, or vocal capacity of the animals, were frogs, dogs, birds, and insects. Stops near low wet areas were 6h omitted some nights due to the chorus of frogs. Noise, more than similarity of the sounds of some frogs to the peent of a woodcock, 'was the disturbing factor. Barking dogs running loose or at’nearby residences often caused.too much noise interference. Some birds, such as the whipdpoordwill,‘Wilson's snipe, killdeer, and great horned owl, were too active at some stops to allow gathering of accu- rate counts. 'Whip-poordwills, especially, were pests since they fre- quently landed near the car and continually called. Insects, particularly mosquitoes, made it uncomfortable to observe woodcock along some routes. The observer was unable to con- centrate on counting woodcock on such occasions. Another source of difficulty in obtaining a set of reliable observations is the individual variation between observers. Some people are capable of hearing woodcock performing farther away than others and obtain higher counts along the same route. The sounds made by the bird may not be loud enough or may be pitched in a.fre- quency range which some people can not hear. Carney (195h) found such a variation between observers making crowing-cock pheasant counts. Marked'Woodcock. Methods involved in banding woodcock were discussed in a previous section. In addition, in 1955, several woodcock were marked on a wing with a spray-glo paint. The paint was sprayed on the wings and allowed to dry before the bird was released. Different colors were used for each woodcock. This marking was done so that subsequent trappings to 65 identify birds would not be necessary. Banding operations, recaptures, and some marked-bird observations revealed several pertinent facts which should be considered in the analysis of singing-ground activities. First, many male woodcock returned to the same singing ground each year. Three woodcock were recaptured in 1955 at the exact spot where they were first taken the previous year (Table 7). In addition, some singing grounds which were occupied each spring were being used every night by the same male bird. One male was recaptured at a stop on the Gratiot-Saginaw Game Area route on April 16, 1955. One wing of the bird was sprayed with a bright red paint. Only this bird was observed using that site the remainder of the season. Often several males used one singing ground. On May 2, 1955, a male bird was caught on a singing ground at Rose Lake during the evening flight. The following morning another male was recaptured on the same singing ground. Another singing ground at Rose Lake seemed to be an "open-house" site. At least three different males used the site in 1951;. In 1955 one of then returned, and at least two others shared the site. Sheldon (1953), with a greater number of trapped birds, found many instances of one singing ground serving more than one male. One maleoften used more than one singing ground on the same or subsequent nights. In a number of instances a woodcock was seen landing at different spots, sometimes 300 feet apart, during the dis- play of one evening. On Ma}, 19513:. a woodcock was taken from a singing ground at Rose Lake. One wing of the bird was colored. Table 7. Rec apture data 66 Band Banding Recapture Distance Number Age Sex mate Localioni- Late Location-l- Bettieenfi 513-96916 A M 545-51. Gratiot- h-lo-ss Gratiot- o Saginaw #10 Saginaw #10 513-96915 H S-II-Sh Rose Lake 5— 3-55 Rose Lake 0 #11 #11 513-96923 H 5-30-51, Rose Lalo 5- 5-55 Rose taken .1; mi. . #3 5-13-55 Rose Lske#3 .1. mi.. 513-96913 u 5- 3-51. Rose Lake s-2u-5h Rose Lake .1, mi. #2 #1 553-38005 F 5- 5-55 Rose Lake 6- 2455 Rose Lake 1.0 mi. #3 ~ #20 553-38006 5- 5-55 Rose Lake 5-13-55 Rose Lake to yds. ' #3 #3 523-33507 M 7- 7-55 Tin Bridge 7-25-55 Tin Bridge 0 523-33518 M 8- 3-55 W.McMaster 8- 9-55 w.MeHaster CK.O Ck. 8-18-55 W.McMaster CK.O 523-33520 F 8-11-55 w.MeHsster 8-22-55 W.McMaster o ‘ Ck. Ck. 523—33526 H 8-18-55 Elk Hill 8-20-55 Elk Hill 0 523-33510! F 8-20-55 Elk Hill 8-26-55 Elk Hill . 0 8-28-55 Elk Hill 0 * The numbers represent the singing ground banded or near which the bird was banded. on which the bird was *1! The distance between the site of bending and the site of re- "0' indicates that the bird was retaken at the same capture . spot 0 " This bird had a broken bill resulting from an injury received The bill was mnded by the time of the first ' recapture and the bird was feeding at the time of the second recapture. 67 Three weeks later the same male was shined on a singing ground O.h mile away. Another woodcock was first taken on a singing ground at Rose Lake in l95h. In 1955 this bird was recaptured twice, once at another singing ground and again at the original site of capture (a distance of O.h.mile separates the two grounds). Banding operations by Sheldon (1953) showed that one woodcock was caught at one site on Tuesday and 3i'miles away on the following Thursday. But, the dis~ tance between first capture and recapture sites exceeded 0.5 mile for only 20 per cent of the repeats. Records also showed that some females return to the same breed- ing area. One female was taken in a blind decoy trap on May 2, 19559 on the Rose Lake Area. On May 8, 1956, this bird was taken in the same manner at the same spot. This singing ground was peculiar in that it was located in a tamarack swamp and was only six by ten feet in dimension. Recapture of a young woodcock gave an Opportunity to observe Juvenile birds. Three young birds, about ten days old, were caught and banded.on May S, 1955. Four weeks later one of them was aflaul in a mist net set on a singing ground one mile west. New measurements were taken and it was found that the weight of the bird had tripled. The young bird weighed more than most adult females in spring. The bill had grown from.3h to 65 millimeters and the bird was feathered completely. Activities such as those discussed above tend to make it imprac- tical to use data gathered on singing-ground counts for anything but an index to population trends. Singingzground Counts. The only method devised and used with any consistency to deter- mine breeding population trends has been singing-ground counts. Generally, this method involved setting up a route at least two miles long, establishing stOps at least 0.1; mile apart, running the route during the time of evening activity, and making two-minute interval counts of different woodcock heard at each stop. Complete instruc- tions for running a woodcock route as set forth by the U.S. Fish and Wildlife Service for spring, 1956, are included in the Appendix. A few changes in the instructions have been made since 1953; the major change was the suggestion to run two or more routes in an area once each‘instead of running one route two or three times. When such counts were begun in Michigan in 1951 only 12 routes were run. Gradually, a better coverage of woodcock habitat was obtained, and in 1955, at least so routes were in operation (Plate 10). Many routes have been run by non-department personnel, but the present policy is to depend more upon Game Division employees. It has been more difficult and expensivevto locate and train people not connected with the Conservation Department than members of the Depart- ment. The present group of cooperators should be retained, however, since they are reliable and familiar with procedures. Data which have been accumulated since 1953 were used to analyze fluctuations of spring populations. From these data an attempt was made to answer the following pertinent questions regarding singing- ground counts: (1) How many times should a route be run during the breeding season to obtain a reliable count of active woodcock along u— 69 that route? (2) How many routes should be run in Michigan to obtain a reasonable coverage of woodcock habitat? ( 3) What are the practical aspects or limitations of singing-ground counts? Earlier instructions by the U.S. Fish and Wildlife Service, based upon findings of Mendall and Aldous (19113), recommended that each route be run at least three times during the main part of the breeding season. The reason for three or more counts was to determine the num- ber of singing grounds used regularly and to obtain a mean figure. Subsequent investigations by workers in Pennsylvania (Norris, gt g. , l9hO) and Massachusetts (Sheldon, 1953) showed that the male woodcock was promiscuous and that it often used more than one singing ground. Consequently, repeated running of oneroute is not as important as was formerly believed. When Kosicky, Bancroft, and Homeyer (19514) made a statistical analysis of data obtained by cooperators over a five-year period in the Northeastern States they concluded there was little ad- vantage to running one route more than once. They recommended that a more complete stratification, based on habitat types, be investigated and that a minimum of two routes should be selected at random from each stratum. It was felt that it should be possible for the same expenditure of funds and time to increase the number of routes and decrease sampling error. Before the analysis of Michigan singing grounds was started data were examined to see if the singing ground counts were distributed in a normal fashion. A comparison of a frequency distribution, formed from results from S6 routes scattered through Michigan in 1955, with that expected with a normal distribution with the same mean and u— w 70 PLATE 10 Locati on of sin sing ground routes wh ich were run in 195 S. —Efl¥— '0” )0 \°fl ' a" so .u x E ' I m . DOMINION or CANADA t'}"\ ALG! _ 0"\.. [m 0 scnoo n . I? . LC A8 Cfllflhfg S 4 o '. .. . . (am 7 v "" c 9 000m }.-’0' ( 3‘ $ I“ I] 3 cmuvon msout mt 6‘ _. .' o I a ’x/ 90 3 ”m. orsmo' ALPEIA ‘I ."§ mum O .’ ~ / V ( " mm °""° ““‘W "Wow oscooa “(on / b nuns: 1 i ‘ = .' a mu‘ wurono hlsnuumnscoum oeum taste a ,/ o o l \ O Aunt. ‘ : unso- uu ascrou cunt so" 0’ z / * 0 .f ..a. -' O o in 52° 00 I ocuu wtcosn m nous : «mo 0 ‘ VMCM‘ SMILE I O “e-Ellw Guano 5‘0"" «1 e W" mum um- 5m! .' Ion-A (muo- mt: cum A A: 'mm mm mow mom imam ° "‘ o “If "an MLAIIAZOd. (HwOuw mum usmuu mm 0 CANADA 3: 0 WM" ransom luau eon-o: ‘. lost». I.,- LAX: I" o . [k N A {.-..---.H—..- .-. ERIE 71 standard deviation revealed no evidence of a statistical difference, using a chi-square test (Table 8 ). Therefore, no transformation was considered necessary in the statistical analyses presented in sub- sequent discussions. Information compiled from frequent runnings of routes on special study areas in the present study revealed that the estimated popu- lation fluctuated considerably during the breeding season. The popu- lation along the Gratiot-Saginaw Game Area route varied less than that along the Rose Lake Experiment Station route; however, the latter route was run more frequently (Figure 3). I Counts from the Rose Lake route were used to determine how many times a route had to be run to obtainvarious degrees of precision at the 95 per cent level of confidence. These sample sizes were computed using the formula (1): n = t2 82/(d i )2 (Snedecor, l9h6). Only counts were considered that were takenon nights which fell within prescribed limitations set forth by the Fish and Wildlife Service instructions for running routes. According to these data, which had a standard deviation of 2.2 and a mean of 7.2 calls, the route ana- lyzed would need to be run 27 times to obtain confidence limits ex- tending either side of the observed mean for 25 per cent of its value. The same route would have to be run 170 times to obtain confidence limits 10 per cent of the observed mean. The number of times a route should be run were investigated further in light of sampling theory, using data gathered from routes scattered through Michigan. An analysis of variance of counts ob- tained from routes run at least-twice for each of the three years Table 8. Data used in a chi-square test to determine if woodcock counts were distributed normally. Average Number of Woodcock Heard per Stop Number of Routes on Which Counts Were Made .00- . 195 .395 .595 .795 .995 1.195 1.395 1.595 1.795 1.995 2.195 2.395 10 10 l—‘U‘lw 72 73 (1953, 1951;, and 1955) yielded a pooled estimate of variance (Table 9). Table 9. Pooled estimate of variance for singing-ground counts (1953-1955). Source of Degrees of Sum of Mean Variation Freedom Squares Squares Among routes 91 67.1 .737 Within routes 170 11.2 .066 Since the number of times a count was made varied with the route, an average number was needed to estimate the various components. This average was obtained by use of the formula from Snedecor (l9h6): k0 . l/n-l (Sk - Skz/Sk) . (2) where, to - average number k a number of times a route was run n 3 number of routes run The average number was determined to be 2.8. With the aid of this average 5 and the mean square (32) of the pooled estimate (Table 9) figures in Tables 10 and 1.1 were compiled. The formula, 32%:1/11 (8123+ aft/k) (3) used in the compilation of data for Table 10 is a restatement of formula 10.3 (Cochran, 1953) which was used to predict the variance expected for other subsampling ratios. In the formula as it is used here: 2 s‘f :- variance of the mean n a! number of routes to be run, or sample size Table 10. Variance of the mean (8%) showing the relationship between the sampling (n) and subsampling (1:) sizes. Routes run: n Onflime Two Times Wee m l .306 .273 .262 2 .153 .137 .131 h .077 .068 .066 8 .038 .03h .033 16 .019 .017 .016' 32 .010 .009 .008 Table 11. Number of singing ground routes required for a given precision. Routes run: Precision* Che Time Two—Times m .50 .6 .5 .5 .hO .9 .8 .8 .30 1.6 1.11 1.1; .25 2.3 2.1 2.0 .20 3.6 3.2 3.1 .15 6.1. 5.7 5.5 .10 has 12.9 12.). .05 58.0 51.6 h9.6 * or confidence semi-interval as proportion of the mean (.92) 75 k a number of times each route is to be run, or subsample size 3: a variance between routes (here, .210) a: a variance within routes (here, .066). Data in both tables reveal there is little benefit gained by running a route more than once. These results substantiate the con- clusion drawn by Kozicky, gt 2‘]: (1951;) that there seems to be little gain in precision from running the same route two or more times in the same year. From the practical or financial view it may be more economical to run a route twice during the first season in terms of information per dollar spent. This conclusion is, reached by examining the present set of data according to the methods presented by Cochran (1953) for determining the Optimum number of items in a subsample for surveys where, as in the present study, a subsampling method is used. The optimum number is kOpt ‘ sw/sb )cu/ce (h) k t = Optimum number of times a route should be run Where, standard deviation within routes 1" standard deviation between routes 6" . cost for locating and setting up a route __ cost for running a route after it is set In: 0 I The required information on costs may be derived from records kept during the present study and estimations based on experience. The experience represented in this study may be illustrated by 76 mentioning the efforts involved in establishing new routes. Since only 13 singing ground routes were run in Michigan the first year of the study a better coverage of the state was desired. To get this increase approximately 100 people were contacted by letter or in person. Those persons who expressed interest were again contacted.by a biologist. Tb determine the optimum number of times a route should be run, considering economic aspects, the costs involved in setting up a route as compared to the cost of running a route after it was set up were considered. In determining the average cost of setting up a,route. the following points were included for 25 routes: correspondence, roundétrip mileage from Lansing to the area, mileage involved in locating a route, mileage involved in traveling to and from the route and running the route in a preliminary way, number of hours involved, and room.and board. No expense was attributed to non-department personnel since the time of such peeple was donated. The hours spent by department personnel were not included since in most instances their time was not taken from.the regular working day. Considering these various aspects the average cost of setting up a route was com- puted as $25.25, a minimum figure. Tb run these routes after they were set up cost comparatively little. In determining the average cost of running 25 routes the major consideration was the local mileage to and from the census area and mileage of the route. Again the expense of non-department person- nel was not included, but the mileage of department personnel was since they would be reimbursed for use of their cars (or would use state- 77 owned vehicles). Time used by department personnel was not figured, as discussed above. A certain per cent of loss had to be considered for those routes which were set up and for some reason were not con- tinued by the cooperators. Thus the average cost for running these routes was $2.10, again probably a minimum figure. Substituting the proper values in formula 1;, the optimum k be- comes 1.82 or 2. Thus in early stages of accumulating a number of routes it is more practical to run each route twice. However, the second year there is no cost associated with setting up an already established route, and it should therefore be run only once. Through- out the years the number of new routes will become smaller. In addi- tion, experienced observers may set up additional routes without costly outside aid, and more observers will be located throughout the state, available to assist new personnel thus reducing the cost'of new routes. As a working rule, therefore, it is suggested that new routes be run twice the first year and only once in each year there- after, increasing the number of routes as the budget permits. Reviewing the field emerience and data in 1955 when 56 routes were run from one to three times each, confidence limits of the mean value of 392 birds reard per stop were computed as being between 1.01; and .80, or about 13 per cent of the mean. The mean value for each route was used above regardless of the number of times run. As shown by data in the above tables and by subsequent discussions, there is a practical limit to the number of routes to be run. Once a certain sample size has been reached the precision increases rela- tively little for additional routes. Thus it may be financially 78 practical to reach and maintain a certain sample size, realizing that stratification, as suggested by Kozicky, 93 91. (1951.) might further reduce sampling error. Nesting A study of nesting habits and success was a minor phase of the investigations. Such information was gathered incidentally to other phases of activity. The discussion represents an accumulation of data gathered from personal observations, use of bird dogs, cooper- ator reports, and Game Division files. Reports of only 31 nests were available. The clutch-size for those nests ranged from two to six eggs with an average of 3.9. Mendall and Aldous (l9h3) found an average clutch in Maine to contain 3.9h eggs. Of those Michigan nests known to have produced.youhg every egg hatched. Knight (19h6) mentioned that few infertile eggs are found. in woodcock nests and that normally there is a 95 per cent success. Mendall and Aldous (191:3) found the fertility to be 98.1; per cent. 1 Seven of the Michigan nests were destroyed later (Plate 11). Man and his activities were the cause of destruction for three nests. Two nests were lost to mammalian predation (red fox or skunk); the remaining two nests were deserted. One of the latter may have been due to adverse weather conditions since the nest was found when snow was still present. Many nests built soon after birds arrive in Michi- gan probably are unsuccessful because of adverse weather. 79 The length of the nesting period is quite long in Michigan since about hOO miles separate the southernmost part of the state from the northern edge. One of the earliest nest records was reported near Grosse Points by G. A. Abbott on March 27, 1918 (Wood, 1951). On March 31, 1951;, Walter Iemein, in charge of the Kellogg Forest near Battle Creek, found a woodcock nest, containing two eggs, surrounded by snow. The latest nesting recorded was June 11, 1936, near Ann Arbor reported by G. A. Mann, (Wood, 1951). On June 9, 1955, a woodcock nest containing six eggs was located in Roscommon County. These records give some idea of the length of nesting season. The major nesting effort during the present study occurred dur- ing the last two weeks of April and the first two weeks of May. The period was somewhat later in Region I and the northern part of Region II. In Maine , Mendall and Aldous (191:3) found that most woodcock began nesting about the third week of April and that most young emerged from eggs between My 11 and 25. They also concluded that ‘ toward the northern limits of the breeding range most nests are con- structed during the latter part of April and the first part of May. Broods Observations of broods also were limited during the study. Enphasis was placed on finding young birds in order to band them. Notes .... taken on habitat types, which will be discussed later, and certain life history aspects. Additional observations regarding broods were made by various Department of Conservation field person- nel, and other data was taken from Game Division files. i 8O PLATE 11 A. Site of a woodcock nest beneath overhanging branches of a tamarack. The nest location is indicated by the arrow. B. A close-up of the above-mentioned.nest showing the results of mammalian predation. 81 Brood searches were made in wooded areas near singing grounds. Bird dogs were used much of the time but some broods were found with- out their aid. The size of 28 complete broods varied from 2 to8 with the average size being 3.6 young per brood. The brood with the 8 young was ob- served by F. M. Baumgartner (1938), then with the Michigan Department of Conservation. He suggested that this brood may have been the result of two females laying in one nest or one female taking over two broods. Since it is most unlikely that one woodcock would lay and could incu- bate 8 eggs, this brood was not included in the determination of the average brood size. There were four other instances when the brood sise exceeded four. One brood of five young was found and three in- stances of six young per brood were observed. A division of the broods into two age classes ("less than 7 days" and "7 dqs plus") showed that for the younger age class the average brood size was 3.6 and for the older age class the average was 3.1; young. In Michigan, apparently, most young are hatched during Key. Of the 31 broods, 17 were hatched in May and 16 of these emerged from the eggs during the first three weeks of the month. The earliest hatching record was April 21;, 1955, near Bad Axe. Seven broods came off after June 1, and one of these hatched about July 6 in Alger County accord- ing to Department of Conservation Biologist Carl Dubovsky. Most of the June reports were probably of renesting attempts. Because of their rapid development, young woodcock from renest- ings are probably not much handicapped by their late birth. A young 82 woodcock grows so rapidly that within a month it can make long flights. One banded Juvenile in Michigan had developed so much in a month that it would have been mistaken for an adult if it had not been ringed. Pettingill (1936) and Mendall and Aldous (19h3) agreed that by 25 days of age the young bird is practically "full grown.” Sex and Age Statistics An accumulation of data from specimens during the past three hunt- ing seasons has provided considerable information concerning the sex and age composition of fall woodcock papulations. Most specimens were collected in the Pigeon River Area, particularly during l9Sh-55. Others were received from various hunters throughout the state. Depart- ment of Conservation employees were especially helpful. Information from these specimens included date, location where killed, cover, type, sex, age, weight, various measurements of the bill and the outer three primaries, notes pertaining to molting patterns, kinds of exter- nal parasites and diseases if any, and miscellaneous notations. Castro-intestinal tracts were saved for study of food habits and in- ternal parasites. g and Age Determination. Techniques for sex and age determination involved external measurements. The internal method for sexing was primarily an exami- nation of carcasses for the presence of testes or an ovary. If the gonads had been destroyed a search was made for an oviduct. Occasion- ally, the latter method was successful in determining the sex. 83 The presence or absence of the bursa of Fabricius was the main criterion for distinguishing between adults and birds or the year. In addition, the width of the oviduct in females was greater in old birds than in young as determined by visual examination. A determination of useful external criteria for sexing and aging woodcock was one objective of the study. Several criteria have been developed.and used by other workers. Pettingill (1936) observed that the female woodcock'usually was larger in.most measurements, including weight. Tufts (l9hO) indicated that bill length.was a.more reliable criterion for sexing than weights but that such measurements were not conclusive. Later Mendall and Aldous (l9h3) studied the relationships of body weight, wing length, and bill length to sex. They concluded that too:much overlap occurred with weights and wing lengths to use either of these criteria, but that bill length was quite reliable. In 1953, Greeley's studies on fall-shot woodcock were published. In ad- dition to repeating some measurements previously tried by others, he used the width of the outerqmost primary (number 10) at a point two centimeters from.the tip. ‘With this measurement he reduced the over- lap error considerably. Harris (1953) proposed the use of the bill length with the combined widths (two centimeters from the tip) of the outer three primaries as a.method.for sexing fall-shot woodcock in Minnesota. In examining Michigan specimens in the present study (80 in l953, 1h? in l9Sh, and 189 in 1955) emphasis was placed upon the measuring of weights, bill length, and widths and lengths of the outer three 8h primaries. The woodcock were weighed as soon as possible after being killed. Measurements were made to the nearest gram on a dietetic scale. Six birds were weighed immediately upon being shot and at various temperatures and intervals of time up to 21; hours with weight losses varying only between zero and two grams. Linear measurements were made to the nearest tenth of a millimeter with a Vernier caliper. Bill length was measured from the base of the nares to the tip of the bill as suggested by Greeley (1953). Measurements were made also from the anterior part of the nares to the tip in 1953, but this measurement was discontinued when a statistical "t" test showed no evidence of a variable difference between the two measurements. Four different measurements were made on each of the three outer primaries of the specimens examined in 1953. One of these, the width, was made two centimeters from the distal end of each primary. Both wings were used the first year but when a "t” test showed no evidence of a difference between the primary measurements of the two wings the procedure was changed. Subsequent measurements were made only on the primaries of the right wing unless the feathers were deformed, in which event the left wing was used. In addition to width, three measurements were made on the length of each primary. One length ("intact") was obtained with the primary in place. This measurement was made from the web of the wing to the distal end of the feather. Then the feather was removed from the wing and two more measurements made: the length of the vane and the total length of the primary. The total length measurements were easiest to make and data in Table 12 and 13 show those measurements to be as accurate as the others. were measured in subsequent years. 85 Therefore , only total lengths of primaries Table 12. Length data computed from primaries 8, 9, and 10 of female woodcock specimens collected in 1953. Intact Vane Total I n s2 s 'x' n s2 s x n 32 s 8 82.3 38 3.85 1.96 711.2 38 h. 78 2.19 95.5 38 14.92 2.21 9 79.1 37 11.146 2.11 73.1 37 14.55 2.13 91.2 37 5.75 2.hO 10 71106 38 h.09 2.02 7009 38 130,48 2.12 8,402 38 (401,4 2003 Sum 236.2 36 3h.8h 5.90 218.3 37 39323 6.26 271.0 37 37.09 6.09 Table 13. Length data computed from primaries 8, 9, and 10 of 29 male woodcock specimens collected in 1953. Intact Vane Tom i 32 s E 82 s E s2 s 8 7h.8 3.1.3 1.85 67.9 3.10 1.76 87.2 mm. 2.11 9 7201i 3021'. 1080 6705 3011 1076 8306 14.013 2.03 10 69.2 3.51 1.87 66.1 3.08 1.75 78.3 b.12 2.03 Sum 216.3 27.72 5.26 201.6 25.27 5.03 2119.1 311.77 5.90 Heights of woodcock were considered as a possible sexing method since female birds frequently were observed to be much heavier than males. During the three hunting seasons 3145 woodcock for which the sex 86 was known were weighed, and although the average weight of females exceeded that of males, the region of overlap was too great to allow practical use of weight as a sex criterion. Pettingill (1936), Tufts (l9h0), and.Mendall and Aldous (1983) all made the same observation and concluded that weight could not be used for sex differentiation with any success. Throughout this discussion "overlap" refers to the actual over» lap while "per cent of overlap" means the calculated per cent to be expected assuming normal distributions of the observed.means and standard deviations. Overlap refers to the proportion of mis-iden= tified animals. This is determined by first setting up a point of discrimination between the two sexes. Then the measurements which fall beyond this discrimination point represent the fraction of- individuals which can not be identified. An analysis of the weights of Michigan woodcock indicated the weight criterion to be less reliable than feather and bill measure- ments, as determined.by the much larger standard deviations and coefficient of variability of weights (Table 1h). Table 1h. Analysis of weight records of woodcock specimens collected during the hunting seasons (1953-1955). Number 'Weights* Sex of Birds Range Average 3 CV Female 155 162-252 207‘ 17.8 8.58 Male 1&8 129-201 16h 13.6 8.28 * In grame. 87 A.pr0gressive weight increase was exhibited by woodcock from spring through the fall season (Table 15). The weight increase dur- ing the fall is more evident than in other seasons. At the beginning Table 15. Progressive weight increases of woodcock in 1953. l95h, and 1955- Males Females Number Average ' Number Average Season of Birds we ht* of Birds weight* Spring 18 129 3 167 Summer 22 1h? 8 176 Fall 1&8 16h 155 207 *-In grams. of the hunting season no fat or very little was found on woodcock specimens. As the season progressed the amount of fat increased (Plate 12). This fat deposition probably accounted for the progres- sive weight increase noted in Michigan woodcock (Table 16). A simi- lar weight increase was noted for woodcock collected in Nova Scotia (Mendalland Aldous, 19h3). Table 16. Fall weight variations of woodcock collected in the Pigeon River Area (l95h-1955).* Males Females Number 'Weight Average Number 'Weight Average Periods of Birds Range ‘Weight of Birds Range weight Oct. 1-15 '55 13h-187 158 h9 162-232 196 Oct. 16-31 32 150-189 172 28 190-2h5 216 Nov. 1- 9 8 17h-201 183 8 216-2h6 231 *Vweights are in grams. 88 PLATE 12 Two pictures showing the distribution and amount of fat on woodcock late in the fall season. 89 Measurements considered to be most important and accurate as criteria of sex were bill length and total width and total length of the outer three primaries (Table 17). Table 17. Comparison of criteria of sex for fall-shot woodcock in Michigan (1953-1955). _~ Bill Length Primary Width Primary Length Male Female Male Female Male Female 1: 179 213 181 199 179 198 Range 56.h-6h.2 61.6-72.8 8.1-12.h 12.6-17.1; 232.7-262.9 257.h-29l.3 i 60.1 66.9 10.3 1h.6 218.9 271.7 Q5: .135 .151 , .051. .061 .h08 .398 g 1.80 2.21 .73 .87 5.1.6 5.60 Ev 2.99 3.30 7.09 6.00 2.19 2.06 ; Overlap 14.116 0.36 1.97 On the basis of data obtained by Mendall and Aldous (19113) and Greeley (1953), the bill length was considered as a possible criterion Of sex. However, according to the conclusions of these same investi- gators, as well as judging by a statistical analysis of data from I Michigan woodcock (Table 17), the bill length alone was not considered to be sufficiently accurate for use. Because bills are easy to obtain from hunters and since simple and rather accurate measurements may be taken from the bills, the use of the organ is desirable. To increase the accuracy of determining sex (or to reduce the aInount of overlap or proportion of animals "mis -sexed") the widths of 90 the outer three primaries were considered, following Harris (1953). He found that with specimens he examined in Minnesota a combination of bill length and total width of the outer three primaries was a very accurate method for determining the sex of woodcock. In addition, in the present study the total length of the outer three primaries was examined. The totals of the widths and the lengths of all three outer primaries were used rather than width and length of one feather, be- cause the per cent of overlap was less for the combined measurements compared to any one of the three primaries. A A scattergram is presented for each twosway comparison (bill length.!§. primary width in Figure 12, primary length vs. primary width in Figure 13, and bill length is, primary length in Figure 11:). In Figure 12 it may be noted that a considerable amount of overlap occurred in the bill lengths but that none existed for the primary widths. ‘When the two characters were plotted on the scattergram a blank space extended diagonally between the two clusters of points. An eye-fitted diagonal line drawn through the middle of this space represents a separation point for these data. Any dot plotted below the line represents a male and any plotted above the line represents a.female. In Figure 13, some overlap is shown between the primary lengths but this overlap is less than that shown for bill lengths in Figure 12; hence the distinction between sexes seems more evident. As in the above scattergram.a diagonal was drawn between the two clusters of dots to represent the separation point between sexes. 18385? as 593 33 mo 8 he 3 mo 3 no No He .8 mm am pm i _ q . _ a e _ _ a A _ S. o H. . .. .H eds-Ia. on enema fin ...33 09:3 heave no fine: 38. one fine: as 8.33 8:332:11 .fi tense (”040nm at) "mum 1° mm Relationship between total width and total length of outer three primaries. Figure 13- 260 (ummnw at) “mm-Id :0 mam 92 300 290 280 270 250 2110 Length of Primaries (in millimeters) 93 Anacoosadnws c3 nausea dam me a d. 2. so 8 S no me so me so a 8 em 3 R q i e a m a 1 _ a 1 q A _ a a _ a n x u u x x x x In x ’x I I J OJN 03h. I an n u x at” K II a“ and "& lax} xxx x I. an a an 1. guns ~x I u x x u: an F xx an x x a an: I § I I. .... xxx x x .3...” xx IN...” at.” H 0mm x x xx x x I x k x x xx x mm xxx x x x an I O x I xx x e 00 x x O x I . . 8w 0 e e e no u e e e e e x x e e 0 ee 0 e“ I 0.0 1 e see. u eases”. eeee 00 e e e e so see a case eeeeeeae essences. e e e e CNN 0 e e. 00‘ see ee 0 e 00000. .he 0 e e“ eeeeee e” en so .0000 J . C . C. O O C O In 8“ e. o 63838 .85 torso no finned, 33 one finned as soon»... aflaosseeom .1: ensure (sammrm! at) "um-Id :0 mm 9h For the third comparison of two characters (primary length.z§. primary widths, Figure 1h) no such separation line could be easily designated since the overlap occurred in two directions. This latter tWO‘WQV comparison apparently is not suitable for use in distinguish- ing sex even though the actual overlap for the 370 specimens was only about six per cent. Of the three comparisons the first two were sufficiently accurate to be considered as sexing methods. Naturally the methods will need to be checked with future sets of data from woodcock of known sex. As the data appear now no overlap exists and the per cent of overlap seems insignificant. The amount of overlap decreases and the accuracy great- 1y increases by using a combination of two factors. However, the amount of work involved in making measurements increases also. The measuring of the widths and lengths of primaries takes time, yet it appears that the increased accuracy is worth the effort. Instead of using the scattergram it might be preferable to have a table of critical measurements which separate sexes. Tables 18 and 19 show these numbers, up to a certain point, calculated from.the di- agonal lines of separation. Table 18 was determined by the formulae, femalez'W'+ B/2.5 >‘37.7 and male: W’+ B/2.5‘< 37.7, where, 'W a primary width B a bill length l/2.5 : lepe of line in Figure 12 The sex of a woodcock is determined from Table 18 by looking up first the bill length and then the corresponding primary width; both criteria are important in the separation; 3.5., if a bill is measured as being 60 millimeters long the woodcock is a male if the primary 95 width is less than 13.7 and is a female if the width is more than 13 . 7 millimeters . Table 18. Critical measurements necessary for sexing woodcock as determined from a comparison of bill lengths and primary widths.* Bill Length Primary Width Bill Length Primary width 50 17.7 63 12.5 51 17.3 6h 12.1 52 16.9 65 11.7 53 16.5 66 11.3 5h 16.1 67 10.9 55 15.7 68 10.5 56 15.3 . 69 10.1 57 18.9 70 9.7 58 ' 1h.5 71 9.3 59 1h.1 72 8.9 60 13.7 73 8.5 61 13.3 7h 8.1 62 12.9 75 7.7 ‘ * All measurements in millimeters. Table 19 was determined by the formulae, female: W + L/lO> 38.5 and male: W + L/lO ( 38.5, where, L a length of primaries 1/10 s slope of line in Figure 13. Again the width of the primary is the critical measurement; _e_.g_., if the primary-length is 250 millimeters and the primary width is less than 13.5 millimeters the bird is a male, but if the width is more than 13.5 millimeters the bird is a female. Following the method adopted by Harris (1953) the succeeding three schedules for sexing woodcock were determined from data on the scatte rgrams . 96 Table 19. Critical measurements necessary for sexing woodcock as determined from a comparison of primary lengths and primary widths.* . ' Primary Lenith Primary Width Primary Largth Primary Width 230 ' 15.5 260 12.5 231 15.1 261 12.h 232 15.3 262 12.3 233 15.2 263 12.2 23h 15.1 26h 12.1 235 15.0 265 12.0 236 111.9 266 11.9 237 1h.8 ' 267 11.8 238 1h.7 268 11.7 239 1h.6 269 11.6 2ho 11.5 270 11.5 281 18.1 271 11.h 2h2 1h.3 272 11.3 283 1h.2 273 11.2 2hh 1h.1 27h 11.1 2h5 1h.0~ . 275 11.0 216 13.9 276 10.9 2h7 13.8 277 10.8 2h8 13.7 278 10.7 2h9 13.6 279 10.6 250 13.5 280 10.5 251 l3.h 281 10.h 252 '13.3 282 10.3 253 13.2 283 10.2 25h 13.1 28h 10.1 255 13.0 285 10.0 256 12.9 286 9.9 257 12.8 287 9.8 258 12.7 288 . 9.7 259 ' 12.6 289 9.6 * All measurements in millimeters. 1. If bill length is 61 m. or less the bird is a male. If bill length is 65 mm. or more the bird is a female. If the length is between 61 m. and 65 mm., measure the width of the outer three primaries and add the 17311168. If wing value is 12.0 m. or less the bird is a male. If wing value is 13.0 mm. or more the bird is a female. If wing value is between 12.0 mm. and 13.0 mm. and bill length is between 61 mm. and 65 mm. the sex can not be determined externally. 97 2. If total length of outer three primaries is 255 mm. or less the bird is a male. If length is 265 run. or more the bird is a female. If length lies between 255 mm. and 265 m. then measure width of outer three primaries and add the values. If width value is 12.0 mm. or less the bird is a male. If width value is 13.0 mm. Or more the bird is a female. If width value is between 12.0 mm. and 13.0 mm. and primary length is between 255 mm. and 265 mm. the sex can not be determined externally. 3. If bill length is 61 mm. or less the bird is a male. If bill length is 65 mm. or more the bird is a female. If bill length'is between 61 and 65 mm. measure length of outer three primaries and add the values. If primary length is 255 mm. or less the bird is a male. If primary length is 265 mm. or more the bird is a female. If primary length is bemeen 255 and 265 mm. and bill length is between 61 and 65 mm. the sex can not be determined externally. Inspection of the present sets of data suggest that a three-vari- able method for distinguishing sex could probably be devised if the methods using two variables are found to fail. This method would include the bill length, primary lengths, and primary widths. Throughout studies on fall-shot woodcock no external measurement was found which would separate birds-of-the -year from adults. But, during the fall of 1955. a close examination of legs and feet from freshly-killed woodcock indicated a color difference between speci- mens. Of 123 birds examined, 66 had light or flesh-colored legs. Sixty-four of these birds were adults while only 2 were juveniles. Fifty woodcock had legs and feet which were of a dusky-grey colora- tion, and of these, 118 were Juveniles. The color or age was question- able for 7 birds. Most woodcock of doubtful age were taken during the latter part of the season; this fact might suggest that the bursa had been absorbed on some birds-of-the -year by that time while the leg color may or may not have changed. 98 These results imply that young birds have darker legs than adults during the fall. The colorations should be checked anotlmr year to determine if such a method may be reliable. In a letter from Dr. William Sheldon of the Massachusetts Wildlife C00perative Unit he mentioned that neither he nor Dr. Fred Greeley could observe any con- sistency in color variation of woodcock legs they had examined. Sex and Age Ratios. There were 381 specimens collected for which the sex and age were determined. For one year, this number of specimens seemed suf~ ficient but once the specimens were separated year by year and into four component parts (adult male, adult female, Juvenile male, and Juvenile female) usefulness of the data decreased greatly. Numbers and ratios are presented to make the data available to other investigators. Over half of the woodcock specimens were taken in the Pigeon River Area in 1951; and 1955. These specimens were handled separately from those taken at various other locations over the state, but when the chi-square test of homogeneity showed no significant differences in sex or age ratios between years or areas, the data for the two areas were combined for further analysis (Table 20). Table 20‘. Sex and age ratios of woodcock collected in Michigan. Adult Juvenile V. Hale-Tend. Ma]; $.er Adulfi‘emale- Year Male Female Ratio Male Female Ratio Juvenile Ratio 1953 18 23 78 :100 13 17 76:100 1 :1.39 1951; 29 38 76:100 30 31 97:100 1:1.63 1955 1.3 56 new us 38 118:100 1:1.h8 Total 90 117 77 :100 88 86 1023100 1:1.51 99 When the data were combined for all three years the sex ratio of juveniles approached the hypothetical 1:1 ratio. The adult melezfemale ratio of 773100 was comparable to ratios found for fall-shot woodcock by several other investigators. ~Mendall and Aldous (19h3) reported a sex ratio of 75 males:100 females, but this ratio included all ages. The ratio of 86:100 given by Harris (1953) was also for all ages. Greeley (1953) gave a sex.retio of 80 males:100 females in both adults and.young. Production. Rates of increase, using the formula: r - l'+ fy (Kelker, l9h9) Where; f a proportion of females in adult population y,. average number of young per adult female were determined for the three years. These rates are shown in Table 21 along with estimates of the spring breeding papulation and the fall ratio of young per adult female. (Data for the spring breeding population and broods were discussed earlier). Table 21. Population levels for different periods. Category 1953 l95h 1955 Rates or increasa 1073 1091 1.8h Sprin breeding levels* .80 1.0h .92 re adult female 1.39 1.63 l.h8 * Woodcock heard per stop per route. Survival rates were determined as 0.75 for the 1953-195h period .and 0.h6 for the l95h-1955 period by the formula: 3 = PQ/Plr (Petrides, l9h9) where, 100 P1 - population for one year P2 - p0pulation for second year r a rate of increase The survival rate of 0.75 seemed rather high but it could be possible that mortality, including hunting kill, was reduced for reasons un- known during that period. The spring breeding counts for Michigan showed an increase in the breeding p0pulation from 1953 to 195h. Half of the other states reporting woodcock counts also had an in- crease according to Aldrich (1955). If this increase did occur in Michigan it could have resulted from the low mortality indicated above, from an increased production in 1953, or from both. A definite decrease in the number of young occurs from.the nest to the fall season as determined by information from this study. The table below shows this decrease as it occurred. Average Average Brood-size Clutch—size 0-7 days 8-11. days Young/adult female (fall) mar 309 396 30,4 105 The figure of 1.5 young per adult female is comparable to the ratio of l.h young per adult female which Greeley (1953) found to exist in fall-shot woodcock in Wisconsin. The low ratio indicates that the greatest number of woodcock died during the first five months of life, before the birds were even subjected to hunting pres- sure. In many species of birds this mortality is expected before the fledgling stage is reached. Here the greatest mortality, in numbers of birds, is indicated as occurring after the fledgling stage. Men— dall and Aldous (l9h3) noted that the rate of mortality up to three weeks of age did not exceed 10 per cent, but they did not indicate 101 what the rate of mortality might be from that period to the fall sea- son. A recent method devised by Allen J. Duvall of the Fish and Wild- life Service for aging woodcock by plumages during the summer may be useful in determining what per cent of the mortality occurs in each month. For the data to be reliable a large number of woodcock will need to be caught, and the birds must not show a differential sex or age vulnerability to the trapping technique employed. Assuming that no mortality occurs in the adult group up to the hunting season then mortality of the total population between that time and the next spring breeding season should not exceed 5!; per cent if the breeding papulation of the previous year is to be maintained at the same level. This conclusion is based on the assumption that the spring breeding population levels represent the true trends in the population. For 1951; and 1955 when the survival was determined as 146 per cent the spring population remained almost constant (195,4 - 1.01;, 1955 - 0.92). From Fish and Wildlife Service banding records the writer deter- mined that the average length of life for 141 woodcock banded as young birds in the spring and later recovered was approximately 10.6 months (Appendix). 0f the bl, only 12, or less than 30 per cent, lived long enough to breed. Twenty-seven of the birds were recovered in the fall season of the year of bending. These records imply that as in most species the younger animals comprise the greater part of the hunting kill. There is a considerable difference indicated between the mor- tality rate determined by population data from Michigan and the (band 102 returns received from the Fish and'Wildlife Service. This difference could be due to one or more of several reasons. First, as frequently is the case, the data are lacking. Second, the woodcock population in the northeastern states on which the band records were primarily obtained, could be different in population characteristics from.the birds in Michigan. Third, the woodcock, being more concentrated in the northeast, could be subjected to a heavier local hunting pressure. Also Juveniles may be less vulnerable to hunting pressure, although this would be contrary to the usual higher vulnerability of young birds. Finally, the method of using spring breeding levels to deter- mine survival might not be appropriate since these levels are indicative only of trends. Extent of Kill Although the northeastern states have long been considered the land of "dyed-in-thedwool" woodcock hunters, Michigan has reported the greatest annual kill. This leadership in woodcock kill may have resulted.from.a combination of more woodcock habitat and a larger pepulation of birds or a greater number of hunters in the woods. For several years, kill of woodcock in Michigan has been.deter- mined on the basis of hunter report cards. In the 1ater years these report cards were supposed to be given to hunters when they purchased a small-game hunting license (at first they were a part of the license). And though the law required hunters to send in the cards 103 after the season ended, the return usually was small, especially in the later'years (Table 22). In 195k, a new method of determining the kill of many game species, including woodcock, was started. This method was the polling of a group of hunters whose names had been chosen by a systematic sampling, the start of which was randomized. Approximately thO cards were mailed. If the hunter failed to return the card as many as.four fol- lowdups were made if needed. Of these two methods the post card survey was considered most reliable since it was based essentially on a random sample. The reli- ability of hunter report cards had long been.qmestioned for several reasons. One such reason is bias due to a prObable tendency for the more successful hunters to report in higher prOportions. However, data from these reports have been somewhat useful as an indication of the trends involved. As shown in Table 22 the average number of woodcock'bagged by hunters reporting on woodcock kill varied between 2.h7 and 2.7h. The largest kills came when the greatest number of hunters were in the woods. A brief examination of the computed woodcock kill in Table 22 might leave the impression that the woodcock population showed a cyclic trend. However, the lows and highs of the kill were a.result of the lows and highs of hunting pressure as shown by the straight- line relationship in Figure 15. In general the various data pertaining to hunting records and population levels are too contradictory to set forth any definite conclusions. Although the data presented in Table 21 show that the 10).; Table 22. ‘Woodcock kill data.from hunter report cards and post card surveys. "Hunter Report Cards” Total Computed Computed Average No. of Percent No. of Percent Woodcock Woodcock Woodcock per Hunting Year Hunters Return Hunters Kill Hunter Reportingg‘Woodcock 1938 515,522 19.h 21,807 55,156 2.53 u.23 1939 538,026 39.6 22,337 60,081 2.69 b.15 l9h0 537,630 32.8 16,259 h0,690 2.50 3.02 l9hl 610,000 25.1 13,882 35,553 2.56 2.28 19h2 573,837 20.1 11,73h 31,905 2.72 2.0a 19h3 510,000 13.h 13,208 32,938 2.u9 2.59 19th Shé,680 10.8 11,882 30,020 2.53 2.17 l9h5 5h7,20h 10.3 15,107 37,258 2.h7 2.76 19h6 685,500 12.h 25,513 69,0h9 2.71 3.72 19h? 535,927 11.0 21,223 58,2h9 2.7h 3.96 19h8 582,000 9.0 29,588 77,871 2.63 5.08 19h9 626,9h1 9.5+ 3h,291 92,110 2.69 5.h7 1950 620,192 9.0+ 30,h18 78,095 2.57 8.90 1951 6h0,010 9.2+ 31,785 79,762 2.51 h.97 1952 66h,728 7.2 32,903 8h,197 2.56 h.95 1953 705,000 5.8 26,823 67,26h 2.51 3.80 195k 71h,000 5.0 21,h88 5h.701 2.55 3.01 1955‘ 727,87h 3.h 25,216 57,985 2.30 3.7h "Post Card Survey" 195k 662,920 9h.0 29,150 h9,320 1.69 h.uo 1955 677.370 95.h 33,700 h6,230 1.37 h.99 105 rates of increase, survival, and spring population levels for 195).: were the highest of the three years, most records show that fewer woodcock were killed in that year (Tables 22, 23, 2b,, 25, and 26). The post card survey (Table 22) showed a higher kill for 1951; than for 1955, but no comparison could be made to the 1953. harvest. This hunting pressure, in addition to natural mortality, has not reduced the woodcock population to a critical level. If the population were being harmed it is unlikely that most of the hunt- ing records would have shown an increase in kill in 1955 over that in 1951;. The limit to which woodcock may be harvested can not be given. However, since the hunting success has remained relatively constant and the kill has continued to increase as the hunting pressure increased the point of overharvesting has not been reached. It is likely that the woodcock may become more popular as a game bird in the southern states as a greater demand is placed upon wildlife as a source of recreation. If this additional harvest starts in the South then probably bag limits may need to be reduced in order to decrease the kill per hunter in Michiganand other northern states. The same number of hunters or more can still hunt woodcock but at a lower daily bag limit. There could be. a greater surplus to (harvest if natural mortality could be reduced in Juveniles. 0f the current crop only 1.5 young per adult female, or about 38 per cent of the young, live to the fall. 106 Assess ooo.H use 2355 xoooeooz Mo .3852 mm am am am mm mm Hm ea NH me me He _ _ 1 _ _ 1 _ _ _ a» 1 _ .Ammmaummmd #32 333003 e5 nonsense messes: soonest oasnsosnsaom .me seamen 0m on on om ON. om OOH (sqrun ooo‘I Ur) parrrx aooopoom JO JeqwnN 107 Table 23. Compilation of reports from those hunter cooperators who flushed woodcock. Number Number of Number of wogggECk Woodcock Year of Hours Woodcock Seen ‘Woodcock Shot 'per Hour Shot per Hour 1953 1389 151.). 1.12 1.11 .30 1951. 1135 1008 266. .89 .23 1955 1109 1281. 31.0 1.16 .31 Table 2h. Compilation of reports from those cooperators who hunted only woodcock. ‘Woodcock Number Number of Number of Seen Woodcock Year of Hours 'Woodcock Seen ‘Woodcock Shot per Hour Shot per Hour 1951. 660.75 1019 271. 1. 51. in 1955 510.50 956 279 1.87 .55 Table 25. Compilation of records from.all ruffed grouse hunters reporting. Number Number of Number of ”8°d339k 'Woodcock Year of Hours Woodcock.Seen ‘Wbodcock Shotggper Hour Shot per Hbur 1953 1:238 15M. 1:35 .36 .10 1951; 3598 121:6 322 .35 ..09 1955 3069.5 1568 1:37 .51 .11; Table 26. Compilation of hunter reports from.the Rifle River Area. Woodcock Woodcock Number of Number of Number of Seen Shot Year Hunters ‘Woodcock Seen ‘Wbodcock Shot Per Hunter Per Hunter 1951; 66 11.6 13 2.21 .20 1955 121 M3 65 3.66 .51: 108 ECOLOGICAL FACTORS AFFECTING DISTRIBUTION, ABUNDANCE, AND ACTIVITIES Vegetation Plant cover seemed to be a major factor limiting the use of an area by woodcock in many instances. Whether the amount and type of vegetation was the limiting factor or whether vegetation was indica- tive of a combination of other factors which actually determined the use of an area was not always clear. It was probable, however, that both relationships existed, either separately or together, at various times. V Information pertaining to vegetation was gathered in several ways. Most data taken in relation to singing grounds, diurnal coverts, nests, and broods were based on a general examination of the immediate area. Records were made of the major species which occurred and the relative abundance of each. The summer and fall coverts were analyzed more systematically; the methods used will be discussed under those categories. Bmedmwas o Such areas included three primary niches: singing grounds, diurnal coverts, and.feeding sites. Which of the three was most im- portant was difficult to say, but it was probable that the singing ground and feeding site played comparable roles in determining whether or not a woodcock would be in the general area. A stand of trees 109 ‘- without any suitable openings was not used by a male woodcock regard- less of the abundance of food. Nor was an area of suitable openings normally used if a scarcity of food prevailed for several miles. ”11 At least 98 singing grounds were checked over the threeiyear period. Of these, h3 were examined closely to determine species come position, relative density, and the height of herbaceous and woody types. Since all three of these characters changed as the breeding season progressed, it was hard to characterize a singing ground for the entire spring. Generally, when the first woodcock returned to Michigan the ground was relatively bare, the only noticeable vegeta- tion leafless shrubs and trees and dead herbaceous plants. Sometimes dead plants played an important part in determining the exact site to be used by a.ma1e bird. If the herbaceous growth had been ex- tremely luxuriant in the previous year, and nothing had occurred to remove the greater part of it, then the dead stems remained dense and " - bothersome to the woodcock. Normally however, spots with relatively little cover were present on sites chosen for singing grounds. As the season progressed new plants started growing, and often by the first week of May herbaceous vegetation.on some singing grounds was from 6 to 12 inches high. Usually one or more spots had shorter vegetation so woodcock continued using the singing ground; however, the rapid growth of herbs on some "grounds" undoubtedly shortened the use of those sites by woodcock. The density of herbs progressed.in much the same manner as height. But, on many of the "grounds" the prevalent forts and grasses were of the low-growing type, such as strawberry, pussy's-toes, panic 110 grass, and blue-grass, and never reached 6 inches until late in the season. Even though the density may have been heavy, as was the case by May 1 for 32 of the "grounds,” the plants averaged about I; to 6 inches in height, and usually some spots were present which had shorter plants, or none. When a combination of heavy densities and average heights of about 8 to 12 inches occurred use of the site was terminated. In any one season the density and height of woody plants changed very little. The ideal situation regarding woody plants seemed to occur when the plants were scattered sparsely over the general location and less than 6 feet high (Plate 13). Dangler and Marshall (1950) found that the cover on 11; out of 17 singing grounds was not higher than 5 feet. 8 Species composition as given in Tables 27 and 28 was confined primarily to new plants. The dead plants which were present on sing- ing grounds when the breeding season began included primarily grasses, goldenrods, asters, and fleabanes. As indicated by the tables a variety of species, both herbaceous and woody, was found: more forb species could have been added if some "grounds" had been checked later in the season. The variety was increased somewhat by including sing- ing-ground records from both the Lower and Upper peninsulas. :f‘ In summarising, it may be said that the singing grounds observed in Michigan most often occurred in areas where an opening existed near woods. These openings occurred most frequently on such places as abandoned farms, cut-over land, old burns, and seldom-used roads. Generally, the herbaceous cover did not become important as a limit- ing factor until the breeding peak was passed.“ The most important PLATE 13 A. The upper picture shows a singing ground which has been used several years by woodcock. Notice the height and distribu- tion of shrubs which are encroaching upon the opening. B. The lower picture shows an opening Just opposite the one above. ‘Woodcock were never Observed in the site. Notice the absence of woody vegetation. 112 Table 27. Vegetation observed on singing grounds. "Herbaceous Plants" Species Frequenqy* Species Frequency* Grasses _- ho Milkweed h Goldenrods 30 Cinquefoil h Strawberry l9 Alsike clover 3 Pussy's-toes 18 Bracken fern 2 Nut grass 11 Bastard-toadflax 2 Mosses 8 Groundsel 2 Orange hawkweed 8 Crowfoot 2 Aster 8 Wintergreen 2 ‘Yarrow 7 Bulrush 2 Sorrel 6 Red clover 1 Chamomile 6 Vetch 1 Smartweed. 5 Purple evens l Lichen 5 Common thistle l Dandelion h Violet 1 ‘Wbrmwood h WOOdefern 1 Hawkbit h Mint 1 "Woody Plants” ‘Willow' 15 Jack pins 1 Quaking aspen 11 Bush clover l Spiraea 6 Black ash 1 Black cherry' 6 Common juniper l Staghorn sumac 5 Buffalo berry l Hawthorn h Bearberry 1 American elm. h Canadian Juneberry 1 Blackberry h Dewberry 1 Red osier 3 'Wild-raisin 1 Grey dogwood 2 Swamp birch l Chokecherry 2 Raspberry 1 Currant 2 Pin cherry 1 Sweet fern 2 Sugar maple 1 Poison sumac l ‘Witch hazel 1 * The number of singing grounds on which the plant was observed. 113 Table 28. Vegetation enclosing singing grounds. Species Frequency-IE Species Frequency!!- Willow 26 White oak l Quaking aspen 25 Multiflora rose 1 Black cherry 9 Indian grass 1 Red pins 7 Raspberry 1 Blackberry 6 Honeysuckle 1 Red osier 6 Chokecherry l Spiraea 6 Buckthorn 1 American elm 6 Bush clover l Staghorn sumac 6 Sugar maple l Speckled alder 6 Black spruce l Hawthorn )1 White spruce l Swamp birch )4 Buffalo berry. 1 Grey dogwood )4 Large -toothed aspen 1 Apple 3 Chestnut l Tamarack 3 Yellow-barked oak 1 Poison same 3 Silky dogwood 1 Sweet fern 3 Witch hazel 1 White pine 3 Balsam fir 1 Jack pine 2 White cedar 1 Pin cherry 2 Currant 1 Black ash 2 Scotch pins 1 4!- ‘Ihe number of singing grounds on which the plant was observed. 11!: PLATE 1h A. Site of a singing ground where staghorn sumac and scattered pine trees have served to break the continuity of an opening. B. Site of a singing ground where bush clover (Lespedeza) has served as a substitute for early stages of woody vegetation succession. «(u \I \ ...“\ [AF/‘07 115 i7 vegetational characteristic seemed to be that plant succession on the openings was in the early woody or low shrubby stage. However, any- thing which.seemed to break the continuity of an opening seemed to suffice. Some apparent substitutes for the earlijoo y stage of succession were staghorn sumac, Indian grass, and.bulh clover (Plate 1h). While examining singing grounds for information on vegetation, records were kept on other possible characteristics such as soil, size, shape, exposure, and distances to water, nests, and diurnal coverts. The influence of soil will be discussed later. In regard to the next three factors, little importance could be placed on any of them.because of the great variations between singing grounds. Some singing grounds contained approximately 25 square feet, the minimum noted by lbndall and Aldous (191.3) in Maine. Normally, a male would restrict»activities to a small portion of a.much larger area which sometimes was as large as hO acres. These smaller portions were dis- tinctive parts of the overall clearing. One male was known to use several landing sites of various dimensions on about 3 acres. It seemed impractical to compute an average size since it was too diffi- cult to define the actual limits of a singing ground. Shapes of the clearing and the more restricted area of activity were of all possible forms. Because of the variation no significance was noted for this factor. Neither were the exposures considered important since all possible types were observed. rDistances to water, nests, and diurnal coverts were recorded for most singing grounds. The average distances to each respectively were 116 300, 320, and 285 feet. The range varied considerably: Water was found adjacent to singing grounds or as far as 925 feet distant. This water was of the type (streams and marshes) which might influence the surrounding area and thus determine the presence of food. Actu- ally, the nearest permanent water supply was often as far as one-half mile. Nests ranged from 25 to 660 feet from the nearest singing ground. It was not determined whether these singing grounds were ones which the female visited. Mendall and Aldous (19113) reported the average 3 distance to the nearest “ground" to be only 115 yards, which was com- parable to the 320 feet mentioned above. ’ Diurnal coverts were often in woOds adjacent to the singing ground or as far as 750 feet away. Two males were observed using the same diurnal cover (Plate 15). One of these males went )5 feet to a singing ground while the other flew 730 feet. The average distance of 300 feet could be compared to the distance given by Mendall and Aldous (191:3). They concluded territories over 100 yards were the exception rather than the rule. Diurnal coverts and feeding areas were often at the same loca— tion. As indicated by Pettingill (1936), the diurnal territory is the principal abode of the male during breeding season; consequently, much of the feeding must take place in day-time cover. During the present study, however, in several instances other feeding areas were located where male woodcock went after the evening performance. These areas were soft, moist sites nearby. Birds may have gone much farther, but such movements could not be followed. 117 PLATE 15 A clump of tamaracks, on the Rose Lake Wildlife Experiment Station, which served as a diurnal covert for two male woodcock, a nesting site for a female woodcock for two years, and a singing ground for a male woodcock. The singing ground was located on the edge of the clump of trees (marked by arrow). 118 Since diurnal and feeding covers were frequently the same a list of major woody plants observed in those areas is included in Table 29. Table 29. Woody plants found in diurnal coverts. Species Frequency* Species Frequency* Quaking aspen 35 Large-toothed aspen 3 American elm. 15 ‘White pine 3 ‘Willow 1h Black locust 2 Speckled alder 12 Viburnum 2 Red osier 10 Juneberry 2 Grey dogwood 9 Red pins 2 Tamarack 9 White birch 2 Oak 6 Jack pine 1 White spruce 6 Hawthorn 1 Black ash 5 Buffalo berry 1 Maple 5 Staghorn sumac 1 Black cherry h Chestnut 1 Poison sumac 3 Hickory l Spiraea 3 Highbush cranberry l Balsam fir 3 Ninebark 1 Balmrof-Gilead 3 Blackberry 1 * The number of diurnal coverts in which the plant was found. Nesting Cover. Although.more than 30 woodcock nest reports were received, only 12 had cover types included. The immediate nesting cover was varied as would be expected with the numerous vegetative types which exist iin Michigan. Generally, most nests were located in relatively Open cover or at the edge of clearings. ~~~~~~ Five nests were found in relatively'youag and Open second-growth woods. ,Four of these nests were in upland woods consisting primarily .of quaking aspen with an occasional white birch, juneberry, white 119 spruce, or balsam fir. The other nest was in a river valley consist- ing of a mixed growth of quaking and large-toothed aspens, white and Jack pines, and white spruce. Five other nests were situated at or near the edge of openings. Three were in a tamarack swamp on the edge of groves which consisted -, mostly of tamarack, grey dogwood, hawthorn, poison sumac, red osier, and elm. Tito of the nests in the swamp were near the end of over- hanging branches of tamaracks while the other was at the base of a hawthorn (Plate 11). Of the remaining two nests, located near open- ings, one was in a strip of willow, quaking aspen, and grey dogwood between an old field and a sedge marsh. The other nest was under a dead quaking aspen on the edge of an opening bordering a stand of aspen and speckled alder (Plate 16). Two nests were built in fields. One nest was under a planted Jack pins in an upland field while the other was in an Open fallow field of grass and forbs. In no instance was a nest well-concealed. Most of the immediate cover was dead vegetation and small green herbs (Plate 17). The ---birds seemed to depend upon their concealing coloration for protection rather than a dense vegetative stand. Brood Habitat . Since brood movements were not traced, the complete vegetative cover used by woodcock could not be determined. Only that vegetation in which broods were found once was analysed. For 11 records, the general habitat type was the edge of relatively open woods adjacent 120 A. B. PLATE 16 The location of a woodcock nest beneath a fallen aspen at the edge of an alder-aspen stand. A close-up of the above mentioned nest showing the eggs and nest construction.' 121 A. B. PLATE 17 A female woodcock on a nest, illustrating the lack of cover and type of vegetation present at that time of the year. A picture of the above nest showing the eggs and again the lack of cover. 122 to clearings. Only one of the broods was in a dense stand of woods, and thoSe woods were bordered on one side by a road. Of the 11 broods found, 9 were in stands of quaking aspen. These stands included other such plants as blackberry, elm, grey dogwood, black ash, maple, willow, oak, hawthorn, red osier, or juneberry. Another brood was seen in a narrow strip of willow, quaking aspen, and grey dogwood which grew adjacent to a clearing. The last record con- cerned a brood found in a patch of timothy on Isle Royals. -_ __Normally, the ground cover was heavier than the cover in which nests were located, but it was doubtful if the birds sought the heav- ier vegetation for cover. Probably by the time most broods were off, the ground cover was more dense due to new spring growth. Several broods were located in woods which had rather sparse ground cover (Plate 18). Tw0 broods were found in dense stands of grass, but these stands had several Open bare spots. Every brood was found in or near wet sites. Summer. In most places in Michigan the summer habitat of woodcock ap- peared to be similar to diurnal and feeding coverts used during the spring. Occasionally, woodcock moved to other places, but when they did the change usually was caused by a decrease in soil moisture with- in the regularly-used habitat. This decrease in soil moisture re- stricted feeding areas and thus was a major factor in causing woodcock concentrations. Such concentration areas were the ones analyzed for vegetational composition. 123 PLATE 18 Two young woodcock, in center of picture, illustrating the concealing coloration and-the sparsity of the surrounding ground cover. 12h These concentration areas were classified tentatively as feeding or loafing places according to the general appearance of habitat, moisture content, location, and food availability. Loafing sites norm: ally were in dryrupland situations overlooking low moist feeding grounds. Earthworms were few in comparison to the number found in feeding areas. ———Evidant1y, little feeding if any took place in loafing sites, but feeding areas provided both loafing and feeding cover. In most places used as feeding grounds, woodcock could be found any time dur- ing the day. At twilight more woodcock could be seen moving into these areas from.adjacent hillsides or slopes and from.seemingly greater distances. During l95h and 1955, nine feeding areas and four loafing areas were studied. The methods used in l95h to determine species com- position and.density of vegetation was a modification of one proposed by'Webb (l9h2). This modified method consisted of locating the ap- proximate center of activity in the area to be analyzed. From.this point, compass readings were taken to locate the eight principal directions. One circular plot was laid out and analyzed at the cen- ter point. Then 2h other plots (three on each of the eight lines) were analysed in the defined area. The center of each plot was on the line, and the distance between the centers of two successive plots on a.line was one-half chain (see following diagrams -t0p of next page). 12S Ground Cover (3.72' radius) Twentysfive'was chosen arbitrarily as the number of sample plots per area. Then at the end of field work in 195h, the data.were ana- lysed and it was believed that additional plots would not improve the precision enough to warrant theextra time involved. The follow- ing tabular information was used to construct Figure 16 which illu- strates the reason for considering 25 plots a sufficient sample .1... Table 30. Data necessary to determine coefficient of varia- bility of the mean or precision desired for various subsample sizes of vegetation in woodcock habitat. Trees Shrubs Ground Cover Feeding All Feeding All Feeding All Areas Areas Areas Areas Areas Areas 2* 2.20 2.09 1.30 1.33 2.71. 2.1.3 8 059 e& 057 e63 .115 056 * Average density of the vegetation in the sampled areas. Each.plot was composed of three concentric subplots, one each for ground cover, shrubs, and trees. The radii for each of these subplots were 3.72, 8.33, and 11.78 feet respectively (see preceding diagram). Ground cover included all herbaceous plants and any woody 126 plant 1 foot high or less. Shrubs included any woody plant between 1 foot and 10 feet high. Woody plants higher than 10 feet were cone- sidered as trees. The density measurements recommended by Webb (19142) were used be- cause of their apparent simplicity and because measurements are repro- ducible by other workers. Following are those measurements; Trace - leaves cover less than 1/80 of the subplot area. One - leaves cover between 1/80 and 1/3 of the subplot area. Two - leaves cover between 1/3-and 2/3 of the. subplot area. Three - leaves cover more than 2/3 of the subplot area. As the above described method was being used in 199., it became apparent that a center of major activity was difficult to define and that too many plots were being located toward this so-called center. In 1955. the old method of locating plots was discarded but the system of plots, subplots, and density measurements was retained. The new method located plots at random. First, the approximate boundaries of the area being frequented by woodcock were located and a starting point designated. Then two numbers were chosen from “s sheet of ran-= don numbers. . The first number represented the distance from the starting point in one direction, or the first ordinate, and the second number represented the second coordinate perpendicular to the first. At the point located a plot was set up and analyzed if the plot fell- within the defined boundaries. If the plot was outside the area it was ignored. Twmty-five such plots were analysed for each area. Data gathered from these plots and areas were analysed by the method, prescribed by Webb, of giving each plant a rating. This rat- ing depended on the frequency of occurrence and the total density as Coefficient of Variability of Mean 12'? Figure 16. Number of vegetation samples required to gather data which will give various degrees of reliability. Feeding Sites 1 l ! 4. I I A Shrubs 5 B Trees ' C Ground Cover Feeding and Loafing Sites 0 l l l l l J I I l J 5 10 15 20 2S 30 35 ho AS 50 Number of Samples 128 indicated by the formula: R = P x D x K , where, R a rating P - number of plots on which the species was found D an average density computed by dividing the summation of the densities (E) by the number of plots on which the species wasfound (P). ' K a a constant as long as the total number of plots remained the same. Its use made the maximum rating 100. Determined by: lOO/total plots 1 3. Since D . E/P, then a . EK. In analyzing the data three points were considered: (1) dif- ferences betmen the results obtained from two methods of collecting information, (2) characteristics of feeding and loafing areas, and (3) costs involved in gathering data. To determine if the data obtained by the two methods of select- ing sample plots could be combined, a ”t” test was made for each of 15 major species of trees, shrubs, and herbs found in the feeding areas, comparing means of data collected by the two methods. Since no significant difference between the two methods could be detected by the test, data gathered by the two methods were combined. Although certain plants seemed to the investigator to be charac- teristic of either feeding or loafing sites, judging by their abun- dance at one type of site and scarcity at the other, it was difficult to test statistically for these apparent differences. Except for the few most comon species of plants, the fraction of plots with zero occurrence was high, producing a clearly non-normal distribution, and making use of the 't" test of doubtful value. Several other reasons accounted for the difficulty in differen- tiating between kinds of areas statistically. First, too few data were available since in comparing kind of area, each area (25 plots) Table 31. Relative abundance of plant species found in woodcock habitat. "Feeding Areas"* 129 No.0f No.0f Total Trees Areas Plots Density(E) Rating Speckled alder 9 20h 251.67 37.50 Quaking aspen 8 123 153.32 22.8h Spruce 7 57 62.99 9.39 Balsam~fir 5 53 53.00 7.90 Willow 9 55 1.2.00 6.26 Balm-of-Gilead 7 51 38.99 5.81 White birch 8 1:3 35.33 5.26 American elm 5 37 28.67 h.27 Black cherry 5 3h 26.67 3.97 Black ash 5 29 25.66 3.82 Tamarack 3 21 2h.67 3.68 Nannyberry 3 1h 10.67 1.59 Juneberry 2 9 h.3h .65 Pin-cherry 3 9 3.00 .h5 White pine l 3 3.00 .16 Red maple 2 h 2.66 .hO Yellow birch l 2 2.00 .30 Red osier 2 2 1.33 .20 Choke-cherry 2 2 1.33 .20 White cedar 1 2 1033 020 Highbush-cranberry 2 2 . .10 Wild’raism 1 1 0 33 005 Shrubs Speckled alder 9 203 187.33 27.91 Red osier 9 156 87.67 13.06 Raspberry 9 109. 69°33 10.33 Nannyberry 5 h9 30.33 h.52 Currant and Gooseberry 7 69 27.01 h.O2 Black cherry 5 b? 25.67 3.82 ‘Wild-raisin 2 3O 2h.33 3.63 ‘Willow 6 h2 23.33 3eh8 Spruce 5 38 23.33 3.h8 Buthhom 6 ’42 22099 30113 Spiraea 8 hh 20.01 2.98 Balm-Of-Gilead 5 38 17.33 2.58 Balsamsfir 5 26 15.33 2.28 Highbush-cranberry 7 28 12.66 1.89 Honeysuckle 6 23 9.67 1.hh Quaking aspen 8 29 9.66 l.hh 130 No.0f No.0f Total Shrubs (cont. . . .) Areas Plots Density(E) Rating_ Choke-cherry 6 - 17 7.00 1.0h Ninebark 3 12 6.00 .89 Beaked hazel l 8 6.00 .89 Virgin°sébower 5 13 5.66 .8h Pin-cherry 2 1h .5.33 .79 American elm 2 12 h.67 .70 Red maple h 1h h.66 .69 Blueberry 2 6 h.33 .65 Fly-honeysuckle 1 9 h.33 .65 Tamarack 1 8 3.33 .35 Swampefly-honeysuckle 3 7 2.33 .35 Swamp-birch 2 S 2.33 .35 Black alder 1 3 2.33 .35 Juneberry 3 7 2033 035 BlflCk 3811 S 6 1099 030 White birch h 6 1.99 .30 ‘White cedar 3 5 1.66 .25 Bush-honeysuckle 3 5 1.66 .25 Common elder 2 2 1.33 .20 White ash 1 2 .67 .10 Poison ivy 1 2 .67 .10 ‘Wild honeysuckle 1 1 .33 .05 Hairy honeysuckle 1 1 .33 .05 Rose 1 l .33 .05 Ground Cover Grasses 9 196 196.99 29th Sedge 9 185 179.33 26.72 Goldenrods 8 150 111.00 16.5h Strawberry 9 lhl 96.3h 1h.35 Dwarf raspberry 7 107 88.00 13.11 Violet 9 51 73.33 10.93 Asters 7 102 62.66 9.3h Bedstraw 9 130 58.67 8.7h Purple meadow-rue 7 87 53.67 8.00 Orange hawkweed 7 57 50.01 7.h5 Bracken h 36 h6.66 6.95 Bishop's-cap 9 58 3h.OO 5.07 Bunchberry 6 38 28.32 h.22 Lettuce 9 73 2h.3h 3.63 Raspberry 8 52 22.3h 3.33 Crowfoot 6 h2 20.67 3.08 Lady fem )4 29 20001 2098 Spotted touch-me-not 3 35 19.66 2.93 Crested.wood-fern h 31 19.00 2.83 Bugle Weed 3 31‘. 170311 2058 Sensitive fern 2 13 16.33 2.h3 Field mint 5 31 15.67 2.33 131 No.0f No.0f Total Ground Cover (cont....) Areas Plots Density(E) Rating Red osier 8 h6 15.33 2.28 Common thistle 7 35 15.32 2.28 Joe-Pyedweed 6 3O 1h.66 2.18 Mad-dog skullcap 3 27 1h.33 2.1h Equisetum litorale 3 29 13.67 2.0h Currant and Gooseberry 7 hl 13.67 2.0h Fringed polygala 7 35 13.66 2.0h Virginas-bower 6 28 13.32 1.98 Common skullcap 3 29 13.00 1.9h Loosestrife h 28 12.67 1.89 Avens 8 26 11099 1079 Sowothistle 5 3h 11.33 1.69 Alsike clover 3 22 11.33 1.69 'Watercress 1 1h 10.67 1.59 Balsamsfir 5 28 10.66 1.59 Black cherry h 30 10.00 l.h9 Twinflower 3 12 10.00 1.h9 Lycopusrrubellus 1 19 9.67 1.hh ‘Woodahorsetail 2 22 9.33 1.39 Shinleafs 6 13 8.65 1.29 Wild sarasparilla 6 17 8.32 1.2h Poison ivy 3 12 7.33 1.09 False lily-of-the-valley 5 20 7.32 1.09 Goldthread 2 17 7.00 loOh Blue flag h 10 6.66 .99 Speckled alder 7 19 6.33 .9h Blueberry 2 9 6.33 .9h Viburnum 3 18 5.99 .89 Spruce 3 17 5.67 .8h Grapeefern 2 16 5.3h .80 SpikO‘I'llSh 2 7 5033 079 Meadowasweet h 1h h.67 .70 Milkweed - 2 12 h.66 .69 Daisy'fleabane 2 10 he 66 0 69 'Willow 2 11 h.3h ‘o65 Pearly everlasting 3 9 h.33 .65 Moomrt 5 13 11032 061‘ American elm. 2 12 h.OO .60 Juneberry 1 11 3.67 .55 Eggggus americanus 3 9 3.66 .55 T leweed 2 10 3.33 .50 Honeysuckle 3 10 3.33 .50 Yarrow 3 8 3.33 .50 Yellow cress l 8 3.33 .50 Marshébellflower h 10 3.32 .h9 Anemone 2 9 3.00 .h5 Blazmg‘Star l 9 3000 0’45 Broadwleaf anemone 2 7 3.00 .h5 Foamflower h 9 2.99 .h5 132 No.0f No.0f Total Ground Cover (cont....) Areas Plots Density(E) Rating_. Buckthorn h 9 2.99 .h5 Ninebark 2 8 2.67 .hO ‘Wing-stem.monkey-flower 3 6 2.67 .hO Teaberry (wintergreen) 1 h 2.67 .hO Pill'Cherry 3 8 2 0 66 0’40 Quaking aspen 3 8 2.66 .hO Stinging nettle 2 6 2 .66 .110 Red.maple 3 7 2.33 .35 Dogbane 2 7 2033 035 Common sorrel 1 7 2.33 .35 Dwarf dandelion 1 7 2.33 .35 Choke-cherry h ' 6 2.00 .30 White bir0h 2 6 2000 030 White cedar 1 6 2.00 .30 Starsflower 2 h 2.00 .30 Selfheal 3 )4 2 000 030 Duckweed 1 1 2 .00 .30 Putndge "berry 2 6 2000 030 Balm-Of‘Gufiad 3 6 1099 030 Bush-honeysuckle 2 5 1.67 .25 Groundsel 1 5 1067 025 Square-stem.monkey-flower l 5 1.67 .25 Beaked hazel 1 5 1.67 .25 Marsh-marigold 2 2 1. 33 .20 ‘Wormwood . 1 h 1.33 .20 Highbush-cranberry 2 h 1.33 .20 Swamp-birch 1 2 l. 33 .20 ‘White lettuce 2 2 1.33 .20 Red baneberry 2 2 1033 020 Silvery spleenwort 1 2 1.33 .20 Rue-anemone 3 h 1.33 .20 Nightshade 1 2 1.33 .20 Toadflax 1 2 1.33 .20 Bulrush 1 1 1.00 .15 Cornelily 2 3 1.00 .15 Cardinal ”flower 1 3 1000 015 W‘ter‘inat 2 3 1.00 015 Blackberry 2 3 1.00 .15 Beech 2 3 1.00 .15 Plantain 1 3 1000 015 Wood-betorw 1 1 1.00 .15 Fringed loosestrife 3 3 .99 .15 Dandelion 1 2 .67 .10 Arrowhead 1 2 .67 .10 White snakeroot 1 2 .67 .10 Black ash 2 2 .66 .10 Trillium 2 2 0 66 010 Fragile fern 1 1 .33 .05 ‘ Northern green orchis 1 l .33 .05 133 No.0f No.0f Total Ground Cover (cont....) Areas Plots Density(E) Rating Poor Rebin's plantain 1 1 .33 .05 Painted-cup 1 1 .33 .05 Tamarack 1 1 .33 .05 False Solomon's-seal 1 1 .33 .05 Turtlehead 1 1 .33 .05 Stick-tight 1 l .33 .05 Common elder 1 1 .33 .05 White sweet clover 1 1 .33 .05 Common.mouse-ear chickweed 1 1 .33 .05 Trailing arbutus 1 l .33 .05 Wild columbine 1 1 .33 .05 Soft rush 1 1 .33 .05 White pine 1 1 .33 .05 Old-field-cinquefoil 1 1 .33 .05 Pussyis-toes 1 l .33 .05 Wild bergamot 1 1 .33 .05 Blunt-leaf orchis 1 1 .33 .05 "Loafing Areas"** No.0f No.0f Total Trees Areas Plots Densi_ty(n) Rating_ Quaking aspen h 77 107.00 35.63 Red maple 3 m. 51.00 16.98 Spruce 1 23 25 e 67 8 055 White pine 3 31 22.67 7.55 Jack pine 3 30 20.67 6.88 Red pine _ 3 26 18.00 5.99 Large-toothed aspen 3 18 15.33 5.10 White birch 3 17 13.33 mm. Balsam-fir 2 111 10.67 3.55 Pin-cherry 3 15 9.66 3.22 Black Chem 11 13 8099 2099 Junabem 3 6 11000 1033 Red oak 2 3 1.67 .56 Willow 1 S 1067 056 Choke-cherry 1 3 1.67 .56 Speckled alder 2 2 .67 .22 Balm-Of ~Gilead 1 2 .67 .22 Beaked hazel 1 1 .33 .11 Shrubs Blackberry 3 61. 67.33 22.142 Sweet-fern 2 31 26.66 8.88 Beaksd hazel 3 35 25.99 13h No.0f No.0f Total Shrubs (cont....) Areas Plots Density(E) Rating_ Black cherry h h2 18.67 6.22 Quaking aspen h 39 16.33 Sold. Willow u 20 13.67 14.55 Blueberry L. 20 12.67 8.22 White pine 3 16 8.66 2.88 Red maple 3 11. 6.00 2.00 Spruce 2 9 5066 1088 Bush-honeysuckle 2 6 2.66‘ .89 Red pine 2 6 2.66 .89 Pin-cherry 2 S 20311 078 White birch 1 7 2.33 .78 Large-toothed aspen 3 7 2.33 .78 Speckled alder 2 3 2.33 .78 Spiraea 2 h 2.00 .67 Chokeecherry 2 3 1.66 .55 Jack pine 2 3 1.66 .55 Juneberry 2 2 1.33 .hh Raspberry’ l 3 1.00 .33 Honeysuckle 2 2 .66 .22 Balm-of-Gilead 2 2 .66 .22 Red osier 1 1 .33 .11 B318NMHfir l 1 .33 .11 wud-r‘181n l l 033 011 R080 1 l 033 all Ground Cover Bracken h 96 172.33 57.39 Graee h 96 89.67 29.86 StraWberry h 8h 5h.32 18.09 Goldenrods h 67 37.67 12.5h Teaberry (Wintergreen) 3 h5 30.00 9.99 Aswrs ll. 71 29066 9088 Blackberry 3 53 26.33 8.77 ‘Blueberry h 22 17.66 5.88 Black cherry h 39 12.99 h.33 Beaked hazel 3 38 12.67 h.22 Sweetsfern 2 28 10.00 3.33 Orange hawkweed h 237 9.67 3.22 Fringed polygala. 2 17 9.66 3.22 Sedge 2 13 8067 I 2089 Violet 3 22 8.00 2.66 Red maple 2 20 6.67 2.22 Bush-honeysuckle 2 l9 6.3h 2.11 Lettuce h 18 6.01 2.00 Dogbane h 16 5.3h 1.78 Baleamsfir 3 16 5.33 1.77 Alsike clover 3 15 5.00 1.67 Shinleafe h 12 h.67 1.56 135 No.0f No.0f Total Ground Cover (cont....) Areas Plots DensityKE) Rating_ Pin-cherry 2 1h h.67 1.56 Dwarf raspberry 1 6 h.67 1.56 Quaking aspen h 13 h.33 l.hh Ground-cherry l 6 h.00 1.33 White pine 2 11 3.67 1.22 False lily-of-the-valley 2 9 3.66 1.22 Honeysuckle 3 8 3.33 1.11 Bedstraw 2 10 3.33 1.11 Wormwood l h 2.67 .89 Daisy-fleabane 2 7 2.33 .78 Pearly everlasting 2 7 2.33 .78 Wild sarsaparilla 1 5 2.33 .78 Cowawheat l 6 2.00 .67 Wood—beteny 1 3 1.67 .56 Common sorrel 2 h 1.33 .hh Foamflower 2 3 1.00 .33 ‘Willow 2 3 1.00 .33 ‘Whita birch l 3 1.00 .33 Red osier l 3 1.00 .33 Cinquefoil l l 1.00 .33 Cil‘Sim Hfllii 1 1 1000 033 mm 1 2 .67 .22 'White lettuce l 2 .67 .22 Virgin's4bower l 2 .67 .22 Large-toothed aspen 1 2 .67 .22 Blazing-etar 1 2 .67 .22 Spruce 2 ‘2 .66 .22 Mbonwort 2 2 .66 .22 Partridge-berry 2 2 .66 .22 Puesy'e-toes 2 2 .66 .22 Viburnum. 2 2 .66 .22 Anemone l l .33 .11 Balm-of-Gilead l 1 .33 .11 Sow-thistle l 1 .33 .11 Milkweed l 1 033 011 Raspberry 1 1 .33 .11 Purple meadow-rue l l .33 .11 Yarrow l l .33 .11 Bucmom l 1 033 011 Lady ram 1 l 033 011 Crowfoot 1 1 .33 .11 Trailing arbutus 1 l .33 .11' Juneberry 1 1 .33 .11 Star-flower 1 1 .33 .11 Late coral-root 1 1 .33 .11 Gromd‘pine l 1 033 011 136 No.0f No.0f Total Ground Cover (cont....) Areas Plots Density(E) Rating Rose 1 l .33 .11 False beech-drops 1 1 .33 .11 Common hound"s tongue 1 l .33 .11 Hieracium.Gronovii l 1 .33 .11 * A total of 9 areas and 223 plots were analyzed in woodcock feeding habitat. ** A total of h areas and 100 plots were analyzed.in woodcock loafrng habitat“) 137 made up a single datum. More than twice as many feeding areas were studied as compared to loafing sites. Second, these two types of area frequently were adjacent to each other with a zone of overlap (Plate 19). It was difficult to draw a definite line separating the two areas; consequently, some species were recorded for'both types of habitat when a random sample fell in those overlapping zones. Finally, Judgments of the writer based on personal experience may have been biased. A possible way of differentiating between areas, using vegetation, could be by the combined results from several species. If a corre- lation existed between certain species or among several these plants could be used as an entity to describe an area. Then if one species was missing from the entity in one area the absence would not be crit- ical. Also the entity might be included in statistical tests. Such a method was not explored during this study. g The frequency of occurrence and the rating did indicate certain plants to be characteristic of feeding and loafing covers. -Mcst plants which were found on the areas were listed in Table 31. A.few unidentified plants were not included. This list shows the great variety of plants which do exist in woodcock habitat and those plants which were most abundant. The scientific names are included in the Appendix. The most common species occurring in each plant category were fcund.mpst frequently in certain types of environment. That is, such plants as speckled alder, red osier, sedge, and some violets normally were found in low moist sites of relatively heavy soils and hence on 138 PLATE 19 A picture of woodcock habitat used during the spring, summer, and fall. The lower area of alder, which serves as a feeding site, blends into the upper habitat of aspen-conifer which is used as a loafing area. A zone of overlap is formed where the two sites blend together. 139 feeding sites. Other plants like red maple, blackberry, and bracken fern were prevalent in the dry upland andivell-drained soils and there- f0re found in loafing areas. However, some species such as quaking aspen, grasses of different species, and strawberry were almost as common in one kind of area as on the other. Since the average density of plants was hidden in the rating, the following table was included to show the relative density for each category for the 13 areas analyzed. The maximum density as discussed previously was 3.0. Class Feeding Area Loafing Area Trees 2009 1088 . Shrubs 1.29 ' 1.25 Ground-Cover' 2.2h 2.11 Ground cover appeared to be dense, but actually sufficient bare or low-density spots occurred to allow feeding and freedom of move- ment. In many sites where the ground cover was dense the foliage was short enough to allow movement over the plants or tall enough to allow movement beneath. Strawberries were a good example of the first type of foliage and ferns an example of the latter (Plate 20). Some sites were so densely vegetated that freedom.of movement and feeding must have been difficult for the birds. In such areas few woodcock were found and when they were present they used the more open sites. Some areas observed.were disappearing as woodcock coverts; such places have been considered in more detail in the section on habitat control. Trees and shrubs were open enough to allow freedom of flight although the amount of cover overhead frequently covered two-thirds of 1170 PLATE 20 A. A view through an aspen-birch-conifer stand showing the apparent density of bracken fern. Woodcock were still quite abundant in this area. B. -A close-up showing the flushing site of a woodcock. Bracken fern, although dense above, is sparse enough at the ground level to allow freedom of movement by woodcock. Table 32. ‘Woodcock summer cover as based on vegetation at 222 flushing sites. Plant Speckled alder Quaking aspen Spruce Balm-of-Gilead 'Willow Balsam.fir ‘White birch Black cherry Tamarack Red.maple Elm White cedar White pine Black ash Large-toothed aspen Jack pine Red pine‘ Nannyberry Bracken fern Juneberry Ninebark Red osier Viburnum Spiraea Beaked hazel Pin cherry Basswood Highbush cranberry Green osier Frequency 1% n7 % 36 3t 33 22 20 m m 15 11 11 HHHHHHHMNwwrrWQE lhl 182 the area. In addition to the systematic sampling, information was collected on the dominant overstory at the site where a woodcock was flushed. Usually, this overstory consisted of trees, but frequently shrubs were dominant and on three occasions bracken fern was the only cover (Table 32). Again the species which occurred most frequently were speckled alder and quaking aspen. ..i-In summary, a feeding area seemed to be coVered with those plant species characteristic of low moist sites. However, all such sites were not used by woodcock, probably because the density of the vege- tation was too great or food availability was too low. Some areas used by woodcock could have had a very different plant composition since vegetation varies to a certain extent between sections of the state. The plant composition of a typical loafing area was also characteristic of many areas which never were used by woodcock. Thus, a loafing area was characterized, not only by certain plants, but also by its relationship to feeding sites. Previously it has been pointed out that a sample of 25 plots per acre gave relatively accurate data by which a particular area might be characterized. However, the economics and practical aspects involved in gathering the data were not considered; so an estimate of the cost involved in locating and analyzing woodcock cover was Obtained in order to determine the Optimum.number of plots per area. ,j’ Since speckled alder was the dominant species found in many areas, the ratings for that plant were used to determine a mean square or variance required for a cost analysis. An arc sine transformation was used on the original densities because of the percentage nature of 11.3 the data and the apparent relationship between the means and variances. Resulting data were included in the following tabulation based on 1955 data: Degrees of Sum of Source of error Freedom Squares Mean Square Between areas 11 [19328.8 1082.2 Within areas 120 22,111.11.2 187.0 An estimate of the cost for locating areas was based upon travel expense and time required to survey the area. Twenty such areas were located and checked. The cost was $13.83 per area. To estimate the cost of analyzing an area, which included travel expense and actual time spent examining the cover, data .from 11; areas were used. That cost was $113.13 per area or 81.72 per plot. 3 Using formula )4 and substituting the necessary cost figures and variances, the Optimum number of plots per area was computed as being 6.5. Thus it seems that it would be more practical and economical to locate more areas and analyze fewer plots per area. Since many areas analyzed in 19514 and 1955 were known beforehand the average cost for locating an area was reduced. Future areas to be studied would be expensive to find since presumably the location of such areas would not be known. Consequently, the ratio between cost of locating and cost of analyzing an area would be greater than the present 8.03 and thus the optimm number of plots would increase. A sample of at least-10 plots per area should be considered if such an analysis of vegetation is continued. With the reduced sample per area presumably more areas could, be analyzed. lhh Fall. In most cases the areas studied during the summer were used by woodcock in the fall (Plate 19). Consequently, plant characteristics discussed for the summer period remained.much the same. The excep- tions were that the composition of cover species changed in regard to abundance oflive plants. Most of the same species remained, but in many cases they prevailed as dead or dying plants. ‘With the dying of plants many sites became more open. The composition of trees and shrubs remained the same, but the density of the overstory decreased because deciduous species lost their leaves. Generally, the major difference between summer and fall cover was the decrease in density. Records were made of dominant woody plants which.were found at 221; flushing sites of woodcock in the fall (Table 33). Again speckled alder and quaking aspen were observed most frequently, indicating that most woodcock were flushed in low moist sites or adjacent to such places. Land-us... -1 Openings used as singing grounds resulted from various causes. To determine these causes, two state-owned land tracts, the Rose Lake 'Wildlife Experiment Station and the Gratiot-Saginaw Game Area, were chosen for study. These areas were selected because of the availa- bility of historical records and their nearness to spring headquarters. Information regarding past land-use practices was obtained.from those records and from.persons familiar with the history of the area. Pre- sent land-use practices on the area also were determined. Twenty~one Table 33. woodcock.fall cover as based on vegetation found at 22h flushing sites. Plant Speckled alder Wumguwn Wflhw .2 Spruce Elm Balmrof-Gilead Blackzcherry Red osier White cedar 'White birch Grey dogwood ‘White pine ,Tamarack Highbush cranberry Red pine Balsam,fir Black ash Red maple Hawthorn Basswood Spiraea . Black locust Frequency 1m l& w m N 15 HHHHwawrrrmWNSF 11.5 1&6 singing grounds at Rose Lake and 29 at Gratiot-Saginaw were included in the study. Of the 21 singing grounds examined on the Rose Lake Experiment Station, 2 were in food patches while 19 were on land lying idle. However, 1; "grounds" were adjacent to food patches and 1; others were in areas that had been planted to pines and staghorn sumac. Ten singing grounds were located in sites famed prior to 19146, while 2 other "grounds" were in areas cleared and broken about 1938 for farming but never used except for pasturing. Nine other places were pastured until about 19140. One of the areas farmed had been a fruit orchard earlier. When the above areas were purchased by the Department of Conservation, most of them were taken out of cultivation of any sort except for a few food patches and the scattered pine and shrub plantings already mentioned. ' Since most of the areas have remained idle for ten or mere years some woody vegetation has returned. In most of the 01d pastures and farms, where plant succession has been slow, only scattered shrubs and small trees have reappeared. In two portions. of the tamarack swamp which were cleared for farming, the invasion of woody plants has (eliminated most open sites. .For the Gratiot-Saginaw Game Area the history of most of the original clearings has been primarily one of fires. In 1871, at the time of the great Chicago fire, many fires were flaring up over much of the Lower Peninsula. On October 10 of that year, the Saginaw Valley was swept with.rire and four days later fires were reported increasing in Gratiot County (Mitchell and Robson, 1950).’ Other large 1h? fires in 189h'and 19h0 swept much of the area, while smaller fires between the 1871 and 19h0 fires were scattered through the area. Much.of the merchantable timber was removed by lumber companies. _ One of the early lumber'mills (Potter's Mill) and a small community’ were located in a clearing which still exists. One edge of this clearing has been used as a singing ground at least for the past four years. Some farming was attempted in various parts of the area, so a few of the clearings were maintained or reopened by farmers. Eight singing grounds were found on sites which were cleared and farmed prior to l9h0. Eight other "grounds” were located on tracts of land _ which had been pastured at some time. One of these tracts of about 20 acres, which was broken prior to 1950 for farmland but was used only for pasture, contained three singing grounds. - Meet state land involved was Obtained through purchases or tax reversion in 19hl$ a few forties were added in l9h2, 19h3. and l9h7. Since the land was taken over by the state, it has remained idle except for a few food patches and occasional pine plantings. Three singing grounds were located adjacent to food patches and at least seven "grounds" were in or near sites of scattered pine plantings. Agricultural practices and fires in the Rose Lake and Gratiot- Saginaw areas were responsible for clearings which developed into suitable sites for woodcock breeding activity. These clearings, in most cases idle for a decade or so, have retained their open aspect because of slow'woodydplant succession. Decreased availability of soil nutrients, and plant competition prObably have been the major 1&8 factors causing the slow succession. Agricultural practices and.fires accounted for many clearings used by woodcock throughout the state, but other factors hate been equally or more important in restricted areas. Lumbering Operations Opened many sites which were maintained as clearings by fires, various agricultural usages, loss of soil fertility, etc. Other clearings were caused by mining operations, road construction (including rail- roads), power and gas lines, and natural stream.erosion. Soil - A.great variety of soil textures were found in habitat occupied by woodcock. These ranged.from.sandy soils to organic soils of muck and peat. Because of this variety soil types alone could not be ‘ considered as a limiting factor to abundance and distribution of ‘woodcock. V to At least 156 sites being used by woodcock were checked for soil texture (Table 3b). A large portion of these soil types were iden- tified by Cash wcnser, land-use specialist of the Game Division, Department of Conservation. In the remaining sites the.soils were identified on the basis of descriptions presented in soil survey reports. ._ Breeding grounds were located primarily on well-drained soils such as Rubicon and.Rose1awn sands, Hiawatha.fine sandy loam, Oshtemo loamy sand, and Onaway loam. Of the 80 singing grounds checked, h2 Table 3h. Types of soil on which woodcock were found. Type of Site Singing ground Nesting Brood Feeding Loafing Totals Sand 29 1 l O 6 37 Loauw Sand 10 l O l h 16 SandyLoam. 16 3 3 8 6 36 Loan 17 O l 6 O 2).; Clay Loan 0 1 O O O l Silt Loan 0 O O l O 1 Muck 2 l 3 1h 1 21 Peat 6 3 o' 11 o 20 Totals 80 10 8 hl 17 156 1&9 150 were on such soils. Imperfectlycdrained soils, including Newton sand and loanw sand, Saugatuck sand, and Brady sandy loam, were the site of 27 other "grounds." Three singing grounds were located on moderately- drained Berrien loamy sands, while the remaining 8 grounds were on poorlyhdrained soils such as Rifle peat. During the past three years at least 5 active ”grounds" have been observed each.year in a large area of Rifle peat on the Rose Lake EXperiment Station. One of these "grounds“ has been occupied all three years. ’ Of ten nesting sites analyzed, 5 were on well-drained soils, 1 on an imperfectly-drained soil, and the other h were on poorly-drained soils. Three of those nests on the latter type of soil were in the Rifle peat area discussed above and thus near singing grounds. Evidently, females did not show any preference for soil textures. 3. Broods remained on or near the heavier and wetter soils, probably because of feeding habits. Four broods were found on undrained muck and loam soils while three others were on the imperfectly-drained Brady’sandy‘loam. .theeding areas studied were on moist soils. Thirty-three of the bl feeding sites occurred in poorly-drained situations which included such soils as Rifle peat, Carbondale peat and.muck, Lupton and Kerston muck, Griffin loam, and Munuscong sandy loam. Of the remaining 8 sites, h were on moderately-drained soils, 3 on imperfectly-drained soils, and l on a well-drained soil. Occasionally, woodcock were reported feeding on dry upland sites considerable distances from.moist areas. The soils on such sites, though unchecked, probably were of the sandy types. If woodcock were 151 feeding on those areas they undoubtedly were after insects and not earthworms. . Loafing sites were more difficult to define since woodcock fre- quently used one area for both a feeding and loafing site. .However, some places were used strictly as loafing areas. These places usually were dry upland sites adjacent to feeding sites. Ten so-called loaf- ing sites were on well-drained sandy soils, 6 were on imperfectly- drained sandy soils, and l was on a moderately-drained soil of thin muck over'sand. Woodcock could be seen flying from those places in the evening to adjacent moist sites of poorly-drained soils. The type of soil found in an area was not by itself a restrict- ing factor to the presence of woodcock. Factors associated with the soil such as moisture content, earthworm abundance, and relation to other soil textures were more important. As a result of the extensive glaciation.which occurred in Michigan, the soil types within an area are quite‘varied and provide most of the necessary combinations of soil factors required by woodcock. ‘Soil moisture was a factor which determined where most of the activity would be. A.majority of the singing grounds were on drier sites while most feeding areas were on poorly-drained soils. If an area became too wet, birds vacated that location. Likewise, as dis- ‘cussed above, birds concentrated near or in the more moist situations in the summer and.fall when drought conditions existed. Soil pH was not recorded for’most areas examined, and many de- scriptions in soil survey reports omitted the chemical aspects. Howe ever, soils ranged from strongly acid to slightly alkaline with acidic 152 soils predominant. Of 62 soil samples taken from 13 loafing and feed- ing sites in the Pigeon River Area, only 5 samples had a pH greater than 7.00. Eleven samples had a.pH of 6.90 or greater. As shown in Table 35 there seemed to be little correlation between pH, and the type of area and abundance of earthworms. Vegetation did not become a limiting factor to the presence of woodcock unless the stands became-too dense or old. Usually, the types of vegetation occurring on a site was a result of the soil factors. Thus vegetation was often a good indication of the soil type which existed at that location. Food 1 The diet of woodcock which were examined consisted.primarily of animal food, particularly earthworms. Pettingill (1936).found earth- worms more frequently in woodcock stomachs than any other food. Approximate percentages given by Pettingill (1939) were: earthworms 86, arthropods 10, and seeds h. Other reports have placed the per- ' centages of earthworms above 85 per cent (Aldous, 1939, and Mendall and Aldous, l9h3). Sperry (l9hO) reported that for 261 specimens. taken from March through.December, earthworms constituted 68 per cent of the total food content. The lowest percentage of earthworms in the diet occurred in.December -- h8. In October, 82 per cent of the food was earthworms. ~The seed content increased to 28 per cent in September while insects were found in greatest numbers in August. 153 Table 35. Relationship betwun earthworms, soil pH*, and woodcock habitat. Type of pH Avg. Earthworms Area Date Site Range EH Per Sq.Ft.** Species of Earthworm I 8/13/Sh Loafing 5.7-6.1. 6.0 1.1 1 1.1. Lumbricus rubellus 11 9/ 1/5h Loafing 5.5-5.6 5.5 1.5 1 2.0 Lumbricus rubellus v 8/21/5h Feeding 7.2-7.3 7.2 6.5 :. h.3 Lumbricus rubellus VIa 8/lB/5h Loafing 6.h-6.7 6.5 1.6 i 2.0 Lumbricus rubellus VIb 8/19/511 Feeding 6.h-6.6 6.5 9.3 f. 5.9 Lumbricus rubellus v11 11/12/51. Feeding 5.2-6.9 5.9 22.7 112.1. Lumbricus rubellus VIII ll/13/ 5h Feeding 6.h-6.7 6.6 .2 1 --- Lumbricus rubellus 11 7/22/55 Feeding 6.0-7.0 6.1. 8.2 1 6.1 Lmnbricue rubellus x 7/25/55 Loafing 5.6-6.0 5.8 1.6 1 1.8 Lumbricus rubellus x1 8/ M55 Feeding 6.8-7.1 6.9 3.8 1 3.9 Lumbricus rubellus x11 8/ 5/55 Feeding 6.6-6.9 6.8 2.2 1 2.8 Lumbricus rubellus XIII 8/30/55 Feeding 5.0-6.9 6.0 19.8 119.0 Lumbricus rubellus XIV 9/17/55 Loafing 5.0-6.2 5.8 1.0 1 1.1 fisiflfitfsfiigg * pH was determined with a Beckman pH meter. H- Based on twelve samples in each area. 15h Although a large number of woodcock specimens were collected in the Michigan study and the gastro-intestinal tracts were preserved, a detailed analysis has not been made of the food content. Generally, the stomach contents included earthworms, insects, seeds, and vegeta- tive parts of plants. Earthwoms composed the greatest volume; this was especially evident when worms were found in the esophagus and proventriculus. If earthworms had reached the stomach, digestive pro- cesses had destroyed much of the body structure. The same was true for other soft-bodied animals such as insect larvae. -..- Insects, particularly beetles, were found frequently, but the volume was small compared to the abundance of earthworms. Other animal forms included sow bugs and spiders. .... Plant material was as abundant as any other item, but it is doubtful that the material was always taken as food. Seeds, which in some cases were numerous and of a variety of species, could have been taken intentionally. It was believed that, in many instances, plant material, including seeds, was taken unintentionally while the bird was probing. Grit occurred in the stomach of every specimen examined for parasites. This grit could have been taken unintentionally also as the bird probed for food or it may have come from earthworms. How- ever, woodcock were observed on several occasions picking In: small particles, evidently grit, in roads and trails during the twilight hours. Earthworm availability studies were made to determine differences between loafing and feeding areas. Several methods were used to _ 155 determine the availability and abundance of earthworms. These methods included various dosages of potassium permanganate, mechanical vibra- tions, electric currents of varying intensities, and the digging of circumscribed plots. All but the last method were discarded early in the study since most soil situations required different dosages, boundaries of the area affected c0uld.n0t be determined, and earth- worms could have gone down or to the sides as well as to the surface. Some earthworms probably went downward as soon as the soil was disturbed by a shovel, but more worms were found when plots were dug than with any other method. The plot was one foot square and three inches deep. Only three inches was removed since most woodcock could not probe beyond that depth. Twelve plots, chosen on the basis of the random number sampling technique employed with vegetative plots, were dug in each area analyzed.l For five loafing sites the average number of earthworms per square foot was 1.h, while the ayerage for seven feeding sites was lO.h, a significant difference as indicated by a."t' value received when a comparison of the two types of areas was made. The data for Area VIII were not included because the sample taken from the area was not repre- sentative. .An analysis of that area could not be made until after a high water level which flooded the site subsided. The water in combia nation with.freezing temperatures may have affected the distribution of worms. Plots dug adjacent to the sampled area, and where high water did not reach, contained a higher earthworm.p0pulation. Earthworm.counts for Area.XI and XII seemed too low for a feeding site, especially when compared to the counts from the other feeding 156 places. However, woodcock were seen feeding at the two locations dur- ing the day, other birds were seen flying into the areas at twilight, and probings were numerous throughout. The areas were not the most productive sites for earthworms, but much of the ground was soft and birds could probe with ease. Even though counts were low, more earth- worms were found there than in any loafing site. weather of the many weather factors the following affected the distri= bution, abundance, or activity of woodcock during the summer and fall: temperature, wind, precipitation, and light intensities. Information regarding these factors was taken from general field observations and records maintained by the U.S. weather Bureau. Temperature. Heat was not known to affect woodcock directly, although it may have been a.factor causing:min0r local movements. Indirectly, heat helped determine the distribution and activity of many birds by dry- ing up some areas. Cold had a greater effect on woodcock. Even in the summer months twilight flights preceding evening feeding were reduced in length of time and number of birds participating when temperatures fell below hS° F. In late summer and fall cooler temperatures affected.local movements and migrations. One reason given for the woodcock movement 157 from the West McMaster Creek area in early September (1955) was the sudden decrease in temperature. The average daytime temperatures remained approximately the same as in August, but night temperatures drapped sufficiently to cause frost conditions. Such a sudden change could have caused some woodcock movement. . Although occasional population increases in the Pigeon River Area and Region II preceded or followed a cold front, more often the in- crease occurred on the same day as the lowered temperature. Tempera- ture and precipitation data from Sault Ste. Marie, Newberry, Mackinaw City, Pigeon River Research Station, and East Jordan were used to indicate critical weather conditions. for the eastern part of the Upper Peninsula and the northern half of Region II (data included in the Appendix). Since the distance between the extreme locations was only about 100 miles, population fluctuations in Regions I and II could have been caused by the same cold spell. As autumn progressed, increasingly colder temperatures seemed to have less effect on woodcock. Several birds were present in the Pigeon River Area in the latter part of the 1951; hunting season even though temperatures fell as low as 1h° F.3 whereas, earlier in the sea- sm pepulation decreases were noticed when temperatures dropped to h0°. Toward the end of both the 1951; and 1955 hunting seasons, a definite flight occurred in the Pigeon River Area, indicating that some woodcock had remained farther north regardless of the lower average temperatures. Wind. Periods of northerly winds or of no wind favored woodcock migra- tion; whereas, moderate or strong southerly winds seemed to hold back 158 movements. As indicated above, movements occurred most frequently Just prior to or during periods of decreasing temperatures. These periods of decreasing temperatures and the northerly winds occurred at the same time. Some population fluctuations took place when tem= peratures were mild and light southerly winds prevailed. Imhoff (1953) showed that during spring migrations in the south, clear warm days with moderate southerly winds were most favorable to migrants while cold fronts were most unfavorable. Others have deduced from various studies and observations that in general spring migra- tion was related to the movement of air masses (Lowery, 1951; Bullis, 1951;; and Douglass, 1955). In fall migration, birds in the north should be affected similarly by opposite conditions. Mazzeo (1955) reported that birds seemed to arrive at Matinicus Rock, Maine, regard- less of the wind"s direction, but that the greatest numbers arrived during very strong northwest winds. His observations were made from August 27 to September ll, 19,49. fiWoodcock activity during the evening and morning flights of summer and fall were correlated with wind conditions. If the wind, regardless of direction, exceeded 10 miles per hour, fewer woodcock were seen flying; when the wind exceeded 15 miles per hour very few, if any, woodcock were observed. Strong winds during the day had no effect on woodcock since birds remained on the ground most of the time. Precipitation. Rain and snow also affected the distribution, activity, and move~ ments of woodcock. The effect on distribution and activity was more 159 evident on local papulations. During the summer and fall normal pre- cipitation helped maintain many feeding sites by keeping the soil .moist. Below-normal precipitation caused a decreased moisture content in many places and thus caused woodcock concentrations. Above-normal precipitation caused a wider distribution of woodcock since more moist areas were available as feeding sites. Much rain fell in the Pigeon River Area in the latter part of the 1951.; summer. As a result, much woodcock habitat, especially feeding areas, was flooded. Consequently, woodcock were found scattered along the edge of flooded sites and in swales of adjacent areas. In the same period of time in 1955, less rainfall occurred and woodcock were more concentrated“. Similar situations occurred in many other locations. The effect which precipitation had on the flushing distance of woodcock varied. Some birds flushed more quickly and at greater dis- tances when the vegetation was moist or when it was raining. Generally, however, woodcock were more difficult. to flush during rainy conditions, their ascent was less lively, and the distance of flight was shorter. Evening and morning flights definitely were influenced by rain. A light sprinkle had less effect on woodcock activity than a rain, but even with the light precipitation fewer woodcock were seen flying in the areas observed at Pigeon River. A heavy downpour or a slow steady rain were most effective in preventing woodcock flights. Very few birds, and frequently none, were seen if a lapse in the rainfall occurred during the normal flight period. Evidently, the wet vegeta- “tion, high humidity, or dark cloudy sky influenced flight activity even though rain was not falling. 160 Precipitation produced the same results in regard to migration; that is, woodcock did not move during rains. If birds were found in . an area one day and then it rained that night, birds could be expected in that area the following day. If they were not there then they prObably had.moved Just before the rain. The major precipitation often followed the change in temperature by one or more days. Therefore, the population fluctuations discussed previously in relation to tame perature decreases and northerly winds took place prior to the pre- cipitation which often was a result of the changes in other weather conditions. However, the population did not always decrease prior to a heavy precipitation. On several occasions the population was at a peak during the rainy spell, which would tend to substantiate an earlier statement that woodcock did not move during rainy weather. Nebulosity and Light Intensities. Neodcock activity, especially evening and morning flights, was influenced by light intensities. 'Uhen skies were clear, woodcock began their evening flights shortly after sundown. Morning flights were begun shortly before dawn. But on days when the percentage of cloud cover was high, flights started earlier in the‘evenings and later in the mornings. The relationship between woodcock activity, cloud cover, and light intensities was similar to that discussed for the breeding papulation except that the moon was less important. Evening flights did not last any longer when bright'moonlight condi- tions existed. 'Woodcock pepulations fluctuated regardless of the cloudiness or 161 light intensities, not including periods of precipitation. Birds may have moved.more on moonlight nights, but there was no indication of such activity at Pigeon River. Areas were found empty of woodcock fol- lowing a cloudy night as well as a clear moonlight night. No compara- tive data, other than incidental observations, were available. Barometric Pressure. During the fall, fluctuations in woodcock populations at Pigeon River seemed to be correlated with changes in barometric pressure. These population fluctuations occurred at or near the time of pressure changes but not necessarily in the same direction. Sometimes the popu- lation seemed to increase as the pressure increased and at other times under similar pressure trends the population decreased; the population fluctuated similarly with high pressure conditions. Even though the woodcock population did seem to fluctuate with pressure changes, the number of birds present in the Pigeon River area did not seem to be any greater with one pressure condition than the other. In case of the European woodcock, Schenk (1931) reported that in the spring this bird generally appeared in Hungary when a low atmos~ pheric pressure existed over England; however, such a relationship was not evident for the fall migration. Other Factors flnpetition o .5“ All intraspecies competition observed occurred in the spring breeding season. At that time some territorialism was displayed on 1.62 the singing grounds. The strong territorial system and defense common to many other birds was not evident among the woodcock observed. Although no actual combats were seen, a few observations were made of certain actions which simulated defense of an area. A male bird was performing on a singing ground at Rose Lake when another woodcock flew directly over. The performer arose and darted after the passing bird for a short distance before returning to assume the courtship display. During the short defense a rough cackling noise was uttered by the defender. On another occasion a male was observed pursuing another bird which had entered the area being occupied. Both birds flew directly upward with the defender literally "on the tail" of the intruder. The two birds appeared as butterflies flitting upward with the defend- ing bird making consistent and continual passes at the other one. Finally, the intruder faded off in the direction of another singing ground. The intruding bird may have been the occupant of a singing ground less than 100 yards away since no activity was heard in that direction while the two birds were in the air. Several singing grounds were located close enough for the male birds to overlap in their flights, but only a few cases of antagonis- tic action were seen. William.Goudy made an interesting observation one evening on a singing ground when he attempted to imitate the peent of a performing male. This bird immediately rushed toward the sound, making low cackling noises and ruffling its feathers. This display was repeated each time an imitation peent was given. 1.63 Mendall and Aldous (l9u3) reported on several observations cone cerning defense of the breeding area. They felt that males were very Jealous of their display site and seemed to be more disturbed by the presence of other male woodcock than by human intruders. Pitelka (l9h3) observed frequent instances of territorial defense and competi- tion. However. much intraspecific tolerance was noted by Sheldon (1953) in studies on wcodcock in Massachusetts. Even though singing grounds were defended at times by male birds.9 defense of diurnal coverts was not observed. Two male birds used the coverts illustrated in Plate 15 consistently in l95h. Frequently these two birds were flushed almost together. and on several occasions both were heard peenting in the coverts prisr to their departure for individual singing grounds. Two other similar snail and compact diurnal coverts were found being used jointly by at least two males. No information was gathered pertaining to nesting territories. but Mendall and Aldous (l9h3) found little evidence of territorial instincts in females.“ Some nests were found only 25 feet apart though the usual distance seemed to be at least 200 yards. Territorial displays were not observed among woodcock on feeding areas. In fact. woodcock concentrated in certain places to feed dur~ ing the summer and fall. In such places there could have been a direct competition for the available foed supply. Several people have mene tioned the voracious appetite of woodcock (Audubons 183S-mgf. Pettin~ gill. 1936; Forbush, 1925; and Robert Whiting, l95b 9“ personal intern view). However. an analysis was made of the abundance of earthworms "I in several places and a large number of worms were still available after birds had deserted the site. Thus the woodcock must not hav been seriously affected by any intraspecific competition for food. The major interspecies competition involved man. In no instance was any other animal observed in competition with woodcock for space or food. But man and his activities often affected woodcock either directly or indirectly. Such activities as farming. construction. etc., mentioned earlier, often interrupted singing ground displays or re- moved the "grounds" entirely from further use by woodcock. Fires. lumbering‘9 agricultural practices. and other land uses destrqyed diurnal and feeding coverts. The period of time for which some sites ‘would.be usable by woodcock was decreased by shrub and tree plantings. However, such activities as those above were not unduly detrimental to the species. except in local situations, since other sites were available or being made available by similar or other man-influenced enterprises. Predation. Predation was not found to be a restriction to the distribution and activity of woodcock. In the three years of field work only a few cases of predation were recorded. The destruction of two nests by red fox or skunks has been mentioned. The only other records of predators catching woodcock concerned birds taken from clovermleaf traps. A weasel entered one trap and caught a woodcock9'while another woodcock was taken by a raccoon. Two ruffed grouse were eaten in the same trap by a raccoon. The predation at that site ended when the raccoon was caught and removed. 165 A bobcat partially destroyed a trap one night in an unsuccessful attempt to catch a woodcock in the trap. No other signs of a wildcat were observed again in the vicinity of trapping operations. That a variety of animals do prey upon woodcock has been shown by the studies of Pettingill (1936) and Mendall and Aldous (1916). Such mammals as the domestic cat;9 red squirrel, fox, weasel, skunk, and dog were listed as actual predators. The domestic cat was rem sponsible for 18 of h2 woodcock listed by Mendall and Aldous as being taken by predators. They considered the cat as the most serious predator on woodcock. Birds of prey listed by Mendall and Aldous included the great horned owl, longneared owl, sharpshinned hawk, goshawk, marsh hawk, duck hawk, pigeon hawk, and redeshouldered hawk. A compilation of the food of predaceous animals by Latham (1950) also included a record of one woodcock taken by a Coopeer hawk in Pennsylvania. Two reptiles, the black snake and a snapping turtle, were men- tioned by Pettingill as preying on woodcock. Two blue racers were observed consistently on two singing grounds at Rose Lake. Neither . snake influenced the activity of the singing male although the bird landed within a few feet of the reptile. Diseases and Parasites. HDiseased.woodcock seldom were noticed. Such a condition may not have been common or it may have gone unnoticed because such specimens were not handled. Diseased or dead birds were not found in the woods. Two specimens were collected which exhibited a diseased condition. 166 ’ One bird had a fungus or tumorous growth on the breast beneath the right wing. The other bird had scattered tumors in the body cavity involving the intestinal tract or the peritoneum and mesenteries. Neither of the two diseases was identified. Parasites were much more common; in fact, the percentage of internally parasitized birds was high. No detrimental effects were observed as a result of such parasitism. ln l9Sh, 50 birds were examined, h6 of which were parasitiaed with one or more of the three major types of internal parasites: cestodes, nematodes, and trematodes. Of the 65 specimens examined in 1955, 63 had internal parasites. In l9Sh, only the readily apparent parasities were removed from the specimens, while in 1955 an attempt was made to take all parasites from the intestinal tracts. The above rate of infestation was greater than that observed by Aldous (1937) and Mendall and Aldous (l9h3). In the former publica- tion, Aldous reported that no parasites were found in 100 specimens studied. In the latter writing, 93 of 258 specimens were mentioned as being parasitized at a rate of one to twenty parasites per bird. ‘Woodcock'were more severely parasitized during the first part of the hunting season. In 1955 specimens, approximately 63 para- sites were found per bird during the first 15 days of October, while an average of 37 parasites were found per bird the remaining part of the season. Nematodes occurred more frequently than flukes or tapeworms, but the latter two (tapeworms especially) were more abundant when they did occur (Tables 36 and 37). Tetrameres was the parasite responsible 167 for the high frequency of occurrence of nematodes. Forty-three of the 8S woodcock with roundworms had Tetrameres. Frequently, two or more of the major types of parasites were found in one bird (Table 36). Parasites occurred.most commonly in the small intestine (Table 38), although almost as many were found in the stomach (proventriculus and gizzard). The parasites found in the gizzard, particularly the tapeworms and flukes, normally do not live there. These animals probably moved into that organ after the bird was dead. Some para- sites found in other organs also could have been out of their normal habitat; this was true especially of the roundworms found in the body cavity. Since species identification of most parasites could not be made, specimens were forwarded to the Agricultural Research Service at Belts- ville, Maryland. Some parasites were identified tentatively before the specimens were sent. Of the cestodes, only the genus fiymenolepis was identified as being in the woodcock; however, at least three different species seemed to be present. Of these species, two were microscopic in size. Aldous (1938) reported finding cestodes of Hymenolepis in woodcock taken in.Maine and New Brunswick. Some other tapeworms reported for the American woodcock have been Anomotaenia variabilis (Ransom, 1909, and Rankin, l9h6) and Anomotaenia stentorea (Fuhrmann, 1932). Nematodes found in Michigan woodcock included the aforesaid Tetrameres sp. (both males and.females) and.Dispharynx spiralis from the proventriculus. Capillaria sp. was found in the small intestine. Roundworms of the superefamily Filaroidea were taken from the body Table 36. Frequency of parasitization in the woodcock. Number of woodcock 1951; 1959 Total Examined 50 65 115 With trematodes 18 hl 59 With cestodes 18 I45 63 With nematodes 36 1&9 85 With other parasites 1 l 2 Without parasites h 2 6 With one type of parasite 2h 16 )40 With two types of parasite 17 22 39 With three types of parasite 5 25 3O Table 37. Abundance of parasites taken from 1955 specimens. 168 Trematodes Cestodes Nematodes Total number 898 2222 301 Maximum number in one woodcock 188 h72 2).; Minimum number in one woodcock l 1 1 Average number per woodcock 13.8 314.2 h.6 169 edeo um. [Heme Ame zoocoomwv Hmuu “mm soonoooxvl Conceeocoo om HseoesmH ememmwamm H: zooaooox. eomem hHHm zooaoooww ammo: mecca: ammo: woman: ammo: mocha: weomdwos stwom Atoesm SOHSm prmmm zoeSm geese chwmm tomes toeam mmoevmmcm o o u o N HH 0 m Ht peo40. so. ooosewoo eoemw >Hmos oosssw. mango Sea. 5 mo .03 mm 8 .30 we on .rom >Hpomms 00.. a H . 0H2 b. mm. H Hm .omu m NV .NNN >Hpomms 00.. * N omonm 00Hses m Hm .Huw a Hm .uum >HHemws 00.. $ u . SmsHHcm V N? .NwN H N: .okm mesons 00.. b.>sme Nw Hm H.¢um wmsmmm 00.. . owems newer V 0 .VV@ mouse 00.. Mmsxoe meewsmm um uo H.Hoo . Ho mo .muw oseoowmms 00.. . ews wswdme Hm Hm H.0um oswopozm 00.. mossmtooo cu mo H.¢ww mH uo H.woo uo mm H.wot OUHUUmzm 00.. mmsww mowse mo wm .mwm o H: .mtu oswppozm 00.. assume m Hm .uoo m Hm .uoo osweemzm 00.. eHHmos moon u H: .uoo osweeozm 00.. womHmao memsmnoo 00.. >smsmdm were 00.. 0detHs HH000 0003m: 00.. 0000stooo woma osfiosmmos 00.. 0 H z. 0050000H woma oseosmmos 00.. 0 N 300002000 620. 000000.00.. * H 0000 A»: 000H0v 000000 00.. 0 N @000 bmwm ma. Amemos deewv meemnso HmHo 00.. 000000 0H00. womoosaos 00.. 0 H boxHow . womooasos 00.. * N ansonos womooasos 00.. 0 0 mosmswos 00x0 tH H0 NH N: 00 H0 N0 .N00 H.Huw H.000 e Vflm 00 or H0 H0 N0 N0 00 N0 H0 N0 H0 NH NH N: H0 .N00 N.H00 N.¢t¢ H.¢0H .000 H.000 .000 .0H0 H.HNO H.N00 00 NH NH H0 H0 H0 H0 5 E HN HH .000 H.000 H.000 .aoo p.05 .08 H.Hmv H0000 :3 .mmo .0H¢ H.000 .0N0 203 00:00Hosmwd 00.. UHHWMm ww<0s Amesee womsmev 00. 000003 00.. mmoHom HN N: .000 40s wssms 00.. >H30sm :N 00 H.N00 N0 N0 .000 H.000 H0 5. m0 m: :0 omww ONom .000 20h Table h2. Records Of woodcock banded in Michigan, 1953*1955. Band Date of Number Age Sex Banding Place of Banding; Sl3-96901it L - 5/10/53 5 mi. N. of Alba, Antrim County 513-96902 L - n u 513-96903 L - n u n 513-96913 A M 5/ 3/51. Rose Lake Exp. Sta., Orange, Shiawassee $13-9691u A M 5/ 6/5h " " " " 513-96915 A M 5/11/5h " n White, Clinton CO. 513-96916 A M 5/15/51. Gratiot-Saginaw Game Area, Gratiot Co. 513-96917 A M 5 16/51. Rose Lake Exp. Sta., Orange, Shiawassee 513 =96918 A M 5/18/ Sh " ' 513-96919 A M 5/21/51. " Brown, Shiawassee $13-96920 L - 5/28/Sh " " 513 -96921 L - 5/28/Sh " " " Sl3-96922 A F 5/28/5h " " " SIB-96923 A M S/BO/Sh " " " 513-9021. A F S/BO/Sh " . " " 513-96925 - M 8/2h/5z. Tin Bridge, Cheboygan County 52303501 - F 9/21/51. " " 52303502 - F 9/25/5h .. " 52303503 «= M 9 27/Su " " 513-96901. A M 6/55 Rose Lake Exp. Sta., Clinton County 553-3801 A M M27/55 Pigeon River Exp. Sta., Otsego County SS3-38002* A M M28/55 " n u 513-96905 A F 5/ 2/55 Rose Lake Exp. Sta., White, Clinton Co. SIB-96906 A M 5/ 2/55 " 553-38003 L - 5/ 5/55 " Brown, Shiawassee 553-68001: L - S/ 5/55 " " 553-3805 L .. 5/ 5/55 " . " 513-96907 A M S/lh/SS " White, Clinton CO. 513-96908 A M 5/1h/55 " Brown, Shiawassee 553-38006 A M /17/55 *1 Orange, Clinton CO. 513-96909 L - 5/1u/55 " Brown, Shiawassee SIB-9691M A M 5/20/55 " .. .. 553-38007 A. M’ 6/ 3/55 " Orange, Shiawassee 523-33505 J M 7/ 5/55 Tin Bridge, Cheboygan County $23-33507* = M 7/ 7/55 . " 52303508 - F 7/11/55 " " 5231-33509 - F 7/11/55 " n 523 -33510 - F 7/15/55 " " 52303511 - M 7/19/55 " 9 52303512 M 7/19/55 " " 5223-33513 . M 7/25/55 " w 52303511; M 7/25/55 " " 523-33515* - M 7/29/55 Old Grant School, Cheboygan County 523-33516 - M 8/ 2/55 w. McMaster Creek, n 5230351? - F 8/ 2/55 .. 523‘33518 "- M 8/ 3/55 '1 n 205 Band Date of Number Age Sex Banding Place of BandinL 523-33519* - M 8/ M55 w. McMaster Creek, Cheboygan County 93-33520 - F 8/11/55 M M 52343521 - M 8/11/55 M M 93-33522 - M 3/12/55 M M 52363523 - M 8/13/55 M M 52363521: - F 8/16/55 M M 523-33525 - M 8/18/55 Elk Hill, Pigeon River, Otsego County 523-33526 - M 8/18/55 M M- M 52363539 - M 8/20/55 W. McMaster Creek, Cheboygan County 523-335u0 - F 8/20/55 Elk Hill, Pigeon River, Otsego County $23-335h1 - M 8/21/55 W. McMaster Creek, Cheboygan County 523-335h2 - M 8/21/55 Elk Hill, Pigeon River, Otsego County $23-33Sh3 - M 8/25/55 Tin Bridge, Pigeon River, Cheboygan Co. $23-33Shh - - 8/25/55 M M M 523-335h5 - M 8/26/55 w. MoMaater Creek, Cheboygan County $23-335h6 - M 9/ l/SS Elk Hill, Pigeon River, Otsego County $23-33Sh7* - F 9/ 5/55 . .. n . $23-335h8 - M 9/11/55 . M M $23-33SM9M - M 9/28/55 M M M * See Table 11 for recovery data. Table 113. Records of woodcock banded as Juveniles and eventually recovered. 206 Banding: Recovery Band Number State Date State Date Lonm 313 39973 Maine 6/ 1/ 8 Mass. 10/25/10 29 373 20151 M 5/2 37 Maine 10/28/37 S 373 20152 M 5/21/37 M 10/27/37 5 373 201.78 M 5/26/37 New Jersey 11/ 6/37 5 373 2081.0 ‘ M 5/10/1.1 Penn. 10/21/111 5 383 17285 M 5/25/38 Maine 10/20/38 5 393 16053 M 5/21/39 Virginia 11/23/39 6 393 1608A M 5/27/39 Maine 10/18/39 s 103 23573 M 6/ 1/10 M 10/21/10 s 1.03 23601; M 6/ M140 Mass. lO/3l/h2 29 103 23655 M 5/29/h1 New Hamp- 10/20/12 17 h03 23676 M S/ 7/u2 Maine 10/21/h3 17 1.23 05008 M 5/21/16 M 10/13/18 29 06 51858 Mass. 5/13/30 Mass. 10/20/30 5 06 51859 M 5/13/30 M 10/20/30 s 563 33003 M 8/ 8/55 M 11/ 9/55 6 1193 09702 Mich. 5/30/51 Mich. 10/ 6/51 1. 513 96901 M 5/10/53 Georgia 2/19/ s 21 02 1111598 Minn. 5/ 8/32 Miss. l/ 73 8 373 20725 New Bruns. 3/28/39 New Jersey 11/15/39 8 373 20763 M 6 9/h0 New. Bruns. 10/ 7/h1 16 123 02139 M 6/ 6/13 M 10/29/13 5 1123 0211.3 M 6/ 6/113 M 10/27/116 A1 173 051.07 M 5/21/50 Maine 10/15/50 5 A73 08u79 M 5/10/51 M 10/ /51 5 503 11202 M 5/19/50 New Jersey 11/13/50 6 503 11,207 M 5/19/50 Florida 12/15/52 19 503 11.221; M 5/20/50 New Bruns. 10/10/50 5 503 lh2hl M 5/22/50 M 10/19/50 5 503 11905 M 5/15/52 Delaware 11/25/52 6 503 1h936 M 5/17/52 New Bruns. 10/ 3/52 s 503 1h938 M 5/18/52 M 10/ 9/52 s 503 1h939 M S/18/52 M 10/ /52 S 503 11959 M - 5/11/52 M 10/10/53 17 uzu 01002 New Jersey 1./29/1.5 New Jersey 11/ 6/115 6 O6 M8502 New York h/28/29 New York 10/25/29 6 h83 59929 M 5/13/52 M 10/28/53 18 39h 2501? Penn. . S/ 8/ 39 Miss. 12/ 27/110 20 513 833h1 M h/29/53 8. Carolina 2/11/Su 9 553 390h3 M 1/27/55 Penn. 10/15/55 6 553 39062 M h/30/SS M 10/15/55 6 2C 7 Table 1114 SCIENTIFIC AND COMMON NAMES OF PLANT SPECIES Scientific names are based on M. L. Femald, 1950, Gray's Manual of Botany, Eighth Edition, American Book Co. Common names are from Fernald and the following additional sources: Billington, Cecil. 19113. Shrubs of Michigan. The Cranbrook Institute of Science. Fassett, Norman C. 19140. A Manual of Aquatic Plants, lst Edition, McGraw-Hill Book Company, Inc. Rydberg, Per Axel. 1932. Flora of the Prairies and Plains of Central North America. The New York Botanical Garden. "Woody plants" Abies balsamea, Balsam-fir Acer sp., Maple Acer rubrum, Red maple Acer saccharum, Sugar-maple Alnus rugosa, Speckled alder Amelanchier spp., Juneberry Amelanchier canadensis, Canadian Juneberry Betula lutea, Yellow birch Betula papyrii‘era, Paper or white birch Betula pumila, Swamp—birch Carya ovata, Shagbark-hickory Castanea dentata, Chestnut Clematis virginiana, Virgin's-bower Comptonia peregrina, Sweet-fern Cornus alternifolia, Green osier Cornus oblique, Silky dogwood Cornus racemosa, Grey dogwood Cornus stolonifera, Red osier Corylus cornuta, Beaked hazel Crataegusspp., Hawthom Diervilla Lonicera, Bush-honeysuckle Fagus grandifolia, Beech Fraxinus americana, White ash Fraxinus nigra, Black ash Ilex verticillata, Black alder Juniperus communis, Common juniper Larix laricina, Tamarack Lespedeza sp., Bush clover Lonicera spp., Honeysuckle Lonicera canadensis, Fly-honeysuckle Lonicera dioica, Wild honeysuckle Lonicera hirsute, Hairy honeysuckle Lonicera oblongifolia, Swamp-fly-honeysuckle Picea glauca, White spruce Picea mariana, Black spruce Physocarpus Opulifolius, Ninebark Pinus Banksiana, Jackdpine Pinus resinosa, Red pine Pinus Strobus, White pine , Pinus sylvestris, Scotch pine Populus x gileadensis, Balm-of-Gilead Populus grandidentata, Large-toothed aspen Populus tremuloides, Quaking aspen. Prunus pensylvanica, Pin-cherry Prunus serotina, Black-cherry Prunus virginiana, Choke-cherry Pyrus Malus, Apple Quercus spp., Oak Quercus rubra, Red oak . Quercus velutina,’Yellowébarked oak Rhamnus alnifolia, Buckthorn Rhus radicans, Poison ivy Rhus typhina, Staghorn-sumac Rhus Vernix, Poison sumac Ribes spp., Currant and Gooseberry Rosa spp., Rose Rubus spp. , Blackberry and Raspberry Rubus pubescens, Dwarf raspberry Salix spp., Willow Sambucus canadensis, Common elder Spiraea spp., Spiraea Spiraea alba, Meadow-sweet Thuja occidentalis, White cedar Ulmus americana, American elm Vaccinium vacillans, Low blueberry Vaccinium angustifolium, Low sweet blueberry Viburnum spp., Viburnum 2C8 2(9 Viburnum cassinoides, Wildaraisin Viburnum Lentago, Nannyberry Viburnum trilobum, Highbush-cranberry "Herbaceous plants“ Achillea Millefolium, Common yarrow Actaea rubra, Red baneberry Anaphalis margaritacea, Pearly everlasting Anemone spp., Anemone Anemone canadensis, Broad-leaf anemone Anemone cylindrica, Thimbleweed Anemone virginiana, Thimbleweed Anemonella thalictroides, Rue-anemone Antennaria spp., Pussy's-toes Anthemis spp., Chamomile Apocynum spp., Dogbane Aquilegia canadensis, Wild columbine Aralia nudicaulis, Wild sarsaparilla Artemisia caudata, WOrmwood Asclepias spp., Milkweed Aster spp., including: A. azureus, A. cordifolius, A. laevis, A. macrophyllus, A. prenanthoides -- Asters Athyrium FiliXsfemina, Lady fern Athyrium thelypterioides, Silvery spleenwork Bidens cernua, Stick-tight Botrychium dissectum, Grape-fern Botrychium.multifidum, Leathery Grapesfern Caltha palustris, Marshsmarigold Campanula uliginosa, Marshébellflower Carex spp., Sedge Castilleja spp., Painted—cup Cerastium vulgatum, Common mouse-ear chickweed Chelone glabra, Turtlehead Chrysosplenium americanum, Water-mat Cirsium spp., Common thistle Cirsium Hillii, Clintonia borealis, Corn-lily Comandra umbellata, Bastard-toadflax Coptis groenlandica, Goldthread Corallorhiza odontorhiza, Late coral-root Cornus canadensis, Bunchberry Cynoglossum officinale, Common Hound's tongue CystOpteris fragilis, Fragile fern Dryopteris cristata, Crested.woodefern 210 Eleocharis spp. , Spike -rush _ Epigaea repens, Trailing arbutus Equisetum spp., Horsetail Equisetum litorale, Equisetum sylvaticum, Wood-horse tail Erigeron spp., Fleabane Erigeron strigosus, Daisy-fleabane Eupatorium maculatum, J oe-pye -weed Eupatorium rugosum, White snakeroot Fragaria spp., Strawberry Galium spp., including: G. palustre, G. trifidum, G. triflorum-- Bedstraws Gaultheria procumbens, Teaberry (aromatic wintergreen) Geum aleppicum, v. strictmn, Avens Habenaria hyperborea, Northem green orchis Habenaria obtusata, Blunt-leaf orchis Hieracium aurantiacum, Orange hawkweed Hierac ium Gronovii, Hieracium venosul, Poor Robin's plantain Impatiens capensis, Spotted touch-me-not Iris versicolor, Blue flag Juncus effusus, Soft rush Krigia spp., Dwarf dandelion Lactuca spp., Lettuce Lemna minor, Lesser duckmed Leontodon sp. , Hawkbit Liatris spp., including: L. spicata and L. Nieuwlandii -- Blazing-star Linaria sp., Toadflax Linnaea borealis, Twinflower Lobelia Cardinalis, Cardinal-flower Lyc0podium tristachyum, Ground-pine Lycopus spp. , Water-horehound LyCOpus americanus , Lycopus rubellus, Lycopus uniflorus, Bugle weed Lysimachia sp., Loosestrife Lysimachia ciliata, Fringed loosestrife Lythrum sp., Loosestrife Maianthemum canadense , False Lily-of -the -valley Melampyrum lineare, Cow-wheat Melilotus alba, White sweet clover Mentha sp. , Mint Mentha arvensis, Field mint 211 Mimulus spp. , Monkey-flower Mimulus alatus, Wing -stem monkey-flower Mimulus ringens, Square -stem monkey-flower Mitchella repens, Partridge-berry Mitella diphylla, Coolwort Mitella nuda, Bishop's-cap Monarda fistulosa, Wild bergamot MonotrOpa Hypopithys, False beech-drape Nasturtium officinale, Watercress Onoclea sensibilis, Sensitive fern Panicum spp., Panic-grass Pedicularis canadensis, Wood-batony Physalis sp., Ground-cherry Plantago sp., Plantain Poa pratensis, Junegraes Polygala paucifolia, Fringed polygala Potentilla spp. , Cinquefoil . Potentilla simplex, Old-field-cinquefoil Prenanthes alba, White lettuce Prunella vulgaris, Selfheal Pteridium aquilinum, Bracken Pyrola spp., including: P. elliptica, P. minor, P. rotundifolia, and P. secunda -- Shinleaf Ranunculus spp., Crowfoot Rorippa islandica, v. hispida, Yellow cress Rumex acetosella, Common sorrel Sagittaria spp., Arrowhead Scirpus cyperinus, Bulrush Scutellaria epilobiifolia, Common skullcap Scutellaria lateriflora, Mad-dog skullcap Senecio sp., Groundsel Smilacina sp., False Solomon's-seal Solanum sp., Nightshade - Solidago spp., including: S. altissima, S. canadensis, S. rugosa, S. uliginosa -- Goldenrods Sonchus spp. , Sow-thistle Sorghastrum nutans, Indian grass Taraxacum sp., Dandelion Thalictrum dasycarpum, Purple Meadow-rue Tiarella cordifolia, Foamflower Trientalis borealis, Star-flower Trifolium: hybridum, Alsike clover Trillium grandiflorum, Trillium Urtica dioica, Stinging nettle Viola spp., Violet 212 nonEo>oz sopovoo ponEoonm a n H on NN 0N aN NN 0N we we as NH OH N om NN 0N aN NN ON m” o _ n twa- a _ __-._ m. mt mo mio.a Iom m. 8 man 1 [on i .Ioc .smma I coapmpm nonwomom peony uo>wm mommam am vouuooop mohspwhoasvv omeno>w one coavmvwgaoOMm .ua shaman itequazqaa 9901800 213 O o SQQOUI u; IIaluTVH "3 H in O H O 1 N. quEo>oz nonopoo ponEoonm m m H on mm 0N 3N NN 0N ma 0H 3H NH 0H m o d N on mm mm 3N NN om - 211111 lion on .nmma 1 nofipmvm nohmomom prone uo>dm :oowam no pocuooou weedpmuomsov oweuo>m one coapmpamwoohm .ma chewah armameg 899.138q 21h C3 ”3 r! c: 89110111 111; 119mg “3 H honEo>oz a m m a on NN 0N sN NN ON NH nonopoo 0H #H NH OH m ...:«i- Ajfi 7.1 w : Monsopmom N on mm mm 3N mm om sees. .. eases .oee paswm 9m bookcase mouspmeanop ommpo>m new coapwpamwoopm .ma madman on on ow atequeaueg SGGJSGQ 215 Bauoul HI IIBJUIMB Ln ‘3 eonEo>oz honopoo m H on mm 0N :N NN om ma 0H :H NH 0H m m 3 N on mm mm 3N mm c an .nmma 1 caps: .opm Hanan pm voouooou mohsumnanOp owmno>m new coupwvamaooum .ON seamen honEOonm d _ u on em ow atequeiqeg seaafleq 216 “D O o SOMOUI "I IIMJMIMB tn 0 quEo>oz honovoo popsopmom a n m H on ON ON ON NN ON OH OH OH NH OH O O N on ON ON ON NN ON 0 H4 1 H H . H . —=-Ifidq_—_—4 l3-1H . — —_1om >2. .amOH . aeHO assesses pd vovoOOA mohdpmhomfiop ammuo>w use coapwpfimfioopm .HN whsmfih croqueaqeg seeafieq 2“? 60110111 at WJU'FBH <2 M) H ‘1‘ C H O A u honEo>oz nma nonopoo ponEopaom cm mm 0N 3N mm om ma 0H 3H NH 0H m o s m on mm 0N 3N mm om .mmOH . .38 usages: pa ooonoooa mousuwpomsop ammuo>w new :OpraHaHooam .NN ousmHh on so name-rm see-18% ROOM USE 0M” APR m1 - RIES "11111111111111.11111111111‘