A zooesocaApHaCAL ANALYSIS or THE HERPETQFAUNA IN NORTHERN WCHEGAN AND ADJACENT 153.5 ROYALE .Thm-Es {For fhe Deg?» of M. S. MICHIGAN STATE UNIVERSlTY Wanda! 10h“ Wagner Johnson 19¢?» ROOM use ONLY W339 )1 246M517) JAN 2 6 2015 THEgV ABSTRACT A ZOOGEOGRAPHICAL ANALYSIS OF THE HERPETOFAUNA IN NORTHERN MICHIGAN AND ADJACENT ISLE ROYALE by Wendel John Wagner Johnson The objectives of this study were twofold: (l) to ascertain more accurately the geographic ranges of the amphibians and reptiles inhabiting northern Michigan and; (2) to examine critically animals from selected areas of Isle Royale and from the Upper and Lower Peninsulas of Michigan in an effort to establish any clinal trends that may exist. The field work consisted of a 13-day collecting trip from August 12 to August 25, 1963, and a more extensive survey from June 15 to August 7, 1964. Specimens were ob- tained from all ecological habitats in the 15 counties of the Upper Peninsula of Michigan and Isle Royale. Seines, minnow traps and dip nets were utilized for acquiring the aquatic forms while most of the terrestrial species were encountered by turning logs and rocks. Pre- served materials from The University of Michigan Museum of Zoology, Michigan State University Museum, and the Michigan Wendel John Wagner Johnson State University Department of Zoology were also utilized. Complete scalation counts were made on all reptiles and pertinent body and leg measurements were taken with a meter stick and vernier caliper on the amphibians. The geographical ranges of existing herpetofauna have been more accurately defined. There are no large geographic barriers to species dispersal in the Upper Peninsula. Temperature appears to be a deterrent to keep more species from entering the northern Peninsula. The various species occur in mosaic patterns in suitable habitats across the Upper Peninsula. The origin of the Isle Royale herpetofauna is from Canada to the north and east. The northern Peninsula was repopulated after the last glacial retreat, primarily from Wisconsin to the west and south, and secondarily from southern Michigan and Canada to the south and east. A ZOOGEOGRAPHICAL ANALYSIS OF THE HERPETOFAUNA IN NORTHERN MICHIGAN AND ADJACENT ISLE ROYALE BY Wendel John wagner Johnson A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Zoology 1965 {J} .7 1 H F v' .' go Page Page Page Page 22 24 24 27 40 41 :. 4h 43 57 59 67 72 76 85 ERRATA SHEET line 22 reads (£2}g§_sp.) should be (ébggg,32,) line 2 reads (1950 divides should be (1950) divides line 18 reads (Eggglg,2§213£ggg) should be (gg3915.292113§g£3) line 13 reads (Qgggggg g2.) should be (Drosera ggn.’ h — Table 2 reads (40.0-47.7) Should be (40.0-46.7) ’ line 20 reads (per. comm), should be (per. cm.). line 8 reads a counting should be a difference in counting Table 7 reads 81.8 spots should be 81.8 stripes line 5 reads Conant, 0p. cit. (Esp 21) should be Conant, (op. cit.. map 21) line 17 reads across to should be across the line 9 reads cological should be ecological Opheodyrs should be Ophcodrya Tabel l7 - Chrysemya picta marginata is not recorded from Bataan county. Tabel 17 - Chryaemys picta belll should be 0 instead of - for Baragn county. line A road: of salamandera, one should be of snakes. Ono species of turtle. two species of snakes, six species of fraga, and three species of salamander: occur on Isle Royale. THEE" , . . . - , 7 I E . ‘ I s .. \ ‘ r | . ~.. 0 . 1 s p f x " 1 l . . -’ ' - . . . 5 . 4‘ a o . . . l . . I “A. _ _ v ,V . ~ . U ' u .v' . u -c a ;- 1 ‘ . It Hr O c f U V . Pi r-w '13 I-i‘! W . ACKNOWLEDGEMENTS Grateful thanks are extended to Dr. Max Hensley who directed the project and was a constant source of advice, information, and aid. Sincere appreciation is also felt for the advice and assistance given by Dr. Jane E. Smith and Dr. T. Wayne Porter of Michigan State University who served as members of my guidance committee. Dr. Rollin H. Baker and Dr. Michael M. Ovchynnyk of the Michigan State Museum very graciously allowed the author to examine and record data from the herpetological collection; and Dr. Charleis Walker of the Museum of Zoology at the Uni- versity of Michigan, Ann Arbor, also extended similar courtesies. The au-nor is appreciative of cooperation and aid ex- tended by uilliam W. Dunmire, Chief Naturalist, Isle Royale National Park and to Dr. Philip W. Smith of the Illinois Natural History Survey, Urbana, for suggestions regarding interpretation of the field data. The numerous courtesies extended by Mrs. Bernadette Henderson, departmental secretary, were also greatly appreci- ated. The enduring patience, aid, and encouragement of my wife Marilyn and my parents are gratefully appreciated. Part of the summer field surveys were supported by a Sigma Xi-RESA Grant-in—Aid. ii TABLE OF CONTENTS ACKNOWLEDGEMENTS . . . . . . INTRODUCTION . . . . . . . . Climatological Features Surface Geology and Soils Vegetational Features METHODS AND MATERIALS . . . . . ACCOUNTS OF THE SPECIES . . . . Class Amphibia Order Caudata Family Proteidae Family Ambystomatidae Family Salamandridae Family Plethodontidae Order Salientia Family Bufonidae Family Hylidae Family Ranidae Class Reptilia Order Chelonia iii PHYLOGENETIC LIST OF THE AMPHIBIANS AND REPTILES OF THE NORTHERN PENINSULA OF MICHIGAN PAGE ii 12 22 29 33 36 38 38 38 39 41 42 45 45 45 49 56 56 PAGE Family Chelydridae 56 Family Emydidae 57 Order Squamata 58 Suborder Sauria 58 Family Scincidae 58 Suborder Serpentes 59 Family Colubridae 59 Family Crotalidae 7O HABITATS AND DISTRIBUTION . . . . . . . . . . . . . 71 DISCUSSION . . . . . . . . . . . . . . . . . . . . 77 Results of Population Comparisons 77 Environmental Correlations with Distribution 78 Post-glacial Repopulation of the Upper Peninsula of Michigan and Isle Royale 80 CONCLUSIONS . . . . . . . . . . . . . . . . . . . . 85 LITERATURE CITED . . . . . . . . . . . . . . . . . 87 iv TABLE 10 11 12 13 14 15 LIST OF TABLES Soils of Northern Michigan . . . . . . . . . Geographic Variation in Ambystoma laterale . Geographic Variation in Notophthalmus viridescens . . . . . . . . . . . . . . . . Geographic Variation in Plethodon cinereus Cinereus O O O O O O O O O O O O O O O O O 0 Geographic Variation in Bufo americanus americanu§_. . . . . . . . . . . . . . . . . Geographic Variation in Hyla crucifer crqu-fer O O O O O O I O O O O O O O O O O 0 Geographic Variation in Pseudacris triseriata maculata . . . . . . . . . . . . Geographic Variation in Rana septentrionalis . . . . . . . . . . . . . . Geographic Variation in Rana sylvatica . . . Five Characteristics Distinguishing Subspecies in Rana sylvatica . . . . . . . . Geographic Variation in Diadophis punctatus edwardsi . . . . . . . . . . . . . Geographic Variation in Elaphevulpina vulpina . . . . . . . . . . . . . . . . . . Geographic Variation in Storeria occipitomaculata occipitomaculata . . . . . Geographic Variation in Thamnophis sirtalis Sirtalis 0 O 0 O O O O O C O O O O O O O O 0 Habitat Preferences of Species Collected . . PAGE 19 40 4o 43 46 48 58 52 54 55 6O 61 66 68 72 TABLE 16 17 PAGE County Distribution of Amphibians in Northern Michigan . . . . . . . . . . . . . 75 County Distribution of Reptiles in Northern Michigan . . . . . . . . . . . . . . . . . . 76 vi THES‘ FIGURE 10 -11 LIST OF FIGURES County Map of Northern Michigan Showing Sites of Previous Herpetological Studies The Mean Monthly and Annual Temperatures at Five Weather Stations in Northern Michigan (1931-1952) . . . . . . . . . . . Comparison of Percentage of Total Possible Sunshine at Green Bay, Wisconsin and Marquette, Michigan (1902-1952) . . . . Annual Precipitation Normals (Inches) (1931-1952) . . . . . . . . . . . . . . . Map of the Surface Features of Northern Mi Chigan ‘ O O O O O O O O . 0 O O O O 0 O 0 Map of the Principal Drainage Systems of Northern Michigan . . . . . . . . . . . . Map of the Soils of Northern Michigan According to Land Divisions . . . . . . . Map of the Original Forests of Northern Michigan . . . . . . . . . . . . . . . . Map of Collection Sites of 1963 and 1964 Field Surveys . . . . . . . . . . . . . . Frequency of Red Lateral Color Pattern in Thamnophis sirtalis sirtalis . . . . . . . Map of the Surface Currents of Lake Superior . . . . . . . . . . . . . . . . . vii PAGE 10 11 14 16 21 23 3O 69 84 INTRODUCTION The herpetofauna of the Northern Peninsula of Michigan has not been intensively studied. The Herpetology 9§_ Michigan, University of Michigan Handbook, Number 3, pub- lished by Alexander G. Ruthven, Crystal Thompson and Helen T. Gaige (1928) is the most recent comprehensive work involving this region. Previous to this handbook there was a similar publication by the same authors (Ruthven, et a1.) in.l912. Ruthven made several studies of specific areas of the Northern Peninsula (1904, 1906, 1909, 1910) and Ruthven and Crystal Thompson (1915). The localities for these previous herpetological expedi- tions are shown in Figure 1. Since 1928 there have been relatively few studies involving the amphibians and reptiles inhabiting this area. Manville (1948) did a sur- vey of the vertebrates of the Huron Mountains and listed the herpetofauna in that area. Logier and Toner (1961) have published and revised a checklist of the herpetofauna of Canada and Alaska but do not include any areas south of the Canadian border. Conant (1958) illustrates ranges for all reptiles and amphibians of the United States and Canada east of the 100th Meridian, but because of the Th'r'; FIGURE 1 PREVIOUS HERPETOLOGICAL STUDIES IN THE NORTHERN PENINSULA Ruthven, Alexander G., 1904 Ruthven, Alexander G., 1906 Ruthven, Alexander G., 1909 Ruthven, Alexander G., 1910 Thompson, Crystal, and Helen Thompson, 1913 Gaige, Helen T., 1915 Gaige, Helen T., 1915a Evans, Arthur T., 1916 Manville, Richard H., 1948 H HMDOHM \J 0.1 P' 1.. I .a S. \ x V ._ . 5... ... . . 1 _. . . II. 5 5. n L ‘\ _ 0. log. a. q. 1. . x . .506 2 : _ . . n .x :- l . \\ - \ \ a If... I. I . .. 7 7 .- 1|.- 3. 1 _ x r p u .03. I I __ v I I nc‘ —- _ a 1 n n to A , . o I .’ .1 . .( x . I... . .__ M n I... I u _ )\ . in — ... I . 1\ . luv. 1. l T .2 c .5"... I. I a ‘1 I... In ., .11 ._ 2 s W .135 A . q §I.lU¢O 4v ....... ¢ u»»u§(l l\/ H u a ~. ‘ _ .\ — . .. .a \ I r .L I _,u I. 1 . _ / O 5 __ n .., I. T x I u . {)1 14 I U A . .. 1a 5 .. l. .1 1) \ 1 :1 ,. .11 1 11/ .. ’(lnui at o» m N . . . W K. , W. L “ _. a .flf" I ~ <1) ~ I \' . cl ( I- < 1 ( C 1 ._ i ' ~b « M .r J a 4”. I. I ,i .19 1. , I 1‘ 1 1 1 1 11111. 1 1 r .- 1 119111111- 1 c, THE? extremely large area covered by the ranges listed for all species found in northern Michigan are necessarily quite general. A note published by Hensley (1964) was the most recent work dealing exclusively with this area. The objectives of this study are: (l) to ascertain more accurately the geographic ranges of the amphibians and reptiles inhabiting northern Michigan and; (2) to examine critically animals from selected areas of Isle Royale and from the Upper and Lower Peninsulas of Michi- gan in an effort to establish any clinal trends that may exist. The areas under consideration are the Northern Penin- sula of Michigan and Isle Royale in Lake Superior. The Northern Peninsula, comprising 16,353 square miles, is bordered on the south by Lake Midhigan, Lake Huron on the east, Lake Superior on the north, and continues into Wis— consin on the west. The maximum eastdwest dimension is 325 miles and it is 240 miles from Menominee to Copper Harbor at the tip of the Keweenaw Peninsula. Isle Royale lies fifty miles northwest of Copper Harbor, and 12 miles south of the Minnesota-Canada shoreline. It measures 45 miles in a northeast-southwest direction, has a maximum width of nine miles wide, and includes 210 square miles. Climatological Features Because of its position between two of the Great Lakes, the Upper Peninsula of Michigan has a more moderate and island-like temperature than any other area in the interior of the continental United States. United States Weather Bureau records indicate that 400 F. is the mean average annual temperature for the entire Peninsula, however aver- age daily temperatures and frost free days vary consider— ably between areas. For example the cities located near the lake shores have more moderate winter and summer tem— peratures. The coldest regions of the Upper Peninsula are in Iron and Gogebic counties. Watersmeet in Gogebic county has a January mean average of 12.50 F. (See Fig. 2); this is the lowest mean temperature observed for Janu- ary according to data from fifteen Upper Peninsula weather stations. In contrast to this St. Ignace, which is bor- dered by Lake Michigan, in Mackinac county, has the highest mean temperature of 21.10 F. for the same month. There is no gradient in temperature from south to north in the Peninsula, but instead the rule of modification of tempera- ture by the Great Lakes is apparent throughout the Upper Peninsula. Incomplete monthly weather records from the Th Fr, THEE FIGURE 2 Graph of the mean monthly and annual temperatures at five weather stations in Northern Michigan, (1931- 1952). 52¢ no so: .0 <1- <1- 1300 <3IIJD Cacao —‘ — "‘1 —1 _ ——‘ .529. Imatxi I I Gonna bonanza»; 1. 3:6. 5.. 1 G 0 356.1 32 .. 0 5238.. no ozmom4 ID oo<1 D. ID oo<1 0Q. N MMDOHm <1- co 0. OH ON on co On Do On araquaaqua 6931830 Mott Island Station on Isle Royale make a complete com- parison with the Northern Peninsula impossible. However, the records from May to November show that mean summer temperatures on Isle Royale are cooler than in the Northern Peninsula by several degrees as would be eXpected by its more northerly location. This location would indicate that Lake Superior is exerting a modifying effect on the climate similar to that seen in the areas along the Great Lakes. Another indication of the modifying influence by Lake Superior is seen in the difference in the number of frost- free days between weather stations located on Lake Superior and those stations farther inland. Areas along the southern border of Lake Superior seldom have killing frosts after May 15 or before October 1, whereas, other areas of the Peninsula may expect killing frost until June 10 and as early as September 1 in the fall. In certain regions of Iron and Dickinson Counties the growing season for plant crops is only 75 to 100 days. Baten and Eickmeier (1955), have shown that Green Bay, Wisconsin, receives as much as 16 percent more sunshine than Marquette, Michigan during various times of the year. This is because of increased cloud cover over Marquette, Michigan. As indicated in Figure 3 the difference in percent of sunshine is greatest in the months of Novem- ber, December, January and February and least in the summer months. Weather Bureau records from 1927 to 1953, indicate a warming trend over the preceding 27 years. During the period from 1901 to 1931, total annual precipitation in the Upper Peninsula ranged from an aver- age of 26 inches in western Marquette county and por- tions of Luce, Chippewa and Mackinac counties to an average of 34 inches per year in Iron County. However records from 1931 to 1952 show that the St. Ignace area had the least rain with an average annual amount of 26.68 inches and the Chatham and Munising station in Alger county had the most precipitation with averages of 33.72 and 33.84 inches re- spectively. Figure 4, shows the isohyeUsof northern Michi— gan. The preceding climatic conditions corroborate with those cited by Kendeigh (1961), for the Pine-Hemlock ectone of the Upper Peninsula. This area is a transition zone between the more temperate deciduous forest of southern Michigan and the cooler boreal forest of Canada to the north. Per Cent 10 FIGURE 3 COMPARISON OF PERCENTAGES OF POSSIBLE SUNSHINE AT GREEN BAY, WISCDNSIN AND MARQUETTE, MIGHGAN (1902-1952) 80 7O 60 50 4o 1 \,=1._ 30 20 10 JFMAMJJASOND Month __ Green Bay, Wisconsin —— Marquette, Michigan (Baton and Eictuneier, 1955) THEE In 55ch Aumafi .saouam possum: .m.=V n... [/11 \\ .0 no...” Item L. a. .29 0- ah 0 .III Fab . s .3 cc...— .cc- 2. .0 1A.”... .00. DI- : .5. mwgo '— Iz—QHJT . I.- ..I . flN-m'hLl .L. a no; \ c0. . I .- HWW.&\ . . :1. o... '20- :1 .19.; :Sso‘c. a. 1......»0...‘ I 6 0 AV Jcotls. :3 L. '03.. 36:0.- a. o. — dual-am, .0 0:05 .1... :0qu I. o 3.... 2.. 2‘. o o. o _ \nwzn .0 ._ 3.. “Mi: . 1:8 3 . 3.21.;U all c& w a... 9.00.0 . .3 .u‘ no: . otu . IL”... 000 910/ .r .0 l 2. .9. I :- k \ a... :6 z .3- Ul — OWOHOO 31.0—- 0 LI...- LAN-II-c— so. I In. . tug-c.01- .1. 1 11 u o 4 _ nun-b: . .l i _ .1 mm :3..- . Q AV moi-.3- / u , . End: 1.. .3... _ Na)“ 1 A * S /¢o . 1V \ Hum-1.5%“. / i (‘1 5T .Ammaa-ammao.buz-- pix. , 1 ammo r _ , 1 mdazmoz oneaeHnoommo Saszza C.‘ Isap'dpl 01.1.“. 722:2: A. q -. n1 -1 f a... a, i / .1 . I l - . 1 1w - -- ’-41<;w a. .ll. .— lug—om... _ _ Iv... co . _ .- - L. .cov . _ u a clinic. _ O. i I .100 _ on... .03 O I O O k \1 No In...‘ //. 11/1\.\ 1 . A2: .31.... :3...- a... mmoaD _ _ . I O .A 5‘ Clo—Io. 12 Surface Geology and Soils The rooks of the western Upper Peninsula, dating from Precambrian times, are much older than those of the eastern sector and are predominantly igneous and meta- morphic with minor amounts of sedimentaries. These rocks are of many types, such as, greenstones, granites, gneissesr. and schists, and have been folded and partly, or wholly recrystallized. They vary in hardness and consequently weather unevenly forming a rugged, hilly relief. Bedrock outcropping is most extensive in the western Upper Penin- sula in the Huron Mountains east of Keweenaw Bay, the Keweenaw Peninsula, the Porcupine Mountains in Ontonagon county on the border of Lake Superior, the Gogebic Range from Ironwood to Lake Gogebic, and the Iron Mountain area bordering Wisconsin. Isle Royale has topography and geol- ogy similar to the-Keweenaw Peninsula fifty miles to the southeast of it. The subsurface geology of the Northern Peninsula, east of a line running from Marquette to Faithorn in Menominee county (Fig. l), is Lower Paleozoic in age and is all of a sedimentary origin. The present topography of the Northern Peninsula like so much of the rest of the Midwest is a direct result of 13 Pleistocene glaciation and post-Pleistocene erosion. Pollen profiles and other paleobotanical evidence dated by carbon-l4 methods indicate that the last substage, the Valders, of any consequence to the Upper Peninsula, of the Wisconsin stage, retreated from northern Michigan approximately 9,000 years ago. As the various glaciers passed over and retreated from the Northern Peninsula drift deposits were left over most of the area. Much of the rugged relief of the western counties of the Pen- insula and Isle Royale has been modified by these thick glacial deposits. 1A series of ancestral Great Lakes de- veloped from the impounding of glacial melt waters. Many of these lakes stood at a higher level than the present ones as shown by ancient elevated shorelines (Hough, 1963), and covered parts of the eastern Upper Peninsula. Today the results of this inundation can be noted by the exten- 'sive swamps and also by lake—clay plains (Fig. 5). .A post-Lake Algonquin stage of the ancestral Great Lakes had the most pronounced effect on the eastern half of the NOrthern Peninsula of Michigan by inundating parts of the area. Ancient beaches and sandy shorelines have been found in parts of Alger, Chippewa, Delta, Luce, Mackinac, and Schoolcraft counties. Since the retreat l4 cmmfisofiz cumnuuoz mo mousummm oommuam «no no as: .m MMDQHm Adamo .38 .5qu o... u... ... a .H‘... ..v. . no sheaf... o... ... .23 :5. 5.; H... .5 4.0:: .wv. in VI to vumwsm m. cham 1.114 s£U.s:E mic—Q 813 .mh‘ufi 35.7.... rams—Q =._.—. ..... -.1.l_ mama—Q .! £33.:C fl 1‘ e 3:352 . 15 of the glacier there has been continual uplift of the ancient lake shorelines in the Northern Peninsula as a result of isostatic rebound of the crust from the release of pressure when the glacier melted. The topography of Isle Royale consists of a series of parallel ridges running in a southwest-northeast direction. Mount Desor is the highest point on the Island, 1394 feet above sea level. The relief of the western counties, generally between 650 and 1400 feet above sea level, reaches a peak of 2023 feet in the Porcupine Mountains in Ontonagon county. In contrast, the generally low re- lief of the eastern Upper Peninsula is characterized by swamps interspersed with morainal ridges and averages 250 feet above Lake Michigan. The drainage of the Northern Peninsula is in two main directions, northward to Lake Superior and to Lake Michigan on the south (Fig. 6). Lake Huron, to the east, receives a relatively small amount of water from this region. The two primary drainages into Lake Superior are the Ontonagon River in Ontonagon county and the Tahquamenon River in Luce and Chippewa counties. The drainage from Ontonagon county consists of a series of small streams running paral- lel to each other into Lake Superior. The drainage there l6 cmmwgowz cumsuuoz mo mEMummm mwuawwun oHdwoawum one «0 do: .0 mMDon AOmmH .meuam MosuGOB .mobv 17 is down a steep gradient whereas the gradient of the Tahquamenon River is not as extreme. The streams flowing into Lake Michigan have much larger watershed areas. The Manistee River in School— craft, Mackinac, and Luce counties is the most extensive drainage system and involves an area of about 1400 square miles. Smaller rivers such as the Menominee, Ford, and Sturgeon also drain into Lake Michigan. Only a few small streams in Chippewa and Mackinac counties flow into Lake Huron. The poor drainage of the Peninsula is apparent when one realizes that 25 percent of the total area is characterized by swamps and lakes. The drainage between ridges on Isle Royale is very poor and has resulted in the formation of extensive peat swamps in these areas. Whiteside, Schneider, and Cook (1959), have found that the major local differences in Michigan soils are associ- ated with variations in the texture of their parent materials and the drainage conditions under which the soils are formed. All soils of the Upper Peninsula are classified as podzolic. However, the western sector of the Peninsula and the southern part of the eastern Upper Peninsula have soils formed from consolidated rock and in these-regions the soils are shallow and bedrock may reach the surface. There is no free lime in 18 the soils west of a line running from Marquette to Faithorn (Fig. l) and the result is a soil that is acidic to a depth of several feet. The soils east of this line contain free lime and are consequently basic in nature. There are thirty different soil associations in the northern Peninsula of Michigan, according to Whiteside et a1. (1959) and these are grouped in seventeen land divisions. Actually there are only eight land divisions of any extent and-these are characterized in a table (Table 1). Figure 7 locates these features. In addition, there are eleven other smaller land divisions defined for the region, included in land division H, but will not be considered herein. THE? TABLE 1. Land Divisions Topography A H Undulating to hilly soils Gently rolling to very hilly soils Hilly and mountainous Level to rolling soils Level to rolling well-drained soils Level to hilly Low, moist, and wet areas A combination of many small land divisions (Modified from Whiteside et a1., 1959) 19 SOILS OF NORTHERN MICHIGAN Locality Western Upper Peninsula at elevations of 1300 to 1900 feet above sea level Western Upper Peninsula at elevations of 1400 to 1800 feet above sea level Rough, steep, and mountainous areas in the western part of the Upper Peninsula; including iron and copper ranges and the Huron and Porcupine Mountains at elevations of 1200 to 2000 feet above sea level; also Isle Royale Drumlin area of Menominee county and east part of Alger county Very extensive in the eastern Upper Peninsula; including Alger, Delta, Chippewa, Luce, Mackinac, Menominee, and Schoolcraft counties Parts of Alger, Luce, Schoolcraft, Chippewa, and Mackinac counties Most of Schoolcraft, large parts of Chippewa, and Luce, and also areas in Alger, Delta, Mackinac, and Menominee counties Scattered small areas throughout the Upper Peninsula 20 Soil Constituents Sandy loams to silt loams, overlaying either reddish, sandy loams, loams, or course sandy glacial drift Acidic Acid, stony sand to sandy loam parent material. Course stony soil Sands, sandy loams, and silt loams over bedrock of granite, basalts, slate, shale, or schist. Soils are excessively stony and bouldery with outcrops common Well-drained limy, sandy loams, and loams Well-drained sands from materials low in lime Dominantly sands, loamy sands, and sandy loams Wet and dry sands with closely associated peats. The mineral soils were developed from sandy parent material sandy etc. loams, sand, Clays, loams, Ammma ..Hu um .mcumouarz scum sundanozv .g 93 .1551... .5 .3. .88 .333 an! ii... in? go _ .. . can 13:3 58 .538 3 33 .33... I .n 5a. , . .. .. Rf. 33:38 8.35. s3 2!. Bl «a 8383-8 4 .3 add- 0» .340. 3.393 3!. is. 5 and a2— 5 ~13. 3°- Iouu cannon 5093305433. one bag 22—: .o .33. 5 has 3 ~83 .u......... ..__ .laoa has: ou 3:3 .93. .30- .: POOH on. g .93 gash—v .30. ban to» ca MESH?— baa-00 an: 3.6 v8.5 ado. van-no.3... .528 3 H25 .Idoa and. use 3 .oluoa Bov- .30¢ UPC ion 0.30- has on gang‘s 332.8 and” o» 3.3.33- nauanofl 8393- we 0:8 22 Vegetational Features A sequence of climatic and vegetational changes followed the retreat of the last glacier in northern Michigan. Potzger (1942), by means of pollen analysis of bog samples in the Upper Peninsula, postulates that after the retreat of the glacier there were four major climatic changes with resultant changes in vegetation. When the glacier was near at hand, the climate was cool and moist and the region was characterized by a spruce- fir dominant forest. A warmer drier period halted gla— ciation and this brought about the pine maximum. It is at the height of this warming trend that the "steepe peninsula" described by Schmidt (1938) occurred. As the climate became warmer, and more moist, the hemlock (Eggqg. canadensis)and many broadleafed genera (Aggr, Betula, £9337 lug, etc.) appeared. Eventually, as the climate deteri— orated again, there was a slight increase in spruce (gigg§_ g2,), fir (Abig§_§2,), paper birch (g, papyrifera), and yellow birch (g, lutea), with a concurrent decrease in the amount of pine present. Despite the pollen evidence suggesting a decrease in pine, the predominant forest in the Northern Peninsula of Michigan at the time of settle- ment by white man was the white pine (Pinus strobus) (Fig. 8). 34- I ORIGINAL FORESTS 87° 90° OF MICHIGAN hold no!" by ounce o! Agnc Rcsoovch Asst, Economncs, Department of Aqncunure COMDIIQU (tom land OHIcc F J MARSCHNER, LEGEND "V F“ _J: tho HARDWOODS a CONIFERS L..- Spruce, Fu, Cedar 8 FLATS (Hemloch_ [—24 ] PINE 5:2 Whl'e Pmo) Nancy and mm. Pines) PINE PLAINS (Jack, 1 J [ ”TD MARSH a WET mums I _.] PRAIRE v t; C] WET a om LAND mxso FOREST (Rugged Lond) B SWAMP (Cedar, Balsam, Tamarack) -m HAROWOODS 23 A D Perejda Redrown By o! Gooquy 8 Geography Don't. College Michigan State FIGURE 8 ”I. 24 Soil types determine the contemporary vegetation dif- ferences of the Northern Peninsula. Braun (1950 divides the flora of the Peninsula into two main association; the maple-beech climax community on the fine-grained and loamy soils of the eastern sector; and the maple-basswood climax association of the more mature soils of the western sector. The western boundary of beech (Eggg§_g£andifolia), coin- cided exactly with the north-south line from Marquette to Faithorn (Fig. 1). West of this line basswood (Elli§_ americana) replaces beech as the co-dominate with sugar maple (Agg£_saccharum). In the former two associations hemlock is often a sub-dominate or even a co-dominate. In many areas secondary deciduous communities of aspen (£222: lg§_tremuloides) are replacing other forest types such as white pine, and white spruce (Eigg§_glauca), and seem to adapt well to areas that have been recently logged—off. Other deciduous communities of poplar (Populus _p,), paper or canoe birch (Betula papyifera), gray birch (g, populi- £2133), yellow birch (g, lutea), and maple (Acer rubrum), and balsam fir (Abig§_balsamea), are found in areas where the conditions necessary for the climax forest have not yet been reached. Two types of coniferous forests exist in the Northern 25 Peninsula. 0n the higher, sandy areas there are locations of jack pine (Pinus banksiana), red pine (P, resinosa), and white pine. These stands are quite extensive along the Wisconsin-Michigan border. Jack pine has increased in acre- age in areas where fires have destroyed the original forest. Red oak (Quercus borealis), white oak (Q, 2123), and red maple may also be secondary species in jack pine-red pine communities on sandy soils. Bog forests are quite extensive in the lower, wetter areas of northern Michigan such as the Manistee and Tah- quamenon drainages. These forests are characterized by a climax stand of white spruce, northern white cedar (Thgj§_ occidentalis), and balsam fir; whereas, the seral stages may support tamarack (Larix laricina), paper birch, and black spruce (g, mariana). The deciduous communities support relatively rich ground cover as compared to the pine stands. The ground cover, in- cluding Trillium grandiflorum, Smilicina racemosa, and such shrubs as Sambucus pubens, and Lonicera canadengig, is abun- dant from May all through the summer. Of all the hardwood communities, the hemlock-yellow birch—red maple has the richest undergrowth with the most species and individuals. The aspen stands have a ground cover of bracken fern 26 (Pteridium aguilinum), and/or sweet fern (Comptonia BEES: grina), that often completely covers the forest floor. The pine stand ground cover, which is represented by such herbaceous species as Agtg£_macrophvllg§, Clintonia borealis, Oryzopsis g2,, and Pteridium §Q,, creates a very barren situation because of the sparsity of individual plants and the dry condition of the forest floor. Baraga Plains, in Baraga county is an outwash plain that supports an al- most 100 percent jack pine stand. The sand deposit there is quite deep and extensive. The soil has very little ability to hold water and is acidic. Another area farther east in Alger county that is of interest zoogeographically despite a barren appearance is the Kingston Plains. This region originally supported a mature white pine forest until lumbering cleared the area of all trees about sixty years ago. Since then there has been virtually no re- forestation and all that stands on the very sandy soils now are scattered white birches, many white pine stumps and a ground cover of low grasses, ferns, and club mosses (Lyco- podineas). Scattered in this area are-several small shallow lakes and ponds. These plains are bordered by a hemlock- hardwood forest with the corresponding rich ground cover. 27 The vegetation.of Isle Royale, as described by Holt in Adams (1908), may be divided into four units: bog societies, shore societies, forests, and burnings. The bog societies are similar to the bogs of northern Michi- gan with black spruce and balsam fir as climax dominants. Bog species here however are different from the Upper Pen- insula in that they tend to be found in places other than bogs. For instance, the black spruce may be found on dry ground away from the bog, and tamarack may be found occasionally on high dry ridges.) This kind of wide rang— ing occurrence is common for these species farther north in the boreal forest. Skunk cabbage (Symplocarpus goggir dug), sundews (Drocera 52,), pitcher plant (Sarracenia purpurea), and several small grasses (Graminea), provide the ground cover. The shore communities are characteristically composed of lichens on the bare exposed bedrock and farther away from shore low junipers (Juniperus _p,) and ground hemlock (Tgxggicanadensis) are present and are followed by northern white cedar and white spruce in the interior of the island. The hardy vegetation along the shore is predominantly Canadian in origin. There are approximately 21 species of trees on Isle 28 Royale and l3-of these are deciduous. There are two cli- max forests, excluding the bog communities. The boreal forest with balsam fir, white spruce, and canoe birch is the primary forest cover on the Island, and also is of Canadian origin. This forest extends around the periphery of the Island. The ground is very damp in this forest and supports a variety of herbaceous plants including, Oxalis acetosella, Aralia nudicaulis, Cornus canadensig, Pyrola virens, Maianthemum canadense, Linnaea borealis, Mitella nugg, and Rubgg,parviflorus the most common shrub cover on Island. Farther inland, the birches become more dominant and purestands of birch are found in several sites such as Ransom Clearing. The second climax community, a forest of sugar maple and yellow birch occurs only in the southwestern part of the Island. This forest, which grows only in the interior away from the immediate influence of Lake Superior, is the northernmost extension of sugar maple. There is a very good understory of sugar maple seedlings, western thimble- berry (Rubus paryiflorus) and several smaller wildflowers and herbaceous plants. METHODS AND MATERIALS The field work consisted of a 13-day collecting trip from August 12 to August 25, 1963, and a more extensive survey from June 15 to August 7, 1964. Both the spring and summer activity periods of the animals were thereby included in the study. All 15 counties of the Northern Peninsula, including Isle Royale, were eventually visited, '(Fig. 9). An effort was made to collect intensively in each of the major ecological habitats within each county. A total of 941 specimens, involving 18 species of am- phibians and 12 species of reptiles were collected during the course of the study. Several collecting techniques were employed in secur- ing study specimens. Seines, minnow traps and dip nets were utilized for acquiring the aquatic forms while most of the terrestrial species were encountered by turning logs and rocks. All road-kills were identified, recorded, and preserved when feasible. Complete locality data was recorded for each specimen. The amphibians were fixed in a 10 percent formalde- hyde solution and stored in 65 percent ethyl alcohol. Reptiles were injected and fixed in formaldehyde and 29 30 mmo>u=m Sauna scan was mesa was you nonunamoos maauomsfioo mo an: .m mmauHm , o l.- 10- . \ e O 0.. I. \. I a In! . .4. h— . ‘c” I. n . {0’1 .. r u o :0 :a l v \ II. ’c I 7 \ . «Ila. p‘ u. . y c be .l I . . ion 3. . I I ‘3 “In a a {.1 I. ‘0: . . K '0‘ .‘ o I :- -— , a II..- 1. I . - A ‘0: a t. .. ‘I In . . " 4 “I. n ‘ i. In. . \l .l'. u, OI). \ r - I. .. I» :58.- . 3:131! I ,L 0 “III: I J O _ I ¢ 0 0 «Sue , _ U32! (I .. §l.:U-O . C‘IU to b . . .0 an“: elastic 2 Y _.. . O. I ., . . o .2 ‘ < o < z < o I I ~>>u3218 .Qmwne L I ~ ' .r. .. .\ .— V . (\IJZJ -. \ .1 II I II I I T SI - II // 32:2:- 2 . - . /. L I L . .s. u 4.0 u “f) _. .1 a ., . r . ./ A 1 » m.HH MO HOQESZ v.5 mQEHH commmnmcm smmsumn mw>ooum Hmumou mo HmQEdz Abanmav mw>oonm Hmumou N.0H MO HOQEdZ m.w¢ numcma Hmuou mo mmmucwuuwm m mm numcma Hams umcflxumz m3wmmfino masmcacwm meQD venomow masmcflswm Home: rectum V numcma ucm>1usosm Ammo >DGDOO mmmumm manusflewm Home: assumes GHmummm HMHUHHUU mswumswu mammwcau cocoguuam aw Godumwum> aflnmmumoow v mflmdfi 44 comparison with the collection from Illinois shows major differences. The Illinois population (Smith, 1961) has a smaller snout-vent length and the tail is a larger percent- age of the total length. Furthermore, in Illinois specimens have fewer vomerine teeth but this is a highly variable character and the number ranges between 10.7 and 13.0 in the Illinois samples (Smith, op. cit.). The difference in costal grooves may be due to a counting technique. All individuals caught were of the redaback phase. So far as is known, there are no lead phase individuals on record for the Upper Peninsula and Wisconsin. Breckenridge (1944) in Minnesota said the lead phase is also not found in Minnesota. Smith (op. cit.) reported that both phases are found in Illinois but.the red-backed phases occur 66 percent of the time. The frequency of the patterns vary from area to area in southern Michigan. Hemidactvlium scutatum Schlegel This very secretive salamander was not collected during this study. There are museum records for only Alger, Chip- pewa and Gogebic counties. Mr. Martin Kopenski, an instruc- tor at Northern Michigan University, states that (per. comm.) the four-toed salamander is abundant in certain areas 45 in the vicinity of Marquette. The species probably occurs throughout the Upper Peninsula where appropriate habitat is available. Its occurrence on Isle Royale has not yet been reported. Order Salientia Family Bufonidae Bufo americanus americanus Holbrook This species is recorded from all 15 counties of the Upper Peninsula in addition to Isle Royale. Table 5 shows five characters contrasted between three selected samples of Ngfg a, amerNcanug, The Isle Royale and eastern Upper Peninsula series agree quite closely but differ with varying degrees from the Northern Illinois specimens described by Smith (1961). The Michigan populations have more dorsal spots with two or more-warts, smaller heads relative to body length, smaller parotoid glands relative to body length, a lighter ventral pattern, and are larger in body length. Family Hylidae Hyla crucifer crucifer Wied The spring peeper is recorded from Isle Royale and all 15 counties of the Upper Peninsula (Table 16). Two 46 TABLE 5 Geographic Variation in Bufo americanu§_americanus Isle Eastern Northern Royale Upper Illinois** Peninsula* (12) (14) (11) Head length as a percentage of body length 24.7 25.2 28.6 Paroid length as a percentage of body length 18.9 18.9 19.4 Percentage of individuals . with one wart/dorsal spot 50.0 54.5 64.0 Ventral pattern*** 3.75 3.71 3.9 * Alger, Chippewa, and Schoolcraft counties. ** Smith, 1961. *** Ventral pattern values: (1) venter immaculate or with one pectoral spot; (2) a few scattered spots on breast region; (3) a zone of dark spots across the breast; (4) entire venter spotted; and (5) venter predominately dark. 47 populations (Isle Royale and mainland) were compared and Table 6 gives the mean values of several parameters measured. Unfortunately, a larger sample was not attained from Isle Royale and therefore a valid comparison could not be made with the mainland populations. However, on the basis of these limited data, there appears to be a slight variation in body proportions. Ny;§_versicolor versicolor Le Conte The gray treefrog is recorded in Alger, Dickinson, and Marquette counties. A. G. Ruthven in Adams (1908) mentioned a record in The University of Michigan Museum (UMMZ 7457) for Isle Royale but was unable to verify it at the time and since then the species has not been recorded from the Island. Conant (1958, map 221) indicates that this species does not inhabit the eastern half of the Upper Peninsula. However, a metamorphosing tadpole (which had both front and hind legs with large disks on the digits) was collected in a seine at the Seney National Wildlife Refuge in Schoolcraft county and was the only individual of this species collected in the field. 48 TABLE 6 Geographic Variation in Hyla crucifer crucifer Isle Eastern Royale Upper Peninsula* (3) (15) Head length as a percentage of body length 30.7 28.9 Tibia length as a percentage of body length 50.3 48.2 Snout-vent length range 19.1-33.6 20.6-29.0 Number of bars on tibia 1.8 2.4 *Alger, Luce, and Mackinac counties. TABLE 7 Geographic Variation in Pseudacris triseriata Isle Southern Royale Michigan* (12) . (ll) Snout-vent length 2.28 cm. 2.66 cm. (1.85-2.58) (2.53-2.88) Head length as a percentage 25.6 25.3 of total length (22.3-29.4) (19.4-26.9) Tibia length as a percentage 40.4 43.5 of total length (37.2-50.4) (39.5-45.1) Percentage with spots or stripes on dorsum 90.9 spots 81.8 spots *Clinton county. 49 Pseudacris triseriata maculata (Wied) Twelve individuals of this species from Isle Royale were compared with eleven individuals from Clinton county in southern Michigan (Table 7). The results show that there are differences in tibia lengths and body markings and there are therefore two subspecies involved; 2, t, maculata on Isle Royale and 2, t, triseriata in southern Michigan. Dr. Charles F. Walker (per. comm.) verified the Isle Royale population as g, t, maculata. There are no specimens re- corded from the Upper Peninsula although Conant (op. cit., map 235) shows the entire Northern Peninsula as a zone of intergradation between these two subspecies. The population on Isle Royale is limited to the rock pools on the periphery of the Island. Family Ranidae Rana catesbeiana Shaw The bullfrog was not seen or collected on either field survey. There are museum records from Baraga, Chippewa, Luce, Mackinac, Marquette and Schoolcraft counties and this agrees quite closely with Conant (op. cit., map 24). Per- sonal conversation with Mr. David Johnson, a Michigan Con- servation Officer, revealed that bullfrogs were introduced 50 into Marquette county by individuals attempting to raise them commercially. The validity of this report has not been proven. Rana clamitans melanota (Rafinesque) The green frog is one of the most abundant amphibians in Upper Michigan. The species is recorded from all 15 counties and Isle Royale. Measurements of 87 specimens from selected localities in the Upper Peninsula and seven speci- mens from Isle Royale revealed no significant differences in tibia and femur lengths as correlated with snout—vent lengths. The Isle Royale population had the smallest snout-vent length (76.6 as compared to 85.9 average in mainland specimens) of all the individuals, and also had the lowest percentage of individuals with the belly unmarked (51.1 as compared to an average of 85.5 on the mainland). The number of specimens from Isle Royale-was quite low and this may account for the variation from the Upper Peninsula samples or there-may be selection pressure for these characteristics. Rana palustris palustris N§_Conte The-pickerel frog is recorded from Chippewa, Marquette, and Menominee counties. Conant (op. cit., map 239) includes 51 the entire Upper Peninsula within the range of this species. No specimens were taken during the field surveys despite the fact that large numbers of Bag§_clamitans and N, pipiens were collected. This may be attributed to the fact that N§g§_palustris has a more restricted habitat which was col- lected less often than those areas preferred by the other two species mentioned. Rana pipiens pipiens Schreber Second only to the green frog in abundance, the Leopard frog (3, p, pipiens) is recorded from all counties but is not known from Isle Royale. Body measurements of 54 speci- mens from selected populations in western, central, and eastern Upper Michigan failed to show significant differ- ences in variation. However, there seems to be a slight increase in dorsal spotting from west to east (average num- ber 14.4 as compared to 16.2) as well as a reduction in the occurrence of the snout spot (94.7 to 81.8%). Rana septentrionalis Baird The mink frog (not found in the Southern Peninsula of Michigan) occurs in all counties of the Northern Peninsula in addition to Isle Royale. Table 8 contrasts three 52 TABLE 8 Geographic Variation in Rana septentrionalis Isle Western Eastern Royale Upper Upper Peninsula* Peninsula** (2) (18) (7) Head length as a percentage of total length 33.9 26.9 28.0 Tibia length as a percentage of total length 49.1 46.5 46.1 Femur length as a percentage of total length 50.6 48.0 47.4 Number of spots on dorsum 8.5 6.7 8.3 * Baraga, Dickinson, Houghton, Iron, and Marquette counties. ** Alger, Mackinac, and Schoolcraft counties. 53 populations which show only slight variational trends. The small number of specimens collected from Isle Royale and the eastern Upper Peninsula may skew the results. Rana sylvatica Le Conte The wood frog is found on Isle Royale and all 15 coun- ties of the Upper Peninsula. Some authors divide this species into several subspecies according to various body proportions and colorations. Schmidt and Necker (1935) identified the specimens from Illinois, Indiana and Michigan as N, g, gag: tabrigensis despite the fact that their specimens correspond to their 3, g, sylvatica body proportions. They based their conclusion on body coloration. Smith (1961) emphasizing slightly different body propor- tions, found that specimens from northeastern Illinois are classified as N, g, cantabgigensig, Others claim that since these subspecific characteristics are part of a general north-south gradient, there should be no formal subspecies named (Conant, 1958; Logier and Toner, 1961). Table 9 provides data on specimens selected from widely diverse areas in Michigan and which represent populations from the presumed ranges of both forms. Utilizing these data in another manner (Table 10) the status of the Michigan 54 mmfiussoo muuwsvumz can mmaucsoo EMSOGH can .souswao .muumm as: mmwuasoo spawn can .m3mmmflau .HmmH¢ as .3mc0030M .Gomcflxofla .mmmumm e ¢.~m 1m.mmum.~¢V m.mm xe.mmIn-mmv mama AH.mmIv.m~V v.ae m.~m 1m.nmuo.nvv m.mm Am.emnm.amv b.mN Am.OMI¢.mNV m.~m Ao.mmum.mee H.mm Am.nmuo.amv e.mm 1m.omua.omv H.Hm Am.emnn.mev m.¢m Am.mmnm.mvv m.m~ A¢.mmum.sme mmfluum Henson £uw3 unmoumm Hon Hmuflnnouwucw ##83 unmoumm cmtha ummmun £ua3 wmmucmonwm numcma soon mo mmmucmonmm m mm numcwa HDEmm spasms mean no ommusmoumm m mm Sumsma Manse spasms seon mo omnusmoumm m mm gunned cmwm Away sescmmwnuwz GHOSusom AHNV «asasmsflcwm HOQQD cumummm Ame emasmswcmm Homes cumummz Abv mammom MHmH muwum>aam mcmm aw m MAM¢B coflumwhm> caeamumoww 55 TABLE 10 Five Characteristics Distinguishing Subspecies in Rana sylvatica Dorsal Breast Tibia Tibia Head Totals Stripe Spotted Short 55%. 33% BL* BL* C S Southern Michigan** S C S S C 2 3 Eastern Upper Peninsula** C C S C C 4 1 western Upper Peninsula** S C S S C 2 3 Isle Royale** C C S C C 4 l C Affinities closer to Rana sylvatica‘cantabrigensis S Affinities closer to Rana sylvatica sylvatica * Body length ** Same counties as Table 9. 56 population is further clarified when the five supposedly diagnostic characters are evaluated. The results show that the southern Michigan and western Upper Peninsula populations agree more-closely with the characteristics of N, g, sylvatica and that the Isle Royale and eastern Upper Peninsula samples agree with the charac- teristics for N, g, cantabriqensig, These results, there- fore, are inconclusive and until parameters are established 'that more clearly distinguish the races, the species is herein regarded as showing only clinal variation. Class Reptilia Order Chelonia Family Chelydridae Chelydra serpentina serpentina Linnaeus Seven specimens of this species were collected during the 1964 field survey including the first reported observa- tion of the snapping turtle in Iron county (5.5 miles west of Beachwood). The species is recorded from all but Alger, Houghton, Keweenaw, Luce, and Ontonagon counties and prob- ably occurs in these counties. It is very abundant at the Seney National Wildlife Refuge in Schoolcraft county. The snapping turtle has not been recorded from Isle Royale. 57 Family Emydidae Chrvsemys picta (Schneider) Two subspecies, Chrysemys picta Ng;;;_and Q, E: mag: qinata and intergrades thereof are present in Northern Michigan. Conant, op. cit. (map 21), indicates that the two subspecies intergrade in the eastern half of the Upper Peninsula with Q, E, §§;;;_coming in from Wisconsin and Q, Q, marginata entering the peninsula from southern Michigan and Canada to the east. There are, however, recorded intergrades from both Chippewa and Gogebic counties (Table 17)(UMMZ) indicating that the entire Peninsula is an area of intergradation. An adult female 9, p, Nell; was col- lected by the author on Isle Royale thus confirming the subspecies present on that Island. The species has not been recorded from Delta, Iron, Keweenaw (except Isle Royale) and Mackinac counties. This again appears to be a case of insufficient collecting surveys. Clemmys insculpta (Le Conte) The wood turtle is recorded from Dickinson, Gogebic, Menominee, and SChoolcraft counties. It was not collected in either the 1963 or 1964 field surveys, however, it pre- sumably is found throughout the entire Northern Peninsula (Conant, op. cit.). 58 Emydoidea blandingi (Holbrook) One skeleton of this species was collected along Stur- geon Creek one-half mile west of Loretto in Dickinson county. This specimen consisted of an old partly disartic— ulated shell (plastron) but positive identification was possible. The Blanding's turtle is also recorded from Marquette county. Order Squamata Suborder Sauria Family Scincidae Eumeces fasciatu§_(Linneaus) This is the only species of lizard which occurs in the Northern Peninsula of Michigan and is recorded only from Menominee and Marquette counties. These records are inter- ‘esting in that the collection sites are approximately 40 and 100 miles respectively north of the previously reported northern limit of this species in Wisconsin. In certain areas near Marquette the five-lined skink is quite abundant. 59 Suborder Serpentes Family Colubridae Diadophis punctatus edwardsi (Merrem) The ring-nedked snake is recorded in Iron, Mackinac, Marquette, and Schoolcraft counties. Museum material ex- amined by the author included fifteen specimens from several areas in the Upper Peninsula and nine specimens from Lee; 1anau county in the Southern Peninsula. In addition nine individuals from Illinois (Smith, op. cit.) were compared to these two populations. Table 11 illustrates the con- siderable individual variation within various populations of the subspecies edwardsi, particularly with respect to the supralabial and infralabial scales. Ventral and caudal scale counts are higher in the northern populations. How- ever, due to an insufficient number of specimens, it was not possible to discern clinal morphological variations in an eastéwest direction across to Northern Peninsula. Elaphe vulpina vulpina (Baird and Girard) The "pine snake" as it is called in the Northern Penin- ‘sula, entered the Peninsula from Wisconsin to the south and is recorded as far east as Alger and Schoolcraft counties. 60 TABLE 11 Geographic Variation in Diadophis punctatu§_edwardsi Northern Southern Illinois*** Michigan* Michigan** (15) (9) (9) Number of ventrals 156.5 155.7 ----- Males ----- 156.7 144.6 Females ----- 155.3 ‘ 152.5 Number of caudals 53.8 49.7 ----- Males ----- 50.3 48.7 Females ----- 49.3 41.5 Tail length as a percentage of total length 21.6 19.2 ----- Males ----- 19.6 20.1 Females ----- 18.9 16.7 Percentage with supralabials 8+8 53.3 66.0 ----- Males 7+7 26.6 0.0 55.5 Females 8+7 13.3 33.0 ..... Percentage with infralabials 8+8 93.3 11.0 41.7 Males 7+7 0.0 55.5 8.3 Females 8+7 6.6 33.3 29.1 Percentage with chin spotting 26.7 77.8 33.3 Percentage with dots on venter 40.0 33.3 ----- * Iron, Mackinac, Marquette, and Schoolcraft counties ** Leelanau county *** Smith, 1961 61 TABLE 12 Geographic Variation in Elaphe vulpina vulpina Northern Illinois** Michigan* (17) (40) Percentage with 25 dorsal scales in anterior body region 68.4 97.0 Percentage with supralabials 8+8 78.9 83.0 Percentage with infralabials ll+1l 33.0 61.0 Number of ventrals: Males 200.0 ----- Females 204,0 ..... Number of caudals: Males 63.0 64.0 Females 56.4 54.7 Tail length as a percentage of total length: Males 15.9 16.0 Females 14.8 13.0 Northeast Northwest Northern Illinois Illinois Michigan (9) (13) (17) Number of 41.4 38.6 38.6 dorsal blotches (37-52) (35-41) (34-44) * Dickinson, Iron, Marquette, and Menominee counties ** Smith, 1961 62 Nineteen individuals of this species were collected from Dickinson, Iron, Marquette, and Menominee counties. In addition to the counties listed above Elapha vulpina is also recorded from Delta county. A Michigan Conservation Officer in Chippewa county mentioned seeing a pine snake near White- fish Point. However, there are no recorded collections from that area. A comparison of specimens from northern Illinois and northern Michigan revealed that the dorsal blotch count in- creased from south to north, present in Smith's (op. cit.) specimens from several areas in Illinois, did not hold true for the northern Michigan populations (Table 14). Smith (per. comm.) suggests that this may be an influence of the eastern subspecies N, y, gloydi which has a higher dorsal blotch count. Comparisons also indicated considerable variance in other parameters such as the anterior scale row count, supralabials, and infralabial counts between the two populations. HeterodonplatyrhinOS Latreille The eastern hognose snake was not collected in the field by the author. However, there are two specimens from 12 miles west of Stephenson (Menominee county) in the Museum 63 of Zoology at the University of Michigan. These are the first specimens reported from the Upper Peninsula and this occurrence extends their range farther north than illustrated by Conant (0p. cit., map 108). The author, while collecting in the Stephenson area, had been told of the "Blow Snakes" near Banat which is eight miles northwest of Stephenson. The Banat area is characterized by dry sandy soils with predominantly white, red, and jack pine forests. These conditions are similar to habitat preferred by the eastern hognose~where it is found farther south. Lampropeltis triangulum triangulum The milk snake was not collected by the author in either field survey, however, the species is recorded (UMMz) from Marquette and Mackinac counties. Conant (1958, map 130) gives the northern limit ofiLampgopelti§_triangulum Egg: angulum in Wisconsin as Green Bay, but does show an isolated record from the Huron Mountains in Marquette county. The species is not present on Isle Royale but is present throughout southern Michigan and Ontario east of Chippewa and Madkinac counties. 64 Natrix gipedon gipedgg_Linnaeus The common water snake appears to be restricted to the eastern half of the Upper Peninsula. It is found in Alger, Chippewa, Delta, Mackinac, and Schoolcraft counties. The northern water snake is reported by a resident of Saint Ignace to be fairly common along certain areas of the Lake Michigan shoreline in Mackinac county. It appears that this species has entered the Upper Peninsula from Canada to the east and perhaps Midhigan to the south. This species was not collected by the author. Qpheodrys vernalis vernalis (Harlan) Ten specimens of this species were examined (eight of which were collected in the field) and all agree with the description given by Grobman (1941) for 9, y, vernalis. No significant variation was observed in the specimens studied. This species is recorded from all counties except Houghton, Keweenaw and Ontonagon. ,Regina septemvittata (Say) The queen snake is not recorded from the Upper Penin- sula mainland. However, there is a record (UMMZ) from Bois Blanc Island (Mackinac county) in Lake Michigan. Conant (op. cit., map 91) indicates the northern limit of this 65 species as the middle of the Southern Peninsula of Michigan, but does show an isolated collection site from the Straits of Mackinac. Storeria occipitomaculata occipitomaculata Storer Recorded from Isle Royale and all 15 counties, the red- bellied snake is one of the most common snakes in northern Michigan. An analysis of 115 individuals from throughout the Upper Peninsula revealed very little geographical vari- ation in most parameters with only the infralabials and supralabials showing some individual and geographical dif- ferences. Four selected populations are compared in Table 11 of illustration. The width of the red-colored venter increases in the more northern samples. However, using the t-test these dif- ferences are not significant at the five percent level. Thamnophi§_sauritus septentrionalis (Linnaeus) The ribbon snake's northern range limit is Bois Blanc Island but it is not found on the mainland of the Northern Peninsula. There has been some confusion among herpetolo- gists about the occurrence of this species in the Upper Peninsula and Isle Royale. However, Dr. Charles F. Walker 66 TABLE 13 Geographic Variation in Storeria occipitomaculata occipitomaculata Roscommon counties *** Chippewa, Luce, and Mackinac counties Southern Isle Northern Eastern Michigan* Royale Upper Upper Peninsula** Peninsula*** (12) (18) (44) (18) Number of ventrals 123.6 122.4 122.1 120.8 Number of caudals 44.4 43.5 44.8 44.3 Number of ventrals + . caudals 167.3 165.9 167.0 165.3 Width of red color on aventrals 3.15 3.8 3.3 3.3 Tail length as a percentage of total . length 21.3 20.8 21.4 20.2 Percentage with supra- labials 6+6 100.0 94.8 82.2 88.9 Percentage with infra- labials 7+7 75.0 66.1 54.5 61.1 * Clinton, Ingham, Midland, Oscoda, Otsego, and ** Gogebic, Houghton, Keweenaw, and Ontonagon counties 67 (per. comm.) has informed me that the specimen from Isle Royale identified as Thamnophis sauritus (UMMZ 7452) is lost and quite likely was actually Thamnophi§_sirtalis. There are at the present time no specimens of Thamnophis sauritus recorded from Isle Royale or the Upper Peninsula. Thamnophis sirtalis sirtalis Linnaeus From numbers collected by the author this is the most abundant reptile in the Upper Peninsula and the species with the widest eological tolerances. The eastern gartersnake is recorded from all but Dickinson and Keweenaw counties on the Peninsula. Individuals from various areas were analyzed and the data is presented in Table 14. The results support the findings of Ruthven (1908) in his extensive work on variation in the genus Thamnophis. The ventral and caudal scale counts appear to be lower in the eastern than in the western ex- tremes of the Northern Peninsula. However, larger collec- tions are needed before an actual geographical gradient may be determined. Superficially, the Isle Royale specimens appear to be influenced by the red-sided gartersnake (T, g, parietalig) found in Minnesota and Canada to the north and west (Fig. 10). Examination of the data, however, indicates that the 68 TABLE 14 Geographic Variation in Thamnophis sirtalis sirtalis Isle Western Eastern Southern Royale Upper Upper Michigan*** Peninsula* Peninsula** 010) (15) (5) -- Number of ventrals --------------- 152.0 Males 149.0 ............... Females 152.0 153.2 150.6 ————— Number of caudals ____________________ Males 80.0 ---------- 73.0 Females 68.2 71.1 64.2 64.0 Percentage with red pattern 44.4 0.0 20.0 ----- Percentage with bars on supra- labials 4+5 55.5 42.8 ---------- * Baraga, Gogebic, Iron, and Ontonagon counties ** Chippewa and Mackinac counties *** 'Eaton county, Sperry (1901) JA‘ 6 .J' FIGURE 10 FREQUENCY OF RED LATERAL COLOR PATTERN IN Thamnophis sirtalis sirtalis 100 ‘T 50 '* l I I I Isle Western Middle Eastern Royale Upper Upper Upper . Peninsula* Peninsula** Peninsula*** (9) (16) (16) (S) * Baraga, Gogebic, Iron, and Ontonagon Counties ** Alger, Marquette, Menominee, and Schoolcraft Counties *** Chippewa and Mackinac Counties 7O similarities are merely a local color variation in 2, g, sirtalis. A specimen exhibiting a blue melanistic phase of E, g, sirtalis was also collected from the Island. This variation was partially described in 1908 by Ruthven and apparently the population has maintained the variation since that time . The sex ratio of g, sirtalis collected was unusual in that of the 46 specimens in which the sex was determined only four were males. 'Family Crotalidae Sistrurus catenatus catenatus Rafinesque This is another species recorded (UMMZ) from Bois Blanc Island and not from the Northern Peninsula mainland. It occurs throughout the Southern Peninsula of Michigan but the range is terminated at the straits. HABITATS AND DISTRIBUTION A particular species occupies a mosaic distribution throughout its range and has a certain preferred habitat to which it is largely limited. Table 15 lists six major habi— tat types from which animals were collected during the course of the study. Collecting time was not evenly distributed among these habitats; therefore the results may be slightly inaccurate. However, the results do show primary habitat preferences. The terrestrial habitats were not exploited as well by the amphibians (Table 15). The importance of water in their daily metabolic needs limits their available habitats. The reptiles, on the other hand, were not limited by water requirements and were nearly equally distributed between aquatic and terrestrial habitats. The coniferous forests described earlier in the Intro- duction support a very limited herpetofauna. The lack of adequate ground cover and the acidity of the soil present eological barriers that are overcome by few amphibians or reptiles. The deciduous forest, which includes a variety of plant associations, harbors several more species than the coniferous 71 72 0H ca mmflommm mo Aeneas Hmuoa + + + + + + + I + + mum>flu Imammnum + + + + + + + + + + I + + Iwuonm .ccom mxma I + + + + I + + + + I I I I I I I + + I I + ummuom mumwuom QEMSm Icamwm msos Imom ammo Inflown owumsv< mummuom macho Inacoo Hmwuumwuuwa mflamuuwm magmoqfimna mumasomeouamwooo sensuoum maamcum> mumcomneo mswmas> mammHm Amuseaman mmeressm wuuwm whammhunu scaucmmumm unchamsu mowum>ahn scum mHHMGOAHucwumwm scum msmwmfim mama mcmuafimao mama Hoaoofimum> mamm Homeosuo Mama museummfluu manomcsomm mscmoaumam omsm mswumsao coconuwam mcmommcaufl> msaamnunmouoz wamumumH mEoumhnE< mmflommm cmuomaaoo mwflommm mo mwocwummmum umuwnmm ma mqm<fi 73 forest. The increased species diversity and species abund- ance of plants of the ground cover provide habitat for such characteristic species as Plethodon cinereus. The open field-forest edge category is actually a com- bination of many habitats, namely, abandoned barns, bedrock outcrOppings and sandy outwash plains, in addition to old fields. The several species (Table 15) that were present in these various habitats all exhibited one common character- istic; that is, the ability to survive under relatively dry conditions. For amphibians, the bog-swamp offered a suitable habitat for four species, all ranids. The lake-pond environment was the most productive collection area; however, it was also the habitat in which most collecting time was spent. Ten species of amphibians and four reptilean species were taken from this area. Three ranids (Naga_g;amitans, pipiens and septentriona;;§) have exhibited very definite spatial rela- tions in regard to each other in this lake-shoreline habitat. §§g§_pipiens was often found well away from water or along the shoreline; N, clamitans was found to stay close to the shoreline either in water or on land; and N, septengrionaiig was usually found well out into the water. 74 The river-stream habitat did not harbor as many species as the quite similar lake-pond habitat. This may be ex- plained by the slightly more rigorous environmental condi- tions of a flowing body of water. Tables 16 and 17 are summaries of the geographical range according to counties of each species as discussed under the species account. The Northern Peninsula has been divided into an eastern and western section in order to distinguish better between species with a primarily eastern or western distribution in the Upper Peninsula. Although counties are not a natural method of defining species ranges, it was the most feasible method in the limited amount of time available. 75 .>DGDOU EGAN cwcuoomu no: u I can "Nnbmz can .Zsz .NSZD um mcuoowu u + nmcsum mswusc cwuumaaoo mwflommm n o * I eICIe I¢+¢ +6 +d§+¢+¢ +¢ I'$$ IOOI l+ IGGI++I I III I |d>¢¢ IiI moaum>ahm mflamcofluucwummn newsman .m mfluumsamm .m muocmHmE msmufiamau mamamnmmumo ++e Ieee + e Ie-I-e Ieoe I¢+¢ Iee+ Iee+ Ie++ Iee+ I Ie+e+ II +¢+¢¢ .> Homeosuo .o mscmoflHmEm .m I eI+ I e-+ ace IIS + + I eI+ + $0 + + + I el+ I + + chcb scam mcmm scum scam scam mcmm Sansone museummauu manomcsmmm mswuwcflo .u coconumam I I mcmummcflua> .> msaamzuamouoz I I I I I mwmsmcmwmflsoa grusaamsufimouoz I I I I I msuowum msmoasomfi mausuomz msmoHDUME .E mausuowz Esumusom Edahuomcflfimm mass mahm x omsm a ..Edcfiumfiu .u maoumhnfi< EsumHSUME mEoum>QE¢ wamumuma meoumhnfid DPUTXDEW ‘+ + emeddrqa .+ I I 195“! e++ee+++ I UOII + + PAIGG eeurmouew ennenblew -+-+ uosurxorq 262123 + . M2uaamsx .+ uoqunoH uobeuoquo + orqabos + + mmwowmm 439131001138 0 212503 21s: a, I masmsaswm HQOD cumummm MHSmcacmm swans cumummz ecmmflnoflz cumnuuoz ca mananflnmam mo cofiusnauumfln hucsoo 0H mqm<9 76 cwcuoomu uoc mwflummm u H uNoam: cam .znmz .uzzo um meuoomu .ccmamH unmam mflom Eoum haco cwcnoomu H mm cam “mucsoo Scum 1+ mm 08mm can mumcflmumfi .Q .0 x aaamn muowm whammmnno + nmcsum mcflusc cmuomaaoo mmfiommm n o * a a + e + o e o e I o + + e e e mflamuuam .m magmosamge mm I I I I I I I I I I I I I I I mwamsofluucmumom msuausmm magmoeemna mm I I I I I I I I I I I I I I I nauseoumo .u unassumam + + + e + e e e + + e + e + + -9 mumasomaouwmflooo .o mfluououm mm I I I I I I I I I I I I I I I mumuua>fimummm scammm + a + + + e + I + + e I I + I + maamcum> .> mhucomseo mm + I + + + + I I I I I I I I I cocmmsm .m xauumz + I I I I I I I + I I I I I I I Esasmcmauu .u mfluaomoumamn I I I I I I + I I I I I I I I I mocfinuhumHm .m socoumuwm I I I + e + e I e e I I I a I I msamas> .> wnmmam mm I I + I I I I a I I I I + I I encumscw msumuocsm manmocmwa I I I I I I + I + I I I I I I I msumaommm mmomfism I I I I I I I I + o I I I I I I Amusecman.mmenoessm I I I + I I + I I + I I I I I + mumasomsfl mmEEwsUA + + I e + I + I + + o I I o I + maflucmanmm mncmawno I m m H I I o I H I o I I I I I mumcfimuma muofim whammhuno I I I I m + o I m I + o I m m HHHOQ maven mhfiwmmunu I 9 memsmmw smmaam.mm o t. o u. T..b u T. 1 o 1 M n o 4 H .d a o 1 e o a .b x. 2 a .6 u o a I .d o 2 1 m n t..b e U. u a. u e T. I. H e u 2 u 1. 2 Tr em. . mm: mm a. m a 1... a e... u .3... J .L 1 a masmcfldmm Amman cuoummm masmcficmm Home: cumumoz *cmmanoaz cum£uuoz CH mmaaummm mo coausnfluuman mucsoo 5H QO