LIBRARY MIchIgan State UnlversIIy PLACE ll RETURN BOXto movathh mum yum TO AVOID FINES "mm on or Honda. duo. DATE one DATE DUE DATE DUE MSU Iommmwomlm mm: AN ECOLOGICAL STUDY OF THE COMPOSITION, STRUCTURE AND DISTURBANCE REGINES OF THE PRE-EUROPEAN SETTLEMENT PORESTS OF WESTEEN CHIPPEWA COUNTY, NICHIGAN By David Lynn Price A THESIS Submitted to Michigan State University in partial fulfillment of the requirenents for the degree of ' MASTER OF SCIENCE Departnent of Forestry 1994 ABSTRACT AN ECOLOGICAL STUDY OF THE COMPOSITION, STRUCTURE AND DISTURBANCE REGIMES OF THE PRE-EUROPEAN SETTLEMENT FORESTS OF WESTERN CHIPPEWA COUNTY, MICHIGAN By David Lynn Price To successfully inplenent ecosysten nanagenent an understanding lust be achieved regarding how different forest conunities and ecosystens function and interrelate at the landscape level. This study explores the conposition, structure and disturbance patterns of the pre-European settlement forests of eastern upper Michigan, by reconstruction of the forests fro: General Land Office Survey notes. Results suggest that the pre-European settlenent landscape was a vast array of irregular patches, composed of different successional stages and forest associations of different age and size classes. The composition and structure of the forest was driven. by fire, windthrow, insect related nortality and beaver (mm: ‘W) floodings. Henlock (m M) was a dominant species in the landscape. The. results of the study provide a foundation for understanding how today's forests differ fron those that dominated the landscape before Europeans began to harvest tinber. ' This thesis is dedicated to Stephanie. iii ACKNOWLEDGEMENTS I wish to thank Bill Enslin and Brian Buckley at the Michigan State University Center for Renate Sensing for naking this research possible. Their programing expertise and advice regarding the use of C-MAP were invaluable. I also wish to thank.LeRoy Barnett.of the State.Archives of Michigan for allowing ne extended access to the cm survey notes, Karen ‘Waalen of the ‘U.S. Forest Service for 'the querying' of conpartnent records on the Hiawatha National Forest, and Dr. Don Dicknann and Dr. Dan Brown of Michigan State University for their review of and connents on this thesis. I an especially in debt to ny advisor, Dr. Kurt Pregitzer who provided advice and funding for this research over the past two years. iv TABLE OF CONTENTS LIST OF TABLES ....................................... viii LIST OF FIGURES ...................................... xi INTRODUCTION ......................................... 1 STUDY AREA ........................................... 9 History and Location ............................ 9 Environnental Characteristics ................... 13 Clinate .................................... 13 Landforn and Soils ......................... l3 Vegetation ................................. 16 METHODS .............................................. 18 GLO Surveys ..................................... 18 Data Collection ................................. 23 Data Analysis ................................... 24 Use of the C-MAP Geographic Infornation .... 24 System Composition and Structure ... ........ ....... 27 Disturbance Regines ........................ 32 RESULTS AND DISCUSSION ............................... 38 Surveyor Bias ................................... 38 Forest Composition and Structure ................ 45 The Northern Hardwood Forest Type .......... 48 The Mixed Pine Forest Type ................. 54 The Jack Pine Forest Type .................. The Mixed Conifer/Deciduous Upland ......... Forest Type The Mixed Conifer/Deciduous Lowland ........ Forest Type The Mixed Conifer Swamp Forest Type ........ Distu rbance Regimes ............................ The Northern Hardwood Forest Type .......... muix“ Pin. ror..tWOOOOCOOOOOOOOOOOO The Jack Pine Forest Type .................. The Mixed Conifer/Deciduous Upland ......... Forest Type The Mixed Conifer/Deciduous Lowland ........ Forest Type The Mixed Conifer Swamp Forest Type ........ Forest Comparisons: Past and Present ........... SUMMARY AN APPENDIX A APPENDIX B APPENDIX C APPENDIX D Forest Composition ......................... rot..t StruCtur. ......OO.........OOOOOOOOOO Disturbance Regimes ........................ DmchIons ..........OOOOOOOOOOOOOOO..... A dictionary of data field codes used in the General Land Office Vegetation Entry (GLOVE) Program. . . . . . . . . Township forest cover type maps of ....... western Chippewa County, Michigan. Diameter distribution graphs for ......... significant tree species in the northern hardwood forest type. Diameter distribution graphs for ......... significant tree species in the mixed pine forest type. vi 59 60 65 74 9. 102 108 112 117 121 127 127 134 137 143 147 150 185 190 APPENDIX 3 Diameter distribution graphs for ......... significant tree species in the mixed conifer/deciduous upland forest type. APPENDIX F Diameter distribution graphs for ......... significant tree species in the mixed conifer/deciduous lowland forest type. APPENDIX G Diameter distribution graphs for ......... significant tree species in the mixed conifer swamp forest type. LIST OF REFERENCES vii 194 201 205 210 LIST OF TABLES Table 1. Average climatic data for Chippewa county, nichig‘n. ......OOOOOOOOOOOOOOOOOO... 1‘ Table 2. Species used by surveyors for bearing tr...’ 00............OOOOOOOOOOO0.0.0.000...O 20 Table 3. Analysis of mean distances of bearing trees for diameter and species bias in the northern hardwood forest type. .......... 39 Table 4. Analysis of mean distances of bearing trees for diameter and species bias in the mixed conifer/deciduous upland forest type. ....... 40 Table 5. Analysis of mean distances of bearing trees for diameter and species bias in the mixed conifer swamp forest type. .................. 41 Table 6. Analysis of mean distances of bearing trees for diameter and species bias in the mixed pine forest type. ................. 43 Table 7. Analysis of mean distances of bearing trees for diameter and species bias in the mixed conifer/deciduous lowland forest type. ................................ 44 Table 8. nasal area, density and mean DBH estimates for the pre-European settlement forest types of western Chippewa County, Michigan. ................................... 49 Table 9. Relative density and dominance of tree species in the northern hardwood forest tm. ......OOOOOOOOOOOOOOO0.0...0.0.0.0...O. 50 Table 10. Relative density and dominance of tree species in the mixed pine and jack for..t t”... O.......OOOOOOOOOOOOOOOOOO0.0.0 55 Table 11. Relative density and dominance of tree species in the mixed conifer/deciduous “pl.“ tor.'t tm. .....OOOOOOOOOOOOOOO0.0.. 61 viii Table Table Table Table Table Table Table Table Table Table Table Table Table 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. Relative density and dominance of tree species in the mixed conifer/deciduous 10'1“ torut tm. OOOOOOOOOOOOOOOOOOOOOOOO Relative density and dominance of tree species in the mixed conifer swamp torut tm. ......OOOOOOOOOOOOOOO000......O. Number of individual disturbance events observed by surveyors in each forest type. .. Total number of estimated individual disturbance events in each forest type. ..... Disturbance frequencies (events/year) by forest type and disturbance regime, based upon 15 and 30 year recording intervals. .... Rstimated return intervals (years) for disturbances in each forest type, based upon 15 and 30 year recording intervals. .... Estimated areas of annual disturbance (ha/yr) and percentages of total area affected in each forest type, based upon 15 and 30 year recording intervals. ......... Estimated total number of disturbance events for each size class in the northern hardwood forest type. ....... ....... lstimated total number of disturbance events for each size class in the nixed p1" for.‘t tYP" ......OOOOOOOOOOOOOOOOO.... Rstimated total number of disturbance events for each size class in the jack p1“ tor..t tm. ......OOOOOOOOOOOOOOOOO.... lstimated total number of disturbance events for each size class in the mixed conifer/deciduous upland forest type. ....... Observed plus estimated nunber of disturbance events for each size class in the mixed conifer/deciduous lowland forest type. ........................... ..... Estimated total number of disturbance events for each size class in the mixed conifer swamp forest type. ix 66 71 78 94 95 96 99 103 109 113 118 123 Table 25. Pre-zuropean settlement forest type areas. .. 129 Table 26. MIRIS (1978) cover type areas. .............. 130 Table 27. Estimates of basal area, density and mean DOM for pre-Ruropean settlement and present day forest types. ................... 135 LIST OF FIGURES Figure 1. Regional Landscape Ecosystems of Upper Michigan, Regions III and IV. ............. 10 Figure 2. Example C-MAP plot of bearing trees and digitized forest type boundaries. ......... 26 Figure 3. Pre-European settlement forest types of western Chippewa County, Michigan. ........ 46 Figure 4. Density of tree species in the northern hardwood forest type. ..................... 51 Figure 5. Basal area of tree species in the northern hardwood forest type. ............ 51 Figure 6. Diameter distributions of tree species in ' the northern hardwood forest type. ........ 52 Figure 7. Density of tree species in the mixed pine for..t tm. ......OOOOOIOOOOOOOOOOOO...... 56 Figure 8. Basal area of tree species in the mixed pim10r..ttypee .0.........OOOOOOOOOOOOO. 56 Figure 9. Diameter distributions of tree species in the mixed pine forest type. ............ 57 Figure 10. Diameter distributions of tree species in the jack pine forest type. ............. 60 Figure 11. Density of tree species in the mixed conifer/deciduous upland forest type. ..... 62 Figure 12. nasal area of tree species in the mixed conifer/deciduous upland forest type. ..... 62 Figure 13. Diameter distributions of tree species in the conifer/deciduous upland forest tm. 000............OOOOOOOOOOOOO00.0.0... 6‘ Figure 14. Density of tree species in the mixed conifer/deciduous lowland forest type. .... 67 xi Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 28. Basal area of tree species in the mixed conifer/deciduous lowland forest type. .... -Diameter distributions of tree species in the conifer/deciduous lowland forest tm. ..................................... H ocm HHH mcoflmom .cmoflSOfl: Momma no mEoum>moom oomomosmq Hosoflmom .H shoves ooodood; 35m m: ..2 . h . _.t 2.. 3.. N: . Z: \ m 3 o. 3: wood hosum 11 the surveyor's township plat maps of the study area along the south shoreline of the Nhitefish Bay of Lake Superior. The extent of the villages ranged from 24 to 41 ha (60 to 100 ac). The first influences of European culture can be traced back to the early 17th century when the region was first penetrated by French trappers and explorers. The first European settlement occurred in 1668 with the establishment of a Jesuit Mission at Sault Ste. Maria. A second missionary settlement was established in St. Ignace in 1671. The region remained under French influence until the close of the French and Indian wars in 1763, when England gained control of much of the French empire in North America. The United States in turn, gained control of the region free England following the Nar of 1812. The region remained sparsely populated and relatively undeveloped until the mid-1800's. The non-native population of the entire Upper Peninsula of Michigan was estimated at 1,300 people in 1840 (Naramanski 1989). The survey of township lines in Chippewa County was conducted in 1840 by Nilliam Austin Burt, the inventor of the solar compass used in the surveys of the period. Bourdo reports in his 1954 dissertation that Burt's work ”was of excellent quality" and that his "integrity is beyond question“. Subdivision of the township by surveys of the section lines was begun in 1845 and was coaplete by 1850. The survey notes show that the subdivision surveyors were Harvey Mellon, Nells Burt, James H. Mullett and Henry 12 Brevoort. The first sawmill in the Upper Peninsula was established in Sault Ste. Maria in 1822 for the construction of Fort Brady. By 1835, when the sawmill was leased to the American Fur Company, limited logging had progressed to encompass an area only 8 km (5 mi) from the town (Karananski 1989). At some date between 1846 and 1849, Janos P. Pendill constructed a sawmill 40 km (25 mi) west of Sault Ste. Marie where the creek that bears his name enters Nhitefish Bay. Nhen Harvey Mellon's survey team reached the site of the sawmill on September 14, 1849, he noted that an area encompassing only 101 ha (250 ac) had been disturbed by logging. Nhen Nells Burt was likewise subdividing the farthest township on Nhitefish Point on 29 June, 1849, he made note of the new lighthouse that became active earlier in that same year. An area of approximately 22 ha (55 ac) was potentially disturbed by the establishment of the lighthouse. By 1850 the non-native population of the Upper Peninsula reached approximately 6,000 persons, with the greatest concentrations probably occurring at Sault Ste. Marie and the copper ranges of the Neweenaw Peninsula. In 1855 the first canal and locks were completed in the St. Marys river at Sault Ste. Marie, and in June of that year commercial logging began on the southern shore of Lake Superior from the Nhitefish Bay to Grand Island at present day Munising (Naramanski 1989). 13 . Based upon this chronology, I have concluded that the study area was relatively undisturbed by any major logging before 1850, when the subdivision surveys of the townships were coapleted. Furthermore, none of the townships in the study area were subsequently found to be fraudulent and thus required re-survey at a later date. Therefore, I have also concluded that the surveys of the study area are at a minimum coaprehensive and accurate in scope. Baviremmeatal Characteristies Climate Bounded by Lake Superior to the north, and Lakes Michigan and Lake Huron to the south, region III has a cool lacustrine climate. The most pronounced effect of the lakes is to moderate the climate such that warming is retarded in the spring and cooling is retarded in the fall. Hence, the growing season is relatively long, although maximum summer temperatures are depressed. Average climatic data for Chippewa County is presented in Table 1. The presence of the lakes also reduces the severity of summer thunderstorms, reduces the frequency of tornadoes to rare occurrences and causes considerable lake effect snow in the winter. Landferm and soils The region encompassing the study area has, like all of Michigan, been distinctly influenced by the Nisconsinan period of the Pleistocene glaciation. It is primarily 14 Table 1. Average climatic data for Chippewa County. Growing season length (days). 135 Growing season heat sum, base 7.2 C, April- October (°C-days) 1860 Total annual precipitation (mm) 787 Annual average temperatures (°C) 5.0 July average daily max temperature (°C) 24.4 January average daily minimum temperature (°C) -13.1 Barnes and Nagner (1981), Albert et al. (1986) 15 characterized by low elevations and flat lake plain topography, with a relatively young bedrock of limestone and dolomite. The following subdistrict descriptions are summaries of Albert et al. (1986): 13.1 St. Ignace Subdistrict. The subdistrict is characterized by sand lake plains and limestone bedrock at or near the surface, with occasional areas of rolling ground moraines and large ridges. Elevations range from 175-315 m (580-1040 ft). The sand lake plain is characterized by both poorly drained depressions and by excessively drained ridges. Drainage is poor where the limestone bedrock is near the surface. 13.2 Rudyard Subdistrict. The subdistrict is characterized by flat, post-glacial clay lake plains, with some areas of ground moraine and sand lake plain. Elevations range from 175-245 m (580-800 ft). The soils of the clay lake plains are generally poorly drained. The ground moraines are well drained, and the sand lake plains are often excessively drained. 14.1 Seney Subdistrict. The subdistrict is characterized by flat, poorly drained sand lake plain. There are also occasional narrow ridges or dunes with excessively drained sand soils. Elevations range from 180-270 m (600- 880 ft). Areas of limestone bedrock and moderately sloping ground moraines, as found in the St. Ignace Subdistrict, are lacking. 14.2 Grand Marais Subdistrict. The subdistrict is 16 characterized by sand end moraine ridges, outwash plains and lake plains. Elevations range from 185-380 m (602-1240 ft), with the highest elevations in the western side of the subdistrict. The well drained and moraines are steep and irregular 30-61 m (100-200 ft) ridges, interspersed with kettle lakes and poorly drained swamps. The eastern edge of the subdistrict is the outwash Raco Plain, with excessively drained sand soils. The steep topography of the subdistrict is in sharp contrast with the flat Seney and Rudyard Subdistricts to the south and east respectively. Vegetation Following the retreat of the Nisconsinan glaciers regions III and IV were reforested by migration of species from refugia in the south. Boreal species such as white spruce (Risen clause). black Spruce (Risen mariana). alder (Alana 899-). eastern larch (Lari: larisina). jack pine (Finns hanksiana) and balsa! fir (Abies balsamea) migrated into region III approximately 10,000 years ago. Nhite pine (Pings fitngng) entered the region approximately 2,000 years later. Late successional species such as maple (Age: spp.) eastern hemlock (Tang; gangfigngig) and American beech (Eggng gxgnfiiiglia) entered the region approximately 7,000, 5,000 and 4,000 years ago respectively (Davis 1981). The tree species currently found in Chippewa County are a direct function of past exploitation and subsequent management practices. Northern hardwood forests are 17 dominated by sugar maple (Age; gagghaznp) and American beech and Yellow birch (fistula alleshanisnaia) . with mullet components of red maple (App; INDEED). balsam fir, black cherry (Emma aeration). ban-wood (Tim m: ironwood (913:1; yinqiniana) and eastern hemlock. There are extensive pine plains, with many intensively managed plantations of jack pine, red pine (Pinup resingga) and white pine. Smaller components of the pine plains are white birch (mule martian). biqtooth aspen (2mm mndidentata). troubling “pen (292nm stemming) and northern red oak (angzgng Inhra). Extensive areas of mixed swamp conifers are dominated by northern white cedar (Tania occidantalis), white and black spruce and eastern larch, with smaller proportions of white pine and white birch. one surveys The surveys of Michigan followed the 1833 instructions of the Surveyor General for the States of Ohio and Indiana, and the Territory of Michigan, and the 1850 instructions for the states of Ohio, Indiana and Michigan (Nhite 1984). In addition to recording the position, species and diameter of corner and line bearing trees (hereafter referred to simply as hearing trees), the surveyors were required to note the ”face of the country”, the character of the soil, the most prevalent timber and undergrowth species, the occurrence of windfalls and swamps, and the occurrence of burned land. The instructions also specified that the surveyors were to draw a plat map of each township, scaled at two inches per mile. This plat map was to be drawn in the field as the survey progressed, to ensure completeness and accuracy. Surveyors were required to draw "the crossing and courses of all streams of water, the intersection, situation and boundaries of all prairies, marshes, swamps, lakes, hills and all other things mentioned in (the) field notes" (Nhite 1984, pages 299 and 370). This included the boundaries of forest types and the location of disturbances noted on the survey lines. 18 19 Both the 1833 and 1850 instructions specified that four bearing trees be established at all township corners and at all section corners on range or township lines. However, only two bearing trees were required for the interior corners which subdivided each township into sections. Surveyors were required to record a minimum of one line tree per section line (Nhite 1984). Table 2 lists the tree species used by surveyors as hearing trees in the study area. The surveyors did not differentiate between some species, such as white and black spruce, bigtooth and trembling aspen and species of willow. Consequently, I will also refer to them as simply spruce, aspen and willow. Some species were referred to by names that are not commonly recognized today, and are listed by other surveyor names in the second column of Table 2. It is probable that surveyor references to black oak were actually pertaining to red oak, and I have recorded them as such. The primary concern in the use of GLO survey records in forest reconstruction is the degree of bias present in the sample. There are two biases of concern, the surveyor's choice of tree species and preference for specific diameter classes for use as bearing trees. Biases are evident by over-representation of tree species and diameter classes, and when bearing trees are consistently reported in opposite quadrants with only two trees established per corner. In general, biases are not present in even-aged stands, but there are usually biases toward medium sized trees in all- 20 Table 2. Species used by surveyors for bearing trees. CM“ MOI. Other Surveyor Name Scientific Name American elm Balsam poplar Basswood Beech Bigtooth aspen Black ash Black spruce Bur oak Cottonwood Balsam fir Hemlock Ironwood Jack pine Mountain ash Northern red oak Elm Balm-of-Gilead Lynn Aspen Spruce Swamp oak Fir Spruce pine Norway pine Black oak Northern white cedar Red maple Red pine Speckled alder Sugar maple Tamarack Trembling aspen Nhite birch Nhite pine Nhite spruce Nillow Yellow'birch Maple Yellow pine Alder Sugar Aspen Pine Spruce Birch Yellow birch 21 aged stands. Smaller diameter trees were generally biased against because space was required to blaze and carve the required township, range and section data. Bourdo (1956) noted that the primary concern is not whether bias is present, but whether it will significantly impact quantitative analysis. The presence of bias can be identified by quadrant analysis of bearing trees and by analysis of mean distances from corner posts to their respective bearing trees (Bourdo 1956). Quadrant analysis will reveal the presence of bias, but cannot differentiate between diameter and species bias. Quadrant analysis of bearing trees is based upon the principle that the tree nearest the corner may occur with equal probability in any of four quadrants. If surveyors expressed no bias toward a particular species or diameter in choosing a bearing tree, then the frequency with which each quadrant was chosen should be nearly equal. Analysis of mean bearing tree distances can be used to detect both diameter and species bias. Detection of diameter bias by analysis of mean bearing tree distances involves computing the mean distances from corner posts to their respective bearing trees for each two inch diameter class within a forest type. For a given type, bearing trees of every diameter class should be located a similar mean distance from corner posts. If bias toward a particular diameter class occurred, then that diameter class will have a higher or lower mean distance than other diameter classes and wide 22 variation in these mean distances would be expected. Detection of species bias by analysis of mean bearing tree distances is similarly accomplished by computing the mean distances from corner posts to their respective bearing trees for each tree species occurring in a forest type. -There are other potential sources of error that may impact quantitative analysis of survey data. The diameter of bearing trees recorded by surveyors may be considered suspect. Surveyors were only required by the instructions to estimate, rather than measure, bearing tree and line tree diameters (Nhite 1984). The experience of the surveyor thus determined the relative accuracy of reported diameters. A “good tape measure” was listed among the equipment required for surveyors in the field (Nhite 1984), but it is unknown whether the tape was used by surveyors to periodically check their diameter estimates. One must also accept the reported diameter of bearing trees as a close approximation of diameter at breast height (1.4 m or 4.5 ft). The surveyors probably estimated diameter at the height at which they blazed and recorded township, range and section corner data, which depending upon the height of the surveyor, would coincidentally be close to 1.4 m. The reported distances from corner posts to their respective bearing trees can also be a potential source of error. The chainage from the corner being established to the more distant bearing trees was documented by Bourdo (1956) to sometimes be paced or guessed, even though the instructions specified that they 23 were to be measured. Bourdo documented this by actual measurement of original bearing trees in western upper Michigan. He found that where pacing or guessing was evident, it occurred primarily for hearing trees greater than 6.0 m (19.8 ft) from corner_posts. Thus, the accuracy of bearing tree distances was very much dependent upon the integrity of the surveyor. In summary, it is known that: 1) the quality of survey work varied depending upon surveyor integrity, 2) the surveys do not constitute a truly random sample, 3) some degree of surveyor bias may be present in the selection of bearing trees, and 4) one must acknowledge some inaccuracies of reported bearing tree diameters and distances. Despite these limitations, the historical data contained in the GLO surveys can still reveal a wealth of information concerning the composition, structure and function of pre-European settlement forests. Previous investigations have concluded that the GLO survey records can be used to reconstruct pre- European settlement forests (Bourdo 1956, Nilburn 1958, Curtis 1959). Data Collection . GLO survey notes are held at the State Archives of Michigan, in Lansing. Data was collected from microfiche copies of the transcribed GLO survey records of Chippewa County, and entered by township into FOXPRO database files using a General Land Office Vegetation Entry (GLOVE) program 24 developed by the Center for Remote Sensing at Michigan State University. A dictionary of codes used in each data field of the GLOVE program is presented in Appendix A. The following information from the GLO survey records was entered into the database files, and served to reference subsequent attribute point data: 1) Township number. 2) Section number. 3) The reference corner of the section line along which the surveyor was traversing. 4) The bearing in which the surveyor was traversing. 5) The distance in chains (to the hundredth decimal place) from the reference corner that the surveyor traversed before setting a line or corner post. 6) The bearing and distance in links from a line or corner post to the post's bearing trees. The following attribute data from the GLO survey records were then entered into the database files: 1) The bearing tree species. 2) The bearing tree diameter (in inches). 3) The year in which the survey was conducted. 4) The presence and orientation of disturbances noted by the surveyors. 5) The topography of the section line, as noted by the surveyor. 6) Notes of the species of trees as observed by the surveyor in order of predominance. Thirty-five database files were created using the GLOVE program, one for each township in the study area. Data Analysis Use of the c-MAP Geographic Information System C-MAP 2.1.1 is a vector-based geographic information system (GIS) developed for use on personal 25 computers by the Center for Reaote Sensing at Michigan State University. c-MAP is specifically designed for use with federal databases in ARC/info and the GIS data foraats of the State of Michigan, which include the Michigan Resource Infornation Systee (MIRIS). The C-MAP programs offer several capabilities which include creating thematically shaded maps, running statistical and data grouping operations, performing database queries on object attributes and neasuring and spatially analyzing feature distributions. Point data referenced in the POXPRO database files required conversion into state plane coordinates before it could be used by C-MAP. Data conversion into state plane coordinates was perforeed by a conversion program called LOCTRIB, also developed by the Center for Remote Sensing. Some error was induced in the conversion of point data to state plane coordinates. The error occurred because point data was placed by true north, south, east and west vectors originating froe the referenced section corners. Such referencing did not take the convergence of lines of longitude into account (Figure 2). However, for the purposes of this study where data was sampled on a transect grid with spacing of approxinately 1.61 km (1 ei) between transects, the anount of error induced by non-convergence of a vector over a maxieum distance of 1.61 km can be considered negligible. A future version of LOCTREE will convert point data in the Universal Transverse Mercator systen, thus eliainating convergence error. 26 I :99 UP : ' I I; JP JP a! asp : Jp 4;: 'UP 1" -..-4.. --;'--.-'-. ..... L-’—-‘f-f.£ ......... !-_f__f--f-4fi.-- 1 1..»- ; I)” I , .‘P ' I 2,»: if u,» L» L? \ I 3’ i I ‘ 28 :. I I I I 5. :.w in b 5" 33 F F” i” is i» .. .. mmmw a... .----'+ .............. - ........ t. Figure 2. Example C-MAP plot of bearing trees and digitized forest type boundaries. 27 Once converted into state plane coordinates, tree species and topography point data were displayed by C-MAP. The topography data indicated changes in forest type as recorded by the surveyors, and tree species were used as indicators of expected forest types (See Appendix A for tree species and topography codes used). A forest type nap of the study area (Figure 2) was created with the on-screen digitizing progran in C-MAP, using the displayed tree species and topography point data and printed copies of the surveyor plat naps (on which the surveyors drew general forest type boundaries) to detersine the placement of my forest type boundaries. Delineation of forest types was made to the greatest detail possible, based upon the criteria outlined above. In some cases the presence of a single tree indicator species (such as a single red pine in a mixed conifer swanp) provided sufficient evidence to define a forest type boundary. The resultant type map was then cleaned to correct for topological inconsistencies and the topological files were built to represent forest type polygons (defined contiguous areas).' Point data were labeled by forest type using the overlay function of C-MAP, and then used for analysis of composition and structure and disturbance according to forest type. Composition and Structure Analysis of composition and structure was conducted on a sub-sanple of the study area. The sub-sample 28 consists of eight adjacent townships in the interior of the study area: Townships 45 north Ranges 3-6 west, and Townships 46 north Ranges 3-6 west. The sub-sample was chosen for two reasons., The first is that human influences in the study area were concentrated along the shorelines. This is to be expected since travel by water was the nest expedient mode of transportation during the era. Thus, the forest in the interior can be expected to be nuch less perturbed by human activity, and more representative of the pre-luropean forests of the region. The second reason for the location of the sub-sample is that it contained substantial areas of each forest type found in the overall study area, and thus represented a sample of each. Point data for the sub-sample were exported to Quattro Pro for composition and structure analysis. Methods for estimating density, relative density, basal area and relative dominance from GLO survey data are reviewed in Bourdo (1954) and Cottam and Curtis (1956). Two :methods are described by Cottam and Curtis: 1) point to plant methods and 2) plant to plant methods. The forner :methods rely upon neasured distances from a point to the nearest tree(s). The later methods rely upon neasured distances between two trees that are nearest to each other, and not necessarily those trees nearest to a point. The 1833 and 1850 survey instructions specify that only those trees nearest to the corner being established shall be chosen as bearing trees. Trees selected as bearing trees 29 ‘may or may not be those trees that are closest to each other. Thus, the plant to plant methods are not suitable for use with GLO survey data. Point to plant methods are better suited for use with CLO survey data, but have been used with varying success. The closest individual nethod is the sinplest to use with survey data. The nethod is based upon the mean area (M) occupied by a tree. The square root of M is a direct indication of the spacing between trees. Thus with this method, the distances measured by surveyors from corner posts to bearing trees can be used to estimate tree density. For the closest individual method the distance from the corner post to the single closest tree is required. The mean of all measured distances has been found to equal 50% of the square root of the mean area M (Cottan et a1. 1953, Morisita 1953). This nethod, therefore, requires multiplication of the average distance by a correction factor of 2.0, before squaring to obtain M. The mean area M can then be divided into the unit area to yield the number of trees per unit area. In order to obtain accurate estimates of tree density using the closest individual method, the tree being sampled lug; be the nearest tree to the corner being established. .Although surveyors were instructed to utilize the nearest trees as bearing trees when establishing a corner, there is no assurance that this was always done. Subjective biases can severely distort the results of point to plant nethods. 30 As previously discussed, bias may be expressed by surveyor preference toward particular tree species and toward certain diameter classes of trees. The closest individual method is particularly susceptible to these biases because only one tree per point is used in calculations. To obtain estimates within 10* of true densities, the closest individual method requires a minimum of 150 sample corners. Furthermore, when trees are clumped together with open spaces between groups, the closest individual method will not yield accurate estimates of stand density (Cottam and Curtis 1956). Spurr (1952) devised a fixed diameter method of estimating basal area, which was reviewed by Bourdo (1954). This method is well suited for use when diameter class bias is present, because it can utiiize this bias to an advantage. Spurr determined that a single, randomly chosen tree from a series of plots could be used to determine basal area, mocording to the equation: (4) (302.5) D 2 (l) Basal area/acre - ----- ;-; ----- where D equals half the diameter of the tree (in inches), R equals the plot radius (in feet, to the center of the tree) and 302.5 is a constant. Because survey trees are reported in quadrants, a multiplication factor of 4 is required so that the estimate represents all quadrants. Bourdo (1954) found that Spurr's formula could be used, with the most frequently selected tree diameter class as D 31 and the mean distance from the corner post for that class as R, to yield good approximations of basal area per acre. Bourdo reasoned that the diameter class most frequently chosen by surveyors as bearing trees was the most representative diameter of the forest type. The process is thus highly dependent upon determining the diameter class that is indeed most representative of the forest type. There are four methods of doing this. The most obvious is a simple count of the number of trees to determine the most frequently represented diameter class. Two other methods are determining the average diameter and the median diameter of all trees of all diameter classes. The fourth method is to determine the mean diameter based upon the mean basal area of all trees in all diameter classes. By using all four of these methods the most representative diameter class can be determined with confidence. As a further precaution against error, Spurr's formula may be applied to the three most representative diameter classes. When this is done the most representative diameter class must be used as the middle value, and the contribution of all diameter classes to basal area must be weighted by their representation in the type. Bourdo (1954) compared basal area estimates obtained using Spurr’s formula to actual basal area :measurements in residual old growth northern hardwoods, and concluded that good estimates of basal area per acre can be determined by using the principles defined by Spurr. Thus, Iourdo confirmed that Spurr's formula is based upon sound 32 scientific theory. Because.all GLO surveys are probably affected to some degree by diameter class bias, I have used Spurr's formula for estimation of basal area per acre in each forest type. Mith basal area known, density was estimated by dividing the basal area per unit area by the mean basal area per tree. Relative density was calculated by dividing the number of individuals of any species by the total number of individuals of all species, and then multiplying by 100. Relative dominance was calculated by dividing the total basal area of a species by the total basal area of all species, and again multiplying by 100. Disturbance Regimes The main limitation in the use of GLO survey records for the estimation of disturbance frequencies and return intervals is that direct evidence of blowdowns and burned over land becomes blurred once forest canopy closure occurs after approximately fifteen years (Lorimer 1980a, Canham and Loucks 1984). Thus, any reference by surveyors to blowdowns and burned over land would indicate that the disturbance has probably occurred within the previous fifteen years. Fifteen years is a fairly narrow period on which to base estimates of disturbance frequency. However, given the range of approximately fifteen years over which the survey and examinations (checks for accuracy and completeness of survey work) of Chippewa County were_ 33 conducted, direct evidence of blowdowns and fires may potentially be distinguished in the survey records over a thirty year period. Therefore, meaningful estimates of disturbance frequencies and return intervals can potentially be derived from CLO survey records. ' The CLO surveys represent a systematic sample of Chippewa County. Surveyors were required by the 1833 and 1350 instructions to record the distance along a transect at which they encountered and departed any disturbance, and to plot its location on the township plat map. I cross-checked the plotted location of each disturbance on the plat map with the associated survey notes for each township in the study area, and found the location of each disturbance to be accurate. The most expedient method of analysing disturbances would be to simply digitize the area of each individual disturbance event from the plat maps. There is a problem with this approach. The location along a section line and the bearing in which an individual disturbance event was running was the only information available to surveyors when drawing the location of disturbances upon the plat maps. Because the survey section lines were 1.61 km (1 mi) apart, the area of each disturbance could confidently be estimated and drawn only for disturbances large enough to intersect a survey line in at least two different locations. Other than deviating from the survey lines, surveyors had no way of knowing the course of disturbances within the interior of sections. Surveyors were not able to draw the 34 boundaries of small disturbances when they covered less than one section in area, and intersected survey lines in only one location. Additionally, any small disturbances that were less than 2.27 km (1.41 mi) in length could potentially remain undetected by surveyors. I attempted to retroactively estimate the area of such disturbances, but concluded that areas cannot be estimated with sufficient accuracy or reliability to be used with any degree of confidence in analysis of disturbance. For these reasons, I have not used area as a basis for analysis of disturbance regimes. I have alternatively conducted analysis of disturbance by measurement of the total length of survey lines impacted by different disturbance regimes, within the different forest types. Area line transect sampling theory was used to estimate the total number of disturbances less than 1.41 km in length from the actual number observed by the surveyors (Canham and Lcucks 1984, Canham 1978, Devries 1974, Warren and Olsen 1964). If a transect of length L is placed through a population of randomly distributed disturbances, then the estimated number of disturbances (X) in each forest type can be calculated by summing sections of size s: n 1 (2) x - 0.7854 2 ----- i-l y where n - the number of times disturbances were noted to intercept a transect. y - the length of the i'th intercepted disturbance (measured in miles). 35 0.7854 - a constant simplifying the area relationship s/2L where the grid system of survey lines provides an average of Lpz miles (3.22 km) of survey lines for the area s-l mi2 (2.59 km?) in each section. Distances are in English units for convenience of calculation only. The only information required for use of equation 2 is the length of each individual disturbance encountered along a transect. Analysis of disturbance regimes was conducted on the entire study area. Disturbance data were segregated by forest type prior to disturbance analysis. This step was necessary because the type, frequency and scale of disturbance is dependent upon the pre-existing stand composition and structure. Disturbance lengths were determined by screen digitizing disturbance point data, 'merging the resultant disturbance arcs with the forest type arcs and than reading the lengths of disturbance arcs within each forest type after the resultant file was cleaned. Equation 2 allows subdivision of disturbances into size classes. Disturbances were classified as those greater in length than 50 m, 100 m, 200 m, etc., up to a maximum of 12,000 m. The estimated total number of disturbances in each size class, with mean length less than 1600 m (one mile), were calculated using equation 2. The numbers of disturbances with lengths greater than 1600 m were determined from the numbers actually recorded in the surveyor notes. The estimated total length of section line in each size class that was affected by disturbance was 31. di 36 calculated by multiplying the number of estimated disturbances by the mean length of the disturbances actually observed in each size class, and then adding the result to the total length of those disturbances actually observed. Analysis of disturbance regimes involves the calculation of disturbance frequencies, disturbance return intervals and areas of annual disturbance. Disturbance frequency is defined as the number of disturbances per unit time in a designated area. The actual number of disturbances observed by surveyors in each forest type, were added to the additional estimated number of non-observed disturbances to yield the total number of disturbances that I used in the frequency calculations. The disturbance frequency was estimated by dividing this total number of disturbances by the number of years over which these disturbances were observed. Disturbance return intervals are defined as the number of years between two successive disturbance events in a designated area (Romme 1980), or the number of years that it would take for an entire landscape to be disturbed. It must be emphasized that the concept of return intervals can be misleading. One must recognize the heterogeneous nature of landscapes and realize that some portions of a landscape are more prone to disturbance than others. In other words, different parts of a landscape are subject to different types, combinations and rates of disturbance. Disturbance return intervals were estimated by calculating the percent distance of surveyors lines in each 37 forest type that were affected by a particular disturbance regime over a fifteen and thirty year survey period. The annual percent affected was then divided into 100 to yield an estimate of the disturbance interval (Whitney 1986). The fifteen year interval is probably conservative with respect to disturbance frequencies and return intervals, and the thirty year interval is probably high. Both recording intervals will be reported in calculations of disturbance frequencies and return intervals. The area of annual disturbance was estimated for each forest type and disturbance regime by dividing the area of each forest type by the return interval for each disturbance regime. RIBULTB IND DIOCUBBION Surveyor Sims Quadrant analysis of bearing trees at 343 corners in four townships (Townships 45 north, Ranges 5-6 west and Townships 46 north, Ranges 5-6 west) shows that 83, 89, 96 and 75 trees were located in the northeast, northwest, southeast and southwest quadrants of those corners respectively. If bearing trees were chosen randomly, with no bias toward diameter or species, than one would expect to find 85.75 trees in each quadrant. A chi square analysis shows that a hypothesis of random departures of the observed bearing tree counts in each quadrant from the expected value of 85.75 trees cannot be rejected “222.7843, 3 df, P > 0.3). Quadrant analysis of bearing trees thus reveals no indication of bias toward specific diameter classes or tree species. In the northern hardwood, mixed conifer/deciduous upland and mixed conifer swamp forest types the low K variances of mean bearing tree distances indicates that there was little bias toward either specific diameters or tree species (Tables 3-5). Initial analysis of the diameter class distribution of the mixed pine forest type showed that the 10, 15 and 20 cm 38 39 Table 3. Analysis of mean distances of bearing trees for diameter and species bias in the northern hardwood forest type. Diameter Mean Mean Class‘ Tree Distance Tree Distance (cm) Count (m) Species2 Count (m) 15 13 7.7 BE 19 8.2 20 15 8.6 H 58 9.2 25 23 8.6 RH 11 9.7 30 24 8.7 SM 45 9.0 35 23 8.2 YB 25 7.5 40 21 10.0 45 13 9.5 50 11 7.9 Sum 143 - 158 Mean 8.7 8.7 Std DOV 0.8 0.9 Variance 0.6 0.8 ' Diameter classes range from 10 to 90 cm, but outlying diameter classes were excluded due to under-representation. 2 DE - Beech, M - hemlock, RM - red maple, SM - sugar maple, Y8 - yellow birch. 40 Table 4. Analysis of mean distances of bearing trees for diemeter and species bias in the mixed conifer/deciduous upland forest type. 51ameter Mean Mean Class‘ Tree Distance Tree Distance (cm) Count (m) Species2 Count (m) 10 11 8.8 BF 11 10.2 15 24 8.3 H 32 8.5 20 24 7.0 'RH 8 8.6 25 18 7.2 SH 6 8.8 30 16 10.0 SP 36 7.6 35 13 9.8 WP 24 8.5 I YB 12 7 . 4 Sun 106 129 Mean S.5 S.5 Std DOV 1.3 0.9 Variance 1.6 0.8 ‘ Diameter classes range up to 100 cm, but outlying diameter classes were excluded due to under-representation. 2 BF - balsam fir, B - hemlock, RM - red maple, SM - sugar maple, SP - spruce, WP - white pine, YB - yellow birch. K 41 Table 5. Analysis of-mean distances of bearing trees for diameter and species bias in the mixed conifer swamp forest type. Biameter Mean Mean Class‘ Tree Distance Tree Distance (cm) Count (m) Species2 Count (m) 5 6 8.8 C 77 7.6 10 46 9.0 K 8 10.3 15 82 8.2 JP 15 8.5 20 46 8.3 SP 83 7.5 25 39 8.0 T 67 9.8 30 28 8.5 WP 8 8.2 35 9 7.8 SUI 256 264 NO.“ 8.4 8.7 Std DOV 0.4 1.2 Variance 0.2 1.3 ‘ Diameter classes range up to 90 cm, but outlying diameter classes were excluded due to under-representation. 3 C - cedar, H - hemlock, JP - jack pine, SP - spruce, T = tamarack, WP - white pine. Di t: 3F 42 diameter classes represented the three highest densities of all the diameter classes. Table 6 shows that the 10, 15 and 20 cm diameter class bearing tree distances can also be considered a sub-sample of the entire diameter class distribution. The 10 through 20 cm diameter classes are grouped around a mean sub-sample distance of 9.2 meters, with a standard deviation and variance of 0.5 and 0.3 respectively. The bearing tree distances of diameter classes 25 through 50 are closely grouped around a mean of 13.0 meters, with a standard deviation of 0.6 and a variance of 0.3. Table 6 also shows the mean bearing tree distance of jack pine to be 8.9 meters. Given the seral nature of jack pine, its dependence upon periodic stand replacing fires for regeneration and the relatively small diameter distribution characteristic of nearly pure jack pine forests, I have separated the mixed pine forest type into a subtype of nearly pure jack pine and a subtype of mixed white and red pine (of which I retain the name of "mixed pine”). Subsequently, the low variances of mean bearing tree distances indicate that there is no bias toward either specific diameter classes or species in the mixed pine or jack pine forest types. The mixed conifer/deciduous lowland forest type shows some moderate bias toward the 15 and 20 cm diameter classes, but there is no indication of any significant bias toward a specific tree species (Table 7). The diameter class bias may also result from a difference in stand structure. If 43 Table 6. Analysis of mean distances of bearing trees for diameter and species bias in the mixed pine and jack pine forest types. Biameter Mean Mean Class1 Tree Distance Tree Distance (cm) Count (m) Speciesz Count (m) 10 27 9.0 JP 71 8.9 15 46 8.8 20 28 9.8 80. 101 71 NO.“ 9.2 8.9 Std DOV 0.5 0 Var 0.3 0 25 18 12.7 RP 63 12.5 30 30 13.2 WP 80 11.7 35 13 13.0 40 9 13.9 45 17 12.9 50 9 12.2 S“. 96 143 Mean 13.0 12.1 Std DOV 0.6 0.6 Variance 0.3 0.3 ' Diameter classes range from 5 to 75 on, but outlying diameter classes were excluded due to under-representation. 2 JP - jack pine, RP - red pine, WP a white pine. ’\ 3 ~Slsv 44 Table 7. Analysis of mean distances of bearing trees for diameter and species bias in the mixed conifer/deciduous lowland forest type. Diameter Mean 3 Mean Class‘ Tree Distance Tree Distance (cm) Count (m) Species2 Count (m) 10 8 7.0 8? 6 10.1 15 8 11.7 H 5 9.6 20 10 11.7 SP 10 8.3 25 6 7.9 T 11 8.2 30 3 6.7 WP 4 7.1 35 4 8.7 SUI 39 40 “CID 9.0 8.7 Std DOV' 2.2 1.2 variance 5.0 1.4 ‘ Diameter classes range from 2 to 40 inches, but outlying diameter classes were excluded due to under-representation. 2 SF - balsam fir, H - hemlock, SP - spruce, T - tamarack, WP - white pine. an: 10‘ 31:1 die to: fr. an Th! Iii th. 8&1 tn 95: rec ex 1' I‘Qt 45 the 15 and 20 cm diameter classes were statistically the two most numerous classes in the type, then one would expect these classes to be used as bearing trees with a greater frequency than other classes. In summary, the only evidence of bias was toward the 15 and 20 cm diameter classes in the mixed conifer/deciduous lowland forest type. This bias is of a moderate nature. Therefore, I have concluded that the CLO surveys for the study area are mostly free of bias toward any specific diameter classes or tree species. Forest Composition and Structure Six distinct forest types (Figure 3) were discernable from the tree species and topography data grid of the study area: 1) Northern Hardwoods 2) Mixed Pine 3) Jack Pine 4) Mixed Conifer/Deciduous Upland 5) Mixed Conifer/Deciduous Lowland 6) Mixed Conifer Swamp The forest types are accurate to the scale of 1.61 km (1 mile). This seemingly coarse scale is unavoidable due to the nature of survey records, where the survey represents a sample grid with a distance interval of 1.61 km between transects. While the broad patterns of forest types are essentially accurate to within less than 1.61 km, one must recognize that a transition zone often (but not always) exists between two adjacent forest types. Thus, these rather coarse-scaled forest type boundaries are actually 46 .csmncOnz .>u::oo esoQQflnu cucumc: no moc>u umcnou ucoscfluuom csoconcmnonm .m ensues .... a... E .. .... ... 2...: \\\\\\\\\\+\W\\\\\\\\\\\\\\\\\ \\\\\\\\\\\Wm#flh a...r..\vn.\\.\\\\\\ \\ \x\ l . . omw . Sin 38m .././ /\../ $\\\&.\\ .. awe. .....w/JZ. 4..., «he . \\\\\w . cunoz rm. 5 ,u can; home—.80 fills-a Mm! Goa—8°... 1u°8fi\:00 ‘SHfin- ‘¢°8\=8 III! S: u.— ‘68—. 8a- ufi ‘8! n! 1008194: Elsie: ”a. E 0.3. NJ me: 0:50.... nxmw ‘ \? ._Mw.hum..______%z t) ed Co; tht det P0: Pat ‘la ““81 47 represented by transition zones. C-MAP 2.1.1 does not have the capability to define and separate out these transition zones and I deemed it too laborious a task to digitize them by hand. Consequently, transition zones exist, but are not visible in the forest type boundaries. Another consequence of the coarse scale of the forest types is that a complete picture of the heterogeneous nature of the landscape is not obtained. Some small pockets of different forest types are undoubtedly imbedded within the fabric of the larger forest types. These small pockets can be caused by changes in edaphic and topographic conditions and by disturbance. While some of these pockets are apparent because good fortune placed them on a survey transect, others are transparent because they are located in the interior of a 2.59 km? section. Thus, inconsistencies such as red pine or spruce within a northern hardwood forest type were found in the sample data, but were attributed to the coarse scale of the sample. 1 The heterogeneous nature of the landscape also confounds any conclusions regarding stand structure that may be drawn from diameter class distributions. The scale of the CLO surveys (with transect spacing of 1.61 km) cannot detect the heterogeneous nature of inter-stand structure. For example, a forest type comprised of small, even-aged patches or stands of different ages (with origins from small-scale disturbances) could be misinterpreted as being uneven-aged when diameter data is taken across all patches 1: of sh be anc TP}. Md f0): 5% Bug.- Dru 48 and then viewed from a landscape perspective. The only forest type where heterogeneity may not confound diameter class distribution is the jack pine type, where large scale fire disturbance probably resulted in a more homogeneous landscape. Maps detailing the forest cover types in each township of the study area are presented in Appendix 8. Density, basal area and mean DBH estimates for each forest type are shown in Table 8. Discussion of each forest type follows below. The Northern Rardwood Forest Type The northern hardwood type had the highest basal area and density values of all the forest types (35 mfi/ha and 348 TPH). Table 9 and Figures 4 and 5 show the relative density and dominance of the tree species in the northern hardwood forest type. The virgin northern hardwood forests of the sub-sample study area were dominated by eastern hemlock and sugar maple in both density and basal area. Hemlock was predominant in terms of basal area (9.8 afi/ha), while sugar maple was dominant in density (99 TPH). Yellow birch and beach were also relatively important in the forest type, but beech (2.1 mz/ha) was not nearly as dominant as yellow birch (6.0 mz/ha) . Other indicator species of a northern hardwood forest such as balsam fir, basswood and ironwood were also present as minor components in the sub-sample area. An understory of hazel was sometimes noted by surveyors in the 49 Table 8. Basal area, density and mean DBH estimates for the pro-European settlement forest types of western Chippewa County, Michigan. Estimated 'Estimated Mean Forest Type Density (TPM) BA (mz/ha) DBM (cm) Northern Hardwood 348 35 33 (141)‘ (154) (13) Mixed Pine 81 S 31 (33) (36) (12) Jack Pine 326 8 18 (132) (33) (7) Mixed Conifer/ 287 21 , 2S Decid. Upland (115) (92) (11) Mixed Conifer/ 269 15 23 Decid. Lowland (109) (66) (9) Mixed Conifer 314 14 20 Swamp (127) (60) (8) ' Numbers in parentheses are density in trees/acre, basal area in ftz/acre and DBM in inches. Sp- BE BF 894‘ IN SH SP “8 50 Table 9. Relative density and dominance of tree species in the northern hardwood forest type. Total Trees Basal BA per Tot Relative Per Area Relative Hectare Species2 I Density Hectare (IF) Dominance (IE/ha) A 6 0.7 2.5 0.2 0.2 0.1 (1.0)‘ (2.0) (0.3) 88 86 10.4 36.1 5.0 5.9 2.1 -(14.6) (53.6) (9.1) BF 55 6.6 23.2 2.0 2.3 0.8 (9.4) (21.2) (3.6) 80 2 0.2 0.7 0.6 0.7 . (0.3) (5.9) (1.0) B“ 4 0.5 1.7 0.5 0.6 0.2 (0.7) (5.4) (0.9) H 170 20.5 71.7 23.5 27.8 9.8 (29.0) (252.3) (42.9) IV 3 0.4 . 0.1 0.1 0.02 (0.5) (0.7) (0.1) R! 68 8.2 28.7 3.3 3.9 1.4 (11.6) (35.5) (6.0) RP 2 0.2 . 0.3 0.4 0.1 _ (0.3) (3.5) (0.6) SM 235 28.4 98.8 19.2 22.8 8.0 (40.0) (206.1) (35.0) SP 30 3.6 12.6 2.0 2.4 0.8 (5.1) (21.3) (3.6) 88 9 1.1 3.7 0.7 0.8 0.3 (1.5) (7.1) (1.2) HP 42 5.1 17.8 12.7 15.1 5.3 (7.2) (136.7) (23.2) YB 116 14.0 48.9 14.4 17.1 6.0 (19.0) (154.7) (26.3) Sun 828 100 348 84 100 35 (141) (906) (154) 4 ‘ Numbers in parentheses are trees/acre, ft2 and ftz/acre respectively. 3 A - aspen, BE - beech, BF - balsam fir, BO - black/red oak, BW - basswood, H - hemlock, IW - ironwood, RM - red maple, RP - red pine, SM - sugar maple, SP - spruce, WB - white birch, WP u white pine, YB - yellow birch. 51 70 TREES PER HECT ARE ¥ 10‘ mm" .... ‘ .... ............ .. .m...“ A BE BF BOBW H IW RM RP SM SP WBWP YB TREE SPECIES (See Table 9) Figure 4. Density of tree species in the northern hardwood forest type. BASALAREAPERHECTARE to o a on o u (b A BE BF BOBW H IW RM RP SM SPWBWPYB TREE SPECIES (See Table 9) Figure 5. Basal area of tree species in the northern hardwood forest type. 52 northern hardwood forest type. There was also a white pine component in the northern hardwood forest type. Although low in density (18 TPH), white pine was a dominant in terms of basal area (5.3 nfi/ha). The two northern red/black oak trees in the sub- sample study area were 51 and 66 cm in diameter. Thus, oak was a minor component in this forest type. The early- successional species aspen and white birch were also present within the forest type. Diameters for aspen ranged from 15 to 31 cm. Diameters for white birch ranged from 13 to 56 cm. A diameter class distribution of the northern hardwood forest type is presented in Figure 6. Appendix C contains “W ... ... .. TREES PER HECTARE 10<----~ I» ~- ~- wwmamawamnwwmnwwfii f 1'36 DIAMETER CLASSES (cm) Figure 6. Diameter distributions of tree species in the northern hardwood forest type. 53 diameter class distribution graphs for all of the significant species in the northern hardwood forest type. It is evident from the diameter class distributions that some very large trees occurred within the forest type. The mean and median diameters of all sample trees was in the 30 cm class. The mean diameter based upon the mean basal area per tree was 36 cm. Diameter classes less than 15 cm were under-represented in the sample because very small trees were unsuitable for blazing and recording of township, range and section corner data. I believe that it is highly probable that advance regeneration of tolerant species was occurring in the understory of the virgin northern hardwood forests of the sub-sample area. Due to landscape heterogeneity I cannot definitively conclude from the diameter class distributions that the northern hardwood forests of the study area were uneven-aged. Disturbance, ranging in size from individual tree fall gaps to larger windfall events, probably created a mosaic of forest patches or stands. The presence of seral aspen, oak, white birch and red pine, and mid-tolerant white pine provide evidence of the heterogeneous nature of the landscape. The composition of stands probably ranged from even-aged seral species to uneven-aged or possibly.even-aged late successional species. Therefore, the northern hardwood forest type cannot be broadly classified as uneven-aged, although large patches of it undoubtedly were. N ha Ii. 01“ Ii} 54 The Mixed Fine Forest Type The mixed pine type had a very low density (81 TPM), and a basal area similar to the jack pine type (8 afi/ha). Table 10 and Figures 7 and 8 show the relative density and dominance of tree species in the mixed pine forest type. For convenience, both the mixed pine and jack pine forest types are shown in Table 10. The mixed pine type was dominated in both density and basal area by white and red pine. White pine exhibited a density of 36 TPH and a basal area of 5.2 Ifi/ha. Red pine had a density of 27 TPH and a basal area of 2.4 mz/ha. Minor components of the mixed pine forest included aspen, balsam fir, northern red/black oak, eastern hemlock, red maple, spruce, white birch and yellow birch. Both large and small areas of disturbance that were thickly stocked with seral aspen and white birch saplings were noted by the surveyors. The oak component was also seral, with the two sampled individuals being 8 and 10 cm in diameter. The mid to late-successional species balsam fir, hemlock, red maple, spruce and yellow birch were found either in the understory, or more likely in small pockets on more fertile sites. Some of these better sites where white pine dominated may have been converting to the northern hardwood forest type. Figure 9 shows the diameter class distribution of the mixed pine forest type. Appendix D contains the diameter class distributions of significant species found in the mixed pine forest type. Once again, any conclusions 2 Esp Table 10 . the mixed pine and jack pine forest types. Relative density and dominance of tree species in Eotal Trees Basal BAper Tot Relative Per Area Relative Hectare Species2 # Density Mectare (m2) Dominance (ma/ha) A 25 3.2 2.7 0.6 0.8 0.1 (1.1)1 (6.4) (0.3) 81" 12 1.6 1.2 0.3 0.4 0.02 (0.5) (3.1) (0.1) 80 2 0.3 0.2 0.01 0.02 0.002 (0.1) (0.1) (0.01) C 3 0.4 0.2 0.2 0.2 0.02 (0.1) (1.6) (0.1) H 20 2.6 2.2 1.7 2.2 0.2 (0.9) (18.0) (0.3) RM 1 0.1 0.1 0.1 0.1 0.01 (<0.1) (0.8) (0.03) RP 255 32.9 26.7 22.8 29.8 2.4 (10.0) (245.0) (10.7) SP 42 5.4 4.4 0.9 1.1 0.1 (1.0) (9.4) (0.4) T 7 0.9 0.7 0.2 0.2 0.02 (0.3) (1.8) (0.1) WB 56 7.2 5.9 0.9 1.1 0.1 (2.4) (9.3) (0.4) WP 339 43.7 35.6 48.4 63.2 5.2 (14.4) (520.0) (22.0) YB 14 1.8 1.5 0.7 0.9 0.1 (0.6) (7.1) (0.3) Sun 776 100 81 77 100 8 (33) (823) (36) JP 378 100 326 8.85 100 8 (132) ‘ (95.11) 1 Numbers in parentheses are trees/acre, ft2 and ftz/acre respectively . 2 A - aspen, BF - balsam fir, BO - black/red oak, C - cedar, H - hemlock, JP - jack pine, RM - red maple, RP - red pine, SP - spruce, T - tamarack, WB - white birch, WP - white Pine, YB - yellow birch. 56 TREES PER HECT ARE A BFBO C H RMRPSP T WBWPYB TREE SPECIES (Soc Table 10) Figure 7. Density of tree species in the mixed pine forest type. BASAL AREA PER HECTARE 1 A aF'ao'c H RM RP S? T wawp YB TREE SPECIES (See Table 10) Figure 8. Basal area of tree species in the mixed pine forest type. 57 16 14 1‘ TREES PER HECT ARE c) : 6.....me ... .. 4f ~ " m 2 ... .. .. .. .. .. 51015202530354045505560657075 80 DIAMETER CLASSES (cm) Figure 9. Diameter distributions of tree species in the mixed pine forest type. regarding age structure are confounded by the heterogeneous nature of the landscape. The long right hand tail of the diameter class distribution (Figure 9) may indicate residual survivors of disturbance. The very low density of 81 TPR suggests that the virgin mixed pine forest of the study area had a very open, park-like structure. Numerous groves of white and red pine were noted by the surveyors. The surveyors frequently referred to areas of mixed pins as ”openings“, which lends further credence to an open, park- like structure. No understory or ground flora data were recorded in these groves and the pole size diameter classes are probably under-represented in Figure 9, so the degree to which advance regeneration was occurring is unknown. 58 However, the thick bark of mature pine in these groves would be very resistant to the periodic fires which occurred within this forest type, while high mortality would be expected in the seedling and pole size classes. Consequently, one would expect the diameter distributions of white and red pine to show large numbers of mid to large sized trees, some very large monarchs and relatively few smaller diameter trees. Indeed, the mean and median diameters of all sample trees in the mixed pine forest type were in the 30 cm class. The mean diameter based upon the mean basal area per tree was 36 cm. I theorize that hemlock may have occurred within the groves of white and red pine noted by the surveyors, and given its shade tolerance was probably uneven-aged in distribution. Since hemlock has very exacting moisture requirements for seed maturation and germination, many of the groves where hemlock was present were probably located in the ravines of stream bottoms, where more moist and favorable microclimates existed. These are also the probable sites in which spruce was located. I suspect that significant regeneration of hemlock in the 5 and 10 cm classes was present in these more fire resistant habitats within the mixed pine forest type, but was not reported due to lack of suitability as bearing trees. Seral aspen and white birch probably existed in even- aged patches within the mixed pine landscape. Large areas of windthrow were noted by the surveyors within the mixed 59 pine forest type, and in addition to the remnant white and red pine it was reported to be thickly stocked with aspen and white birch saplings. This fact, together with the open park-like structure, the tendency of white and red pine to occur in groves, the fire resistant nature of mature pine and the ravine microclimates, suggest that the mixed pine forest type was very heterogeneous. This heterogeneity was probably manifested in a mosaic of both even and uneven-aged stands within the forest type. The Jack Pine Forest Type The jack pine forest type was often referred to as thickets by the surveyors. The density of the forest type confirms why jack pine stands were referred to as thickets. They had a high density of 326 TPH and a low basal area of 8 Ifi/ha, and were comprised of nearly pure jack pine (Table 10). The very few monarch red and white pine that occurred within the jack pine forest type were excluded from the density and basal area calculations. Figure 10 shows the diameter class distribution of the jack pine forest type. I do not believe that landscape heterogeneity has confounded the diameter class distribution to a significant degree, and an even-aged distribution is suggested. An even-aged structure is expected given the seral nature of jack pine, its predisposition for stand replacing fires and its dependence upon fire for regeneration. The mean and median diameter of all sampled wfizxkomwz CUE W721 v.0(fi ji be hc 10 (3 re 60 160 14G 120 100 JACK PINE PER HECTARE 8 60 40 20 °_ 5‘ U 8— 5 10 15 20 25 30 35 40 45 DIAMETER CLASSES (cm) Figure 10. Diameter distributions of tree species in the jack pine forest type. jack pine trees was in the 15 on class. The mean diameter based upon the mean basal area per tree was also in the 15 cm class. Very few jack pine apparently survived periodic fire disturbance to reach the larger diameter classes, and the jack pine landscape appears to have been quite homogeneous. The Mixed Conifer/Deciduous Upland Forest Type The mixed conifer/deciduous upland forest type also has a relatively high density (287 TPH), but a significantly lower basal area (21 mz/ha) than the northern hardwood type (35 mz/ha). Table 11 and Figures 11 and 12 show the relative density and dominance of tree species in the mixed 61 Table 11. Relative density and dominance of tree species in the mixed conifer/deciduous upland forest type. TEtal Trees Basal BA per Tot Relative Per Area Relative Hectare Species? # Density Hectare (IF) Dominance (IE/ha) A 12 1.6 4.50 0.5 0.9 0.2 (1.0)‘ (5.1) (0.0) BA 8 1.1 3.00 0.3 0.5 0.1 (1.2) (3.0) (0.5) BE 2 0.3 0.7 0.1 0.2 0.02 (0.3) (0.9) (0.1) BF 78 10.2 29.4 2.8 5.0 1.1 (11.9) (30.2) (4.6) C 64 8.4 24.0 3.5 6.3 1.3 (9.7) (38.1) (5.8) H 131 17.2 49.2 13.7 24.3 5.1 (19.9) (146.7) (22.4) JP 31 4.1 11.6 1.0 1.7 0.4 (4.7) (10.2) (1.6) RH 38 5.0 14.3 2.2 3.9 0.8 (5.8) (23.8) (3.6) RP 11 1.4 4.2 1.1 1.9 0.4 (1.7) (11.6) (1.0) SM 25 3.3 9.4 1.4 2.6 0.5 (3.0) (15.4) (2.4) SP 151 19.8 56.8 5.7 10.1 2.1 (23.0) (61.2) (9.3) T 38 5.0 14.3 1.3 2.3 0.5 (5.0) (14.1) (2.2) W8 21 2.8 7.9 0.9 1.7 0.3 (3.2) (10.0) (1.5) HP 83 10.9 31.1 15.7 28.0 5.8 (12.6) (168.8) (25.7) YB 70 9.2 26.2 6.0 10.6 2.2 (10.6) (64.2) (9.8) Sun 763 100 287 56 100 21 (116) (603) (92) ‘ Numbers in parentheses are trees/acre, ft2 and ft‘z /acre respectively. 2 A - aspen, BA - black ash, BE - beech, BF - balsam fir, C - cedar, M - hemlock, JP - jack pine, RM - red maple, RP - red pine, SM - sugar maple, SP - spruce, T - tamarack, WB - white birch, WP - white pine, YB - yellow birch. 62 TREES PER HECT ARE I ‘ ...I O... I... ... W”””WCU ... m”... ... W0... ‘ A BABE BF C H JP RM RPSMSP T WBWPYB TREE SPECIES (See Table 11) Figure 11. Density of tree species in the mixed conifer/deciduous upland forest type. BASALAREA PER HECTARE A BABEBF C H JP RM RPSM SP T WBWPYB TREE SPECIES (See Twle 11) Figure 12. Basal area of tree species in the mixed conifer/deciduous upland forest type. 63 conifer/deciduous upland forest type.' Spruce, eastern hemlock, white pine, balsam fir and yellow birch were the predominant tree species of this forest type, with red maple also present in significant proportions. White pine and hemlock were dominant in terms of basal area (5.8 and 5.1 Ifi/ha respectively). Spruce and hemlock had the greatest densities of any species in the forest type (57 and 49 TPH respectively). Other species that were found within the forest type were aspen, beech, jack pine, red pine, sugar maple and white birch. The mixed conifer/deciduous upland forest type was apparently a mosaic of poorly drained, well drained and excessively drained soils, with well drained sites predominating. This would explain the black ash, cedar and tamarack, versus the jack and red pine components of the forest type, which typically occur upon poorly and excessively sites drained respectively. The diameter distribution of the mixed conifer/deciduous upland forest type is shown in Figure 13. Appendix B contains diameter distribution graphs for the most significant species in the forest type. The smaller diameter classes are probably under-represented in the sample due to non-suitability as bearing trees. The diameter distribution is again confounded by the spatial heterogeneity of the landscape, thus precluding any ' (definitive conclusions regarding stand structure. The mean and median diameters of all sample trees was in the 25 cm (class. The mean diameter based upon the mean basal area per 64 TREES PER HECT ARE mm- ... .. .. 10W -- ~~ ~- 51015202530354045505560 7075 90 100 DIAMETER CLASSES (cm) Figure 13. Diameter distributions of tree species in the conifer/deciduous upland forest type. tree was 30 on. 'It is evident from the diameter class distributions that some very large trees occurred within the mixed upland type . Since hemlock often occurs in groves and has very exacting regeneration requirements it is possible that some hemlock existed in stratified cohorts, with at least two distinct canopy classes. Those cohorts may have been even- aged. However, hemlock is very tolerant, grows extremely slowly in its early growth stages and can persist in a suppressed canopy position for up to 200 years (Tubbs 1977). Thus, any apparent even-agodness of the stratified canopy layers may be deceiving, and some of the hemlock stands 'were, therefore, probably uneven-aged. 65 Red pine and jack pine probably existed in even-aged cohorts on smaller pockets of excessively drained soil within the mixed upland forest type. Cedar and tamarack probably occurred upon poorly drained microsites within the mixed upland forest type. Sugar maple probably occurred upon well drained, morainal features throughout the forest type, as a late successional species. Aspen was probably predominant within areas of disturbance as a seral species, in even-aged cohorts. Overall, the wide variety of tree species (each with very different physiological characteristics and site requirements) suggest that the mixed conifer/deciduous upland forest type was quite heterogeneous in composition and structure across the landscape in which it occurred, probably being a compilation of even-aged and uneven-aged stands of various combinations of species. The Mixed Conifer/Deciduous Lowland Forest Type The mixed conifer/deciduous lowland type has a slightly higher basal area (15 mz/ha), but is also less dense (269 TPH) than the mixed conifer swamp type. Table 12 and Figures 14 and 15 show the density and dominance of tree species in the forest type. The dominant species in the forest type were white pine, spruce and tamarack. White pine was the clear dominant, with almost twice the basal area (4.6 mz/ha) of any other species in the forest type. Spruce and tamarack were dominant in terms of density with 66 Table 12. Relative density and dominance of tree species in the mixed conifer/deciduous lowland forest type. Total Trees Basal BA per ‘ Tot Relative Per Area Relative Hectare Spec iesz # Density Hectare (m2) Dominance (mz/ha) A 10 5.2 14.1 0.3 2.5 0.4 (5.7)‘ (2.9) (1.7) BA 2 1.0 2.7 0.04 0.4 0.1 (1.1) (0.4) (0.2) BF 18 9.4 25.2 . 4.4 0.7 (10.2) (5.0) (2.9) BR 1 0.5 1.5 0.01 0.1 0.02 (0.6) (0.1) (0.1) C 17 8.9 24.0 1.0 9.4 1.4 (9.7) (10.9) (6.2) H 12 6.3 16.8 1.0 8.9 1.3 (6.8) (10.3) (5.9) RN 4 2.1 5.7 0.2 1.5 0.2 (2.3) (1.8) (1.0) RP 5 2.6 . 0.8 7.6 . (2.0) (0.0) (5.0) 8H 2 1.0 2.7 0.2 1.4 0.2 (1.1) (1.6) (0.9) SP 57 29.7 80.1 1.9 17.3 2.6 (32.4) (20.0) (11.4) T 39 20.3 54.6 1.3 11.8 1.8 (22.1) (13.7) (7.0) 88 3 1.6 4.2 0.1 1.3 0.2 (1.7) (1.5) (0.9) WP 17 8.9 24.0 3.3 30.8 4.6 (9.7) (35.7) (20.3) YB 5 2.6 6.9 0.3 2.7 0.4 (2.0) (3.1) (1.0) Sum 192 100 269 11 100 15 (109) (116) (66) ‘ Numbers in parentheses are trees/acre, ft2 and ftz/acre respectively. 2 A - aspen, BA - black ash, BF - balsam fir, BR - bur oak, c- cedar, H - hemlock, RM - red maple, RP - red pine, SM - sugar maple, SP - spruce, T - tamarack, WB - white birch, WP - white pine, YB - yellow birch. 67 .5. z c m u: a. g m 10- 2...... I -------~° 0- 4 A BABF BR C H RMRPSMSP T WBWPYB “TREE SPECIES (See Tale 12) Figure 14. Density of tree species in the mixed conifer/deciduous lowland forest type. 6 4.5 & fa) 0| (ID ‘ O G BASAL AREA PER HECTARE p a ‘ p (m ? Figure 15. A BABFBR C H RMRPSMSP TWBWPYB use m M... TREE SPECIES (See Table 12 Basal area of tree species in the mixed conifer/deciduous lowland forest type. 68 80 and 55 TPA respectively. Northern white cedar, balsam fir and eastern hemlock also occurred in significant densities. Other species that occurred within the forest type were aspen, black ash, bur oak, red maple, sugar maple, white birch and yellow birch. Aspen, yellow birch and red maple were the most significant of these minor species. The major difference between the mixed conifer/deciduous lowland and mixed conifer swamp forest type was not the species composition of the types. Both forest types were composed of similar species, with the jack pine component of the mixed conifer swamp being absent in the mixed conifer/deciduous lowland type. The major difference between the two forest types was the relative proportions of species within both types. There was much greater parity in the density of species in the mixed conifer/deciduous upland forest type. Cedar was not nearly as dominant, and hemlock and balsam fir were relatively more dominant in terms of density and basal area. Most significantly, the relative density of the hardwood deciduous component of the forest type was much higher than in the mixed conifer swamp forest type. Figure 16 shows the diameter class distribution of the mixed conifer/deciduous lowland forest type. The moderate bias toward trees in the 15 and 20 cm diameter classes is evident in the distribution. Appendix F contains the diameter class distributions of significant species within the forest type. The smaller diameter classes are again 69 7 E a “.....- - c: ‘1‘ w 30 us In E 20------- .- 10....mm- . . .... .... .... I II I T. 51015202530354045505560 75 00 DIAMETER CLASSES (cm) Figure 16. Diameter distributions of tree species in the conifer/deciduous lowland forest type. under-represented in the diameter distributions due to non- suitability as bearing trees. It is evident that some very large trees were present in the mixed lowland forest type. The mean and median diameters of all sample tree species ‘within the forest type were 23 and 20 cm respectively. The :mean diameter based upon the mean basal area of all trees was 25 cm. The wide variety of sample tree species indicate that ‘the mixed lowland forest type was probably very heterogeneous in structure. The dominant species, white pine, spruce, tamarack, cedar and balsam fir probably existed in uneven-aged and even-aged patches scattered throughout the landscape. Seral aspen likely occurred in 70 even-aged patches. Hemlock may again have existed in stratified, uneven-aged and even-aged stands. Red pine probably occurred upon small, excessively to well drained upland islands, while mesic species such as sugar maple and yellow birch occurred upon well drained, more fertile sites imbedded within the overall landscape. The Mixed Conifer Swamp Forest Type The mixed conifer swamp forest type had a moderate basal area (14 mz/ha), but was quite dense (314 TPH) . Table 13 and Figures 17 and 18 show the density and dominance of the tree species in this forest type. The forest type was dominated by northern white cedar, spruce, tamarack and white pine in terms of both basal area and density. Cedar had the highest basal area (4.0 mfi/ha), and spruce had the highest density (97 TPH) of all species in the forest type. Of the dominant species, white pine had a relatively low density (15 TPA), but had a relatively high basal area (2.5 nfi/ha). Other species occurring within the forest type were aspen, black ash, balsam fir, eastern hemlock, jack pine, red maple, white birch and yellow birch. Of these species, only hemlock and balsam fir were significant in terms of density and basal area. Alder and briars were sometimes noted by the surveyors in the mixed conifer swamp forest type. . Figure 19 shows the diameter class distribution of the :mixed conifer swamp forest type. The diameter class 71 Table 13. Relative density and dominance of tree species in the mixed conifer swamp forest type. Total Trees Basal BA per Tot Relative Per Area Relative Hectare Species2 # Density Hectare (IF) Dominance (IE/ha) A 6 0.5 1.5 0.8 1.5 0.2 (0.6)‘ (0.4) (0.9) BA 6 0.5 1.5 0.2 0.3 0.1 (0.6) (1.7) (0.2) BF 40 3.4 10.6 1.0 2.0 0.3 (4.3) (11.2) (1.2) 30 1 0.1 0.3 0.02 0.04 0.01 (0.1) (0.2) (0.02) C 292 24.7 77.3 15.0 29.0 . (31.3) (161.7) (17.4) H 29 2.5 7.7 2.0 3.8 0.5 (3.1) (21.2) (2.3) JP 55 4.7 14.6 1.1 2.2 0.3 (5.9) (12.1) (1.3) RM 8 0.7 2.2 0.3 0.5 0.1 (0.9) (2.9) (0.3) RP 11 0.9 3.0 0.7 1.3 0.2 (1.2) (7.2) (0.0) SP 366 30.9 97.1 11.2 21.5 2.9 (39.3) (120.0) (12.9) T 287 24.2 76.1 9.3 17.9 2.4 (30.0) (99.9) (10.7) WB 16 1.4 4.2 0.6 1.2 0.2 (1.7) (6.7) (0.7) ‘WP 58 4.9 15.3 9.5 18.2 2.5 (6.2) (101.7) (10.9) YB 9 0.8 . . 0.6 . (1.0) (3.4) (0.4) Sum 1184 100 314 52 100 14 (127) (550) (60) ‘ Numbers in parentheses are trees/acre, ftz and ftZ/acre respectively. 2 A.- aspen, BA - black ash, BF - balsam fir, BO - black/red oak, C - cedar, H - hemlock, JP - jack pine, RM - red maple, IRP - red pine, SP - spruce, T - tamarack, WB - white birch, ‘WP’- white pine, YB - yellow birch. 72 1 80 ‘5‘ 7 “ 0 .. - z: a: u: a Hi In E; 30 m ...l l 1 - A BA arao c H JP PM RP SP T WBWPYB TREE SPECIES (See TabIe 13) Figure 17. Density of tree species in the mixed conifer swamp forest type. 4 3.5 2.5 BASAL AREA PER HECT ARE A) 0.5 .. A BABFBO C H JPRMRPSP T WBWPYB TREE SPECIES (See Table 13) Figure 18. Basal area of tree species in the mixed conifer swamp forest type. 73 90 7 E :“E som- 35 :1 “be" m - {3 E ”“W' “ " ”W .. ... .... 1mm - ~ - -~ swamammwaw m mm m DIAMETERCLASSES (cm) Figure 19. Diameter distributions of tree species in the mixed conifer swamp forest type. distributions of the individual species within the forest type are presented in Appendix C. The smaller diameter classes are again sis-represented due to their lack of suitability as bearing trees. The mean and median diameter of all tree species in the sample was in the 20 cm class. The mean diameter based upon the mean basal area per tree was 23 cm. There did not appear to be many very large trees within the mixed conifer swamp forest type. Landscape heterogeneity again confounds the diameter class distributions, precluding any definitive conclusions regarding stand structure. However, many of the stands within the mixed conifer swamps may have been of fire origin, and were probably even-aged in structure. It is 74 highly probable that almost pure patches of jack pine became established on upland islands after significant fire events within the forest type, and were almost certainly even-aged. Remnant white pine may have added a dimension of uneven- agedness to the structure of the mixed swamp type, due the fire resistant bark of the mature trees. Disturbance Regimes Disturbance regimes noted by surveyors in the study area were fire, windthrow, fire associated with windthrow, dead timber from unknown cause (possibly insect related mortality) and beaver floodings (ponds or meadows). There were 285 individual disturbance events recorded by the surveyors in the study area. Figure 20 shows the distribution of all survey lines affected by these disturbance events. The thickness of the affected survey lines are exaggerated for purposes of clarity. Figure 21 shows a very approximate estimation of the total area covered by these disturbance events in the study area. It is immediately apparent from these figures that natural disturbance affected large portions of the study area. The *widespread, and in several cases very extensive, occurrence of disturbance clearly played an important role in :regulating the composition and structure of every forest type. Tables 14 and 15 show the distribution of these events by forest type and disturbance regime. Table 14 jlists only those disturbance events actually observed by the 75 .eman on oewn some cmmnsonm .sucsoo mamcsnso :ncuwcz E." musc>c cocmnnsumnp he consumes mused >c>nsm .om unseen T... ..u... TH Lace)“. 3%! emN . womua Odsom nunoz seeds—.558:— . Roam-.00 in x:- .831. ICI—ISJ TuCSfl\S°0 ‘Ciuha ‘uSSfi\:°0 Sflufi £085 Sflmfi 108a! ZDE Tool-Fad: Pal-obs! "1:080..— h‘. Dubs—z. . \\\\ H .x. L® asamaaa ..4ceam I z 4. x , a .3 ... .. ®..______._,‘r/./f ‘ \NwA. If. Iii-ruin 76 .vmma CU coma EOHH cmwwSOflz 4%»:900 c3omnnnu oneness ca coccnnsumwp no moons pcusaeumm .HN shaman ......1. .-...~ «u. m.“ entrain. Sum: www.moeun canon suuoz SOSSAASouaO Rehucoo A8880 Mu! mop-Idle..— $u08\¢8 figufifl QuOécs Stem 8065 Scr— mogul EIEIIE IIIIEQII 1008188: SASSoRcz "washed it Onhgxh \ \ (x. . \\\. IMAM/I I / .- J IAutxexfl v x is. . x r. .mfihfl? , ...:u/ i 8 \ . . ... .\ \ \x x. .... \.\.\.\“.x. e s x N I I \ . \. \ .I \‘ \‘ o ., . . .\. \ \ \ . .\ x j SQIRXr \ . 't .\ \ .Wx\\\“\\\&\\\\\\x‘ \. x. . .. ... ,. . \\\4 \ ~\ J \ 040/. .\\\ \a‘ l \, .X. \x\\ \.\\. J .\ \\ (In 77 Table 14. Number of individual disturbances events observed by surveyors in each forest type. North Mixed ’3ack Mixed Mixed' Mixed Hdwds Pine Pine Upland Lowland Swamp Total % Area (ha) 66,129 27,490 9,632 44,585 21,974 103,427 Total Transact Length (km) 898 378 121 657 314 1,407 Fire 4 12 7 14 9 24 70 24.6 Wind 6 6 - 33 3 39 87 30.5 Wind/ 2 1 1 30 16 35 85 29.8 Fire Insect - 5 4 3 - 14 26 9.1 Beaver 1 - - 8 1 7 17 6.0 Flooding . Totals 13 24 12 88 29 119 285 100 8 4.6 8.4 4.2 30.9 10.2 41.7 100 Table 15. events in each forest type. 78 Total number of estimated individual disturbance North Mixed Jack Mixed Mixed Mixed Hdwds Pine Pine Upland Lowland Swamp Total 8 Area (ha) 66,129 27,490 9,632 44,585 21,974 103,427 Total Transact Length (km) 898 378 121 657 314 1,407 Fire 48 56 59 131 56 139 489 24.1 Wind 58 8 - 196 19 131 412 20.4 ‘Wind/ 53 1 5 186 83 227 555 27.4 Fire InsoCt - 15 8 51 - 121 196 9.6 Beaver 25 - - 206 15 128 374 18.5 Flooding frotals 184 80 72 770 173 746 2025 100 8 9.1 4.0 3.6 38.0 8.5 36.8 100 79 surveyors of the study area. Table 15 lists the observed disturbances plus an additional number of estimated disturbances that the surveyors did not detect in section interiors. All of those estimated disturbances are smaller than 1600 m in length. Given the large number of disturbances actually observed by the surveyors and equally impressive area covered by these disturbances, I believe the total estimated number of disturbances in Table 15 to be reasonably accurate. The overwhelming number of disturbance events occurred within the mixed conifer/deciduous upland (38.0 percent) and mixed conifer swamp (36.8 percent) forest types. The northern hardwood (9.1 percent) and mixed conifer/deciduous lowland (8.5 percent) forest types were approximately equal in their proportion of disturbance events. The mixed pine and jack pine forest type had 4.0 and 3.6 percent of all disturbance events respectively. Windthrow was the dominant disturbance regime in the study area landscape. Figures 22 and 23 show the incidence of windthrow (without subsequent fire) in the study area, and Figures 24 and 25 show the incidence of windthrow with associated fire. Windthrow (without subsequent fire) accounted for 20.4 percent of all disturbances, but this figure increased to 47.8 percent when additional windthrow events that subsequently burned were also taken into (account. Since it is logical that windfall events created heavy fuel loads on or near the forest floor, I have assumed 'that fires occurred subsequent to the incidence of 80 .eman on came some caveman: .xucsoo czocmfico :ncumcs :H zoncuccns >n couscous momma >c>n=m .NN shaman a.1_..r_...~ L11 21.. DWI u. omNLwowuH Tamom SUHOZ 209565: I Roam-.80 .8880 In! mu ecu—:62 scoos\coo a Mini: ‘u08fl\68 scat JOSH Sperm 10! :- EDE- ‘OCI-vfilz £10802 33:030.. an! Unhiw—P— .«mma on coma scum cmensonm .>ucsoo masseuse success on soncuosws he pmuocmaw cons oeucawunm .nN masons 81 :1.._._....n..__.u LII. ..H. =qu www.moeun canon nuuoz censuses: soeacoo azure.xux see—son caoosxcoo scans: mac-execs scam moss scam sexes El E! E! I! El E3 II 1008184: Shofiohcz “sconce an! 05—32.— .4',.\\ .\ .. ...“ :?NAM \, N, \ \ , \6 \\ , 82 .vmma on oema Bonn savanna: .mussoo cacmmwsu snoumoa on when use saucepan: an ocuocuuc mcswa >o>nsm .vm shaman 35.; $088188 mag ifillb 3m: www.mocun canon nuuoz 2.: s 260.315... I IIHd-boo g8“ you: E mes—36.. 366.30 I mes-s: 366.30 E 95.— nosn E 2:.— ..83... H— : 691996: "5:080; hCI DubGIUSF . ..s. .. . \ 0 . .4 1m . . ,\ \\ \s \\.\\~ . . . . ‘ . .. . .\\.\ . .\\..\’ . \ \..\ 4‘. . . \\ \\ \\ . . 1 ._ h . \. xx _ . . . . . . a . . .. . . ._ v T“ Surname/J . fl \ - ..s. e we a ._ :.JI_- ; . \\ \ , \. , _.1.. . \ . \\ to. C \ .\\ \\\\\\.\\\ I vu.“vM.“8um\\\\ \ £7 . x I.\\ L . (wwwvwmxxnw I .\\ 1- . \“.\.. ACQWVF ’1 0M.\ . \\>.$& . b / ., /0.mw.\xxv.,rw. \\ . \, __= ./ L7,, .4 IA... .3 ....,..:./ a 4 . I\. \, .emma ou coma some scmwnoflz .mussoo cscmafino :ncumcS :a when use zoncposws he ocuocmaw cons ocucawumm .mn shaman 83 fi.¥».—PH$._Q Amdwqu. DUN; hfiwmf. cm~.mooun canon .AsdaS. he‘s. \ \\I\\\\\\\. \ \\x .\ . . __. [or We nunoz .mom1_ . a.». x a . I 2.: s 0922.5: I 2.2... /_// RIUu-ba in ”a: a .\\.. 1¢8~8°J 1nCSfl\COO $68—53 1eOSfl\¢°0 Sflmfi 188‘! 8008888: 698—0902 E— ee... .54.. E H— I "isobad ICI Ouhclfixh 84 windthrow, rather than vice a verse. I do recognize that down-drafts caused by large crown fires can also result in fallen timber, but the impact of this phenomenon is impossible to determine. Two very large windthrow events were observed by the surveyors in the study area, with one of these events apparently burning after it blew down. Figures 22 and 24 show that these events occurred across two to three adjacent forest types. The greatest incidence and extent of windthrow was in the mixed conifer/deciduous upland and mixed conifer swamp forest types, with significant numbers of events also occurring within the northern hardwood and mixed conifer/deciduous lowland forest types. Relatively few windthrow events occurred in the mixed pine and jack pine forest types. It is interesting to note that windthrow disturbance appeared to be associated with the proximity to the shoreline of Lake Superior. There was a pronounced lack of windthrow events in the southwest portion of the study area, in the lee of the landmass to the west of Whitefish Point. An analysis of historic wind patterns and velocities would undoubtedly prove interesting, if a positive correlation existed between the wind patterns and this observed "snapshot” of windfall disturbance. Some windthrow events can likely be attributed to severe thunderstorms and tornadoes, although the incidence of such storms is suppressed in the region due to the moderating effect of the Great Lakes. Snow and ice loads from winter storms may also contribute to downed woody debris within the 85 forests of the study area. The survey records show that fires not associated with windthrow accounted for an additional 24.1 percent of all disturbances within the study area landscape. Figures 26 and 27 show the incidence of fire in the study area. Fires were most common in the mixed conifer swamp and mixed conifer/deciduous upland forest types, with fire also occurring within the jack pine, mixed pine, mixed conifer/deciduous lowland and northern hardwood forest types (Table 15). In the northern portions of Michigan it is not uncommon for organic matter decomposition to be greatly reduced due to the cold climate, thus sometimes resulting in significant accumulations of organic matter on forest floors. Such accumulations can, over time, result in high fuel loads, which increase the chance of fire occurrence. There are many factors which play a role in determining the degree to which a particular forest type, or stand within a forest type, is susceptible to fire. The length of time since the last fire event is the most important, since time has a direct influence on successional status, the degree of biomass accumulation, the presence of fuel ladders, the incidence of natural mortality and the probability of other disturbance such as windthrow and insect or disease related :mortality creating increased fuel loads. Another factor is anthropogenic influences. It is impossible to determine from the surveyor records the influence of the native- .American population upon the incidence of fire in the study .eman 0» oven souu :mmnzonz .sucsoo czcmmflco cucumcz so onflu he pouoouum nosed >o>nsm .em shaman ....._..r in 2.1.5. u. 3%.! www.mooun canon nunoz Shah Beau—.00 98880 Xe! ‘88-'04 ‘aOSfl\CS° mm- 1¢8nfi= 1a°Sfl\¢°U SCuh 8°85 filial! Sflmh ‘Squ $808198: Shea—ohm: z "flcobuJ kc: UnhCIfilh .emwa on coma scum cmmnnonz .sucsoo cscmmficu snoumoz on when me concede“ cons poucswumm .hm enemas 87 ....m—era. Leicu. am! an \ \ h s . “we... \ .mx , om~ woman onaom , shame , \ ; ///2,.xmh . . ,S. I // \\x . s\\\\.\ \\N\\\. I . ‘ . C no , ,/3s;§ ‘ , >§$§88sv- . I I a. . .\ .\. ‘\ I/ I/ z. e: .x I I z . . nuuoz an... IUQmp-a 3,“ meat ‘68-'84 ‘mOSfl\8°0 EIIIEJII U¢8~A= ‘uOSfl\¢80 88mh 8°85 SSuh QSX«I HIEIEH 100828: Shep—ens! "flowMUA hc‘ UnhCIUSh 88 area. It is possible that some fire events were deliberately initiated by native-Americans to improve game habitat for hunting purposes. I cannot discount the possibility that they may have had a significant impact upon the frequency and magnitude of fire disturbance in portions of the study area. Areas of dead timber accounted for 9.6 percent of all disturbance events in the study area landscape. Figures 28 and 29 show the areas of dead timber noted by the surveyors. The most significant number of events occurred within the mixed conifer swamp forest type, with additional numbers of events occurring within the mixed conifer/deciduous upland, mixed pine and jack pine types (Table 15). No incidents of dead timber were noted in the northern hardwood or mixed conifer/deciduous lowland forest types. Tree mortality may have been caused by induced stresses from endemic insect infestations. Insects such as the spruce budworm (We funifsrana) . jack pine budwor- (Chainsaw minus) and the «stern larch butl- (Dendmcnnus simplex) may have been the causal agents of stress. Balsam fir and ‘white spruce are the preferred tree species of the spruce Ibudworm, although they will also secondarily feed on black spruce, tamaracks, hemlock and pines if they are growing nearby. The jack pine budworm will also feed on red and ‘white pine if they are growing under or adjacent to an infested area of jack pine. The eastern larch beetle jprinmrily affects tamarack (Wilson 1977). Insect induced 89 .emmn on come Bonn cavemen: .xucsoo usmddnno :ncuwcs on human» some an convenes mocha >c>nsm .wm unseen m... Tanmn 111...”. :MI wm~.w0bna canon a. . .... 1. 3.. .5 , as? 18, x . “xx“\ xx \xxxx \\ \ . \ no .\ :3» \ \ \ SH 2 c 9 ¢ 3 . xx xx x \ \ x \x xxxxxxxx xx xx xx xx xx.xx x . Cx. x. x. x. x») . \ \ \ \ .. .\ \ \ \ \ \ .:¢\:¢\\S \ \ fit‘lmh $8.9 xx xx x Achesoo AS8SM Xe! E II 1:8u804 ‘uSSfl\¢80 Vfllufia 1w°8§\¢°0 ewe-uni gnaw-wh- icufi 182:! IE filial! 1008188: Sfiofiohoz “53:050..— hCI Unb¢lfl8h 90 .emwa ou oven some savanna: .aucsoo czcmmflno oneness on human» poop he concede“ cons ucucaflumm .mN enemas .. . I 8 Au“: :cuimmfi QII-U 2%! Q8 .. _. _ . . ... -6... K a) .- \\x\\: .. .... .38 n /.,. . ex . . umfiwfi. ..w.\\\.\\ \\\\\\m\\\11 \W x. . / .. a? . \ . \\ xx.\\.....\ \ . \ n \ .- cmm . m8: 080m . .3 §\\\ \x \WX. J- . .\.x. . .r?f .* \ 1: fiWEeu ... \ \.x — o n. .- 1 .3 \\\\\\\\\ \\\\&¢\ uh .. .. .. ___1\.___.. \ .\\\ x x \, / \\\\\.\\\I.K\Kw\xvl' - '......II . .\.v . .\ . .x i/\ \\4«\ .\\\.\x\xx\x\xu\\\\\\\x\\ l. \-\\\\ /\.x x .\\.\\x . / \\\\ a . .x, .. x\...,\ .\. x\\ x .. . \x. \.\.\. r .a/ . . x, ... .4..xx \ \x x 81. \\.\ I \x\ .\\. . x.\x. \x/ \x x. o 1 .xx/ , . x v . \c \x .. .f .x I v...‘ . Y x:\a .. ., — - _, Illa , 4%...- nuuoz ASAIuh 18Sfl a Rome—.00 88880 Xe! 168—884 finOSfl\¢80 8881-: 12.6800 SBuh 8°85 Sflnh 188a! Ill: EIEZII EIII 80881888 68080802 "ficufiwd ...... n... ..V 27% kc! Uuhclfllh SI TIC 10 Ii I'lh. 0f 91 mortality can be widespread during major outbreaks, but outbreaks are typically patchy in nature, affecting small stands or portions of stands throughout a landscape. Because the recorded instances of dead timber were observed in forest types dominated by tree species that are susceptible to insect infestation, I have concluded that such areas of tree mortality probably resulted from insect infestation. All subsequent discussion of dead timber will be referred to as insect induced mortality. Areas affected by beaver (Castor ganadgnsig) floodings accounted for 18.5 percent of all disturbance events in the study area. Beaver disturbance events were noted by surveyors as ponds or meadows. The felling of trees for use as food and dam building material is also a factor in beaver related disturbance, but no mention of this was made in the survey notes. The areas affected by beaver flooding are shown in Figure 30, but are hard to distinguish due to their small size. The greatest number of events occurred within probable imperfectly drained depressions and streams in the :mixed conifer/deciduous upland forest type, with a significant number also occurring within the mixed conifer swamp forest type (Table 15). Some beaver disturbance was .noted in the northern hardwood and mixed conifer/deciduous lowland forest types, but no events were recorded in the :mixed pine or jack pine forest types. By the period in ‘which the surveys of the study area occurred, the incidence of beaver related disturbance was probably significantly 92 .emwa ou owed some sameness .mucsoo cscmmaco abounds cw manpooHu no>ccn an pmuocmsfi some pcumaflumm .on unseen «.1. .—.~.I an QTI: U :5... www.mowua OHmOm nuuoz IIUm-bs 3’“ we" MU sea—364 4.0-excoo mes—s: sac-execs 90.1 zone sens 1.2.x 1602694: anemone: "mashed hcl UnhCIUSb use—vocab 909803 I xx x \ . xvx...“ w .\xx\ \ . \. . .I. .f . . \ ...; s\\\\xuaw\\\a\\wh . . \ .\\ .x / .. x x .\\\X\\\\\\\\V\\.\x .8 x..\\\\\x\\rx ... «KL x... ......‘P. I’. \«WCK xxxx. 09/ x x\..\.x\\.x . . . \ .\ . .\\ 8’ . . . . xxx . x .x 2. s a . M, m \. .\\ \7 I I . , . xxx xxxxxl. .\ \a I . . .. x. n\\“\.\ . .1 x x , x. . \\\ \ \ \ x / *2 .\ / / \\x\\\\x§ // Ir 4 \ \\,\\\.\\ \\\ /I \\\ I ./ . v I. \.\ \ \ \ \ . \ I / I. 4 . ‘ . ,é.nn: . . ... . fl: r - xx.\\x\ \xvxxwx ./ J x ‘ ..xxxxx\,x\ x\ s../ .7 x ,\\.x\.\ \ .\\ x .9 l I . .\\ xx \\\\\\.\\.\ .l I. f. «by x. \\ \\ x\.\\ \\\\\\x /.I I. ’4 x x x x x .0 mm/ 1'87. \.\\ \\ \\ \\.\\ . \. I - . \.\.\\x.\\\\\\\\\ 9 I4. . \\ \\\\\ ..\.\ \\ / I (“I .. : ... . x . x x a . A\ \x \x \\ \\ I .1 I . . ..x.x \1. x/ a). .. , .‘Nn . II 4 1.! . . cpl Vfimxutug a. A . . In . les tre1 of l and for The dist 30y 18 r beav tipa land were 93 less than its original extent. This was due to the fashion trends of the era, and the resultant deaand for and harvest of beaver skins. Estinates of disturbance frequencies are based upon 15 and 30 year recording intervals. Disturbance frequencies for each regine in each forest type are shown in Table 16. The estinated return intervals and areas of annual disturbance for each forest type are also based upon 15 and 30 year recording intervals, and are shown in Tables 17 and 18 respectively. Return intervals have no leaning for beaver floodings because floodings are restricted to riparian areas, and cannot possible occur over an entire landscape. Therefore, return intervals for beaver floodings were not calculated. 94 Table 16. Disturbance frequencies‘ (events/year) by forest type and disturbance regine, based upon 15 and 30 year recording intervals. North Mixed fie): nixed Mixed nixed Hdwds Pine Pine Upland Lowland Swanp Area (ha) 66,129 27,490 9,632 44,585 21,974 103,427 Total Transect Length (kn) 898 378 121 657 314 1,407 Freq Freq Freq Freq Freq Freq Eire 15 yr 3.2 3.7 3.9 8.7 3.7 9.3 30 yr 1.6 1.9 2.0 4.4 1.9 4.6 Hind 15 yr 3.9 0.5 - 13.1 1.3 8.7 30 yr 1.9 0.3 6.5 0.6 4.4 W 15 yr 3.5 0.1 0.3 12.4 5.5 15.1 30 yr 1.8 0.0 0.2 6.2 2.8 7.6 Insect 15 yr - 1.0 0.5 3.4 - 8.1 30 yr 0.5 0.3 1.7 4.0 Beaver flooding 15 yr 1.7 - - 13.7 1.0 8.5 30 yr 0.8 6.9 0.5 4.3 Totals 15 yr 12.3 5.3 4.7 51.3 11.5 49.7 30 yr 6.1 2.7 2.5 25.7 5.8 24.9 ' Frequencies are representative of the entire area of each forest type within the study area. Table 17. disturbances in each forest type, based upon 15 and 30 year recording intervals. Estinated return intervals (years) for 95 North nixed Jack Mixed Mixed nixed Hdwds Pine Pine Upland Lowland Swanp Area (ha) 66,129 27,490 9,632 44,585 21,974 103,427 Total Transect Length (kl) 898 378 121 657 314 1,407 [111 15 Yr 2252 247 142 429 292 490 30 Yr 4504 493 283 858 585 979 fling 15 Yr 1387 284 - 145 633 231 30 Yr 2778 568 291 1266 463 Wind/, 15 Yr 5128 1760 1133 180 122 314 [1:1 30 Yr 10204 3521 2266 359 243 628 Infiggt 15 Yr - 451 218 2119 - 835 30 Yr 901 435 4237 1672 Beaver 15 Yr N/A N/A N/A N/A N/A N/A flood 30 Yr 96 Table 18. lstinated areas of annual disturbance (ha/yr) and percentage of total area affected in each forest type, based upon 15 and 30 year recording intervals. North nixed Sack Hdwds % Pine 4 Pine 4 Area (ha) 66,129 27,490 9,632 Total Transect Length (kn) 090 370 121 fire 15 yr 30 (73)‘ 0.04 111 (275) 0.41 68 (167) 0.70 30 yr 15 (36) 0.02 56 (138) 0.20 34 (84) 0.35 Hind 15 yr 40 (119) 0.07 97 (239) 0.35 - 30 yr 24 (59) 0.04 49 (120) 0.10 - nindtzirs 15 yr 13 (32) 0.02 16 (39) 0.06 9 (21) 0.09 30 yr 6 (16) 0.01 8 (19) 0.03 4 (10) 0.04 Insect 15 yr - 0 61 (151) 0.22 44 (109) 0.46 30 yr - 0 30 (75) 0.11 22 (55) 0.23 Isa11r_zloodins 15 yr 6 (16) 0.01 - 0 - 30 yr 3 (8) 0.005 - 0 - TQSIII 15 yr 97 (239) 0.14 285 (704) 1.04 121 (297) 1.25 30 yr 48 (119) 0.07 143 (352) 0.52 60 (149) 0.62 ‘ Nuabers in parentheses are areas expressed in ac/yr. 97 Table 18 (Cont'd). Bstinated areas of annual disturbance (ha/yr) and percentage of total area affected in each forest type, based upon 15 and 30 year recording intervals. Mixed Mixed Mixed Upland 4 Lowland 4 Swanp 4 Area (be) 44,585 21,974 103,427 Total Transect Length (kn) 657 314 1,407 fire 15 yr 104 (257)‘ 0.23 75 (186) 0.34 211 (521) 0.20 30 yr 52 (120) 0.12 30 (93) 0.17 106 (261) 0.10 flind 15 yr 307 (759) 0.69 35 (86) 0.16 446 (1106) 0.43 30 yr 153 (378) 0.34 17 (43) 0.00 224 (552) 0.22 fiindllire 15 yr 248 (612) 0.56 180 (445) 0.82 329 (813) 0.32 30 yr 124 (307) 0.20 90 (223) 0.41 165 (407) 0.16 lanes: 15 yr 21 (52) 0.05 - 0 124 (306) 0.12 30 yr. 11 (26) 0.02 - 0 62 (153) 0.06 15 yr 49 (122) 0.11 6 (16) 0.03 44 (119) 0.05 30 yr 25 (61) 0.05 3 (0) 0.01 24 (59) 0.02 Totals 15 yr 729 (1802) 1.64 296 (733) 1.35 1160 (2865) 1.12 30 yr 365 (900) 0.81 148 (367) 0.67 581 (1432) 0.56 ‘ Munbers in parentheses are areas expressed in ac/yr. th- 98 is 98 The northern larlwoed Forest Type Disturbance regines noted by surveyors in the northern hardwood forest type were fire, windthrow, windthrow with subsequent fire and beaver flooding. six to twelve disturbance events occurred per year in the forest type (Table 16). The transect size class distribution for the estisated total nuaber of disturbance events in the northern hardwood forest type is shown in Table 19. As expected, windthrow and windthrow events associated with fire were the doninant disturbance regines in the forest type, accounting for 60.3 percent of all disturbances in the type. Fire was associated with windthrow in 28.8 percent of the disturbance events. Fire (not associated with windthrow) accounted for 26.1 percent of all disturbance events within the forest type. The three snallest size classes (< 200 n in length) accounted for 91.9 percent of all disturbances events. Disturbance frequencies for windthrow events in the study area were estinated to be between 4 and I events per year (Table 16). This estinate does not take individual tree fall gaps into account, since such snall events were not recorded and were probably ignored by the surveyors. Half of the windfall events burned subsequent to the windthrow. Most windthrow events were also in the three slallest size classes, with occasional events occurring in the larger size classes, up to 1600 n in length. The estinated return interval for windthrow events (not associated with fire) was between 1389 and 2773 years, with 99 Table 19. Bstinated total nunber of disturbance events for each size class and disturbance regine in the northern hardwood forest type. Transect Size Wind/ Beaver Class (1) Fire Wind Fire Insect Flood Total 4 < 50 - 30 30 -, - 60 32.6 50-100 25 - 23 - 25 73 39.7 100-200 19 17 - - - 36 19.6 200-400 - 6 - - N/A 6 3.3 400-800 4 3 - - N/A 7 3.8 800-1600 - 2 - - N/A 2 1.1 Totals ' 48 58 53 0 25 184 100 4 26.1 31.5 28.8 0.0 13.6 100 an estinated area of annual disturbance between 24 and 48 ha per year (Tables 17 and 18). The return interval for ny study area is slightly longer than that in the pre- European settlenent northern hardwood forests of Rosconlon and Crawford Counties (1220-2439 yrs) in the lower peninsula (Whitney 1986). The estinate is sinilar to the 1210-2420 year estinate calculated by Canhan and Loucks (1984) for the henlock-northern hardwood forests of northeastern Wisconsin. The return interval for windthrow followed by fire was between 5128 and 10,204 years, with an area of annual disturbance between 6 and 13 ha per year. Overall, between 0.05 and 0.09 percent of the total northern hardwood forest was impacted by windfall on an annual basis (Table 18). 100 Eastern henlock was a doainant in both density and basal area in the forest type and is known to be shallow-rooted and nuch less wind fire than other species in the northern hardwood type. Many heslock trees were doainants in the canopy, and probably tended to grow singly or in snall groves. It is possible that aany of the ssaller disturbances in the size classes less than 100 n in length occurred in healock groves. Fires not associated with windthrow occurred at less than half (2-3 events per year) the frequency of windthrow events (Table 16). Figure 27 shows that the fire events observed by the surveyors in the northern hardwood forests of the study area occurred in the transition zone adjacent to nore fire prone forest types. Most fire events occurred in the 50 to 200 neter size classes, with a few in the 400 to 800 a size class (Table 19). The estinated area of annual disturbance was between 15 and 30 ha per year (Table 18). The return interval for fire in the northern hardwood forest was between 2252 and 4504 years (Table 17). This is Inch longer than Whitney’s (1986) interval for Roscoanon and Crawford Counties (1389-2778 yrs).' Overall, the results indicate that nost fire events in the northern hardwoods were shall in size, and occurred prinarily in the transition zones adjacent to other forest types. This night be expected in a forest type that is relatively fire resistant in nature. Beaver flooding was a relatively ninor disturbance 101 regine in the northern hardwood forest type. The area of annual disturbance was a very low 3 to 6 ha per year (Table 18). In nost occurrences, it is probable that beaver related disturbances did not physically occur within what would be considered northern hardwoods, but rather were restricted to seall pockets of lowland types inbedded within the overall northern hardwood forest type. Even so, this reenforces an overall pattern of landscape heterogeneity within the northern hardwood forest type. Overall, disturbance affected only 0.07 to 0.14 percent of the northern hardwood forest on an annual basis. This was the lowest percentage of any forest type. Windthrow and fire were the doainant disturbance regiaes, with fire occurring in the transition zones between northern hardwood and other, nore fire prone, forest types. Snaller events less than 200 a in length were lost col-on, with larger events Inch less frequent in occurrence. The return intervals for disturbance were such longer than the naxiaun potential lifespans of the doninant tree species within the northern hardwood forest. The shortest estinated return interval was 1387 years for windthrow. The naxinua potential lifespans for sugar naple, eastern henlock, beech and yellow birch are 300-400 years (Godnan, Yawney and Tubbs 1990), 400-990 years (Godaan and Lancaster 1990), 370 years (Tubbs and Houston 1990), and 300-370 years (Erdnann 1990) respectively. Consequently, once succession to the northern hardwood forest type occurred on a favorable site, the 102 forest type was likely to persist and doninate the site through the natural recruitnent of its component species. This does not nean that disturbance had no inpact upon the conposition and structure of the northern hardwood forest. Snall disturbance events created pockets where seral forest types could conpete within the overall northern hardwood forest natrix. I believe that snall windthrow events and occasional burns helped to perpetuate the existence of henlock and yellow birch in the northern hardwood forest type. Both species regenerate well on windthrow hunnocks and where burns severely reduce inpenetrable, desiccating litter layers or expose nineral soil (Godnan and Lancaster 1990, and Erdnann 1990). The Mined Pine rerest Type Disturbance regines observed by the surveyors within the nixed pine forest type were fire, windthrow, windthrow with subsequent fire and insect induced nortality. Three to five events occurred per year in the forest type (Table 16). The transect size class distribution for the estinated disturbance events in the nixed pine forest type is shown in Table 20. As expected, fire was the doninant disturbance regine in this forest type, accounting for 70.0 percent of all disturbance events. Insect induced nortality accounted for 18.7 percent of all disturbances. Windthrow accounted for 11.3 percent of all disturbances, with only 1.3 percent of windfall events also associated with subsequent fire. Table 20. 103 lstinated total nunber of disturbance events for each size class and disturbance regine in the nixed pine forest type. Transect Size Wind/ Beaver Class (n) Fire Wind Fire Insect Flood Total 4 < 50 - - - - - 0 0.0 50-100 - - - - - 0 0.0 100-200 31 - - 8 - 39 48.8 200-400 14 - - - N/A 14 17.5 400-800 6 - - - N/A 6 7.5 800-1600 2 5 - 7 N/A 14 17.5 1600-3200 2 l - - N/A 3 3.8 3200-6400 1 2 1 - N/A 4 5.0 Totals 56 8 l 15 0 80 100 (4) 70.0 10.0 1.3 18.7 0.0 100 104 The disturbance frequency for fire in the nixed pine forest type was estinated to be between 2 to 4 events per year (Table 16). Most fires occurred in the 100 to 800 n size classes, with sons larger events being observed in the 800 to 6400 n size classes (Table 20). The return interval for fire was short, at 247 to 493 years (Table 17). Uhitney's (1986) estinates for fire return intervals in the pre-Buropean settlenent nixed pine forests of Rosconnon and Crawford Counties were between 129 and 258 years. Fires in ny study area therefore occurred with alnost half the frequency of fires on outwash plains in the northern lower peninsula. The estinated area of annual disturbance was between 56 and 111 ha per year, representing between 0.20 and 0.41 percent of the forest type per year (Table 18). This was the highest rate of any disturbance regine in the nixed pine forest type. Heinselnan’s (1973) classic study of fire in the nixed pine forests of Minnesota's Boundary Waters Canoe Area showed that 0.82 percent of the virgin forests burned on an annual basis. Meinselnan estinated a return interval of approxinately 100 to 200 years for nixed pine stands (1981a). Again, this indicates a rate of fire disturbance at least twice that of ny study area. However, it nust be recognized that Heinselnan’s estinate is based upon physical fire scar evidence over a 141 year period fron 1727 to 1868, which included five exceptional fire years. Exceptional fire years occurred in 1727, 1759, 1801, 1824 and 1864 (Heinselnan 1973). During these exceptional years 105 Beinselnan docunented that fires consuned areas ranging fron 8.3 to 44.1 percent of the virgin forest. The thirty year period fron 1824 to 1854 enconpassed only one exceptional fire year (1824), the least severe of all the exceptional years during the period of 1727 to 1868 (Heinselnan 1973). By the years 1845 to 1850, when the subdivision surveys of ny study area were conpleted, any evidence of fires that occurred in the 1824 exceptional fire year would be 20 to 25 years old, and rapidly fading if evident at all. Seral forest cover would be well developed after 20 to 25 years. The prinary evidence of large scale fire events in 1824 would be substantial areas of pole sized tinber, as typified by the pure, even-aged stands of jack pine on the Raco outwash plains. The 15 cn nean and nedian dianeters of these stands suggest that they could possibly have their origin after the 1824 exceptional fire year. Incidents of insect induced nortality in the nixed pine forest type occurred at a low rate of 1 or fewer events per year (Table 16). The estinated return interval was between and 451 and 901 years (Table 17). The area of annual disturbance was between 30 and 61 ha per year, representing 0.11 to 0.22 percent of the forest type on an annual basis (Table 18). Insect related disturbance events occurred in two classes, the 100-200 n and 800-1600 n size classes (Table 20). A likely causal agent of extensive areas of nortality in the nixed pine forest type is the jack pine budworn. Presently, snall outbreaks of jack pine budworn 106 occur frequently in the Lake States. They attack jack pine of all sizes, preferring trees with abundant nale flowers (Uilson 1977). They will also attack adjacent stands of red and white pine. Free the evidence presented, it appears that insect infestation caused infrequent, and noderate areas of tree nortality in the nixed pine type. Windthrow occurred with a frequency of less than 1 event every two years (Table 16). The return interval of windthrow alone was rather short, estinated at 284 to 568 years, with an area of annual disturbance between 49 and 97 ha per year (Tables 17 and 18). Whitney's (1986) return interval estinate for windfall in the virgin nixed pine forests in Bosconnon and Crawford Counties was nuch longer at 1587 to 3174 years. This indicates that windfall events were nuch nore extensive in ny study area, as shown in Figures 22 and 23. The area of annual disturbance represented 0.18 to 0.35 percent of the forest type on an annual basis. Windthrow events were quite large, with all events falling within the 800 to 6400 n size classes (Table 20). Thus, although windthrow events in the nixed pine forest type were not very frequent, they were typically large in extent. The return interval for windthrow with subsequent fire was between 1760 and 3521 years. Table 18 shows that the estinated area of annual disturbance of windthrow events with subsequent fire was between 8 and 16 ha per year, representing a snall proportion of the forest type area, 107 between 0.03 and 0.06 percent annually. Only one windthrow event with subsequent fire occurred in the nixed pine forest type, but it was in the 3200-6400 n length class (Table 20). Overall, disturbance affected 0.52 to 1.04 percent of the nixed pine forest on an annual basis. This was a noderate rate of disturbance, with 3 to 5 events occurring per year. Fire, windthrow and probable insect induced nortality were the doninant types of disturbance, with fire and windthrow affecting the greatest annual percentage of the forest type. Midbsized events between 100 and 1600 n in length were the nost connon, with larger events nuch less frequent in occurrence. The return intervals for fire (247- 493 yrs) and windthrow (284-569 yrs) were less than the naxinun potential lifespans of the two nost doninant tree species in the forest type. The naxinun potential lifespans for white and red pine are 450 years (Mendel and Snith 1990) and 300-400 years (Rudolf 1990) respectively. Thus, regeneration of tree species within the nixed pine forest type was potentially dependent upon disturbance. Fire and windthrow, rather than species recruitnent as in the northern hardwood forest type, had the capacity to regulate the conposition and structure of the nixed pine forest type. However, species recruitnent still played a role in the regulation of forest conposition. Major fire events scarified the soil by burning away-nuch of the organic natter covering nineral soil. Such scarification provided an excellent seedbed for the gernination and growth of white 108 and red pine via seed dispersal fron residual trees (Wendel and Snith 1990, Rudolf 1990). Conversely, windthrow favored the growth of seral species such as bigtooth aspen, trenbling aspen and white birch which have high potential asexual reproduction rates, and did not have to depend upon favorable seedbed conditions for sexual reproduction. The desk Pine Forest Type Disturbances noted by the surveyors in the jack pine forest type were fire, windfall with subsequent fire and insect induced nortality. Only 3 to 5 disturbance events occurred per year in the forest type (Table 16). The transect size class distribution for the estinated disturbance events in the jack pine forest type is shown in Table 21. As expected, fire was the doninant forn of disturbance within this forest type, accounting for a total of 81.9 percent of all disturbance events. Insect induced nortality and windfall with subsequent fire represented only 11.1 and 6.9 percent of all disturbance events respectively. Fire occurred with a frequency of 2 to 4 events per year in the jack pine forest type (Table 16). The estinated return interval for fire was 142 to 283 years (Table 17). This interval conpares favorably with Whitney's (1986) estinate of 83 to 167 years for Rosconnon and Crawford Counties. Meinselnan (1973) estinated a return interval of approxinately 50 to 100 years for jack pine stands in the Boundary waters Canoe Area in Minnesota. The estinated area 109 Table 21. Bstinated total nunber of disturbance events for each size class and disturbance regine in the jack pine forest type. Transect Size Wind/ Beaver Class (n) Fire Wind Fire Insect Flood Total 4 < 50 - - ' - - - - 0 0.0 50-100 47 - - - - 47 65.3 100-200 - - ‘- -' - O 0.0 200-400 - - 5 - N/A 5 6.9 400-800 6 - - 3 N/A 9 12.5 800-1600 6 - - 4 N/A 10 13.9 1600-3200 - - - 1 N/A 1 1.4 Totals 59 0 5 8 0 72 100 (4) 81.9 0.0 6.9 11.1 0.0 100 of annual disturbance was 34 to 68 ha per year, representing 0.35 to 0.70 percent of the forest type annually (Table 18). This was the second highest percentage of any disturbance regine in any forest type in the study area, indicating the inportance of fire in the jack pine forest type. The vast najority of fires in the jack pine forest were in the 50-100 n length class, with a significant nunber of events in the 400 to 1600 n size classes (Table 21). Figures 26 and 27 show that nost of the area affected by fire could be attributed to the larger size classes. My results show that large areas of the jack pine forest type burned on a regular basis, resulting in a very low return interval. The 15 on 110 . nean and nedian dianeters of jack pine, and the even-aged dianeter distribution of the forest type suggest that the fires in jack pine were stand-replacing in severity. My results support the hypothesis that fire dependent connunities burn nore frequently and with greater severity than do other connunities, because natural selection has forced evolution of physiological characteristics in their conponent vegetation that nake then inherently nore flannable (hutch 1970). Jack pine has evolved physiological attributes such as early naturation and persistent,‘ serotinous cones that enable it to thrive as a species on the xeric, fire prone sites of the Raco Plains. Insect induced nortality occurred at a rate of approxinately 1 event every two years within the jack pine forest type (Table 16). The jack pine budworn was probably the causal agent of this type of disturbance in the forest type. The estinated return interval for jack pine budworn infestation was only 218 to 435 years (Table 17). The area of annual disturbance was between 22 and 44 ha per year, or at the high rate of 0.23 to 0.46 percent of the forest type annually (Table 18). Table 20 shows that all of the budworn disturbance events occurred in the largest size classes (400 to 3200 n in length). The results indicate that although the frequency of jack pine budworn outbreaks was relatively low, the events were typically large in scale. ‘I have no nethod of deternining whether areas of dead jack pine were positively correlated with the frequency of fire within the 111 jack pine forest type. However, I hypothesize that extensive patches of stressed, dying or dead jack pine would have dranatically increased the fuel load in large areas of the jack pine forest type, increasing the intensity and probably the frequency of fire within the type. Windthrow with subsequent fire'occurred at a very infrequent rate (<1 event every 3 years), and affected only noderately sized areas (between 4 and 9 ha/yr) and a very low percentage of the jack pine forest type annually (0.04 to 0.09 percent). The few windfall events nay have served to increase fuel loads in snall pockets, and nay have had an effect on fire frequency. However, I have concluded that windfall with subsequent fire was a relatively ninor disturbance nechanisn in the jack pine forest type. Overall, disturbance inpacted 0.62 to 1.25 percent of the jack pine forest on an annual basis. (This was a high rate of disturbance. Jack pine equaled the nixed pine forest type for the lowest frequency of disturbance of any forest type, at 3 to 5 events per year. Fire and probable insect induced nortality were the doninant disturbance events. Snell-sized events in the 50-100 n length class were the nost cannon, with larger fire and insect infestations nuch less frequently recorded. The return interval for fire (142-283 yrs), which was the nost doninant node of disturbance, was less than the 185-230 year naxinun potential lifespan of jack pine (Rudolph and Laidly 1990). Thus, regeneration of the jack pine forest type appears to 112 have been dependent upon disturbance. Fire drove the life cycle of jack pine, and was the doninant regulator of the conposition and structure of the forest. Major destructive fire events opened the persistent, serotinous cones of jack pine, and exposed the nineral soil to provide a seedbed for copious jack pine reproduction. Fire allowed jack pine to persist and doninate on nore xeric sites that were predisposed to fire. The Mixed Conifer/Deciduous Upland Forest Type Disturbance regines noted by surveyors in the nixed conifer/deciduous upland forest type were fire, windthrow, windthrow with subsequent fire, insect induced nortality and beaver flooding. Twenty-six to fifty-one disturbance events occurred within the forest type per year (Table 16). The transect size class distribution for the estinated disturbance events in the nixed upland forest type is shown in Table 22. Windthrow was the doninant forn of disturbance within this forest type, accounting for a total of 49.6 percent of all disturbance events. Beaver floodings and fire were also significant disturbance regines within the nixed upland forest type, representing 26.8 and 17.0 percent of all disturbance events, respectively. Insect induced nortality accounted for only 6.6 percent of all disturbance events. Snaller size classes of disturbance were doninant within the forest type, with lengths of less than 400 n accounting for 87.9 percent of all events. Table 22. 113 each size class and disturbance regine in the nixed conifer/deciduous upland forest type. Bstinated total nunber of disturbance events for Transect Size Wind/ Beaver Class (n) Fire Wind Fire. Insect Flood Total 4 < 50 49 30 - - 83 162 21.0 50-100 14 36 50 48 101 249 32.3 100-200 30 60 65 - 22 177 23.0 200-400 23 29 37. - N/A 89 11.6 400-800 12 19 19 3 N/A 53 6.9 800-1600 3 17 15 - M/A 35 4.6 1600-3200 - 4 - - N/A 4 0.5 3200-6400 - 1 - - N/A 1 0.1 Totals 131 196 186. 51 206 770 100 4 17.0 25.4 24.2 6.6 26.8 100 114 Windthrow occurred with a total frequency of between 12 and 25 events per year, the largest frequency value of disturbance in any forest type (Table 16). Fire occurred subsequent to windthrow in about half of these events. The return intervals of windthrow were anong the lowest of any disturbance regine in any forest type. The return interval for windthrow was only 145 to 291 years (Table 17). The area of annual disturbance was between 153 and 307 be per year, and represented a very high 0.34 to 0.69 annual percent of the forest type area (Table 18). The return interval for windthrow followed by fire was slightly greater at 180 to 359 years, with an area of annual disturbance between 124 and 248 be per year. The areas represented slightly lower annual values of 0.28 to 0.56 percent of the forest type. Windthrow events were distributed in the snaller size classes, with nost occurring in the 50 to 400 n classes (Table 22). A few large windthrow events occurred in the larger (1600 to 6400 n) size classes. Spruce and eastern henlock were doninant in density in this forest type. Both species, but especially henlock, are shallow- rooted and subject to windthrow. Thus, nany snaller to noderately sized windthrow events occurred within the nixed upland forest type, with about half of the events subsequently burning. Sons windfall events were quite large. Beaver floodings were also very frequent in the nixed upland forest type. Although beaver floodings are 115 restricted to riparian areas, they occurred in the upland type with a frequency of 7 to 14 events per year. This frequency was the highest of any forest type. The estinated area of annual disturbance for beaver floodings was between 25 and 49 ha per year, representing between 0.05 and 0.11 percent of the forest type area on an annual basis (Table 18). Mot surprisingly, nost beaver floodings were less than 100 n in length (Table 22). White birch, aspen and red naple are well represented within the forest type, and are also preferred for food and construction naterial by the beaver. This fact, as well as a greater proportion of strean habitat within the forest type, nay explain the high incidence of beaver flooding found here. Fire (not associated with windthrow) also occurred quite frequently within the nixed upland forest type, at a rate of 4 to 9 events per year (Table 16). The return interval for fire was 429 to 858 years. The area of annual disturbance was between 52 and 104 be per year, or 0.12 to 0.23 percent of the area on an annual basis (Table 18). Most of the fire events occurred in size classes less than 400 n in length, but a few were as large as 800-1600 n in length (Table 22). Insect induced nortality occurred at a low rate of 2 to 3 events per year in the nixed upland forest type, and affected only 0.02 to 0.05 percent of the forest type annually.' The causal agent of tinber nortality in the forest type nay have been the spruce bud worn, since spruce 116 was a doninant in both density and basal area within the type. Tanarack and jack pine were also present within the forest type, so the eastern larch beetle and the jack pine budworn nay have been contributing factors to snall pockets of tree nortality. The nixed conifer/deciduous upland forest type had the highest frequency of disturbance of any forest type, at 26 to 51 events per year. Disturbance affected 0.81 to 1.64 percent of the forest on an annual basis, the highest percentage of any forest type. Windthrow, with or without subsequent fire, and beaver floodings were the doninant types of disturbance. Snaller sized events, less than 400 n in length were the nost cannon, with larger events occurring with rapidly decreasing frequency. The return interval for windthrow, between 145 and 359 years, was less than the naxinun potential lifespans of the doninant tree species. The naxinun potential lifespans for henlock, white spruce and white pine are 400-990 years, 250-300 years (Mienstaedt and Zasada 1990) and 450 years respectively. Thus, windthrow had the potential to regulate the conposition and structure of the nixed upland forest type. Regeneration of tree species within the nixed upland forest type was potentially dependent upon disturbance. Windthrow events which were not associated with fire probably favored the reproduction of spruce, aspen and white birch. Exposed nineral soil fron windthrows provide one of the best seed beds for white spruce regeneration and can result in 117 stocking levels approaching 100 percent (Mienstaedt and 2asada 1990). As discussed previously, aspen and white birch have a conpetitive advantage due to their asexual reproduction capacities. When fire followed windthrow events the reproduction of white pine and henlock and spruce were probably favored on the expOsed nineral soil,' especially when subsequent noisture conditions were favorable for seed gernination and seedling growth. The Mixed conifer/Deciduous Lowland Forest Type Disturbance regines observed by the surveyors within the nixed conifer/deciduous lowland forest type were fire, windthrow, windthrow with subsequent fire and beaver floodings. Six to twelve disturbance events occurred within the forest type per year (Table 16). The transect size class distribution for the estinated disturbance events in the nixed lowland forest type is shown in Table 23. Windthrow and fire were the doninant nodes of disturbance within the forest type, accounting for 59.0 and 32.4 percent of all disturbances respectively. Windthrows associated with fire accounted for 48.0 percent of all disturbance events. Beaver floodings represented the renaining 8.7 percent of all disturbance events. Disturbance events with lengths between 50 and 800 n accounted for 90.2 percent of all events. The 50-100 n size class was doninant, representing nearly half of all events. Windthrow events occurred at a rate of 3 to 7 events Thbl. 23. 118 each size class and disturbance regine in the nixed conifer/deciduous lowland forest type. Bstinated total nunber of disturbance events for Transect . Size Wind/ Beaver . Class (n) Fire Wind Fire Insect Flood Total 4 < 50 - - - - - 0 0.0 50-100 31 16 23 - 15 85 49.1 100-200 - - 37 - - 37 21.4 200-400 14 '- 4 - N/A 18 10.4 400-000 7 - 9 - M/A 16 9.3 800-1600 4 2 6 - N/A ‘ 12 6.9 1600-3200 - - 3 - N/A 3 1.7 3200-6400 - 1 - - N/A 1 0.6 6400-12800 - - l - N/A 1 0.6 Totals 56 19 83 0 15 173 100 4 32.4 11.0 48.0 0.0 8.7 100 119 per year, with fire following windthrow in every 3'to 5 events (Table 16). The return interval for windthrow alone was between 633 and 1266 years (Table 17). The area of annual disturbance was between 17.and 35 be per year, representing 0.08 to 0.16 percent of the forest type on an annual basis (Table 18). With fire subsequent to windthrow the return interval was an extrenely low 122 to 243 years (Table 17). The area of annual disturbance was between 90 and 180 ha per year, representing a substantial 0.41 to 0.82 percent of the forest type. All together, windthrow affected between 0.49 and 0.98 percent of the nixed lowland forest type annually. Mbst (90.2 percent) disturbance events occurred in the 50 to 800 n size classes, with the size class distribution rapidly decreasing above 800 n (Table 23). The largest single length of disturbance of any regine in any forest type, was a windthrow and burn in the 6400-12800 n size class. Spruce, northern white cedar and eastern henlock are significantly represented in density and basal area within the forest type. These species are also shallow rooted and prone to windthrow, and nay account for nany of the windthrow events within the forest type. Apparently fire followed windthrow in nost windthrow events. These events occurred in snall to nediun sized patches throughout the nixed lowland forest type, and affected noderately large areas. Fire events not directly associated with windthrow occurred with a frequency of 2 to 4 events per year in the 120 nixed lowland forest type (Table 16). The estinated return interval for fire was a relatively short 292 to 585 years (Table 17). The area of annual disturbance was between 38 and 75 ha per year, representing 0.17 to 0.34 percent of the forest type on an annual basis (Table 18). Alnost half of all burn events occurred in the 50-100 n size cless, with significant-nunbers also occurring in the 200 to 1600 n size classes (Table 23). Thus, burns occurred regularly in the nixed lowland forest type, but were noderate in size and probably sonewhat patchy in distribution. Beaver floodings occurred at a rate of 1 per year in the nixed lowland forest type (Table 16). The estinated area of annual disturbance was a very low 3 to 6 ha per year, representing an equally low 0.01 to 0.03 percent of the nixed lowland forest type (Table 14). All flooding events occurred in the 50-100 n size class (Table 23), although this estinate is-based upon only one observed event. Overall, disturbance affected 0.67 to 1.35 percent of the nixed conifer/deciduous lowland forest on an annual basis. Windthrow and fire were the doninant types of disturbance. Mid-sized events between 50 and 1600 n.in length were the nost cannon, with large to very large events nuch less frequent in occurrence. .Windthrow associated with subsequent fire accounted for alnost half of all disturbance events, and the annual area of disturbance was nore than twice that of the other types of disturbance in the type. 121 The return interval for windthrow with subsequent fire was between 122 and 243 years, and the interval for fire was between 292 and 585 years. The naxinun potential lifespan of black spruce, tanarack, white pine and northern white cedar are 200-280 years (Vireck and Johnston 1990), 150-240 years (Johnston 1990b), 450 years and 400 years (Johnston 1990a) respectively. The return intervals are less than or equal to the naxinun potential lifespan of the four nost doninant tree species in the forest type. Thus, regeneration of tree species within the nixed conifer/deciduous lowland forest type was probably dependent upon disturbance. Windthrow and fire had the potential to regulate the conposition and structure of the nixed lowland forest type. Conplete renoval of organic surface layers and exposure of nineral soil by windthrow and fire favors the successful regeneration of black spruce, tanarack, white pine and cedar (Viereck and Johnston 1990, Johnston 1990b, Wendel and Snith 1990 and Johnston 1990). Thus, windthrow and fire were probably doninant factors in the perpetuation of forested wetlands. The Mixed Conifer swanp Forest Type Disturbance regines observed by the surveyors within the nixed conifer swanp forest type were fire, windthrow, windthrow with subsequent fire, insect induced nortality and beaver floodings. Twenty-five to fifty disturbance events occurred per year in the forest type (Table 16). The 122 doninant node of disturbance in the type was windfall, accounting for 48.0 percent of all events. Fire, insect induced nortality, and beaver floodings were approxinately equal in representation, accounting for 18.6, 16.2 and 17.2 percent of all events respectively. ‘The transect size class distribution for the estinated disturbance events in the nixed conifer swanp forest type is shown in Table 24. The snaller size classes less than 800 n were doninant in the forest type, accounting for 89.6 percent of all events. Windthrow occurred with a frequency of 12 to 24 events per year (Table 16). Windfall with subsequent fire accounted for two-thirds of these events. The estinated return interval for windfall alone was a short 231 to 463 years (Table 17). This return interval is alnost one-sixth of the 1316 to 2632 year interval calculated by Whitney (1986) for the pre-Buropean settlenent swanp conifer forests of Rosconnon and Crawford Counties in northern lower Michigan. The return interval for windfall with subsequent fire was longer at 314 to 628 years. This indicates that windfall was nuch nore prevalent in the wetlands of Chippewa county. The area of annual disturbance for windfall was large, at 224 to 448 ha per year, representing 0.22 to 0.43 percent of the nixed conifer swanp forest type on an annual basis. This was the highest area of annual disturbance for any regine in any forest type. For windthrow with subsequent fire the area of annual disturbance was snaller, between 165 and 329 ha per year, or 0.16 to 0.32 percent Table 24. 123 Bstinated total nunber of disturbance events for each size class and disturbance regine in the nixed conifer swanp forest type. Transect - Size Wind/ Beaver Class (n) Fire Wind Fire Insect Flood Total 4 < 50 - - 35 71 26 132 17.7 50-100 18 14 53 19 94 198 26.5 100-200 62 22' 53 - 8 145 19.4 200-400 34 36 34 7 N/A 111 14.9 400-800 13 22 38 10 N/A 83 11.1 800-1600 11 28 10 12 M/A 61 8.2 1600-3200 1' 6 4 2 N/A 13 1.7 3200-6400 - 3 - - N/A 3 0.4 Totals 139 131 227 121 128 746 100 4 18.6 17.6 30.4 16.2 17.2 100 124 annually. Windthrow events occurred in every size class (Table 23). Most windfall events fell within the 100 to 1600 n size classes, with sons events up to 3200-6400 n in length (Table 24). Windfall events with subsequent fire were well distributed throughout all size classes less than 800 n, with the largest events occurring in the 1600-3200 n size class. It appears that fire followed windfall with a high frequency, affecting slightly less area and occurring in sanewhat snaller patches than windfalls not associated with fire. Fire (not associated with windthrow) occurred at a rate of 5 to 9 events per year in the nixed swanp forest type (Table 16). The estinated return interval was between 490 and 979 years (Table 17). The estinated return interval of fire in the virgin conifer swanp forests of Rosconnon and Crawford Counties was 2959 to 5917 years (Whitney 1986), again indicating a nuch greater inpact of fire in ny study area. The area of annual disturbance was 106 to 211 ha per year, representing 0.10 to 0.20 percent of the forest type per year (Table 18). Alnost all of the fire events occurred in the 50 to 1600 n size classes (Table 24).~ The one exception was an event in the 1600-3200 n size class. Upland islands of jack and red pine often occurred within nixed conifer swanps in the central portion of the study area. In the late sunner and early autunn the nixed conifer swanps and these upland islands can becone very fire prone, especially in years characterized by drought. It is 125 therefore very plausible that relatively large and frequent fires occurred within the nixed conifer swanp forest type, affecting both snall and nid-sized patches of the forest. Disturbance by beaver flooding in the nixed conifer swanp forest type occurred at a rate of 4 to 9 events per year (Table 16). The estinated area of annual disturbance was between 24 and 48 ha per year, representing a low 0.02 to 0.05 percent of the forest type annually (Table 19). All events were less than 200 n in length, with nost occurring in the 50-100 n size class (Table 24). There were alnost half as nany beaver flooding events as occurred in the nixed conifer/deciduous upland forest type. This can probably be attributed to the lesser density of preferred food and building naterial tree species (white birch, aspen and red naple) in the nixed conifer swanp forest type, rather than a lack of suitable riparian habitat. The incidence of insect induced nortality in the nixed conifer swanp forest type occurred at a noderately high rate of 4 to 8 events per year (Table 16). 'The spruce budworn and the eastern larch beetle were the probable causal agents of extensive tree nortality within the forest type. The estinated return interval for insect infestation was between 835 and 1672 years (Table 17).. Snaller pockets of infestation were nost cannon and nay have been the result of relatively ninor insect outbreaks, or nay have sinply occurred in areas with high densities of spruce or tanarack. The occasional larger areas of tree nortality are likely the 126 result of sore severe insect infestations. In general, insect infestations occurred witha noderately high frequency. Although the area of annual disturbance was noderate in size, the total annual percentage of the nixed conifer swanp forest type that was affected by insect infestation was rather low. With the exception of the nixed conifer/deciduous upland forest type, the nixed swanp conifer forest type had the second highest rate of disturbance, at 25 to 50 events per year. However, disturbance inpacted only a noderate 0.56 to 1.12 percent of the nixed conifer swanp forest on an annual basis. Windthrow and fire were the doninant disturbance regines, with fire occurring subsequent to windthrow in two-thirds of all windthrow events. Relatively snall-sized events less than 800 n in length were the nost cannon, with larger events progressively less frequent in occurrence. With the exception of northern white cedar, the return intervals for disturbance in the nixed conifer swanp forest type were greater than the naxinun potential lifespans of the doninant species in the forest type. The lowest return interval was for windthrow at 231 to 463 years. The return interval for windthrow with subsequent fire was longer (314 to 628 years). The naxinun potential lifespans for black spruce, northern white cedar and tanarack are 200-280, 400 and 150-240 years respectively. 127 Forest Conparisoas - Fast and Present Various neans of canparison nay be used to exanine the differences between the pre-Buropean settlenent forests and the forests that exist today in western Chippewa County. I will discuss differences in terns of area coverage, conposition, structure and disturbance. I have classified the pre-Buropean settlenent forests into six distinct forest ecosysten types. These were the northern hardwood, nixed pine, jack pine, nixed conifer/deciduous upland, nixed conifer/deciduous lowland and nixed conifer swanp forest types discussed above. Each of these ecosystens differed in terns of conposition, structure and disturbance frequency. Although beyond the scope of ny study, I believe that other internal functions such as carbon and nutrient cycling were also undoubtedly different anong the six different ecosystens. The variation observed anong the six ecosystens probably stenned fron inherent differences in the edaphic characteristics of the soils on which the ecosystens occurred. Bdaphic characteristics would at a nininun include substrate conposition, acidity, cation exchange capacity, nutrient availability and water retention capacity. Further exanination of the study area for correlations between the forest types and edaphic soil characteristics would undoubtedly prove interesting. Forest Conpesition Changes in land use, forest conposition and the inpacts 128 of hunan nanipulation upon the landscape are readily apparent by conparisons of the area covered by each forest ecosysten. I have canpared the areas of ny six forest types to the land use areas contained in the 1978 MIRIS survey, which was based upon aerial photography. Sone error in scale is present between ny data and the MIRIS survey data, as is evident fron the difference in total area reported. This nisregistration error nay have been due to distortions in the aerial photos. A nore serious source of error in the MIRIS survey data is inaccuracies concerning photo- interpretation of the different land use types. The Land and Water Managenent division of the Michigan DWR has indicated that wetland and lowland forest types were often very difficult to distinguish fran upland forest types. Consequently, the reported area coverages can only be considered as rough estinates, with an unknown degree of error. For lack of a better data source, I have proceeded with using the MIRIS data far ny analysis. I acknowledge that sone conparisons nay be rough.in nature, but general trends are still apparent. I have nade conparisons of conposition and area by conparing the percentage of area covered, thus nininizing the error associated with scale. Pre-European settlenent forest type areas (ninus the area of lakes in the study area) are shown in Table 25. The 1978 MIRIS cover type areas are shown in Table 26. The inpact of hunan nanipulation upon the landscape is readily apparent. Nearly 2 percent of the study area is now 129 Table 25. Pre-European settlenent forest type areas. MIRIS . Code Forest Type Area (ha) Percent 411 Worthern Hardwoods 65,590 24.21 421 Mixed Pine 27,393 10.11 4213 Jack Pine 9,622 3.55 37,015 13.66 414 Mixed Conifer/ 21,917 8.09 Deciduous Lowland 422 Mixed Conifer/ 44,245 16.33 Deciduous Upland 423 Mixed Conifer Swanp 102,112 37.70 SUI ' 270,879 ' 100 130 Table 26. MIRIS (1978) cover type areas. Miris ' Percent Code Forest Type Area (ha) Percent Change 31 Open Herbaceous 3,405 1.31 +1.31 32 Open Shrub 6,836 2.57 +2.57 100 Urban. 4,990 1.00 +1.00 200 ' Agriculture 26,153 9.84 +9.84 411 Northern Hardwoods 70,929 26.70 +2.49 413 -Aspan/White Birch 13,001 4.89 +4.89 414 Lowland Hardwood 9,815 3.69 -4.40 421 Pine 50,538 19.02 +5.36 422 Other Upland Conifers 17,362 6.54' -9.79 429 Christnas Tree 36 0.01 +0.01 ' Plantation 423 Lowland Conifers 33,072 612 Shrub Wetlands 14,971 621 ' Aquatic Bed Wetlands 823 622 Mnergent Wetlands 13,634 62,500 23.53 -14.17 Sun 265,653 100 0.0 131 classified as urban, whereas only 0.04 percent of the study area was occupied by indian villages in the 1840's. Alnost 10 percent of the study area is now under conventional agriculture, or in Christnas tree plantations. An additional 4 percent is classified as open herbaceous or shrub land. An aspen/white birch association now covers alnost 5 percent of the landscape.. These early successional species becane nore predaninant after the widespread exploitation of tinber in the study area, particularly in the pine and nixed upland ecosystens. Portions of the study area are naintained in this early successional state by intensive nanagenent of jack pine and aspen for pulpwood. There has been an estinated 14 percent loss of wetlands in the study area (where wetlands are defined by ny nixed conifer swanp and the MIRIS cover types), and over a 4 percent estinated loss in the nixed conifer/deciduous lowland forest type. It is estinated that alnost 10 percent of the original nixed conifer upland forest type has been lost. Sane of the swanps, nixed lowlands and nixed uplands nay have been converted to urban areas and agriculture. The percentage of northern hardwoods actually increased by over 2 percent, possibly through succession on sone of the better sites previously occupied by pine or nixed uplands that were not extensively logged. There has also been an increase of over 5 percent in the area covered by pine. It is likely that sons of the nixed upland forest type has been converted to pine plantations. Large areas were aggressively 132 afforested with pine in the 1930’s, and continue to be nenaged as pine plantations by the U.S. Forest Service and the Michigan Departnent of Natural Resources. Since the MIRIS survey did not differentiate between the species of pine, I cannot deternine what changes have occurred in the percentage of area covered by nixed pine and jack pine. Other than the genesis of the aspen/white birch forest type, the nest striking change in the conposition of the forest within the study area is the pronounced decline in the doninance of eastern henlock. Menlock has been severely reduced-fran its position as a doninant tree species in the northern Lake States, and now occupies a nere 0.5 percent of the landscape (Bckstein 1980). I do not have definitive data on present day densities of henlock in the study area, but fron extensive personal observation I know that henlock is not present at anywhere near the densities that I have calculated. I have found that henlock was fornerly a doninant (and possibly Sh! doninant) species in both the northern hardwood and nixed conifer/deciduous upland forest types of ny study area (Tables 9 and 11). There are several reasons for the decline of henlock. Menlock was logged heavily in the region both for lunber and for its bark, which contained the tannin used for nany years. in the leather tanning process (Marananski 1989). This. heavy cutting and the frequent highly destructive fires that followed in the wake of lunbering severely reduced residual. seed sources. Menlock also has rather exacting gernination 133 requirenents. Menlock seeds and seedlings are highly susceptible to desiccation, and do not survive without forest cover that is at least pole sized in growth (Jordan and Sharp 1967). As discussed previously, fire that exposes nineral soil can enhance henlock regeneration. Hough and Forbes (1943) described sane henlock stands that were of clear fire origin. Henlock regeneration also occurs on decaying nurse logs and stunps, and upon windthrow hunnocks where warner surface tenperatures and better noisture regines exist (Godnan and Lancaster 1990). However, the second-growth forests of today have not yet natured to the point where large anounts of fallen and decaying naterial are again present upon the forest floor, and present day fire suppression policies severely restrict the scale of fire in all forest types. Mladenoff and Sterns (1993) hypothesised that fire suppression has created alnost pure deciduous forests, which have resulted in the replacenent of conifer litter nutrient cycles by deciduous litter nutrient cycles that discourage the regeneration of henlock, and provide a positive feedback that favors the regeneration of deciduous species. Mladenoff and Sterns also suggest that browsing by the current high population densities of white tailed deer (ngggilgns yirginignns) is not the critical factor suppressing the regeneration of henlock. They theorize that regeneration is prevented by the previously discussed inhibiting factors, and therefore renders the issue of browsing noot. 134 _ I believe that the lack of residual seed sources is a prinary factor inhibiting the regeneration of henlock, because without seed sources regeneration cannot possibly occur. Where significant seed sources do exist the above discussed problens affecting regeneration then becone a concern. Perhaps when second growth forests nature to the point where natural nortality and windthrow again produce significant quantities of down and decaying naterial and windthrow hunnocks on the forest floor, within range of residual henlock seed sources, significant henlock regeneration will occur. Managenent practices which encourage dawn and decaying naterial, the use of prescribed burns near residual seed sources and downward pressure upon deer populations can assist henlock regeneration. Regardless, it will take hundreds of years (if it is even possible at all) for henlock to regain the predoninance that it once possessed in the northern hardwood and nixed upland forests of ny study area. Forest structure Changes in forest structure are also apparent. Table 27 shows the estinated basal area, density and nean DBH for the pre-Buropean and present day forests of the study area. The basal area and nean DBH data are fran 1994 U.S. Forest Service conpartnent records. Bstinates of density were not readily available, and were calculated as described in the nethods section above. The resultant densities for the Table 27. 135 Bstinates of basal area, density and nean DBH for pre-Buropean settlenent and present day forest types.1 Basal Bstinated Mean Forest .Type Area (nz/ha) Density ('rpn) DBH (cn) 18§Q___1221 18§9___1221 1§§Q___1221 Northern 35 21 340 563 33 23 Hardwood (154)2 (92) (141) (220) (13) (9) Mixed Pine 0 -15 01 570 31 10 (36) (67) (33) (231) (12) (7) Jack Pine 0 15 326 054 10 15 (33) (64) (132) (346) (7) (6) Aspen/ N/A 16 N/A 601 N/A 10 White Birch (68) (276) ' (7) Mixed Conifer 21 -17 287 578 28 18 Upland (92) (75) (116) (324) (11) (7) Mixed Conifer 15 21 269 520 ' 23 V 23 Lowland (66) (92) (109) (214) (9) (9) Mixed Conifer. 14 17 314. 963 20 15 Swanp . (60) (74) (127) (390) (0) (6) ‘ Present day estinates are based upon 1994 U. S. Forest Service Conpartnent Records. 2 Nunbers in parenthesesare basal area in ftz/ac, density in trees/ac and DBH in inches. 136 northern hardwood and jack pine forest types canpared favorably with figures achieved using stocking curves (Tubbs 1977, Benzie 1977), so I believe that the procedure used to calculate the densities is accurate. The structural data of the present forests in Table 27 was canpared to the data of the pre-Buropean settlenent forests in Table 8. The density of every forest type is significantly higher in the present day forests of the study area. The nean DBH is significantly less in every type except the jack pine and nixed conifer lowland types, which are only slightly less. The estinates of basal area per hectare were significantly higher in the northern hardwood and nixed conifer upland types of the pre-Buropean settlenent forests. The higher densities and lower basal areas and dianeters in the present day northern hardwood and nixed upland types are indicative of forests that are still relatively young. This would be expected in the second growth forests of the study area, where all stands are probably nuch less than 100 years old. Sonewhat higher basal areas were found in the nixed pine, jack pine, nixed conifer lowland and nixed conifer swanp types of the present-day forests of the study area. These snall increases in basal area can be attributed to the trenendous increases in density and associated snaller dianeters within these forest types. Few truly nixed pine stands exist in the present day forests of the study area. Most white pine, and particularly red and jack pine stands are intensively 137 nanaged in plantations, where densities far exceed even the dense pre-Buropean settlenent jack pine stands. Although the nixed conifer lowland and nixed conifer swanp forest types are not nearly as intensively nanaged as the pine types, high natural densities have occurred in the regenerated stands. In contrast, the pre-Buropean settlenent forests of the study region were characterized by large and doninant trees of several species, relatiVely low tree densities associated with relatively high basal areas, probable nulti-layered canopies, and widespread windthrow events which created dead snags, large fallen logs, windfall hunnocks and variably sized pockets of seral species. In the northern hardwood forests of the study area seral species such as aspen, oak, white birch and white pine nade up only 6.9 percent of the bearing trees selected. All of these characteristics are rare or lacking in young second growth forests, and they can be considered indicators of old growth forests. Disturbance Megines Disturbance regines are very different in the present day forests of the study area. The pre-Buropean settlenent forests of ny study area were disturbance driven and dependent ecosystens. Disturbance by fire, windthrow, insect related nortality and beaver flooding had profound inpacts on both the conposition and structure of the pre- luropean forest. Disturbance does not have nearly the sane f 138 inpact today. I Since the 1930's wildfire has been rigorously suppressed in the study area. Sane controlled burning has been utilized in recent years, to assist in the regeneration of jack pine for instance. However, the widespread and catastrophic fires that frequently occurred, particularly in the pine forest types, are likely never to be experienced again. Interests concerned with the protection of private property have greater political weight than forest professionals in deternining fire suppression policies. Bornann and Likens (1979) reported the present-day incidence of fire in both the eastern and western halves of the Hiawatha Mational Forest. On a one-nillion acre (405,000 ha) basis, an average of 19.5 fire events occurred per year. Of these, 2.5 events were caused by lightning, and 17 events were caused by hunans. .Weighting the total nunber of events, to take the acreage of the study area into account, yielded a present day average of 13 fire events per year in the study area. This is significantly less than the estinated 16 to 32 fire events and 19 to 37 windthrow and fire events per year for all pro-European settlenent forest ecosystens conbined (Table 16). Furthernore, the average area burned in the present-day forests is only 65 ha per year. The estinated area of annual disturbance for fire in all pre-Muropean settlenent forest types conbined was between 296 and 591 ha per year. For windthrow with subsequent fire the area of annual disturbance was even 139 greater at 392 to 783 ha per year (Table 18). Thus, fire plays an insignificant role in deternining the conposition and structure of the present day forests of the study area, as canpared to the forests of the 1840's. Managenent practices of forestry professionals have taken the place of fire in regulating the conposition and structure of the present day forests. I believe that the exclusion of fire has altered nany facets of the original forests of the study area, particularly the conposition of the forest. Today's white and red pine plantations on the better sites of the Raco outwash plains often have deciduous canponents in their understory. Bigtooth aspen and red naple are present in jack and red pine stands in densities that could not have existed in the virgin forests (Palik and Pregitzer 1992). I believe that the nixed conifer/deciduous lowland forest type has a nuch greater relative density of hardwood species in its present day conposition, hence it is referred to in the MIRIS codes as lowland hardwoods. In the pre-Buropean settlenent forests these deciduous species, and seedling and pole sized conifers as well, were periodically destroyed by fire. Fire reduced conpetition by deciduous species in forest types daninated by conifers, or essentially every type but the northern hardwoods. Such a regine favored the survival and growth of tree species that were adapted to periodic disturbance by fire, such as the pines, henlock, the spruces, cedar and tanarack. Data on the present day frequency and scale of 140 windthrow events in the study area'are not available. However, an excellent study of natural windthrow patterns in the henlock-hardwood preserves of the Porcupine Mountains, Sylvania Wilderness Area and the Huron Mountains in western upper Michigan was conducted by Frelich and Loriner (1991). Their estinates of return intervals range free as low as 69 years for events with nore than 10 percent canopy renoval, to a high of 1920 years for events with greater than 60 percent canopy renoval. This return interval is within that calculated for the virgin northern hardwood forests of ny study area (1387-2778 years). It nust be enphasized that these results apply to virgin, old growth northern hardwood forests. The relatively young second growth forests currently in ny study area do not have the sane structure as old growth forests, and I suspect that they therefore are not inpacted by the sane intensity of windthrow disturbance. Fran personal observation, I certainly have not seen any evidence of windthrow events anywhere near the fraquency, size and scale of those noted by the surveyors in the virgin forests of ny study area. Windfall is likely a noderate disturbance regine in the present day forests of the study area. However, I believe that the inpact of windthrow will increase as nore forest area natures and begins to develop old growth structural characteristics. A field study to deternine the current extent of windthrow in ny study area would undoubtedly be very useful in assessing the current inpact of windthrow upon the conposition and structure of 141 the forest types of the area. Despite the best efforts of forest nanagers, insect related nortality renains a doninant influence upon forest conposition and structure, especially in conifer daninated types. It is very likely that insects have a greater role and have nore inpact (relative to other disturbance regines) in regulating the conposition and structure of today's forests, than they did in the virgin forests of the study area. There are several reasons why I believe this is true. The 5.4 percent increase in the area.covered by pine provides nore forest area that can potentially be inpacted by insect outbreaks that favor such tree species. The predoninantly pure conposition, and unnaturally dense structure of nany pine stands provide few effective buffers to slow.and disrupt the dispersal of insects. In the case of jack pine, current nanagenent practices exclude fire with the objective of artificially extending the lifespan so as to produce nerchantable tinber. Such stands are nuch nore dense and older than historic stands and therefore nore susceptible to stresses such as those induced by insect infestation. Consequently, when such infestations occur they tend to be severe and devastating in scale. In 1991 the jack pine budworn defoliated approxinately 20,250 ha in the Upper Peninsula of Michigan (USDA 1993). The Raco ' outwash plains of ny study area has, in a two year period since 1991, experienced extensive jack pine bud worn defoliation on over 6,480 ha of forest land. Typical 142 nortality rates are between 10 and 40 percent, and will. result in 648 to 2592 ha of dead tinber. This is a rate of between 324 and 1296 ha per year, far exceeding the estinated 55-109 be per year area of annual disturbance calculated for the pre-Buropean settlenent jack pine forests of the study area. The incidence of insect related nortality in the present-day lowland hardwoods and lowland conifers nay not be as extensive as in the virgin forests due to the severe reduction of their area coverage. I know of no data regarding the current inpact of beaver floodings upon ny study area. The genesis of the aspen/paper birch forest type nay provide nore suitable habitat for beaver. The decline of the fur industry has sinultaneously reduced the negative pressure that for so nany years repressed the populations of beaver. Given these facts, I suspect that the incidence of beaver flooding in the study area is increasing. A field study to deternine the current extent of beaver floodings in the study area would be very useful in assessing the current inpact of flooding upon the conposition and structure of the forest types of the area. IUIIII! IUD OOICLUBIOUI It is inportant to realize that this study represents a nere snapshot in tine. Despite this fact, zone relevant conclusions regarding the conposition, structure and disturbance dynanics of the study area nay still be drawn. I believe that the overall structure of the pre-European forests of western Chippewa County fit the shifting-nosaic steady state nodel of Bornann and Likens (1979). The landscape was essentially a vast array of irregular patches, conposed of different successional stages and tree species associations of different age and size classes. The forest ecosystens were variable in scale, and also contributed to the heterogeneity of the landscape. They were also variable in conposition and structure, probably being a conpilation of uneven-aged and even-aged stands. The jack pine ecosysten probably had an even-aged structure. The conposition and structure of the forests were driven by disturbance. Disturbance regines in the study area were fire, windthrow, insect related nortality and beaver floodings. The northern hardwood ecosysten was a windthrow- dependent systen. The nixed pine, and jack pine ecosystens were fire dependent. The nixed conifer/deciduous upland, nixed conifer/deciduous lowland and nixed conifer swanp 143 144 ecosystens were windthrow and fire dependent systens. The northern hardwood and nixed conifer swanp ecosystens had disturbance return intervals that were greater than the naxinun potential lifespans of their doninant tree species. The nixed pine, jack pine, nixed conifer/deciduous upland and nixed conifer/deciduous lowland ecosystens had return intervals less than the naxinun potential lifespans of their doninant tree species. The regeneration and persistence of these forest types were probably dependent upon stand- replacing disturbance. . The pre-Buropean settlenent forests of the study area were characterized by large and doninant trees of several species, relatively low tree densities with relatively high basal areas, probable nulti-layered canopies, and widespread windthrow events which created dead snags, large fallen logs, windfall hunnocks and variably sized pockets of seral species. All of these features are characteristics of old- growth forests. I do not use the tern old-growth to inply that each forest type was honogeneously daninated by late- successional 'clinax' species associations. It is clear that disturbance regines caused seral heterogeneity within every forest type. Such heterogeneity varied in scale, fron relatively snall patches in the northern hardwood forests to large-scale patches of seral jack pine on the Raco outwash plains. The area within each forest type possessing old- grawth characteristics was likewise variable, ranging fron large contiguous areas in the northern hardwood type to 145 essentially none within the jack pine type. The conposition, structure and predoninant nodes of disturbance have changed dranatically since the period of the GLO land surveys in the 1840's. Most notable is the decline of henlock fron its doninant status in the northern hardwood and nixed conifer/deciduous upland forest types. A greater percentage of the landscape is now covered by seral species, such as the aspen/white birch association. The forests of today are characterized by very high densities, lower nean dianeters and relatively lower basal areas. Because of active suppression, fire is no longer a significant factor in regulating the conposition and structure of forest ecosystens. The inpact of windfall has been reduced as well, probably by changes in forest structure. The tinber harvesting aspect of forest nanagenent, and insect related nortality have becane the doninant disturbance regines in the study area. _' Since it is highly unlikely that large areas of western Chippewa County will be designated wilderness areas and be allowed to grow and function according the natural processes, forest nanagenent will continue to be the doninant influence upon the conposition and structure of the forests. Gaining an understanding of how the ecosystens of the study area were once structured and naturally functioned is paranount to the attainnent of a nanaged condition which naintains the long-tern health and sustainability of these ecosystens. Managenent nust begin at the landscape and 146 ecosysten levels, and then filter down to the connunity and stand levels. In order to successfully inplenent ecosysten nanagenent as the broad franework in which to nanage the forest resources of the study area for nultiple-use, while also naintaining the long tern integrity, health, productivity and biodiversity of the forest ecosystens, we nust recognize the role that disturbance has historically played in the function of the ecosystens. We nust understand the role that disturbance played in regulating the conposition and structure of the forest ecosystens if we can ever hope to successfully enulate the disturbance regines through ecologically sound nanagenent practices. Managenent practices can be inplenented at the stand level. However, such actions nust be taken with a full understanding of how individual stand dynanics interrelate with overall connunity, ecosysten and landscape structure and dynanics. I believe that the historic conposition, structure and disturbance dynanics revealed by this study can provide a substantial foundation upon which sound ecosysten nanagenent practices nay be based, not only for the study area, but also for nuch of eastern upper Michigan. 1323.01! 1 A dictionary of data field codes used in the General Land Office Vegetation Bntry (GLOVE) progran. ASPECT A two digit character field available for recording slope aspect data. Bl . A one digit character field (associated with DISTl) recording the bearing (N or S) fron a line or corner post to a bearing tree. 82 A one digit character field (associated with DIST2) recording the bearing (N or S) fron a line ' or corner post to a bearing tree. CNTY‘ A two digit nuneric code representing the county in Michigan to which subsequent field data pertains. ' CSB A two digit nuneric code representing the orientation (or course) of any disturbance event recorded by surveyors, where: 1 - North-South 3 - East-West 2 - Northwest-Southeast _4 - Northeast-Southwest DIA A two digit nuneric field recording the bearing tree dianeter (in inches). DIR A one digit chnracter field recording the bearing (M, S, B or W) in which the surveyor was traversing. DIST A six digit, 2 decinal nuneric field recording the distance (in chains) fron the reference corner that the surveyor traversed before setting a line or corner post. DISTl A five digit, one decinal nuneric field recording the first distanCe (in links) fran a line or corner post to a bearing tree. 147 DIST2 DRNG DSTRB GEOL NOTES RBCNUH 148 A five digit, one decinal nuneric field recording the second distance (in links) fron a line or corner post to a bearing tree. A two digit character field available for recording soil drainage data. A two digit nuneric code recording the presence of any disturbance noted by the surveyors, where: 92 - Fire 95 - Beaver Pond 93 - Dead 0 Windthrown 97 - Dead Tinber 94 - Beaver Meadow 98 - Burned 8 windthrown A two digit character field available for recording quaternary geological data. A text neno field for recording any supplenental data of interest. A sequential nunbering of records in the database. A two digit character field referencing the section corner (NE, NW, SE or SW) fron which the surveyor was traversing. A two digit nuneric field for the section (1-36) to which subsequent field data pertains. A two digit character field representing the general topography of a section line, where: ES - Enter Swanp Depression SD - Swanp Depression Enter Open Pine Land Leave Open Pine Land Enter Pine Grove Leave Pine Grove Enter Jack Pine ThiCket Leave Jack Pine Thicket LS - Leave swanp Depression EB - Enter Alder Bottons SB - Strean Botton LB - Leave Alder Bottons EM - Enter Marsh OM - Open Marsh LM - Leave Marsh LP - Level Plains WP - wet Level Plains SR - Side of Ridge TR - Top of Ridge RU - Rolling Uplands FU - Flat Uplands SW - Swale 55538888 149 A two digit character field tree species, where: A - Aspen IW AL - Alder ' JP BF - Balsan Fir BP BW - Basswood ‘MA BE - Beech RM BA - Black Ash RO BC - Black Cherry RP BO - Black Oak SP BR - Bur Oak SM C - Cedar T CW - Cottonwood WB E - Eln WP GA - Green Ash W H - Henlock YB recording the bearing Ironwood Jack Pine Balsan Poplar Mountain Ash Red Maple Red Oak Red Pine Spruce Sugar Maple Tanarack White Birch White Pine Willow Yellow Birch A six digit alpha-nuneric field recording the township tier and range (ie. 40W10W) to which subsequent field data pertains. A two digit nuneric field recording the year (ie. 40 - 1840) in which the survey was conducted. IDPIIDIX I Township forest cover type naps of western Chippewa County, Michigan. 150 APPENDICES 151 .umoz H among .nuuoz we nannnaoa non neg» uo>oo weapon wag-04.52- coomousmloh—m .Hd Gunman nip—Law?“ 11: U 20.: .;- A4 , .M . p . , H: \uww‘woflmwu om~rmoonfl macaw A .. .3. . ox «‘_ + .r._ N X4 .. SuHOZ \xx‘c . .\ : \\\~\\\\ 5% \\\ “\x N“? AFV\\ ..3 . - \,\\ Esau—.60 ...-Ila .3! E e? 15:33 amok—>50 I with: 132—>30 I . 0:: 32:. E I/ , § ‘\ u 0: «VA 0. 3 sex... H— .3// ANN woos—:3: ...—0.3.3: 5 /, ,\\m«e _~ §x "1:337. to: 0:36.: . . fin? ‘. .933 a 093m .532 3 3:258. now «093 uo>ou uneuou noun—«Haven anamouaunoum .ud 2:5: 152 :+_;n.a ;z: o :m: 4 mmm . wow u .n wamom Suuoz sxdd heuacoo It‘la Wu! 1:1 —!°.— 1 a 0.9\C8 “duh...“n.‘ ‘57-: Qu°8\p—OO _ . \u¢\\ 2.:— xoaa :— . .. , , \:\ .\\\x\» \ ow a: :— 1..st E soon—Zn: ...-0.3.3: fl . .. \L-\)\|.L>-LLL-> “hung. 34:93.01. \. x . \ \st q\ . .9 \\.~.\ ~ \ m x. \\V\ ,C \\ .\V\, “x \\ . ‘ , \\\ x\ \\x \\,\\\\ ‘\ s\ \ . . . . . .\ D QR: ‘ .5 :S:\ a , ~ ~ ~, \ \ 1 :x x x x x x \\ \\\\\. : \\\§.\\\ \ \\.\ \\ \ ‘~ \\ \\\s\ \.\ x \ \.\. s x . K \xthsxmwxx YER xx» hunan?» \xx 3» \ \\. \x.\\. x ‘A\.\\ \\\\\\ Cx \. .\\ ‘V..\\\ \\,\\.\\ \ x x ‘C b 3£¢wV \\\\ Cr tsex. \‘V‘ \‘x \C. , . . .\. .é\\\\\.\ \\ \\ \\ s \\ \.\ \ \ \ xx CaC‘ \\‘\.\\\\\\\\.\.\\ \\\\\o N\\\\ “\W\ \ \\.\\\~<\w\ \\\.\“\\ \\A\ \x\x \ \\ uv\~\\ \ \\.\ \\ \\\ \ \‘~ 9‘va \ \\ \ o . I; Oaks-PF RUDY“; .. , 153 .uuos n yucca .suuoz v. ngsacsoa you scan 9260 unouou ucoaoauuos coomouaflueum .nd 8502 imitfifu 1.1! u. :MI mm~.momna wamom .h p. ;H.M.ga s. . . .; A ._ .vwwo , . \ . , ,,_ . \ . . , .. .,. u ._,_ . ,v““““umw , , \.\\\.W\ ~ \ \ . V\\\\ \V \\ \ :uuoz 1\W\\.\\»m‘ \ \“x.\ \\\.‘\q\ \\.\ \\\ \ \ \s‘ \ \ .. x . W¢\\x \» \“\\\\ \X \\ \x S \ 8.1.. fisuqcoo A!!!“ la! sea—394 1_ooo\:eo icflufia ‘a00fl\:60 Uflufi ‘9.” _¥\\\ \\ v~,\ \ . x, s C \ x. \J\M\\\ \ , \WV‘fi ‘ \H \x , .\ \ .¢ . \ .\\~\\\ \\x\.s\\\\\ 2:.— oox a: \\\\n a% :\\.\\ A \thv o \wauwxmvaxmxwé \.\\\%\\m\\\ \K.\\ x“\ \ . x \ x‘ »\ \ woos—23.. ...—05.32 1:0304 . N\ \ . . \ \\\\\ ‘4 v \x“ “\\\“\\“V , ‘X \ \\ ~\\ \“x\\\\\3 . . \\ \\ x \\ C A. \ \ “x““\\“\v“\\“m\wv “x“\ \;n\“\\ \ R\\\H\Ns\ .\ \‘ w \ \ .nl , \ \k!\¢ x a .a\\\.\ .\\ .. flux x'\\\‘\\\\\“Cp.\ ~ ~ « kc: Uubcluzh 154 .uuos v oocum .nuuoz qv nqnucaoa you hogan uo>oo uuouou ucoaoHuuou cnomounmloum .v.n eusmwh D1. —...m..._D .4211 h. 2%: omm.oowud Oamom suuoz can; souucoo oldie was 151:... 1:03:00 1caum: smoooxcoo o:_m noun econ 1.x“: 192.73.. ...—Iago: Sup-56.— 155 .uuos m omega .nuuoz vv manunaoa you neg» uo>oo uuouou woo-3300s :oonouoflueum .n.n cannon :¢.:poo amtzo :m: www.mowua OHMOM nuuoz can; .- 3256 5.1...» 1:. E 1ca.aog woo-excoo & wan—n: 1.oeo\:oo team sous acme sexur 102.73.. ...-0.3.3: "1:OVO4 EC! Unhiuzh I...- I l v I. 1 i 33.0““ \WCVCW‘\\\W~\\\\ \ x. 3 xx, \~ \\\\\n\ x \\ \x\x\\ \ \ . ‘ .s\ \\ :47 . \ \ \~\\\ «A A; J€ : xx x} 3 , ~\uxxfix\“dx“\ \ \ x I CC .31}! I xxx {Ax \ w v ‘ / . \\.\\\ \ hf, / C \ 00/1; ’1 \xn¢\\\\ a. ,,/./\wm«m «fix . z ./ rare/zit /// \y“\.‘\\ z // //I “\\m.\\\ .\\ x \ ; u, II t;/, «t 11 \ \\n“\.\“\\\ _ / , \ \ \ \ \ I / \“vxx\ \“\““ x \H\\ \‘ \x\ _T\\ ,\\ t it A«VA \A .\ \x. x. \ , \ \‘x\‘~\‘.\x\\“ ‘ \ \ \“xx \\\“\,\u\\w\\\xx \ \ \ \ x \ \.\ \ ,0. xx \ \.s \\\.\x.\\\ \,s\.\\ Cb, \~\\\\\\MHM\V\\\C x)“: \. \ \ NV~V\\4\V _= x nh_ _ I. ‘u I l \- l Mlt1 WV“. \\ru.\\\\“\\hx.\\,\\\x“\ C .‘\\ \\\\\\ \\.‘\ xxx \ \ §¢3¢FV¢¢ \ . \ \u,‘\\\\.\ \\\\.‘\ x» \ \xV\\ \\ \\ \x.. . (xx .\. \ \uk? \x\m\\. \. CVC \x \\ \x \\ \ \\.\\\~\ \\ ‘ \x\ . v \X ,. \~ \ \ \. \lexXVth \ \\\\\~\\\ \ \ t w%/; \3. / 1., \ \ \ \ ‘\ \\\.a\\\\\\\\\\\.\\\ t x x \\.\ \x\ b.— .owéhr\\: \ xx \\ \\ \st \\.\ \\\\\\\~\\x\ \\\\\\\\\ xx ‘ . _ \\ x. x \ ‘— \\\w\“\\“\\\ m\\ _ _ \.\ .\ s. _~w%«\ . \\\3 \ s\“\\\~n\\ . \ .\\~\x\x \.\ \ \\ \ \I\ . . ~_:\V1Vv\\\ . , \ _ \\\w\ \“\\w\ _ . Gxx \\ \_ . _ _ ~ _ 5:: . -—__-I.ux. \‘ .\\.\\.\\.\\ 156 Jun: o 093m .532 3 neon—Boa you coma» uo>oo unouou ucosoauueo coomousmuoum .o.n shaman 13,7.an ...:mt .. 2er 0mm . wow u H mdmom suuoz extd fisuacoo A!!!“ km! ‘fiuicu— Van-£9.00 «v HEW—Sq— Qa°.fl\:8 scum 80.5 Oflufi ‘OXul H— 1033: Eln-£02 a Sigma). an; Ouhfuzh ‘ 1 . v. 359.31 CC \\ \W\ \\x w.\\“\\\\\\\ \x \ .‘x \\x\\\ \ \\,~\ .\ \\\\\\\\\‘r . \~\\ \ \\\‘\\\.\\ \\\ \\\ \\ x x . . \ s x .\ \\ \\ x \ \K‘sxx s \s\\\\ \ \CK \ s \x 0va \ x xx \ .C x .v C\ s .x x \ {/1 x xx. \x\\. . . I / I ‘ 3 \xx\\ ~ I I \x . / z/I . 1 VI .c\\. I . 1.. xx \\\\x V\\\\ \\ \~ \\ \x. ,\\ \x.C \ \\ \\ \x . 2; \s\\\ x , . \ . \ .— _ \ I: z ‘ s x x \ x . . \\\‘\\V\\ \\\\\\V \‘\\\‘ \ \\\\\‘\A\\\‘\ \\\\ \\ \ \ .\\\\\\C.\\ \\\\\.\\\\ \\\\\\ ,\ \\ s\~.\\\\.u<\\\\\\\\ \ \\.\\‘\\\~w\\\ \\\‘ , ._ ct»««<$$ == \\\ \ \xfiCx‘ \ .\ \\ \\ x \ x .x. .~ .~ \ \ V ._ \C\.\\.\n\\ \ .\ \\\\\ C . \\\ s\\\.\\\\\\ \ ~ \\ x \ \ .x.\x.~ \ \\ C HENQ AFR!“ \ x \ \ \\ x \ \\u \ (AC: ,3 x . , \x\\x . ,\>\ v _ . x x ‘ A: , u _ \kw&w%x& \ \\\.\ xx AAfiA \ x xx ~ \.\\ x... \\\ \oxty\\“ \ ‘ \\. \ \ \ x x . \x x “\x” \.\\x t \\ x x . xx \ \«C\\ \ \ \A\\\\\ . ‘ \xxm“ wx \xxx x“..\\\“uaw\ \“Avfiwx , ~\ ,\\.x \ \ \\ x. \ x C \\\m\\N.\\\ \\\\\\ :— , ~\ _\\\.\.— \\\ x“ \ \ \ \ \\.\x s\\ ~\\\s\\. \\\A\.\\\.\\\ \ x .\\x ~\ \H .\ \ \\V\\N. 3 _ o\ » wx s x \ \\\\\\\ x \ .x . , , \\ \ s \“.\ww\\ x\\ \\. \ x \\ \\\ \ x s \\ \x x a \\ \\\\\ \~ \ . A \ , ARR : \ \ «Rx nhv‘xamxvio \A\ \ .\\.\\.\ .C.\\.\\. ‘ . \ ~\ _\ .\fiio%x CX\\~ \3 3x A .4‘ C . \ \\\\.w\€\ \ \\\\\m\\ \\\‘ \\\\\\\\\\“~\ w\\\\>\\\s\ \MKXJ \ \ \\\\ \C \ V\ t t L. x \ xx? xx .39? .\\ \xxx‘ \T 157 .unoz H 093% £3.32 me 3:250 no mug» uo>ou umouou unosoHuuou common—Mien“ (Sn 8:5,: 31...}. . .H mm! ..J/.\.\x\xx /-; x x x?) . ‘ .. FF, V x .. \xuxxxxwfvxxxxwxxxux r x.x\\xxxx “xxxxxhxwxxxxx . J/é \x \x\.x xxxx 1 I a , axxxx xx“x\\.\xxxx I . I z . x .a/ x \“xquuxMxxxx z/zzl/o/ / /‘ x7 x 4 z / xxxxxx \xx / / ix x r // \ \ \x.x\\x / I z I AMI/I Ira xix lull/A ’04 1‘ [VI . z z t z I / I z I I./ . O/ I l . I I / z x z z z, z I z z I 3 x , z x 1 I. a; / z z I z x I g / // z /. I; / z . nuuoz \ x“\x\\\ It I \ \x . \uxxqux. \xxx / x {xxxx x x.x. \ \ .x x / t. Ce \ \.\ Ctxxr x . \ \ \ 2;;xr/z z/ z // In xxxx xnxxuxxxxx ... x s // z / . . z / 274 fish a :00 hid)“ ... «I x. .x\ . \ , Aw «fl_,¢x&§m&. / z /x\ x , . xxxx . 1.3-lea 12.1—)- v A. v- “A? “\Q \ / afl/ r/z/x/ / x“ \wnxxxvxxu xx“. r , . \ c 3 ,/ \vkaR / / It I 4; x . h , , z x x.\ \ \.x x , 17L ; r. / {Mxxxx x‘. \ xx“x\x\\xx \ . z . , x . \\ x ,\ I z , x \x. xx xx ,x . xxxxxvx\x\\ / xx. \x x x x x xx / u‘xvxx \‘.\x xx\x.\x \\ , .x z / v\x\ xx x. x \\\x ‘5 uh: ‘ a 06:00 \xxxx s: x x‘ at Josh. // ARV Afi\3\\¥n«» O \\\\R\“& I ‘ \ \xx \ \ x x x x z \. xxxx.xx“\~\x xxx\oo unouou ucoaoauuou Eamon—5:25 .od Gunman 158 me.an_n 21: 4 :mr . u u no ‘_ \ “we? . omN moo H a m A rxx. x xx xx,x x ~.xx xx. x , . “x“wmxxxNxxWK x xx $ yew x . nuuoz x . , .x . . ANFSNX: / , \x x . zanwxf ,«xxéxmfi V \x x . u . . VAR I... 08.4 vx. N\ \xxfixx\\m\. \ “Chin-8 ‘5'“ xi: x\x .\\\Wfix 1:0 ~I°d ‘uotfi\¢°0 ‘5 fl“: ‘d°3\=8 \H“xw&x x x x V xxfi$® “x xx. x\ \x .x m \ \ . . ’ “NA“. \ . x . x . . . x x x \a xxxxx , \ x x .a- we" v.0.” DE."- tflnfi ”Quad! x . >xxxz . . xxxo. x ‘ xxwxxwxwfimx“ x xx \ 10°31fil: cisfiufioz “fie-aha.— nx .., . ., \é \xn , , X. . z .11 It .x\\-nW“\\\%\xxx x x 1“ V.‘/ x xx“\\x\xxx xxx \\.xx x xx x. “xxx x x A“ x x. x xxxxe\x\x ‘ I \ x x xx \ \ \ \ \ x\xx .xx.‘ . 2 xxx? x.< I 1 I / x xx! Alia/t 1r 1 {IV 1 I as. 255:: .. . h ‘ .. . . .. A,/4A,O,.V,,.AE//., . 159 woman .uuus n ounce .nuuoz m. uo>ou uaouou useloHuuea coemouomneum nacncaoa non .e.n encouh hshucoo val—Boa 1:! 7n: :1, Tim:— wm~.womua maaom nuuoz .814 £1.39 Wu! 10031:: gin—0&0: L .1! ...x 3V2 \ x. x xx.xx \ ‘vxxxxxwxw, \ “\xxxxx\x xxx x xxxxx .\x x.xAxx x\xxxxx.xx xxxxxxxx xxx x x x x xxxxxx . . “x xxx xwxxxxx . xxxxxcx. xxxxxx . xh xx x,\ xx x x x.lxx\.\.wixx x xx x .x x xx x x . x ‘wxxsxmxx xx , , . . . xx ,xxVx A . ,x x . 1 xxx xuNxxvfix Wxxxxx - .. . , . . . x x .u‘uu‘u . xx , x\xxx xnxxxVxxaxWxn x . ,\xxx, x, x {Rx x xxxxxxxxx x x xxxx \xx\\x “x x\ x. x xxNxxxA xxxxxxxx “xxx xxxss::mxxmwx&:xo . \ . §xxx 4\V\x xxvx xwxux—Nx xxxxxxx“, ‘ x \xx x\ xxx xx I V I , ,xxxMxfinx 1. \\\.x\\\\\\ x K xav x ix . xru W“ .x xx xxx \x\\x\xxx.xx. V . xxxxxxx xxx , . xx x \x. .x x \ x.xx\x.x x .x xxx “\HxwaAxxnx. , . wxxmxx. . x x , \.x\xmx\\xxxw“ . “w ,xxxxwwmv&w\_u \ \x x x l x. , x. xxxxxxxx \ \xxx “x Rx xxx x\\“x. ,.x x x \ xxxu‘xxyuwxx xx x . x x x .x . x \\\\ \ \“\ ...ooexcoo x x xxxNx\\x \.x\\ x x . . \ \ 339x50 xxmxxxx \o x“ \\ xxd ....x..1x\ xxMRxw . .x x o: E :2... :— pawax \x\ x . x .. a .l zikk§ xv. 2:.— 183: H— , c.& .x . 53% x x . R , . xxxxwpxxxxxx. xx xx \xx\ A xxxx x.x.xxxx\s . x 3 xx xx x\\\.x xxx “xxx. a . . xxx. xxxxxxxxxuxx. . x.xxx.x“_\ .\xxx\‘ \.x.\ . . \fixxmxxxxxx xx“ ‘ .xx x x x, xx xxxxxxxx. xx xx x.x \x, x . "docuhmj xxaxx“xx“x\x x E! Oaks-US... xxx x x I ‘2. x.\ / xwx, xxx x\ x x x“ I xxxvx.x\x\\\\x\\\x. x\Hx xWxxxxxx—“xvx. \xxnxx“xx x\.xxx x\xx x xx xx xxxxnxx x xxu\\xx.xux . xx x _\x xxxxxxlx L .xx.x.x mnwy/xv / / //, 160 the: v 00:3. £9.32 3 manages you neg» uo>oo unouou Houses»... csomoununehm .o." .n 933..— 21.Lm_c 21: u zuz wmmxooona mamom nan; 50hucoo fl!!!“ 1.! ”GI—304 ‘nOOfl\¢00 16‘ufifl §u009\¢°0 scam xoso Utah 10an sooa1fis: enogosoz "#:wv~4 161 .uumz m mono“ . nonxu uo>oo uuouou u: guuoz mv macaques you «Beauuou cuomousmloum .Hd.n enough .F . , .. , x... E. :I x 2...: ...pEIEE \ \ \ :m_—__.—— x me\ xxxx\x x x x e \A\\\ mmw mow...” OHmom “:Nxx\x\x$xxxx xx” xaxxx x.\ xxx. xxxxxx~x\xxxxx xx,x xaxxx.» xx x 1/ xxx xx \.\\ x\x xx xxx ..x; / “Qxx .xxxMxxx xNWHxxM—(x ‘_“‘—._—_1x. I . x \“xxmxxuxrx .x xxx). x3333 nunoz .‘QJ I .. “ v \\\\ at. x , 2:00 alaaa x3. mu x\.:\x\ u “x .x x? __ 1.3—3o... incl—>30 I Rx... x I. 1607:: 1a00fi\:00 I ~ w.._ .v x x x. ., 2:.— «nos... :— xx x — .— \xxwxxx x . \\..\ xx a.-. .- .....x . .x. I. xixxer. xx. .x..V\tx\.\xx.-\ ..x ...-Mm $.Xfl' a .axx xxxxxx x9 x \x. V. a . x “x xxx xxxxvxxxxéx x“. xxxflmmmxnx-x \xr vwx x x,x x xx $00313: Fug—0562 a xxx\xxx.x..xx “xx\x.xxm“m .x u my— FF ix vab Jar—ova... . -x u . is. Unhguth . . . _ .n I. . 162 duos o omega :33: no mv 53.3.50 u 93 uo>oo uuonou unaguuuoa zoomou5%nowm .«Hd Pagan :m_am_n gxr-u :wr ’\ \ ’ \ \x7} xx“ 33 x} x mmN wow u H mamom \ \ . . \\ bi“ x x11. . xx xxx xx x x\x x\x xxx.x,x \x xxxx x \x xxx\ xxx xx x xxux\\x x.x~xx xx x xx .x x.xxx xwxx . . x. .x . x x . x x“wx“ka“x\ xx xx xx . xxx x xxx.xx. x xv, xxx x x x x x xx xxx \x xx. xxW\ «xxxxxxx xx xxx. x zuuoz ti..— .utk a :00 ill)“ x u! wed—neg 1 . Amwwma «OX—\GOO . . . Ax x Nx «.5 uh..- uo a 05:00 0: Tu lot—c .... Tm ‘OX :- 1093158 ell—«hot / . x. of \ xxxxxxxx \xx xx W xxxxmxxx x// A x x .0} .1 ’ .xxmnxxxx» ...: 4 xxx . MNNKV xnxwxx xxxfixx xxx x\x“ J, xVxxx\ .x.xA\x x mm» x - ..uax V.) x xexxyxxxcr x 163 .uuuz x. 093m .5302 mv manages you mag» uo>oo umouou £33.3qu :aonouauamum .35 0.30?" :1, 5+me «in! xu. :MI mm~.momufl oaaom nuuoz 9:3 4 htumcou aldla xu! mull—304 1m°tfi\:00 ISO—an: 1u°s\:8 o:«& 80.5 0: «an 1.! ml IIIUEEI wooldflflz :fioflafioz Jar—ammu- fi¢l cabin-FF x9xx x .x x\xxxx .x fix)? x\ xxxxxxxmx \\ x .x. , “x\N\xxxxxxax x xx x x x xxx Axx .x xr/xxn xx x .xx xxxtVWx\ x xxxx. xxx“.xxx xx\x~x“x\d xqmmfifim xxxxxVNXxxm.x\.xxx.xxx\x“x . “.xxm\xxxxxxx “xxx 1 . . x . x x 4’1 ///1 x xxx . a x \x xx xx x\x x \\ r z x I x .xx x.xxx x x xvxxx . x x .x \ xx x xux‘.‘ xxx xx .xx.x x. x\\ xx. xx xx . x xmm..x\ x xvxxxxxx x xxxxxx x .x x . . x \A\xx xxxxgxx x \x.x.xx.x x\ . .x_ xxxxxxxxxx. xx x. x““\x , \ x \ \\ x \H-x x xx x xx .x xxx. x \xwxxtxxx xmxxxxxwxx x x x x .x.x .x ax x xv xx xx x x x x xx x x Vx.x\\xx.%x\‘xt\n“ x xx c\\\&xx\x xx .xxx.x xx \ \ xxWx\ . . x xxx x x x .. .x,x . .xx xx‘.xx.x.xx \\.\\x\\ xx x . x \ x x x x x xxxxxxx “\xx xx \x‘ .x x x x \ x\x ,xxwxWxxKxxx xx xx . x xx xxxxx,xx.x x ,x. \ xxxxV-x , x xxx. \“ ¢0Nx . x V x xx. .xxxxxx xx x .\\x““\\x“\\\ x$x\x\x“Wux x . x “\x x xx .. x“ , . x .\ . xx x , x.x.x .x \ x x x\x , x , x . x . x x x xxxx x .x x .\xx“wwm“xxww.x\“xv“.\xxx xx “WW“? x x xx xx xx: xvxmmuxxxx ..xxxx. xnx. .v x \xx xxxxxxmxx. x: .A xxxxxxx .xx .xx‘ x . .x x .xx .. x xxxx. .x. xx xx\x 0?.me . .. x kaxxxxmx x \r.\ . xx. xxx xx\xx x xxx} x xxx? \x\ \ x x x ““xx“x“u x xx.xxx xxx. x x\xxx§xx. xx .xxa\x“x\“xx“ x xxx \xx x xx xx x\x . xxWxx xxxx \wawxxwg .x\. A . x xx4 x .xxmxxxxx. x x x x xx x xx x x.x.\xxxxx.xx.x.xxxx.$ 164 .uuoz a cocoa .nuuoz wv aanunzoa you nag» uo>oo unouou 2.85392. nuaousnuoum .35 0.5.5: :1.Am_o amz-u 2m: omwxwooud «doom nuuoz .814 fituuceo in x:- 1cn—394 sac-excoo 1.3—h: $2.655 0.:t 30:. 0:7— Gila! $00323: ...-080...... I! E: E5 E3 II E3 II "1:6..qu fis- Uuhgzh .x x a? x xx x.x xx x\x x x.xxx x“r\ x\xxxxx v x xx xw.xx“\xux xx—xxx x '\x x xxx xxx \x xckca“ xxxxx x xx x xx. . xx:xxx , xxxxxxx {..xx \ \ .\.\ x.xx“x x .x xx x . x. xxxx.x“ xxxxxx x. x x.x. . .xxxx x\“xxx x x \ x.x x .xx x x x «\x“ \x . “x.x.xXx efi.Aw&W xxx x , x \ x , .xx “..x” xxwxx“x\xx x x x. xx xx \ x xx xwxxwxxwxxxxn“ x. x.xxx\x x . x x xxx \ x \ x\. Vi .x x xxxxxx.x “x xx x x\ ...\xxxx. x \ .xx x x,\\\ x x x xx. x.xxxxx \\xx. xx.xxx. xx. xxx—xx xxx xxxx. xx x x\x bxxxx xxxxxxx x xx xxx\x x \Vm“\ x . .. x ... \....x.\xx A“ x xxx x xxxwa\“x ..Wumxxxxxxx“ xx“ x “xxx—“x . . . x x x x x . xx x\x .xwwxxxx xx xxxx. “Nx xx.\ x x x x xxx xxxx. .x xxx x xxxxx xxxx xx x\xx xx\\x..x—x“ x x\“xx.xx“.\x x x \xxx xx ,xx\\\ xxx x\. x.x x \\ x xx \x\§x xxxx x xxx A x \ xxxxaxx. xx x .x x xxx. Ax xxx \x xxx A.x\x.xxx x xx“.x.\.xxx \xx x. x. xxx. ..xxxxxxxx bxxx xx .xx x x. xxx \xx ¢x xx . txnh . x.“ xxx m.“ x x x . x x x.x .x\x“\ .xx x\x x\xx x .. xx .x \\x x x xxxxxxxx x.x x.xx\x x xx xxx\x.xxxxx\xx xdxxxxx x xxx. .xx.\xxx.xx .xxxxx . x x x .xx x xxxx .xx xxxxxx x\ x x.xx\x““x .xw. x“ . , x x xxxx xxx§xvxxvxxxx\xxx J .xxx.\.uxx \.hxxx xx x x. .xx . Vx x x ..xlx\xx\xV\x“\w.x xx x\x xxx\x\xx xxx.xx xxxx .xx .x x x\xxxxx “xx“x xcwxxx x . x x “xxxxxxxmx\xx xwxrwxxx x \ xx .x . xxxx,x\xxx xfiWVVxxxAxxxwxfiv x x x. Y\. xx . xx x..x,x .xx% xxxx. Rx. xx‘x .xxxxxx—x x x .xx x x .x xx . \xx xx xx.xx\xax \ \ \ .\ \ \ \ \ \ §x xx xx x x\xxvx x.x.x x.x\,\\x xxx x.\\ ..x \ .\. \ \ \xx x\éxx .uuuz u wanna .nuuoz ov mananaoa you away um>oo uaouou unaluauuoa :aomouannaum .3" .n 253E 165 ......FE :1... u. 2...... . x x xxx om~.mooua mamom .. xxxxx xx xx—xxxxx x. xxxxx x.x x.xx.x scxsxxx x¢:\¢ x. x x.x x x xxxxxxxxi. xixxxxx . x x. x x . x . ..xxxxr..xx..x$ xxnx xx xx xx xx Rx xx xxx xx xxxx, x xx—xx x pmwx nuuoz ./ av / .X'J . //// ~ . I , /,oo unouou use-adv»!- cuomougIoum .0” .n Chanda Ix.iwia :mE U :wi mmm.woouH manom :uuoz 0:14 E3 II hthucoo I!!!“ Km! 1fll~364 ‘uOUfi\:°° n x wan—a: «.oonxcoo e:.m moan Uth Qtiul E! E3 :5 100819‘: Chofiofioz "fitmwga he! U-hCIUSH \,\ xxxx xx.x x.xxx . . I / 75,. / I 167 mama» uo>oo vacuum unaaoauuoa :aomousuuoum .ha.n ousmwh .uuo: v oocum .nuuoz.ov nanncaoa uou 5.7;.5 :2... u. 2?; .5/ z o a ..x r s o Nix/H /fi 41;. me wow..." mamom ...7, . . , 4,7,. . . z: .= aux,” n - - . - _ a _ nuuoz h; u — _ n U 3— .. l .. fithmcoo I!!!“ 1:. E u g . - _ " 1:133 133x56 x ,n. .-4 i w : u . . . _ 13...: goonxcou I A? _ n . “ 2:.— ..ona _ _ . 2.: nexus” L.... __ : “ _1_ _ Vogfififlz gonna-n02 n w m 7.1.7327. —‘ . : .. /fl_ __ _ I . . ...... 25a: wavy/x -/ l'l _ ‘1 "l 168 .umuz n mason .nuuoz ov masonsoa Hon mama» Mo>oo uuouou unaaoauuon coomousmuaum .mn.n Gunman 31.;n_n :1: u :w: www.mowua «doom can; al.flu:8 ‘5'" X“: 1:133 3on.30 sea—a: wuoonxcoo scum scan ocum 1.x“: woos—:3: 23.3.3: 32.0%“... 169 .0503 o 093“ 50.3: no hunan—boa .uou nag» ...—«>00 uuouou 0.330.300: acumen—5:98 .3 .n 0.50.: I... a«... . . . x. r.“ x... . . 31;; 1.: .. 2m: ,. I x? «3... 3.3x... . 4 . , xx x\“xxmmflx. fl , f x x xxx x x . . . , xxx mxxu .xxxxx s o . .. . \ xxxxxx .x, . I own now . a 33m . . N . xxx .....i ..xxiix. ......xxx xix ..xx..xxxx..i x.x...xxxxx xxiixxxxx ..xx... xxx «xxx» .xx... \ xx xxxx xx xx x .xxxxxx .xm \xxxixxx. xx: 3“ 3. ..xxx. \ \\ \ xxxxfixxnxxxxv x xx x xx x xx x . .xxhx A“ xxx \xvx xmxuumxxflx . xx xx x\xx xxx xx. xx\ xv.x , . x,\ . x. xxxx x xxxxxxxxvxxxxxxx xxxxx,\xxx .x . . , , x\xvxx .xx \xxxvx x.x-x.xx x\xx.x.xx,x,xx\x xxxxxxx xx.\.xxx\a.x . xx xx xxxxxxxxxxxn . x xx x\xxxx\\xxxx x\—.xx xxx.xx xx xx xxxxxxxxx. xxxxxxxxx..x\ x xx xx xx. xx.xx. .xx vx.xx,xx.xx x .x,xx xxx xxxx ..vn xxx x\x xx..x :uuo x33:3: xxx???» x...“ xxx». .... x.x“..x...x..mxxx.v. ,. z x Ixx:.vx*xl\.x . xx. x.x ..x xx xx xx. x x.x x ..x xx. xx x \.\x\, xxxx. x. x\xxxx.\.»x.x ex xxxxxxxxx xxxxxxxxxxxxxx xxx x.xx xxx xxxx. xxxxxxxxx x.x .\xxxxx.x.xxx. xx x xx .xxxxxxxxxxvxxxxxxxxxx x \xxxxxxxx xxxxxxxxxxwx. xx . . x x xx .xx xx\xxxx-.xx,x.xx xx x xx .xxxx\ x x xxx x x. x . xx —\ . xx x\x . x x x x x .x . ,. .x. x x.x.xxx. xxx . xx x“\ x xx. . ,x. xv\\ x.x x x.x. x.x x» x . xx“ x x\ x.xxm\x xxx\x.x x xx\\x x.xx xwxx“.xxmxxw.x\ xxxwxxwxxn xxx 1 x\xxx.‘ x, .xxx xx.x \ xx.x\xx.x xxxx \ x (“\“x xxx\“x\\x . . x\xxx\xx x x x xwxxxxxxAxxxxx x \xxxxxw—x, .\x.x. ...xxxixxx .\ . x.xxxxxxix x.xi..xxx ..xx x . .x .xxxxxaxxx. xx xx\xxx\ mx\ x \x_x.x xx x xvxxxx xxxxxxxx xx xxwxxx\ xAx x . xx ‘ . . \ . .xx.xx.x.xxxxx.x x , xx xx xx \Mx xxx x 31.. ‘Ohmfls g,“ x“! x . \ xx \x N \‘x xx :x x.x x.x ”GI—’9.- 1u08\¢8 xxx, xxx Nxx‘xx x xxxx\xxs.x. \xxx x“\naxx\xx\x “xx x. x xx xx xv» x .xx x xxx x.xx\xxx x , . . xMxxkxx xxx xx r.x KPx ..x .x x.x..ixxxx .xxixxivxaxi. xx . ......xxxxxxx . _ .x. xxi. x\x.“xix...\..ix.x..\.x..xxw.\. xxxxxx x... . . .. ..x . .xixxxxx..xxxvi....\.x.x.xxx. .\x\.xx. 1.3 ...: gal—x50 xx .mmxw..®xwwfl.3“\mx\x.xx\xx x wxxxxiwxxxxxxxxmxxxxxx xx , _ .. ..x xxx .x\xxxxx x x x.,x¢x\\h. . x . ». mimimiiiiwxx“ xwxxxxifixxxkaxx x. xw:xx\\x\\mx\.\\x\xxx..\xx x.» . _ ..x \ DE- .n-d“ ‘0.” \x x\xv , \x x x \xx x \xx “x. xx.\ xxxxx\x x, xxx“ xx xx x xxx.xAx x\\xx. xx Rx x x a. x x x\xxax ... .....x x. . x xxx . xx x .xx ..xx _. ..x .xx. ix .xx\.xxxxxxx .4. xxx: ....xxixfikxix. .xxxx .ixixixx\.x.. ....iixx... 16031:: £080.33. a mex “x xmx .x\x\x§x.mmwxxxmx\xx .x \“WwvAxme. ..xxxx/ .. \ xxxx “Mk xxxxx\“.xu“\\\xx x xx\M¢x“xx \ x x“ x x \\ \ \ “.\ 7... Cum .UJ \xx‘ . x“m“\x“xx x\xxAx k . x . x x x x x x x\xxx xx\x\\\ x xxx . \.x.x,..x\\x.x .\xx x xix“ . . xx. x.x“.ux x\nxxm\wmm\mxwx.n&m“x\xxxnxuw\\x\xxx\ A xx \ \ fixvxxvx\xx x x ,. «j/// /././..x./ m E‘- Ouhifllh . flaw? xx. x. x xx.x\xxxx. . x.x,xxnnn , h 170 .933 x «9.3 :33: 3 35:58 no... nag» Ho>oo uuouou Una-Banyan cannon—Each.» .cud Hannah $1.7”... TH 07;... U omm . moo: 33m \ zuuoz m: x.xx\\xx ... ,. , ...“..umxduxumwufi xxx. , \xx x“ . x x“ xx xn‘xxxx xx x . Xxx. xx. \\x\ V “xxxx x x .x xx , xx\ x“. xx\“\xxx.xxx x.x . xxxxx >\xx.x:x x\.xxx.x x\x\xx. xxx\x\ x, xx xx xx x x . \ x \ lxmxxx \xxx xxx .x \ x xx. x xxvxx x, x .x xxx \xxxxx. xxx xxxxx x . . x \ \ . ..xx '\ \\\’x :x 1‘ . \ x. xx xx \\ ~ \: 3:. . x xi :6 , x \x \ x\x\ 3i \ \\ \ R3 '\ x xxx \ x \.x \x x\‘ x xx x \ \ . x \ \ . \ xx x“ “\x“xx\ xx .xx .\. x . \x“ “x \\ xx x \xxwx,\xx x x.x . \ . I \ rim—9’ \ xxx xx» 1 x xxfixxxx“: xx x \xx. xx,\ x.x xxxx.x \. xxx xx \s xxx.x x. .. \ ..x.\\rxx\. . . .x..\xi.x~xxxxx 0:4 x x\xxxxx hthucs $3330 3:. x xxxxx.xx x.x nus—led Um°a\:8 . \ \\ .xx x :l .- OIA\:°0 3:33“ 4. n = 1n xxxVxxxxxxx x xx x x \flvvxmx tfluh «no.5 “xxxxxx .x xxxx xx / xnixxV ...—uh ileum: x x.x xx x fixx 10032.: slut-ache: xx x x , .. . . .xx xxx xx. x xx.xx x x . xxx\H““.xx\xax“Hx“Mx“mwaxx\xxwa , .\x 35333 . . x.xxiii..x.x x.xxx.x.. _ . .. ,. . “xixx «..xx i i .x.“..iixi .“xxiixi. . xx .x axxxxx x .x x x H.“ x _ xfix xfixxvxnx. .xxuxxxvmxvxxxxxxxx x ks. OahiUSF x . ......»XQ . . \vxxxmmxflmvmxwx. 171 .umos H manna .nuuoz av mwsunaoa Hon woman uo>oo uuouou unoaoHuuun anomounuloum .H~.n Gunman :x—maifi lll‘w DWI wmmxoooua 0Hm0m zuuoz Old... hOchs hi0 X3. ‘BI—IOJ Igosxceo 16.—i: iuozxcoo 0...?— 88—. oema 1.x.3 El E3 E5 E3 II II II ‘eiun EEO-n2 ”scam“: h! Onhsah x x x x x .x x .x x x xxxxxuxxxxxxxxxxx x xx .xxxxPxxxxxwxxx xxxxxx xxxxx. x xxxxx x xx vx xx x xx.x x . xxx x \x x x x xxxxxx xx xv x . x ,xv x. .. .s xx , xxxxxxx .xxgxwxfixx xxx hxx\xx xx \x xx “xxx x xwxxxxx x xxx x\“x. xxxxx,.w.x x\x x xxxxxxx xxxx\ xxx xx xx .xx xx x x xx .xx xxxxxAx/fmzixxx I . n r, / / / xxx/,1; . / /rt ;/ 1H/ 1.1 L. . xx x x‘ .x.‘ xxvmfixx xx x . x A x x A . E%&&Rx xmwxxxx .xx x xxx x x\x...xxx xx ,x xx.M“mxvxxx xxx? )rxx\x®ax¢\ :mxxxA V x . x x x x x x ‘ x x \xxxxxxxxxvxxxxx xx x x .wxrrwwxwxmmwx x. // I x .x \\x\\\ x .xxxxx“ . x. x .x\ \,x x x x.x\xxnx,xuxn\ xx xxxx x xx “xxwvxnxfixxfi xxxxx xx xx xx x ,xx x\xx .umm: n omcnm .nuuoz 5v masucaoa you «00%» u0>oo unauou ucoaoauuou cacaouauloum .«~.n unavah :r72 :1_zm_a am£,u an: . Nx “\W‘\\n\\é..\u.l\‘xi‘ r .fixhwx. \‘ mmmxmomud mamom nunoz OSIJ fithacoo $8.80 .3! «eagle; «.oonxcoo VGI-ha ‘uOIfi\B°0 Oflufi 80.5 06a& fillat 100315.: 65.30508 ID CIEBIZ IIEQII HiawRMJ EC! Onhclfizh 173 .933 n 0.93. .532 5 92258. you nomhu uo>oo umuuou acoaoauuou :oumousflloum .n~.m Gunman ntiflmdfl Qtriv DWI mmmxmowua wdmom 37.4 hob u :8 ill)“ I =- 166-364 fiuOOQ\¢OU ‘flflufia ‘mOtfl\:°0 Oflmh 80'” Ule 1.2m: icoiififlz :fitfiofioz “ficuhad 174 .uumz x mason .guuoz xv nflzuczoa non mwnau uo>oo ummuou usuauauuoa :aanouamnoum .¢~.m Gunmam ?_1—......TH L115; 2.2.; It .7- mm~.momua oamom nuuoz oat; ficu u :00 a!!!“ X at men—:94 5.9-excoo ‘5 «an—.- 6 u Oép-OO .8 «h 80:. 0.- mh ‘0! «I I! E: E3 E3 ll E3 II 10032... 6810.2... ”ficuMUd fin! Oaks-ah 175 .uumz m ooznm .nuuoz 5v n«:u:309 you mama» uu>oo umouou acosoauuon :aomou:nuoum .mn.n Gunman ridfimgfi 112 u. DWI . wmm x wow u H mHMOm c zuuoz can... I flown-.00 $8.30 maul mu wan—......— 3ooaxcoo x.x val—fl: incl—\coo H 2:.— ..od. :— OCT- 13:! H— aooauhaa ...-0.3.32 5 u we law an. h! Uuhglh 176 .ummz w omcnm xnuuoz 5v magmnaoe Mon mama» uo>oo ummuou unoaaauuoa :aonounnnoum .w~.n ounuah al.;r_a a1: u :v: wm~.wooua mamom c Spuoz Said alCUMA-Onv “in and: IE. 16‘~3°J 1u°.fi\¢00 Uta—I: 1a0.fi\¢°0 ll ...- GE ‘0', Oflufi 1.8a! EIEIIE 1903152.:ioSoaa‘ ”ficutud EC! Ouhclfixh v/ / x\x xxx? duos x. 09.3. :33: 3 3.3869 you mug» uo>oo uuouou acoloauuou :uomounuuoum .36 0.30: 177 m... 339 3.1-0. :m: .xxxvxxxx x.xxx x . x.x“ xx. . xxx“x. xVxx xx“xx. x xxax x x e . x\xxxxxxx.~xx.x\x xxxxxxx x xx xx “fix x“x\xfixxwxxx x\ x\ . x , \xx x.xxxx x\“x xxx.xxx xx x \x .. . . 2.3 x ”8 u H 38m .xxwxmmxkaaoxxwvxxxwxv\xmxwxx. x\ “x\x . . x xxx\ x c x , xxxxxxx xxx\ xx xx xx x. xx x \ , . . .x . ..x xx \Amxx xxx \ xxx.x.xxMMx.k\ x x.x‘ux x xx \ xx x , x . x\x xxx\ . x x \ x ”xx: xxxxxxx xx. x\xxxxxxkxxxxxVxxxwwxxx ,x xxx .xx xx xxxx x xx .xx xxwmwxxxxxxx“ \xxxxvxe \x xx. /. .&.¢\\M®N®N® x\xxx.\xxxm\\ x xx x.x.x x xx x x x . , . x\“xxnx x.xxxxrxx\\xx“wx.\xxxx.x»\“x xvm.“ ““xxwxxxn“xvxx xvmvvxm‘x . x xxexxxxx x xxx x \ xxxx“x. xxxNxx .x “xxxxxx , “xx . .. xxxxxvxxxxxxx\xxx xxx“xx , .xxxxxxxx xx\x x xx xxxx lxx xx xx xxx xx x.x xxx .xxxxxxvm x\‘x ,xx\x xxx\ \x .x \ xxx. x x . .AxAx x x6? x“ x I xx x xx a\xxxx x . x x x x . . , . xx.xxx ./ .I x . .xx x .xx x. xx . x\“xxxwu‘xxxxxxuxx xx 5.82 / v . ,_ x xxx“. .xx xx xxx». .xxxxxx . x \ x \\3\ x x\x“ xRxx“mvmxxwxxwxx“x , . .. xx xx x xx . xxxxxxxx .x . . . , . x \ x x x . \ x x xxx\ “x.xxxx“xx..x\“xx“x x xxx\,xx\.x t . x xfixxxx. xxxxme x #14 .xx\ x xxx xx fits a :00 hie .x =- 5 lo .I . x x.x\l..n W n ..- ‘«9638 xx x\x\ x“ J \ x x x.x x, . ,. x x xWx\xx xxxHxxx x x xx“.x.xx . xx x \ \xxx x/ . x . erx x l. x . ‘6! 7n: 1 a 00A\€°0 0: a.— 80‘s. 06 Tu 10! «I IIIDEEI moon—:3: cal-«...! "...—.3an h! Unhs-USH 178 .uaoz o omcnx .nuuoz a. nunansoa uou «max» nm>oo umuuou unoaoauuou commousunoum .n~.n Gunman :1.:n_¢ 213.0 DH! ommxmoona QHMOM nuuoz 0:14 hthucoo h!!!“ In! fifllu'OJ duo.fi\:°° I! EJII $61—53 1n0.‘\¢00 ll lflmh Iota .:u& ‘01“! HIEIEE 1eclfihl: Chofiosez "ficwmaJ EC! Unhcluzh x .xxx xx.xx\\\x .—-.- . ,..=________.___u_____ \ xxmxwxxxxxxxx\.- x xxx““x\ xxx xx“m“xa x\x\x“ xxxxxvxxxxxx \xxxxx xxxxx xxxwxxxv\xx x x xmx\ \xx.\ x\(x\\xxxxxxanxxxAxxx x xx xxxxxx xaxxw x\x xx.x\xx~\x.xn.x. xxxmxux xxxvxxx x“xxxxwxxx x xx\mx xxxxx\xx x xxx xx xxx \xx \xxx\x..H.N x x x x x. xx\ xx : xw¢¢< ix. .x xxx\ xmxxxxxxxx e \xxxkxxbx xxxxxwaéxxxx \xx\xxxnxx.\x\ 5.x\\.xx\x xx x\xxxmxx\xx xxx fixvxxxwvx x.xxxxxx.xx“.xxN-x J. xxxlexkxx x\xexvx .3 :flxrxxx 179 .uuos x. 093m :3.qu av 3:253. you «an»... uo>oo unouou acasoauuou :aomounmloum .ond 05.5.: :._;«_a git-» :mx omm x wow u H manom a zuuoz Sal‘— hthufloo 51.30 In! 1¢IAIOJ Qu°.fl\¢00 IBI ICC—fl: Quatfl\=°0 I: a“ 3°.” tfluh GOXuI IDS soot—Zn: ...—23.3: “yer-@391— kcl O—hcthh 180 .233 o 093m 23,32 av manages Mow mug» uo>oo unouou unusuauuun somounmloum .ond 0.30.: 3..., .11 Tu 11.! -u. an“! mmmxwowua oaoom zuhoz 03‘4 Ethucoo filalu Xul wen—3o; 1aooaxgoo w:a_a: cue-Axeoo ocaa sons scam cox“: vooaqan: chososoz a... cut .u... 181 .uuos n «mama .nuuoz ac nwsucaoa you mama» u0>oo umouou vcoamauuun :nomousmooum .Hn.n ouamah wmflxwowua maaow xx.x x.x xxx x x xx xx x x xx.xxx . x\\ \.\. \ 3.xx xx :uuoz .864 houucoo a!!!“ Xul 1:0 1.0.. V a DOA—\COU IN EJII 1:! mi: 1 u OOQ\C8 l x.xxxx .x x. xxxxuxxx\xx.xx“,xxxxx \x ...-u.— ‘OIHo E .xx“xxx x axx. x xx. xx xx x , xx x x.x.xKAx .xbxxx .hxflx I: a“ ‘0! m: a . x xxx-xxxuxxuxxx 1°33: 22“; a \wwxvx x\x ssstx e xx.xxm.xx“xxx-x.x“,x,x“xxxx.xx“ x . 55893 «x\xmxxxxnxxw : x x x x . x x x x x xxxx\xxxxxxxxxx .x xx.xxxx QR...“ J x x... x x \“\x x x\ \ . x\x , xx x\xx x.x.xx xx x x xxx\ x\xxxx xx x \\. 1 x xx x xxx\‘xx\ x x , x 1.71 xxx. .x,x x.x . . . , x I x xx x .xx x . , . . .. x x // // J/ / xxxvxxnxwfiwxwxxxx . . . . x xxxwx xx/7H/ I; .//// 1V / \xxuxxx . xx 5 .x. xx x xx x . xx x / . 4.. x .x ..x u.p¢xu=h xxxsxexx. ..x /,,,./,nvwxxwx%w .1: , txrtkrr 0x1 #viq/I. x x..x. . x.x. xxx x x . 0 xx ..x anxx‘““&x\x\xxxxxxxx . x“ xwx .“xxxxxxxvxx“vxx. .xxxxx.%. \.x xxx. \ - .xxJ «mtmmx\&km,nix x A x x x xxx \ \xxxxxxwxx\xx xx . .x . x x x xx . . \ x x x x x .x : «xxxxxxxxxxx . xxx“ xmwwxxw x . «xxx\ x x xxxxx . x x . .x x.xxxxxxxxxxx xx. xxx .\x\\xp xx x. x x x x xx. x x x\ x x.x x xx x.x. xx xx x\“xx x\ f“? x“.xx“Nx\ xxx x x\x“ «xx» “mun \uxxxwxm xx®x®a¢xxx>xxx , xx. xxxxx x x“ xx xx. xxx x xx xrxxx xx xx. xxxxx xxvxxxx \x xx . ,x xxx xxx.x,\ \xm? x xx xxxx xxxMx x x xxx xxxvxx.x\\x. x x xx xx xx x xx WWVNWxxx x“xx.x wxxxxV xx .x x xx.xxx .\ x xx x K x \xv x x , xx x .xx xx.x x\x\ “m"xvvxxnx . .xxxwxxxxxx\\ xxxx xxx xxxx\xx x\\.\\xxxxx.x xxx xxx 50.x «x.. \- axxwx\ ,x x xx x x .xLVMxxxx x.%x . x , ..xx xx x x x VXxw‘RC xx x . xx.xxmx xxx\xxxw— x x\xxxx\xx. x xx.x xfixxxx x “xxx . x. xx N\\\x \ 182 .uuwz w 093m .430: on 9.22508 now may um>oo uuouou ucofioavvou acumen—annoy." .36 250.: ..1—LrTu. ..wI. .. :5: www.mooua wamum zuuoz 0814 fithucoo Illlu Xul val—Ind inatflxccu icinfi: inIA\:00 tcam X615 ocua fioxul wooldfiflz eht£ofioz ” ..J p. uxv a... 183 1.303 x. 093m £3.32 on 3:253. new woman ...—960 uuouou ucoauuuuwu cacaousuuoum .26 0.53.." :1~;w_n ;tz-v am: wm~.womua mdmom and; gouacoo Alana x“: sea—:04 wflooaxcoo weigh: smooaxueo ocam so.» 0:.m 1.x“: 10931.3: ...-0.3.32 2.. :03. 9.. 184 .una: mum naucnm .nuuoz an umwnucaoa can .unos m coca“ .zuuoz on nasucsoa you «Gnu» uo>oo uIGuou unoaoauuou cacaouamuoum .vn.m Gunman 7+7;_a.€£.J_EI k www.mowud mamum nuuoz .419 sou.eoo $1130 x.x 1.3 .33 328x56 wen—m: 1.ooaxcoo '6‘“ ‘0.” .6q& QOXqI intala IIIIEQII wool—rut: chop—aha: "inflamd EC! 0nh¢lfl8h ”mu 0 Diameter distribution graphs for significant tree species in the northern hardwood forest type. SALSA“ FIR TREES PER HECTARE o 15 20 25 30 36 DIAMETER CLASSES (cm) Figure C.1. Dianeter distribution of balsam fir in the northern hardwood forest type. 185 186 BEECH TREES PER HECT ARE 0 I) 1015202530354045505560 DIAMETER CLASSES (cm) Figure C.2. Diameter distribution of beech in the northern hardwood forest type. 16 I. HEMLOOK TREES PER HECT ARE a 2 n c I I I . . I .L.‘ I wwmamwwwmwwamnwwm DIAMETER CLASSES (cm) Figure c.3. Diameter distribution of eastern hemlock in the northern hardwood forest type. 187 (D '0 RED MAPLE TREES PER HECT ARE 0" I 10 15 20 25 30 35 4O DIAMETER CLASSES (cm) Figure c.4. Diameter distribution of red maple in the northern hardwood forest type. 25 30 35 40 45"— DIAMETER CLASSES (cm) CD l0 SPRUCE TREES PER HECT ARE Figure c.5. Diameter distribution of spruce in the northern hardwood forest type. 188 18 18 14 121-.....mmu... ....... ...... ......" ....... 1o.................... .... ...... ....... mm mm 8‘ IWMOO m m ...-... me 0...... SUGAR MAPLE TREES PER HECT ARE 1015 20 25 so 35'40'45'50 as so 65 7o DIAMETER CLASSES (cm) A m 00...... O 0...... ‘ i Figure C.6. Diameter distribution of sugar maple in the northern hardwood forest type. 3.5 (D 2.5 l0 1.5 WHITE PINE TREES PER HECT ARE ... | | | l l | 10152625 30364045505560'65'70 75'ao'aseobaioé I35 DIAMETER CLASSES (cm) Figure c.7. Diameter distribution of white pine in the northern hardwood forest type. 189 CD fl) YELLOW BIRCH TREES PER HECT ARE I , I I I LI 1 wwmamswwmhwk%fihhh DIAMETER CLASSES (cm) Figure C.8. Diameter distribution of yellow birch in the northern hardwood forest type. ”Pill”! D Diameter distribution graphs for significant tree species in the mixed pine forest type. 1.2 1 E g .. I C E 0.8 3 In E 2 0.4 Ill l ‘4! 0.2 0" T 5 10 15 20 25 DIAMETER CLASSES (cm) Figure D.l. Diameter distribution of aspen in the mixed pine forest type. 190 191 ILL 15 2O 25 DIAMETER CLASSES (cm) 0.7 0.6 0.5 HEMLOCK TREES PER HECT ARE Figure D.2. Diameter distribution of eastern hemlock in the mixed pine forest type. lb RED PINE TREES PER HECT ARE 0 I . I 10 15 20 25 30 35 40 45 50 55 60 65 7O DIAMETER CLASSES (cm) Figure D.3. Diameter distribution of red pine in the mixed pine forest type. 192 (D 25 II) SPRUCE TREES PER RECTARE 3 I I I - 1 5 20 25 30 DIAMETER CLASSES (cm) Figure D.4. Diameter distribution of spruce in the mixed pine forest type. CD p on l0 ‘ WHITE BIRCH TREES PER HECTARE E; p on 5 10 15 20 25 DIAMETER CLASSES (cm) Figure D.5. Diameter distribution of white birch in the mixed pine forest type. 193 GD '0 WHITE PINE TREES PER HECT ARE 0 51015202530354045505550557075508590 Figure D.6. DIAMETER CLASSES (cm) Diameter distribution of white pins in the mixed pine forest type. ”DWI! I Diameter distribution graphs for significant tree species in the mixed conifer/deciduous upland forest type. ASPEN TREES PER HECT ARE 1O 15 20 25 ' so 35 4o DIAMETER CLASSES (cm) Figure 3.1. Diameter distribution of aspen in the mixed conifer/deciduous upland forest type. 194 195 BALSAM FIR TREES PER HECT ARE on '1) Tl ,I rL 5 10 15 20 ' 25 so 35 4O DIAMETER CLASSES (cm) Figure 3.2. Diameter distribution of balsam fir in the mixed conifer/deciduous upland forest type. lb CEDAR TREES PER HECT ARE 0 10 15 20 25 30 35 40 45 5O DIAMETERCLASSES(Cm) Figure 3.3. Diameter distribution of northern white cedar in the mixed conifer/deciduous upland forest type. 195 HEMLOCK TREES PER HECT ARE a I M I l mwmamwwwmamhwfi%kb DIAMETER CLASSES (cm) Figure 3.4. Diameter distribution of eastern hemlock in the mixed conifer/deciduous upland JACK PINE TREES PER HECT ARE forest type. 4.5 4 3.5 25 2 1.5 1 0.5 c 5 10 15 2O fil25 DIAMETER CLASSES (cm) Figure 3.5. Diameter distribution of jack pine in the mixed conifer/deciduous upland forest type. 197 3.5 CD 1.5-W ‘ . I“”W .... some ....“ ”I. RED MAPLE TREES PER HECT ARE M 1015202530354045'50'5550 DIAMETER CLASSES (cm) Figure 8.6. Diameter distribution of red maple in the mixed conifer/deciduous upland forest type. o. 8 ..... 0.5 RED PINE TREES PER HEC'I' ARE 15 20 25 so 35 40 45'50'55reo DIAMETERCLASSES(cm) Figure 8.7. Diameter distribution of red pine in the mixed conifer/deciduous upland forest type. 198 SPRUCE TREES PER HECTARE a 0‘ 5101520253035404550 DIAMETER CLASSES (cm) Figure 3.8. Diameter distribution of spruce in the mixed conifer/deciduous upland forest type. g 25 § 1 A m 5 III a. 3 E 1.5 ‘.‘I 3 1 c I l I r 10 15 20 25 30 35 40 DIAMETER CLASSES (cm) Figure 3.9. Diameter distribution of sugar maple in the mixed conifer/deciduous upland forest type. 3.5 199 25 1.5 1 0.5 . 0" I I r F fl— 5 10 15 20 25 30 35 DIAMETER CLASSES (cm) (I, I) TAMARACK TREES PER HECT ARE Figure 8.10. Diameter distribution of tamarack in the mixed conifer/deciduous upland forest type. g 25 .5. z A m ‘ I.“ O. a 1.5 E 2: 8 I ID i 05. *. III II 0' r l 1 1O 15 20 25 30 35 40 DMMHHERCAASSESkm» Figure 3.11. Diameter distribution of white birch in the mixed conifer/deciduous upland forest type. 200 4.5 3.5 1.5a--—---~ ~- - ~- ~- 1........... .. .. .. .. .. . .. o . f . wwmamxwwwwwwmfiwwmwm DIAMETER CLASSES (cm) (ID II) WHITE PINE TREES PER HECT ARE [0 on Figure 3.12. Diameter distribution of white pine in the mixed conifer/deciduous upland forest type. III I) 1.....W .. ..... .. .... .... .... YELLOW BIRCH TREES PER HECT ARE ‘1" I I 1015202530354045505550557075 DIAMETER CLASSES (cm) Figure 8.13. Diameter distribution of yellow birch in the mixed conifer/deciduous upland forest type. ”mu 1' Diameter distribution graphs for significant tree species in the mixed conifer/deciduous lowland forest type. {I (D a) ASPEN TREES PER HECT ARE 10 15 20 DIAMETER CLASSES (cm) Figure P.1. Diameter distribution of aspen in the :ixed conifer/deciduous lowland forest YPG- 201 202 4 l l l 10 15 20 25 30 DIAMETER CLASSES (cm) ‘ h) .5 O BALSAM FIR TREES PER HECT ARE o M Figure F.2. Diameter distribution of balsam fir in the mixed conifer/deciduous lowland forest type. '0 CEDAR TREES PER HEC'T ARE 15 20 25 30 35 40 45 DIAMETER CLASSES (cm) Figure F.3. Diameter distribution of northern white cedar in the mixed conifer/deciduous lowland forest type. 203 4.5 HEMLOCK TREES PER HECT ARE DIAMETER CLASSES (cm) Figure F.4. Diameter distribution of eastern hemlock in the mixed conifer/deciduous lowland forest type. o'15'2o'2sraoras DIAMETER CLASSES (cm) SPRUCE TREES PER HECTARE 3 l0 5 1 Figure F.5. Diameter distribution of spruce in the mixed conifer/deciduous lowland forest type. 204 II. ‘ i L 10 15 20 25 30 35 4O 45 5O DIAMETER CLASSES (cm) TAMARACK TREES PER HECT ARE Figure F.6. Diameter distribution of tamarack in the mixed conifer/deciduous lowland forest type. I) WHITE PINE TREES PER HECTARE c» 25 so as 40 45 so ssreo'es'7o'7s'eo'esr9?951oo DIAMETER CLASSES (cm) Figure F.7. Diameter distribution of white pine in the mixed conifer/deciduous lowland forest type. APPENDIX 0 Diameter distribution graphs for significant tree species in the mixed conifer swamp forest type. BALSAM FIR TREES PER HECT ARE 10 15 20 25 DIAMETER CLASSES (cm) Figure 6.1. Diameter distribution of balsam fir in the mixed conifer swamp forest type. 205 206 CEDAR TREES PER HECTARE 3 lb q» 10 15 2O 25 30 35 40 45 5O DIAMETER CLASSES (cm) Figure 6.2. Diameter distribution of northern white cedar in the mixed conifer swamp forest type. 1.61 HEMLOCK TREES PER HECTARE p on 15 20 zs'aoTas 4o 45 so DIAMErERCLASSES(cm) Figure 6.3. Diameter distribution of eastern hemlock in the mixed conifer swamp forest type. 207 0| ,5 0| A {a} GI CD 1.5 JACK PINE TREES PER HECT ARE [0 on , , l 5 1o 15 20 25 so DIAMETER CLASSES (cm) Figure G.4. Diameter distribution of jack pine in the mixed conifer swamp forest type. SPRUCE TREES PER HECT ARE s 1015 20 25 so ss'4o'4s'so's?oo DIAMErER CLASSES (cm) Figure G.5. Diameter distribution of spruce in the mixed conifer swamp forest type. 208 ‘ 5.1 ......u................ ......- 1oqm-ooeeeeeeeeeeeeeee ....... ....... TAMARACK TREES PER HECT ARE I l 51015202530354045505560 DIAMETER CLASSES (cm) Figure 6.6. Diameter distribution of tamarack in the mixed conifer swamp forest type. 'é’ E I E g E I 3 D I‘:’ 10 1520 25 so as 40 45 DIAMETER CLASSES (cm) Figure 6.7. Diameter distribution of white birch in the mixed conifer swamp forest type. 209 25 h) 1.5 «- WHITE PINE TREES PER HECTARE wwmzmwwwmsmwmndwm DIAMETER CLASSES (cm) Figure 6.8. Diameter distribution of white pine in the mixed conifer swamp forest type. LIST 01" REFERENCES LII! OI EI’IRIICII Albert, Dennis A., Shirley R. Benton, and Burton V. Barnes. 1986. School of Natural Resources, University of Michigan, Ann Arbor. 32 p. Barnes, Burton V., and warren H.Wagner, Jr. 1981. nignigan IreeEl.A_gnids_t9_the_tresa_9f_nichigan_and_ths_§£sat Lak§§_rggign. University of Michigan Press, Ann Arbor. 383 p. Benson, E. 1976. Logs and lumber: the development of the lumber industry on Michigan's Lower Peninsula, 1837- 1870. Ph.D dissertation, Indiana University, Bloomington. Benzie, John R. 1977. ’ the_nortb_£sntral_§tates. USDA Forest Service. Gen- Tech. Report NC-32. North Central Forest lxperiment Station, St. Paul, Minnesota. 18 p. Bormann, Frank 3., and Gene I. Likens. 1979. Catastrophic disturbance and the steady state in northern hardwood forests. Algz‘_figi. 67:660-669. Bourdo, Eric A., Jr. 1954. A validation of methods used in analyzing original forest cover. Ph.D. dissertation, University of Michigan, Ann Arbor. 207 p. . 1956. A review of the General Land Office survey and of its use in quantitative studies of former forests. Ecology 37. 754-768. Brown, J. H., and B. A. Maurer. 1989. Hacroecology: The division of food and space among species on continents. figigngg 243:1145-1150. Canham, Charles D. 1978. Catastrophic windthrow in the hemlock-hardwood forest of Wisconsin. H.S. Thesis, University of Wisconsin, Hadison. 94 p. Canham, Charles D., and Orie L. Loucks. 1984. Catastrophic windthrow in the presettlement forests of Wisconsin. agglggg 65(3):803-809. 210 211 Cottam, Grant, and J. T. Curtis. 1956. The use of distance measures in phytosociologic sampling. Ecology 37(3): 451-450. Curtis. J. T. 1959. Ths.xsastation_ef_nisconsino University of Wisconsin Press, Madison. Davis, Margaret B. 1981. Quaternary history and the stability of forest communities. Pages 132-153.1n D. C. West, I. H. Shugart and D. B. Bodkin (eds), m anccasaien1_consenta_and_annlication Springer-Vcrlag. New York. Devries, P. G. 1974. Multi-stage line intersect sampling. Bckstein, R. G. 1980. Eastern hemlock (Tang; gangdgngig) in north central Wisconsin. Research Report 104. Wisconsin Department of Natural Resources, Madison, Wisconsin. Erdmann, G. G. 1990. 331311 allgngnignlig Britton. Yellow Birch. Pages 133-147 in R. M. Burns and B. H. Honkala, coordinators- fiilxics_ef_north_AmeriEa USDA Forest Service, Agricultural Handbook No. 654. Washington, D. C. Finley. R- W- 1976. Qri9inal_xssetatien_soxer_ef_niaconain ' . USDA Forest Service, North Central Forest Experiment Station, St. Paul, Minnesota. Map. Frelich, Lee 3., and Craig G. Lorimer. 1991. Natural disturbance regimes in hemlock-hardwood forests of the upper Great Lakes region. figgl,_flgng. 61(2):145-164. Godman, RiChard M., and Kenneth Landcaster. 1990. 113g; ganadgnnig (L. ) Carr. Eastern Hemlock. Pages 604-612 in R. M. Burns and B. H. Honkala, coordinators. fillyig§_gfi . USDA Forest Service, Agricultural Handbook No. 654. Washington, D.C. Godman, Richard M., Harry W. Yawney, and Carl H. Tubbs. 1990. Age: gggghgxn. Marsh. Sugar Maple. Pages 78-91 in R. M. Burns and B. H. Honkala, coordinators. £1111g§_gf . USDA Forest Service, Agricultural Handbook No. 654. Washington, D.C. Heinselman, Miron L. 1973. Fire in the virgin forests of the Boundary Waters Canoe Area, Minnesota. gngtgzngxy Research 31329-382. 212 . 1981a. Fire and succession in the conifer forests of Worth America. Pages 374-405. in D. C. West, H. H. Shugart and D. B. Bodkin (eds), Wigs; WW Springer-Verna. New York- Holling, C. S. 1992. Cross-scale morphology, geometry, and dynamics of ecosystems. IQQL‘JIQDQ. 62(4):447-502. Hough, A. F., and R. D. Forbes. 1943. The ecology and silvics of forests in the high plateaus of Pennsylvania. lggl‘_lgng. 13:299-320. Johnston, William F. 1990a. Innis QQQAQIDLAIAI (L.) Worthern White Cedar. Pages 580-589. in R. M. Burns and I. H. Roman. coordinators. Win- USDA Forest Service, Agricultural Handbook Mo. 654. Washington, D.C. . 1990b. Ln: 3 lgxiging (Du Roi) K. Koch Tamarack. Pages 141-151. in R. M. Burns and B. H. Honkala, coordinators. W. USDA Forest Service, Agricultural Handbook Mo. 654. Washington, D.C. Jordan, J. 8., and W. M. Sharp. 1967. W. USDA Fore-t Service. Research Paper NE-83. Northeastern Forest Experiment Station, Upper Darby, PA. 17 p. Karamanski, T. J. 1989. - . Wayne State University Press, Detroit, Michigan. Kilburn, P. D. 1958. Historical development and structure of the aspen, jack pine and oak vegetation on sandy soils in northern lower Michigan. Dissertation. University of Michigan, Ann Arbor. Lorimer, Craig G. 1977. The presettlement forest and natural disturbance cycle of northeastern Maine. Egglggy 58:139-145. . 1980a. The use of land survey records in estimating presettlement fire frequency. in M. A. Stokes and J. H. Dictation (eds) . W W. USDA For. Serv. Gen. Tech. Report RM-81. 142 p. Mortimer. 1'. J. 1930. W- USDA For. Serv., North Central Exp. Sta. St. Paul Minnesota. Map. Redrafted 1974. 213 Mladenoff, David J., and Forest Sterns. 1993. Eastern hemlock regeneration and deer browsing in the northern Great Lakes region: a re-examination and model simulation. QQDIIIYL_3121- 7(4):1-12. Morisita, M. 1954. Estimation of population density by spacing nethod- W. Sor I. 1:1:7-197. Hutch, Robert W. 1970. Wildland fires and ecosystems - a hypothesis. Icclcgy 51(6): 1046-1051. Mienstaedt, Hans, John C. Zasada. 1990. 21cc; glcnca (Meench) Voss. White Spruce. Pages 204-226. in R. M. Burns and B. H. Honkala, coordinators. Silyicc_cf_flcrth Amazicc. USDA Forest Service, Agricultural Handbook No. 654. Washington, D.C. Palik, Brian J. and Kurt S. Pregitzer. 1992. A comparison of presettlement and present-day forests on two bigtooth aspen-dominated landscapes in northern lower Michigan. W- 127:327'338- Robertson, F. D. 1992a. Ecosystem management of the national forests and grasslands. USDA For. Serv., Washington, D.C. . 1992b. Ecosystem management, public involvement, and clearcutting. USDA For. Serv. Washington, D.C. 3 p. Romme, W. 1980. Fire history terminology: Report of the ad hoc committee. Pages 135-137 in M. A. Stokes, and J. H. Dictorich. (eds). W W- USDA For. Serv. Gen. Tech. Report RM-81. 142 p. Rudolf, Paul 0. 1990. Bing; :ccinccc (Ait.) Red Pine. Pages 442-455. in R. M. Burns and B. H. Honkala, coordinators. 811!AQ§_QI_HQIID_AISIIQI. USDA Forest Service, Agricultural Handbook Mo. 654. Washington, D. C. Rudolph, T. D., and P. R. Laidly. 1990. Bing; hankcigng (Lamb. ) Jack Pine. Pages 280-293. in R. M. Burns and B. 8- Hookah. coordinators. W USDA Forest Service, Agricultural Handbook No. 654. Washington, D. C. Sears, P. B. 1925. The natural vegetation of Ohio. Qh19_11 321- 25:139-149. Spurr, Stephen H. 1952. EQISSL_1DYBDLQIY- The Ronald Press Company, New York. 476 p. 214 Tubbs, Carl H. 1977. ' USDA For. Serv. Gen. Tech. Report NC-39. North Central Forest Experiment Station, St Paul, Minnesota. 29 p. Tubbs, Carl M., and David R. Houston. 1990. Eccnc crcncifclic (Ehrh.) American Beech. Pages 325-332. in R. M. Burns and B. H. Honkala, coordinators. Silyicc_cf Ncrch_Anczicc. USDA Forest Service, Agricultural Handbook No. 654. Washington, D.C. U.S. Department of Agriculture. 1993. HQISDIAIIIID_AIII [crcc§_flccltn_ncpczc. USDA Forest Service, NA-TP-03-93. 57 p. Veatch. J- 0. 1959. 2rasettlenent_foreata_of_nichigan- Department of Natural Resources, Michigan State University, East Lansing. Map. Viereck, Leslie A., and William F. Johnston. 1990. 21cc; marlin. (Mill.) B.S.P. Black Spruce. Pages 227-237. in R. M. Burns and B. H. Honkala, coordinators. fillyicc_cfi W. USDA Forest Service, Agricultural Handbook No. 654. Washington, D.C. Warren, W. G., and P. F. Olsen. 1964. A line intersect method for assessing logging waste. £921_fi£1- 10:267- 276. Weaver, J. E., and F. E. Clements. 1929. 213nc_£cclcgy. McGraw-Hill, New York, New York, USA. Wendel, G. W., and H. Clay Smith. 1990. Bing; cczchnc (L.) White Pine. Pages 476-488. in R. M. Burns and B. H. Honkala, coordinators. Silxisa_ef_ugrth_Amsrica- USDA Forest Service, Agricultural Handbook No. 654. Washington, D. C. Whito, C. A. 1984. 123:3]. Bureau of Land Management, United States Department of the Interior, Washington, D. C. Whitney, Gordon G. 1986. Relation of Michigan's presettlement pine forests to substrate and disturbance history. Ecclccy 67: 1548-1559. . 1987. An ecological history of the great lakes forest Of liabigen. 11.3921- 78: 443-458. Wilson, Louis F. 1977. A_suide_t9_inaect_in1urx_of_conifera 1n_tnc_L§kc_§;cccc. USDA For. Svc. Agri. Handbook No. 501. 218 p.