“ .1 ."X . . . ‘ . ~ ’ ‘4‘ N v2.4: u- ‘ ‘ n ‘3" «.117 A. —.Ad-:< ‘-".‘m~'r 1' “1"" ,u 1.. ' .o :23"? ”it 33" a . '3"; 4“ \v. \ ‘. W.‘ c.-. a“... , ...—-— w my.” ’ I a .9? 3. -5 L— ‘vtf‘fif‘ tfirgr- - ' 13a $7.3"! mm. “z. m"$dfiu‘ .. "-4. any“- ~‘gth‘“'kf.A'LW-. .. .-‘-'r.~§' M ‘11 233 ' ‘ ‘0. ~. lo :1; ." alts: ‘Pz‘fiz.. L,- «S ”in rm; “- u 552 v‘nx \— 3 x ”1»: J “M; - wt lllllllllllllllll This is to certify that the thesis entitled A Comparison of Pre-European Settlement and Present—Day Forests in the High Plains Region of the Huron National Forest, Northern Lower Michigan presented by Michael Joseph Leahy has been accepted towards fulfillment of the requirements for Masters degnwin Forestry // //j/ \MMJfA f1 *7/ Major professor 0.7639 MSU is an Affirmariw Action/Equal Opportunity Institution LIBRARY MIchIgan State University PLACE II RETURN BOXbromovotN-chockomfiun macaw. :I'O AVOID FINES Mum on or bdoro duo duo. DATE DUE DATE DUE DATE DUE L. a 'f" , l h“. I j J r! u - 2 11995 W157 :3 5:19 L348; * 31313 0' 4 W1 Wm“.- ... . . .. 1 n .' Wm! SUPPLEMENTAR A i MATERIALY IN max OF BOOK A COMPARISON OF PRE-EUROPEAN SETTLEMENT AND PRESENT-DAY FORESTS IN THE HIGH PLAINS REGION OF THE HURON NATIONAL FOREST, NORTHERN LOWER MICHIGAN BY Michael Joseph Leahy A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Forestry 1994 ABSTRACT A COMPARISON OF PRE-EUROPEAN SETTLEMENT AND PRESENT-DAY FORESTS IN THE HIGH PLAINS REGION OF THE HURON NATIONAL FOREST, NORTHERN LOWER MICHIGAN BY Michael Joseph Leahy General Land Office survey records of 1838-1846 were used to reconstruct the composition, structure, and disturbance regime of pre-European settlement forest communities along an edaphic gradient in northern lower Michigan. These data were compared to plot data of second-growth forest communities on similar sites. A Rings banfisiana cover type still dominates a third of the landscape as it did in the pre-European settlement period. Age; .gaggharum, Eopulus species, Qggrgug species, 21111 amezigana, Age: rubrum, and fietula papyrifigra_have supplanted Rings s' os , Einus strobus, Isgga canadens's, and Eggpg grgngifiglia as forest dominants along most of the dry-mesic, mesic, and wet-mesic areas of the edaphic gradient. Forest communities have smaller trees and higher tree densities today than in the pre-European settlement period. Likewise, disturbance regimes have been substantially altered. Human- imposed disturbances since 1870 have interacted with physical factors to produce the forests we see today. Dedicated to my parents. iii ACKNOWLEDGEMENTS I wish to thank.my guidance committee for their efforts. The committee was chaired by Dr. Kurt Pregitzer, and included Dr. Don Dickmann and Dr. Peter'Murphyu Dr. Pregitzer provided generous financial support throughout.my graduate program, as well as insight into the scientific process. Drs. Dickmann and Murphy both provided thorough reviews and an honest evaluation of this manuscript. I would like to thank.Dr. Carl Ramm for providing data on the present-day forests of the Huron National Forest. Dr. Lee Barnett of the State Archives of Michigan helped with the acquisition of historical records. Mr. Loren Berndt of the Soil Conservation Service provided assistance in updating old soil maps. Mr. Sherm Hollander of the Michigan DNR provided MIRIS cover-type data. My fellow graduate students, Andy Burton, David Price, and Jill Fisher provided helpfull advice throughout my program. iv TABLE OF CONTENTS page LIST OF TABLES..........................................viii LIST OF FIGURES............................................x INTRODUCTION...............................................1 Background............................................1 Ecological theory and land management............1 Previous studies of the pre-European settlement forestOOOOOOOOI00.0.00...O0.00.00.00.0000000000003 Knowledge gaps...................................5 Problem statement.....................................6 Hypotheses.......................................6 Objectives.......................................7 METHODS AND MATERIALS......................................8 study area............................................8 The Huron National Forest........................8 Climate..........................................9 Glacial geology..................................9 Soils...........................................12 Human history...................................12 The General Land Office (GLO) surveys................16 Data collection.... ..... . ................... .........19 D‘t‘ ‘n‘ly.i’0.0...0.0...00......00.0.00000000000000022 Species-site relations of the pre-European settlement forestOOOOOOOOOOOOO0.0.0.00.00000000022 Pre-European settlement tree species aSSOCiationSOO0.0...OOOOOOOOOOOOOOOOOOOO0.0.0.0.24 Reconstructing the pre-European settlement forest. ...... O00....OOOOOOOOOOOOOOOOOOOOOO ...... 24 Forest cover type map....... .............. ......29 Disturbance regimes O O O O O O O O O O 0 O O O O O O O O O O ....... 0 3o Present-day forests: composition and struCture O O O O O O O O O O O O O O O I O O O O O O O O O O O O O O ...... O O O 31 Analyzing the changing forest ....... . ......... ..32 RESULTS AND DISCUSSION .......................... . ...... ...33 The character of the pre-Buropean settlement forest..33 Witness tree species-site relations.............33 Witness tree associations of the pre-European settlement forest...............................43 Pre—European settlement forest communities across the edaphic gradient.. ................... .......44 The pre-European settlement landscape patterns..85 150 years of forest change...........................94 Changes in forest communities across the edaphic gradient.....O0..0......00......0.00.00.00.0000094 Changes in the landscape patterns since Euro- American settlement...I...0.0...0.0.0.0000000000109 CONCLUSIONSOOI....0.00000000000000000000000.0.0.000000000113 Evidence for/against hypotheses..... ..... ...........113 A verbal model of the study area's vegetation.......114 Management implications..... ......... ...... ........ .115 vi APPENDICESOOOOOO0.0...IOOOOOOOOOOOOOOOOOOOOOOOOO0.0.0....117 Appendix A: Bar graphs of witness tree species importance values by site type.................117 Appendix B: Structural data summaries for pre-European settlement and present-day communities.........127 Appendix C: Present-day communities: structure and comPOSitj-on...OI.00...0....00.000.00.0000000000129 Appendix D: Bar graphs of compositional changes since Euro-American settlement. 0. O O O O O O O O O O O O O O O. O O. .132 LITERATIIRE CITED-0.. OOOOOOOOOOOOOOOOOO 0.0.000000000000000138 vii LIST OF TABLES page Table 1: Scientific, common, and surveyor names of tree speciesinthestudyarea...................................21 Tablez: Sitetypenames...................................25 Table 3: Standardized residuals (Haberman 1973) expressing the degree of association between witness tree species and Site factorSOOOOOOOOOOOOOOOOOOOOCOOOOO0.00.00.00.0000000000042 Table 4: Composition and tree species diversity of pre-European settlement forests (trees 2 10 cm d.b.h.) on xeric SiteSOOOOOOOOOOOOOOOOOO0....0....0.0.0.00000000000000048 Table 5: Composition and tree species diversity of pre-European settlement forests (trees 2 10 cm d.b.h.) on dry- m381c SiteSOOOOOOOOOO0.0.0.00.0...OIOOOOOOOOOOOOO0.0.000000053 Table 6: Composition and tree species diversity of pre- European settlement forests (trees 2 10 cm d.b.h.) on mesic SiteSOOOOOOOOOOOOOOOOOOOOOOOIOOOO ...... 00.0.000000000000000060 Table 7: Composition and tree species diversity of pre-European settlement forests (trees 2 10 cm d.b.h.) on wet- meSic SiteS0.0..0.0.00......OOIOOOOOOOOOOOOOOOOOOOOO0.0.0.0069 Table 8: Composition and tree species diversity of pre-European settlement forests (trees 2 10 cm d.b.h.) on hYdric SiteSOOOOOOCOO...0.0000000000000000000.00.00.0000000076 Table 9: Pair-wise comparisons of compositional differences between pre-European settlement communities using the chi- squaredtestofindependence................................86 Table 10: Species composition by percentage of the study areaOOOOOO0......0.0....000......00....0.0.0.00000000000000088 Table 11: Pre-European settlement extent of vegetation cover types forthestudyareaOOO0.00000000000000IOO0.0.0.0000000090 viii Table 12: Soil and landform composition of the study area. .90 Table 13: Extent of the different soil drainage types in the study areaOO0.0.....0.00..........O0.0.0.000...0.0.00.00000091 Table 14: Disturbance regimes of the pre-European settlement forestOOOOOOOO00.............0.........OOOOOOOOOOOOOOOO0.00.92 Table 15: A comparison of pre-European settlement forest structural attributes (trees. 2 lo cm. one hectare (Canham and Loucks 1984) and crown fires (Lorimer 1980a). The GLO survey provided a systematic sample of the study area along a one mile (1.6 km) square grid system. I used line transect theory to determine the pre-European settlement disturbance rotations for windthrow and fire disturbances in each of the forest cover types. The following steps allowed me to elucidate the disturbance rotations: (1) I totaled the number of miles of survey lines in each cover type. Then (2) I used the section line summaries to calculate the total length of windthrown or burnt section lines in each cover type. The presence of particular vegetation types (aspen, oak, and pine "thickets") were also included as burned areas (Lorimer 1980a: Palik and Pregitzer 1992). Thirdly, the total length of disturbed section lines in each cover type was divided by the total length of survey lines in that cover type: this gave the percent of land affected by disturbance for a cover types I divided the above figure by fifteen years 31 to yield the average % area disturbed annually in the cover type. Lorimer (1977: 1980a) deemed fifteen years to be a conservative estimate of the number of years a prior burn or windfall would still be visible to the surveyors. Lastly, 100 was divided by the average % area disturbed annually in a cover type. This yielded the pre-European settlement disturbance rotation for either fire or windthrow for the cover type in question. 8 - ests: Com os'tio t ct re Present-day forest stand data came from twenty nine stands sampled within the study area for the development of an ecological classification system (ECS) of the Huron NF (Michigan State University, Forestry Department data file). All stands selected were well-stocked, naturally regenerated, at least one hectare in size, and free from recent disturbance. Trees were defined as stems 2 10 cm d.b.h. Stand ages ranged from sixty to ninety years. The ECS sample stands failed to represent the study area’s xeric and hydric sites. I grouped stands by soil drainage class as for the pre- European settlement forest data to permit then-and-now comparisons by site type. I pooled stand data by site type and determined present-day absolute and relative density and dominance by species. As before, I also calculated an importance value for each species by site type. Tree species diversity was calculated by site type using importance values 32 in the Shannon-Weaver formulation (Peet 1974) . For each present-day forest site type I determined its mean basal area/ha, density (trees/ha) , and mean d.b.h. by averaging across stands within a type. a in o s I compared the species composition of pre-European settlement forests to today's forests by site type with the chi-squared goodness of fit test (Gill 1978) and Bray and Curtis' (1957) index of community similarity; It could not statistically test for differences in mean basal area/ha and trees/ha between pre-European settlement and present-day forests“ To test for changes in mean tree d.b.h. between pre- European settlement forests and existing forests I used paired t-tests (Gill 1978). I assessed changes in the areal extent of forest cover types between present-day and pre-European settlement forests. To determine the areas of pre-European settlement forest cover types I used a dot grid and the forest type map. Values for the areas of present cover types were obtained for the study from the Michigan Forest Resource Inventory Program (MIRIS) and compared to pre-European settlement data. 33 RESULTS AND DISCUSSION THE CHARACTER OF THE PRE-EUROPEAN SETTLEMENT FOREST Witness Tree Spggigs-Site Belgtigns In the correspondence analysis which included wetland species, 80% of the variation (total inertia) was accounted for by the first two dimensions. The correspondence analysis in which wetland species were excluded had 82% of the variation (total inertia) explained by the first two dimensions. Figures 5 and 6 are biplots of the coordinates of the species (column coordinates) for the first two dimensions of each analysis. The species' patterns in Figures 5 and 6 demonstrate the response of the witness tree species to a multidimensional environmental gradient likely comprised of soil moisture, nutrient availability, light availability, microclimate, and disturbance factors. The species probably responded individualistically in the pre-European settlement forest, creating a joint distribution of successive replacement (Austin 1985). Hence, the arch evident in both biplots represents an inherent property of the data (Wartenberg et al. 1987). Tamarack lies on the far right side of Figure 5's biplot along with white cedar. Both species have the competitive advantage on sites with.high soil moisture (Burns and Honkala 1990). Jack pine falls on the far left side of both biplots (Figures 5 and.6) probably due to its drought tolerance (Burns and Hokala 1990). A facile interpretation of the patterns in 34 Figure 5: Correspondence analysis of all witness tree species: axes 1 x 2 (80% of the variation). Axis 2 o 35 I l 1 Sugar maple Beech '. 'Haulock I White pine . Red oaks I" .Red pine - Aspen . . Jack pine — White m - Tamarack 1 J 1 -1 O 1 2 36 Figure 6: Correspondence analysis of upland witness tree species: axes 1 x 2 (82% of the variation). 37 1.0 r I ' Red pine 0.5 — y ' White pine (\J -92 x <( ' Red oaks I Aspen 0.0 *- IHanlock . Jack pine . Beech ' Sugar _05 1 1 maple —1 o 1 Axis 1 38 Figures 5 and 6 would state that soil moisture dominates the environmental gradient observed. However, soil moisture was highly correlated with nitrogen mineralization (Pastor et a1. 1984) and fire regimes (Clark 1989) in the upper Midwest. These correspondence analyses only give a limited view of the factors which engendered the pre-European settlement forest. In both biplots the red oak complex and red pine are grouped together (Figures 5 and 6). From this I infer that most members of the pre-European settlement red oak complex in the study area were either black oaks or northern pin oaks. Black and northern pin oak’s niche dimensions are closer to red pine’s than are northern red oak’s (Burns and Hokala 1990). The nonmetric multidimensional scaling (NMDS) of witness tree species produced.a:regular'pattern among the species best described.as.a curvature (Figure:7). I interpret the curve as broadly defining a' soil moisture gradient, as in the correspondence analyses. White cedar and tamarack lie at the lower left side of the curve and jack pine lies on the lower right side (Figure 7): indicative of a wet-dry gradient. The ordination configuration became constant after 44 iterations, with a stress of 5%, which according to Kruskal (1964) is fairly good. The results of the binary discriminant analyses (BDA) gave direct indication of the significance of species- substrate interactions (Strahler 1978) . Sugar maple and beech 39 occupied sites with loamy sand textures on well-drained, ice- contact landforms (Table 3). Hemlock competed best on loamy sand, sandy loam, and loam textures on well-drained, morainal landforms (Table 3). These species have likely evolved in areas with moderate nutrient availability: dense shade during juvenile stages: and gap-phase disturbances. Grime (1977) would classify them as stress-tolerant competitors. Tamarack was restricted to very poorly drained organic soils in depressional features of ice-contact and outwash landforms (Table 3). Somewhat poorly drained and very poorly drained.organic soils and.silty clay loam soils wereldominated by white cedar predominately on outwash plains (Table 3). White cedar competes better on more basic and nutrient rich sites than does tamarack. Tamarack requires full sunlight for optimal competitive ability, while cedar needs much less sunlight (Burns and Honkala 1990). Both occupied similar sites in the pre-European settlement forest probably because of their tolerance to saturated soil conditions. While all stress-tolerators (Grime 1977), the pines exhibited a range in their adaptation to xeric, infertile, fire-prone sites. Jack pines occurred on extremely well- drained, sandy outwash plains (Table 3). Red pines occupied somewhat excessively drained, sandy end moraines (Table 3). The nutrient, energy, and.water-conserving habit of evergreen species (Monk 1966) fits well with the pre-European settlement distribution of these pine species. However, white pine 40 Figure 7: Ordination of witness tree species by nonmetric multidimensional scaling. Axis 2 41 Axis 1 I Sugnrmufle I Reignne m . iflutegnne Helnlock ' Red oaks I Jack pine I Aspen I -— White cedar _ . Tamarack L J O 1 4&2 ._o>o. an o. scoo_._co_. o_u._».u. a . T H .m..ao .Ne.oo. on.. .oo..~ .o~.~. .oo.nk .oo.oo~ .oo.oe_ .Nk..~. .ms.oo o:..> a as... on.» - o~.n No.9 e..~. oo.e. o°.n . . aces sxuax .o.~- km.o~ - ec._ .e.~- 85.“. mm.o.- Po.e~ om.n- ~o.~- aqua uca sun: - o..~ - on.» - - - - . . nae. >a.u so._m .o.e - sn.o- o~.n eo.~ .~..- - - en.o- an.n noon ao.~_ oo.o me... o... .o.~ 65.“. n~.ns- - om.~. e..o goo. saga» n..n .~.~- oe.o mm.~- .o..- so.e. - - oa.~. os.e. ucau sauce oo.~.. om.~.- so.o nk.~. oe.o- km.o. 95.x. ~o.n~. me.n.. 93.... scam >uo.ooo .e_u_~cam .n~.-. .oo.~o~ ~e.n .me.on .n_.n~ .oo.eo ..m.a~m .~6.e~. .~..o~_ .oo.uo. o:..> a -.n. k..k~ .m..- .o.~ o..~- No.8- ~o.o- o..o~ oo.n- no.~- soon >Lo> mn.k oe.e - m~.o on.~ mc.n. kn.o- ~e.o- n~..- o~.. Loon sogzusou a» room on.~. en.n- .~.o n..~- as.» 96.“- oo.o~- . o..o~ n°.o. ..o: o» ..o: .uo: o~.e- eo.e- o_.~ om.o ee.n ~n.o. e~.~. - k..~- k..n- o>_¢uouxo aagaaeom e_.~.. nu.o- m~..- .k.e- om.m. no.8 ~e.k~ o~.m_. .6..—- o~.o. u>_auouxm ooac_oeo .eo._m .nn.c~ em.n so.“ .o_.n~ .kn.om .ao.co~ .oe.o .oo.oo ..o.om on..> u nk.o- so.“ om.~- -.n ao.o- e..o- oo.n~ eo.~ nu.~.- o~.o. - guosuao 88.0 so.~. oe.~ so..- on.“ ~o.o no.k.. o..n. om.m 65.4 .c_..ox on.~ m°.e- om.o on.~. k~.~ eo.o ke.o- o~.o .m.o. no.5 couscou-ou_ acosuco. .o_o..u I I 3.8.5.. 5.8 9.8 9&2 2:. E... 8X 3.2: 68.. 38.. «o_es no“ ou_=3 6oz guns ..aa» rues» .mcouua~ au_o use «o_ooea out» unocu_s cassava comua_oomme so cocoon as» oc_aaaeaxo anno— caaconezv a.o30_uac van_ucoucaum "n 0.9.» 43 competed best on more mesic sites. These sites ranged from somewhat excessively drained to somewhat poorly drained sandy loams, loams, and mucky sands on ice-contact and morainal landforms (Table 3). The red oak complex failed to show a significant association with any of the site factors (P50.05, G-statistic, Table 3). Curiously, aspen species mainly occupied wet sites in the pre-European settlement forest. Aspen species were found on somewhat poorly drained, silty clay loam, loam, and mucky sand soils (Table 3). The surveyors often described aspen’s lowland position: "...dense aspen swamp...aspen and alder swale...wet aspen flat...swampy aspen thicket..." Witness Tr_e_¢_a Associations of the Pre-Euronean Settlement Fotgst In the NMDS biplot, white cedar and tamarack form a loose, but discernable group (Figure 7). Beech, hemlock, and sugar maple cluster closely together (Figure 7). Jack pine forms a distinct group unto itself as does white pine (Figure 7). Red.pine, the red oak complex, and the aspen species form a loosely joined. group (Figure 7). These five groups represent hypotheses of pre-European settlement forest associations. Associations here indicate tree species with similar ecological amplitudes. Cluster analysis finds the "natural grouping" in a data matrix, but will produce clusters from random.data (James and McCulloch 1990). Using the Euclidean distance metric, both 44 the average linkage method and.Ward’s minimum variance method produced five similar clusters of pre-European settlement witness tree species (Figures 8 and 9) . For the average linkage dendrogram (Figure 8), the cophenetic correlation coefficient (Sneath and Sokal 1973) was 0.9973 The cophenetic correlation coefficient of Ward's dendrogram (Figure 9) was 0.996. Both dendrograms were in good agreement with the original distance matrix. As in the correspondence analysis (Figure 5) and the nonmetric multidimensional scaling (Figure 7): beech, sugar maple, and hemlock form a distinct group in both dendrograms (Figures 8 and 9). The dendrograms show a distinct tamarack—white cedar group. From these numerical procedures I surmise that at least three forest associations occurred in the pre-European settlement forest: jack pine, beech-sugar maple-hemlock, and tamarack-white cedar. Not surprisingly, these three associations form the xeric, mesic, and hydric nodes of the soil moisture gradient. -- o-e.n Set lement _o est 0 nities 1. o s e g=-;,. Gradient The site types (Table 2) chosen to group GLO corner points represent diffuse, but definable forest communities of the pre-European settlement forest. Hereon I consider forest site types and communities synonymous. In describing the composition of these forest communities I defined dominant tree species as those with an importance value 2 10%, and 45 Euclidean Distance 0.000 500.000 Jack pine 209.821 Red pine ————— 42.053 HeIlock 11.490 Beech 9.132 Sugar IIple __. 22.158 Uhite pine 10.604 Red oaks 9.136 Aspen 18.653 Tamarack :}. 8.238 Uhite cedar r (cophenetic correlation) = 0.997 figure 8: Dendrogram showing possible associations of witness tree species. Clusters were fonIed using the average linkage Iethod (Sneath and Sokal 1973). Euclidean Distance 0.000 500.000 Jack pine — 358.758 Red pine -——————‘ 60.456 ueIlock 12.275 Beech 9.132 Sugar Ieple . 47.136 Hhite pine 11.094 Red oaks 9.136 Aspen 31.733 TaIerack 8.238 Uhite cedar r (cophenetic correlation) = 0.996 Figure 9: Dendrogram showing possible associations of witness tree species. Clusters were formed using the Hard linkage method (Hard 1963). 46 associate (subdominant) species as those with an importance value of 1-9%. Current ecosystem studies in northwestern lower Michigan have determined that soil moisture availability and related nutrient dynamics constrain forest production and composition in a probabilistic manner (Host et al. 1988: Zak et a1. 1989: Host and Pregitzer 1992). Hence, the forest communities I have defined for the pre-European settlement forest, while artificial, do have a functional basis. Below I describe the structure and composition of the pre-European settlement forest communities in relation to interacting site and disturbance factors. The Xeric Forest Community The dominant tree species of this community were jack pine, red pine, and white pine (Table 4, Appendix A: Figure 21) . Associates included the aspen species, the red oak complex, and hemlock (Table 4, Appendix A: Figure 21). Jack pines were small and numerous (146 trees/ha, 5 ufi/ha, Table 4) . Red pines were scattered and large (35 trees/ha, 6 mZ/ha, Table 4). White pines were infrequent and very large (9 trees/ha, 2 mZ/ha, Table 4) . I defined this community as the jack pine-red pine-white pine forest type for the cover type map (see maps). The numerical procedures consistently separated jack pine (Figures 5-9) from the other species along multidimensional environmental gradients. This community occupied droughty, sandy outwash plains and drainageways (Table 3, see maps). Surveyors described this community as: 47 ...land...very slightly rolling. Thinly timbered with a little yellow and spruce pine and stunted oaks: spruce pine plains: very handsome, high [and] nearly level...have a good view of Au Sable [river] for a long distance. Community-level overstory basal area (trees 2 10 cm: d.b.h.) was 13 nf/ha (Table 4). Similar sites in northern lower Michigan had low potential net N mineralization rates (Zak et al. 1989) . The surveyors often mentioned the presence of Minn species in the understory of this community. yntgininn species were indicative of low N availability and low soil moisture in northern lower' Michigan (Host and Pregitzer 1991). Jack pine’s pre-European settlement size structure (Figure 10) shows a large, likely even-aged cohort of very small trees. This even-aged structure fits well with jack pine’s autecology. Jack pines are extremely shade intolerant (Burns and Honkala 1990). For jack pines to establish themselves as a canopy dominant, they need to regenerate under conditions of full sunlight on a mineral soil seedbed. Jack pines cannot survive for long under canopies of any tree species and therefore typically do not develop an uneven-aged structure. Crown fires were a frequent occurence in this community. The fire rotation (White and Pickett 1985) was z 100 years. Stand-replacing windthrows were infrequent, their rotation period (Canham and Loucks 1984) was s 2000 years. Whitney 48 Table 4: Composition and tree species diversity of pre-European settlement forests (trees 2 10 cm d.b.h.) on xeric sites. Percent Percent Relative Relative Tree species Density DoIinance IV Density DoIfinance m ggnksians 68.38 34.73 51.551 146.202 4.583 mm 1951023.! 16.23 41.99 29.11 34.70 5.54 m m 4.19 16.76 10.48 8.96 2.21 m species 2.90 1.35 2.13 6.20 0.18 m m-velutina- 3.43 0.78 2.11 7.33 0.10 glligoidalig mm mm 1.14 1.29 1.22 2.44 0.17 m grandifglia 0.99 0.83 0.91 2.12 0.11 m); m 0.61 0.67 0.64 1.30 0.09 195.11.: species 0.46 0.36 0.41 0.98 0.05 mg pgpzrifera 0.38 0.27 0.33 0.81 0.04 Guercus an; 0.30 0.23 0.27 0.64 0.03 m m 0.23 0.12 0.18 0.49 0.02 111_uj_p gttjdentalis 0.15 0.15 0.15 0.32 0.02 nun 161-1519; 0.15 0.15 0.15 0.32 0.02 gm; species 0.15 0.08 0.12 0.32 0.01 T_i_l_j_g m 0.08 0.12 0.10 0.17 0.02 M m 0.15 0.03 0.09 0.32 0.00 ignipgtng yirginiana 0.08 0.09 0.09 0.17 0.01 Totals 100.00 100.00 100.00 213.80 13.20 Diversity (8') = 0.6 1 Importance Value = (relative density + relative dominance)/2 * 100 2 trees/ha 3 basal area (mg/ha) 49 (1987) found the pre-European settlement fire rotation period for jack pine stands in nearby Roscommon and Crawford Counties was z 83 years. Simard and Blank (1982) determined that the pre-European settlement fire-return time (average time between successive disturbances at a particular site) for "major" fires (fires burning 4047+ ha) was z 35 years on the Mack Lake outwash plain (the dominant outwash plain in the study area). With a crown fire burning over this community at the minimum of every 100 years, it is safe to say that new forest stands in this community were established after major fire events. Jack pine’s longevity is from 80-100 years and its cones are predominantly serotinous (Burns and Honkala 1990). Jack pine depended on fire to reproduce and maintain itself in this community; The presence of large red and.white pines in this community indicates that fires probably burned in a stochastic pattern across the outwash plains: allowing certain juvenile red and white pines to mature. I conceptualize this community in pre-European settlement time as a slightly undulating "sea" of jack pine. Small "islands" of white pine, red pine, and hemlock occupied areas protected from fire (the lee side of lakes and wetlands) or lucky enough to miss several fires in a row. Small aspen clones and pygmy forests of oak grubs occurred as patches in the jack pine matrix in areas that had experienced repeated intense fires. 50 Figure 10: Histogram showing distribution of jack pine witness trees by 5 cm d.b.h. classes. 51 020 some: .6 E8 ESEEQ mm mm To NO 0.0 v.0 0.0 ©.O NO SGGJi SSGUIIM )0 uouJodOJd 52 The Dry-Mesic Forest Community Red pine, jack pine, and white pine were the dominants of this community (Table 5, Appendix A: Figure 22). Associates of this community included: species of the red oak complex, hemlock, aspen species, beech, sugar maple, and white oak (Table 5, Appendix A: Figure 22). Red pine was classified as a separate group in both cluster analyses (Figures 8 and 9). I defined the red pine-jack pine-white pine cover type of the forest type map (see maps) from this community. This community had many, large red pines (91 trees/ha, 17 nfi/ha, Table 5) and many, small jack pines (112 trees/ha, 4 nfi/ha, Table 5). White pines were sparsely distributed, but just as large as the red pines (32 trees/ha, 6 mz/ha, Table 5) . When surveyors mentioned the understory in this community they described it as either " free from undergrowth" or dominated by "hazel" (gotylus species) and "dwarf oaks." Surveyors described this community as: Handsome pinery...mostly yellow pine, some white pine: yellow and white pine not large but thrifty and tall: timber a few large pines and dense thicket of aspen. This community was found on sandy end moraines and coarse-textured ice-contact features (Table 3) typically directly east or northeast of the xeric forest community (see maps). Community-level overstory basal area (trees 2 10 cm d.b.h.) was 31 mZ/ha. From red pine's size structure (Figure 53 Table 5: Composition and tree species diversity of pre-European settlement forests (trees 2 10 cm d.b.h.) on dry-mesic sites. Percent Percent Relative Relative Tree species Density' DoIfinance IV Density DoIinance Lime m 29.00 53.97 41.491 91.412 16.683 Zing; banksiana 35.67 11.98 23.83 112.43 3.70 fling; gtngtng 10.17 20.51 15.34 32.06 6.34 gggttgg tgtgg-velutin - 5.33 2.12 3.73 16.80 0.66 gllipsoidalig Igggg ggngggngig 3.50 3.71 3.61 11.03 1.15 Egggtng species 5.33 1.88 3.61 16.80 0.58 gngng grandifolia 4.83 2.20 3.52 15.22 0.68 AESL saccharum 2.00 1.72 1.86 6.30 0.53 Qggttng git; 1.67 0.59 1.13 5.26 0.18 M m 0.83 0.19 0.51 2.62 0.06 10215 gtgidentalis 0.50 0.37 0.44 1.58 0.11 2122! species 0.33 0.31 0.32 1.04 0.10 ggtgtg pgnyrifgra 0.33 0.24 0.29 1.04 0.07 Frgxinnt ric 0.17 0.09 0.13 0.54 0.03 323212 gllgghaniensis 0.17 0.06 0.12 0.54 0.02 59135 tglgamea 0.17 0.06 0.12 0.54 0.02 Totals 100.00 100.00 100.00 315.20 30.90 Diversity (8') = 0.7 1 Importance Value = (relative density + relative dominance)/2 * 100 2 trees/ha 3 basal area (ma/ha) 54 Figure 11: Histogram showing distribution of red pine witness trees by 5 cm d.b.h. classes. 55 00¢ sci Ba .6 EB smLmEsQ 00 00 we. mm 00.0 0r.0 000 00.0 888)) sseuum )0 UOIIJOdOJd 56 Figure 12: Histogram showing distribution of white pine witness trees by 5 cm d.b.h. classes. 57 wow 9:0 9:2, .5 Es 55800 00 00 we 0N r 00.0 1 070 1 000 8881i sseuum Io uouJodwd 58 11) I infer that its age structure was composed of two even- aged cohorts. White pine’s size structure (Figure 12) indicates a possible multi-age class structure. Up to five different cohorts of white pine may have existed in the pre- EurOpean settlement forest (Figure 12). The fire rotation period for this community was 3 100 years. The rotation period for catastrophic windthrow was s 900 years. The structure and composition of this community resulted from an interaction between edaphic factors and the fire regime. IRed pines need a mineral seedbed and the absence of competing plants to reproduce themselves (Burns and Honkala 1990). Red pines therefore depend on fire to maintain themselves as canopy dominants and did so in thiSjpre-European settlement community. Both white and red pine's longevities (200-450 years and 200-300 years, respectively) exceed the fire rotation period. This provides further evidence of the importance of fire in this community. The hypothesized even- aged cohorts of red and white pine described above (Figures 11 and 12) probably originated after major fire events. I visualize this community as the classic Great Lakes pine forest as described by Curtis (1959). Park—like stands of large white and red pines towered over a permanent understory of red oak species. Where intense lightning fires occurred frequently enough to eliminate the red and white pines, jack; pines and aspen. maintained. themselves. In protected microsites hemlock managed to reproduce and persist 59 along with beech and sugar maple. The Mesic Forest Community Forest dominants here included: hemlock, beech, sugar maple, and white pine (Table 6, Appendix A: Figure 23). Associates of these dominant trees were red pine, the red oak complex, basswood, aspen, white cedar, black ash, and jack pine. Medium sized hemlock (57 trees/ha, 7 nfi/ha, Table 6), beech (60 trees/ha, 4 nfi/ha, Table 6), and sugar maple 39 trees/ha, 4 mz/ha, Table 6) made up the bulk of this community. However, white pines occurred throughout this community and attained their largest sizes here (24 trees/ha, 5 mZ/ha, Table 6) . From this community I defined the hemlock- beech-sugar maple-white pine cover type for the forest type map (see maps). Surveyors mentioned pine, beech, hazel, red maple, aspen, sugar maple, and hemlock as common understory species. One surveyor described this community as: ...land rolling, good 2nd rate soil, clay and sand...Sugar, Beech, Elm, Maple, Lynn, Ironwood, and some first rate white pines. Hemlock, beech, and sugar maple were grouped together in all of the ordinations and classifications (Figures 5-9). I surmise this reflects their similar environmental tolerances. Physiographically this community occupied end moraines, ground moraines, and ice-disintegration features (Table 3, see maps). Topographically these areas have short, steep hills, swell and Table 6: Composition and tree species diversity of pre-European settlement forests (trees 2 10 cm d.b.h.) on mesic sites. Percent Percent Relative Relative Tree species Density Dominance IV Density Dominance 15.992 m; 23.79 27.00 25.401 57.412 6.863 Egan; grandifglia 25.02 14.88 19.95 60.37 3.78 5235 ggttharum 16.03 14.16 15.10 38.68 3.60 fling; gttgtgg 9.83 19.81 14.82 23.72 5.03 Pings resinosa 7.24 12.06 9.65 17.47 3.06 Qggttng gntgg-vglutin - 3.79 3.68 3.74 9.15 0.93 gllipgojdalig 11113 gngricana 2.07 2.09 2.08 4.99 0.53 Egggtgg species 2.93 0.87 1.90 7.07 0.22 lngig gttiggntgtig 1.38 1.14 1.26 3.33 0.29 Frgxinng nigtg 1.38 1.02 1.20 3.33 0.26 fling; tnntginng 1.72 0.55 1.14 4.15 0.14 figtntn pgnyrifera 1.21 0.45 0.83 2.92 0.11 593; 2292!! 1.03 0.55 0.79 2.49 0.14 fling; species 0.86 0.53 0.70 2.08 0.13 figtgtg species 0.52 0.40 0.46 1.25 0.10 Qggttggpgttg 0.69 0.04 0.37 1.66 0.01 flgtgtg gllggngniensig 0.17 0.40 0.29 0.41 0.10 [gagingg gngricana 0.17 0.33 0.25 0.41 0.08 ggttyn virginiang 0.17 0.04 0.11 0.41 0.01 Totals 100.00 100.00 100.00 241.30 25.40 Diversity (fl') 8 0.9 1 Importance Value = (relative density + relative dominance)/2 * 100 2 trees/ha 3 basal area (mg/ha) 61 swale features, and.small hollows. 'This community occurred.on loamy sand, sandy loam, and loam textures (Table 3). .Present- day communities on similar sites in northwestern lower Michigan had relatively high potential rates of N availability (Zak et al. 1989) . Community-level overstory basal area (trees 2 10 cm d.b.h.) here was 25 nfi/ha (Table 6). I surmise that hemlock’s pre-European settlement size structure (Figure 13) indicates a multi-age class structure of about four cohorts. Sugar maple's size structure (Figure 14) also appears indicative of a four cohort, multi-age class structure. It infer that beech’s size structure reflects a multi-age class structure of three to four cohorts (Figure 15). The fire rotation period for this community was z 5600 years. The catastrophic windthrow rotation period was z 1200 years. Longevities of hemlock, sugar maple, and beech are 250-800 years, 300-400 years, and 300 years, respectively (Burns and Honkala 1990). Catastrophic windthrows occurred far too infrequently in this community to greatly affect stand structure and composition. Hemlock, beech, and sugar maple are all very shade tolerant and respond rapidly to release (Burns and Honkala 1990) . I assume that single-stem treefalls and gap-sized windfalls (Runkle 1982) provided the mechanisms whereby the shade-tolerant members of this community established themselves in the canopy. White pine may have established itself in infrequent community-wide patch 62 Figure 13: Histogram showing the distribution of hemlock witness trees by 5 cm d.b.h. classes. 63 0.8 - l T N r". o 0 seal) sseunm )0 uoguodmd 95 60 25 Diameter (cm) of Hemlock 64 Figure 14: Histogram showing the distribution of sugar maple witness trees by 5 cm d.b.h. classes. 65 1 C9. 0 l i 0! ‘—. 0 o 888J_|_ SSGUllM )0 uoguodmd 35 45 55 55 75 25 Diameter (cm) of Sugar Maple 66 Figure 15: Histogram showing the distribution of beech witness trees by 5 cm d.b.h. classes. 67 i i I (*3. N 1‘. o 0 O 8081) 8880))M )0 uoguodmd 75 55 55 45 85 25 Diameter (cm) of Beach \i 68 blowdowns. I associate this pre-European settlement community with the misguided idea of the "climax community" so prevalent in early ecological work (McIntosh. 1985). I surmise that community dynamics here resulted mainly from competition for light and space (above and below-ground) resources. This community was a thorough mix of medium-sized hemlock, beech, and sugar maple in the canopy. Large white pines probably formed a "super-canopy" in some places. Hemlock germinates best on rotting logs (Burns and Honkala 1990). Because of hemlock’s predominance in this community I hypothesize that coarse-woody debris loads (Harmon et al. 1986) were substantial here. The lat-Mesic Forest Community Hemlock, white pine, red pine, and aspen species were the dominants of this community (Table 7, Appendix A: Figure 24). The many associates included: white cedar, jack pine, paper birch, sugar maple, birch species, beech, balsam fir, black ash, tamarack, red oaks, yellow birch, elms, and red maple (Table 7, Appendix A: Figure 24). This community was highly heterogeneous in composition. The most apparent pattern was the dominance of many large hemlock trees (64 trees/ha, 9 nE/ha, Table 7). Scattered large red and white pines were also present (Table 7) . However, after hemlock the most frequently encountered tree would have been small aspens (58 trees/ha, 2 nfi/ha, Table 7). The numerical procedures 69 Table 7: Composition and tree species diversity of pre-European settlement forests (trees 2 10 cm d.b.h.) on wet-mesic sites. Percent Percent Relative Relative Tree species Density Dominance IV Density Dominance lgggg tnngggntig 17.61 26.47 22.041 64.362 9.483 sings gttgtng 9.26 19.64 14.45 33.85 7.03 2109! tgginggg 6.02 16.14 11.08 22.00 5.78 ggpgtg! species 15.74 6.15 10.95 57.53 2.20 15g); gttiggntgtig 8.33 8.16 8.25 30.45 2.92 21092 tnnksigng 8.80 1.60 5.20 32.16 0.57 5352;; pngyrifera 5.09 4.47 4.78 18.60 1.60 52;; saccharum 5.09 3.74 4.42 18.60 1.34 agggtg species 4.63 2.43 3.53 16.92 0.87 {gang grgndifglia 3.70 2.65 3.18 13.52 0.95 5912; tgtggngg 3.70 1.09 2.40 13.52 0.39 Fraxing! 913;; 1.39 2.31 1.85 5.08 0.83 Lg;1;,)grigigg 2.31 0.78 1.55 8.44 0.28 Qggttgg [gtt_-velutina- 1.39 1.69 1.54 5.08 0.61 gtlipsgigplis ggggtp gllggngniensis 1.85 0.85 1.35 6.76 0.30 91893 species 1.85 0.49 1.17 6.76 0.18 553; [gttgn ' 1.39 0.67 1.03 5.08 0.24 Frngnng gngricana 0.93 0.39 0.66 3.40 0.14 315;! species 0.46 0.24 0.35 1.68 0.09 Qggttgg 9190 0.46 0.04 0.25 1.68 0.01 Totals 100.00 100.00 100.00 365.50 35.80 Diversity (fl') = 1.1 1 Importance Value = (relative density + relative dominance)/2 * 100 2 trees/ha 3 basal area (ma/ha) 70 (Figures 5-9) failed to produce any clear grouping among the dominants of this community. Thus, the hemlock-white pine-red pine-aspen cover type (see maps) defined by this community must be viewed with caution. Site factors were also heterogeneous in this community. Physiographically this community was usually on the interfacebetween morainal complexes and outwash plains (Table 3, see maps): probable seepage areas“ ‘The wet, sandy flats of this community also exhibited intermittent, dry, sandy ridges (surveyor descriptions). So, while the majority of this community occupied wet-mesic sites (hemlock, white pine, aspen species, and white cedar: Table 3), some small, dry-mesic sandy ridges supported red pine, jack pine, and paper birch (Table 3). Interestingly, aspen was associated with somewhat poorly drained sites having loam, silty clay loam, muck and peat, and mucky sand surficial materials (Table 3) . Surveyors decribed these aspen stands as: "Swampy aspen thicket[s]: wet aspen flats: [and] dense aspen swamps.“ Community-level overstory basal area (trees 2 10 cm d.b.h.) was 36 mZ/ha (Table 7). Nutrient availability was probably moderate in this community being limited by low soil pH and seasonal wetness. Hemlock's pre-European settlement size structure (Figure 13) suggests multi-age class stands. Red and white pine’s size structures (Figures 11 and 12, respectively) indicate even-aged stands. Aspen’s size structure (Figure 16) shows a likely even-aged clonal cohort. 71 Figure 16: Histogram showing the distribution of aspen witness trees by 5 cm d.b.h. classes. 72 55 45 85 25 I i i i st C0. N i". o 0 0 o san) ssauum )0 upmodmd Diameter (cm) of Aspen Species 73 Figure 17: Histogram showing the distribution of paper birch witness trees by 5 cm d.b.h. classes. 74 war”; I T I l Vt 0. N "i O o o O seeii sseuum ,io UO!1JOdOJd C35 45 55 55 75 25 Diameter (cm) of Paper Birch 75 Paper birch appears to have had.two even-aged.cohorts in this community (Figure 17). The fire rotation period for this community was 3 900 years. The catastrophic windthrow rotation period was s 200 years. On the wettest portions of this community I propose that the catastrophic windthrow rotation was even shorter than 200 years. TTees grown on soils with high seasonal water tables have shallow, spreading root systems (Kozlowski et al. 1991). Such trees are prone to windthrow. Large, frequent windthrows in this community could explain the significant presence of aspen. The white and red pine which occupied the dry sandy ridges and lower morainal slopes of this community would have had a difficult time establishing without fire. These species probably were established here after major fires (fires during subcontinental drought episodes). The Hydrio Forest Community This community was dominated by white cedar and tamarack (Table 8, Appendix A: Figure 25) . Associate tree species included: jack pine, spruce species, aspen species, balsam fir, black ash, birch species, paper birch, white pine, and red maple (Table 8). The multivariate analyses consistently showed.white cedar and tamarack.as a natural group (Figures 5, 7, 8, and 9). The white cedar-tamarack forest cover type was created from this community (see maps). White cedar (113 trees/ha, 6 mz/ha, Table 8) and tamarack (85 trees/ha, 5 mz/ha, Table 8) were fairly numerous and of small-medium size. 76 Table 8: Composition and tree species diversity of pre-European settlement forests (trees 2 10 cm d.b.h.) on hydric sites. Percent Percent Relative Relative Tree species Density Dolinance IV Density Dolfinance Ingig pggigegggiis 30.83 33.54 32.191 112.682 6.143 ngig lgrigina 23.31 29.07 26.19 85.20 5.32 2130; pgpgsisns 9.39 5.34 7.37 34.32 0.98 3193, species 7.52 7.17 7.35 27.49 1.31 Eggglgg species 7.14 7.07 7.11 26.10 1.29 5912; gglggggg 5.64 3.64 4.64 20.61 0.67 Fraxinus 91352 3.01 3.60 3.31 11.00 0.66 figgglg species 3.01 2.82 2.92 11.00 0.52 ggggig eggygijggg 2.63 1.87 2.25 9.61 0.34 giggg agggggg 2.63 1.58 2.11 9.61 0.29 Age; [3953! 1.50 1.59 1.55 5.48 0.29 Iggpg ggggggggig 1.50 0.78 1.14 5.48 0.14 giggg [ggigggg 0.75 0.93 0.84 2.74 0.17 yugps species 0.38 0.55 0.47 1.39 0.10 Age; saccharum 0.38 0.31 0.35 1.39 0.06 figggg grandifolia 0.38 0.14 0.26 1.39 0.03 totals 100.00 100.00 100.00 365.50 18.30 Diversity (fl') = 1.1 1 2 Importance Value = (relative density + relative dominance)/2 * 100 trees/ha 3 basal area (Hz/ha) 77 Surveyors mentioned cedar, hemlock, ground hemlock (m gggaggnsig), mosses, and.wintergreen (gaglthgriggprgggmbgns) as common understory plants in this community. The pitted outwash plain of the study area was thiscommunity's most common landscape position (Table 3, see maps). Here, this forested wetland community occurred along river and stream drainages; in kettle features: and in larger areas of the outwash plain with a lack of integration in the drainage system. This community occupied muck and peat, and mucky sand soils (Table 3); where soil drainage was very poor. Community-level overstory basal area (trees 2 10 cm d.b.h.) was 18 nfi/ha (Table 8). Nutrient conditions vary with soil acidity. More neutral, nutrient-rich sites with flowing water probably had a larger component of white cedar than tamarack. Tamarack probably dominated sites with acidic, stagnant water, and low nutrient availability (Burns and Honkala 1990). Structurally this community seems to have had both even and uneven aged stands. White cedar’s size structure (Figure 18) appears to resemble an "inverse-J" indicative of an uneven-aged stand (Smith 1986). Tamarack and the spruce species (mainly black spruce) both exhibited size structures reminiscent of even or two-aged stands (Figures 21 and 22). The fire rotation period for the hydric forest community was z 1100 years. The catastrophic windthrow rotation period was z 700 years. Longevities of white cedar, black spruce, and tamarack are: 400, 150-200, and 150-180 years, 78 Figure 18: Histogram showing distribution of white cedar witness trees by 5 cm d.b.h. classes. 79 0.4 — l l T C9. N ‘_. O O o seeii sseuiim io UOI1JOdOJd 35 45 55 55 75 25 Diameter (cm) of White Cedar 80 Figure 19: Histogram showing distribution of tamarack witness trees by 5 cm d.b.h. classes. 81 i I\. O I I I I I i <0. L0. 5t C’). N “i O O O o o o SGBJi sseuiiM io UOIUOdOJd 55 45 85 Diameter (cm) of Tamarack 82 Figure 20: Histogram showing distribution of spruce (primarily black spruce) witness trees by 5 cm d.b.h. classes. 83 $5me mozaw to E8 56:50 on we 0v mm Om mm _ 4.0 NO m0 v.0 DO ©.O seeii seeuiim io uoniodmd 84 respectively (Burns and Honkala 1990) . The disturbance rotation periods above appear correct for stands dominated by white cedar. White cedar was likely uneven-aged in this community and rotation periods longer than its longevity would have fostered this age structure. However, tamarack and black spruce most likely had even-aged stands. Such long rotation periods as above could not have created such age structures. Tamarack requires full sunlight for germination and establishment (Burns and Honkala 1990) . In the case of tamarack and black spruce, either of two conditions may have favored their development in this community. One, they could have been bog succession pioneers; or two, their establishment was related to large fires during severe drought episodes. I picture this forest community as composed of two sub- types in the pre-European settlement forest. The first sub- type was dominated by white cedar stands. Stand replacing disturbances here had rotation periods greater than species' longevities. Gap-phase disturbances (Runkle 1982) and the shade tolerance of white cedar (Burns and Honkala 1990) allowed for an uneven-aged structure. Given white cedar’s nutrient demands and the surveyor descriptions, this subtype was likely located on near neutral stream-side swamps. The second sub-type was dominated by even-aged tamarack-black spruce stands. These sites were likely highly acidic and nutrient-efficient species such as tamarack and black spruce were most competitive here. These species were able to 85 establish here due to the lack of competition and infrequent fires which prepared the seedbed. - e t s 5 Witness Tree Compositional Trends All but one of the ten pair-wise comparisons of pre- European settlement communities showed significant compositional differences between communities (P50.05, chi— squared test of independence, Table 9). The dry-mesic and xeric communities did not significantly differ in tree species composition (Table 9). In general, the pre-European settlement landscape was compositionally heterogeneous. Tree species diversity (H’) increased along the environmental gradient from a low H’ value in the xeric community (Table 4) to the largest H' in the wet—mesic and hydric communities (Table 8). Table 10 shows that the pre-European settlement landscape was dominated by conifers, particularly the pines (61% of the study area's relative density, Table 10). Angiosperms made up only 21% of the study area’s relative density (Table 10); in fairly close accordance with the fact that 17% of the study area’s soils are Alfisols. Jack pine alone accounted for 39% of the study area’s relative witness tree density (Table 10). Pine dominated forest types covered 57% of the study area (Table 11). The best evidence for .pine’s pre-European settlement dominance lies in the physical site factors of the study area. 58% of the study area is covered by outwash plains (Table 86 12). 64% of the study area has either excessively or somewhat excessively drained soils (Table 13). 80% of the study area's surficial geology consists of coarse-textured materials Table 9: Pair-wise comparisons of compositional differences between pre—European settlement communities using the chi- squared test of independence. Community Xeric Dry-Mesic Mesic Wet-Mesic Hydric Community Xeric - Dry-Mesic 19.4# - Mesic 116.1* 83.4* - Wet-Mesic 95.7* 71.8* 46.2* - Hydric 148.1* 145.7* 157.3* 92.7* - # q-value (test statistic) * Significant difference between communities at PS0.05. (Table 12). Entisols and spodosols make up z 57% and 17% of the study area, respectively. These droughty, infertile soils created conditions in the pre-European settlement forest that gave pines the competitive advantage (Aber and Melillo 1991). In turn, the pines evolved in tandem with a regime of frequent surface fires and less frequent crown fires (White 1979). Fires maintained the dominance of pines which in turn maintained the poor litter quality and nutrient status of these sites. The soil, conifer overstory, and fire regime created a complex nexus of feedback mechanisms which created and maintained the pre-European settlement pine forests. Pine barrens (Curtis 1959) occupied 3% of the study area 87 (Table JJJ.. These areas demonstrate pine's xeromorphy and fire-adaptiveness. Surveyors described the pine barrens as such: Yellow pine rather scattering and some spruce pine...A thin growth of spruce and yellow pine...burned plain covered with grasses... From surveyor descriptions these areas contained small jack pines and black oak grubs and widely scattered large red pines. A graminoid ground layer was frequently mentioned. Witness Tree structural Trends Overstory basal area (trees 2 10 cm d.b.h.) of the pre- European settlement forest followed a predictable pattern. Xeric and hydric forest communities had the lowest basal areas: 13 nfi/ha and 18 nfi/ha, respectively (Table 15). The mean witness tree diameters of these two communities while not significantly different from each other, were significantly (PS0.05) smaller than in the other communities (Table 15). Jack pine had the smallest mean d.b.h. of all witness trees (Table 16). Its presence as the dominant species of the xeric community helps to explain this community’s low basal area and mean d.b.h. White cedar and tamarack dominated the hydric sites and their small-medium mean d.b.h.'s (Table 16) created this community’s low basal area and mean d.b.h. In both of these communities the presence of very low or high soil moisture and/or nutrient 88 Table 10: Species composition by percentage of the study area. Species % Ahies balsnnea 0.87 Age; rubggm 0.71 Ace; saccharum 4.20 figtula alleghaniegsis 0.20 netula papyrifera 1.08 Betula species 0.91 Eagus grandifolia 6.59 Eraxings americana 0.13 Eraxings nigra 0.64 guniperus virginiana 0.03 Larix laricina 2.32 Qstrya virginiana 0.03 Eicea species 0.84 Einus banksiana 39.17 Einus resinosa 14.93 Einus strobus 6.73 Eopulus species 1.95 Quercus alba 0.64 Quercus rubra- 3.43 velutina-ellipsoidalis Thuja occidentalis 3.80 Tilia americana 0.44 Iggga canadensis 7.26 fllmgs species 0.34 Total 100.00 89 availability influenced species composition and restricted tree growth. These factors in turn may explain the low pre- European settlement overstory basal area of these sites. Those communities along the middle of the edaphic gradient all had basal areas within the range found for old- growth upland hardwood stands in northern Michigan (McIntire 1931). Trees were indeed large in all three communities (dry- mesic, mesic, and wet-mesic). Red pine achieved large sizes in the jpre-European settlement forest (Table 16). Its preponderance in dry-mesic communities may explain the substantial pre-European settlement basal area and mean d.b.h. there. Species competitive in these communities were often not restrained from achieving their genotypic maximum sizes. Density trends in conjunction with basal area trends allow for a conceptualization of the pre-European settlement forest’s structure. Xeric communities were open, with small trees. Enough light probably made it to the forest floor here to support a rich herbaceous community. Dry-mesic communities had small groves of large trees interspersed with dense patches of small trees. The mesic community was nearly a uniform matrix of medium sized trees with occasional patches of densely spaced small trees and occasional large super- canopy trees. The wet-mesic community consisted of small ridges of large trees within a dense matrix of small trees. Dense patches of small trees, small groves of medium- sized trees, and small open patches comprised the hydric sites. My- 90 Table 11: Pre-European settlement extent of vegetation cover types for the study area. Type Area (ha) % of study area Jack pine-red pine- 46,583 38.43 white pine Jack pine barrens 3,442 2.84 Red pine-jack pine- 18,461 15.23 white pine Hemlock-beech-sugar 26,000 21.45 maple-white pine Hemlock-white pine- 5,248 4.33 red pine-aspen Swamp conifers 18,436 15.21 Harsh 1,248 1.03 Lakes 1,794 1.48 Totals 121,212 100.00 Table 12: Soil and landform composition of the study area. Soil Type % of study area Sand 74.85 Loamy sand 5.45 Sandy loam 8.81 Loam 1.48 Silty clay loam 0.20 Muck and peat 7.73 Mucky sand 1.48 Total 100.00 Landform Type % of study area Ice-contact 16.34 Morainal 25.82 Outwash 57.83 Total 100.00 91 Table 13: Extent of the different soil drainage types in the study area. Drainage Class % of study area Excessive 44.16 Somewhat excessive 20.17 Well and moderately well 19.50 Poor to somewhat poor 7.26 Very poor 8.94 Total 100.00 Disturbance Regime Patterns Annually, z 6% of the study area was in a burned-over condition in the pre-European settlement period (Table 14). Whitney (1986) obtained a similar value of 7% for nearby Crawford and Roscommon Counties. In northeastern Maine Lorimer (1977) found 9% of the pre-European settlement landscape in burned land. Only z 2% of the study area was in windfall in the pre-European settlement period (Table 14). For the study area as a*whole, fire likely played an important role in structuring ecosystems. Large windfalls seemed infrequent enough to have a minimal impact on species adaptations. The intensity, frequency, and extent of fires in the pre- European settlement forest seems to be directly related to vegetation types, soil types, and landscape configuration. As (described earlier, pine dominated ecosystems on dry, flat, :infertile sites burned most frequently and with the greatest 92 Table 16: Disturbance regimes of the pre-European settlement forest. Forest Type Hemlock- Jack pine- Red pine- beech- Hemlock- red pine- jack pine- sugar white pine- uhite pine white pine maple- red pine- Swamp Study white pine aspen conifers area x area affected by windthrow 1 2 1 7 2 2 est. rotation period 2000 900 1200 200 700 900 (yr) for windthrow 1 area affected by 12 11 0 2 1 6 fire est. rotation period 100 100 5600 900 1100 200 (yr) for fire 93 Table 15: A comparison of pre-European settlement forest structural attributes (trees 2 10 cm d.b.h.) by forest community or site type. Site Mean d.b.h. Basal area Type (cm) 1 SEM nF/ha 1 SEM Trees/ha : SEM Xeric 23 1 0.4 a‘ 13 1 0.7 214 i 11.2 a Dry-Mesic 30 i 0.8 b 31 i 1.5 315 t 15.0 a,b Mesic 33 i 0.7 C 25 i 1.4 241 i 13.3 a,b Wet-Mesic 30 i 1.2 b,c 36 i 11.8 366 1 120.4 a,b Hydric 23 i 0.7 a 18 i 3.1 366 i 62.1 b ‘ Column means with different letters (a,b,c) significantly different at PS0.05, Tukey’s HSD. 9 SEM = Standard Error of the Mean Table 16: Mean witness tree diameters (trees 2 10 cm d.b.h.) by species. Species Mean d.b.h. (cm) 1 SEM Jack pine 18 i 0.3 a* Aspen 19 i 0.8 a Red oaks 19 i 1.4 a Tamarack 21 i 0.9 a Beech 26 i 0.6 a White cedar 27 i 1.3 a,b Sugar maple 31 i 1.1 b Hemlock 36 i 1.1 b Red pine 44 i 0.8 c White pine 46 i 1.7 c * Means followed by different letters (a,b,c) significantly different at Pso.05, Tukey's HSD. 94 intensity. In some instances the fires started in the pine dominated ecosystems went beyond the limits of the droughty, infertile outwash soils and created pine communities on fairly mesic morainal and ice-contact features (see maps). In such cases the position of these pine dominated morainal and ice- contact features was just east of extensive outwash plains (see maps). Strong westerly and northwesterly winds associated with fire spotting in the study area (Simard.et al. 1983) could have spread fires to specially situated moraines and ice-contact landforms. 150 YEARS OF FOREST CHANGE I118“: 'i o -: Csuh 't -s :_ ass i- oao_i s-i-i_ Changes in the Xeric Forest Community Although stand data were unavailable for today’s xeric sites, the literature indicated that compositional changes have been minimal since the pre-European settlement period. The Mack Lake outwash plain (Figure 3) was composed of the following timber types in 1980 (Simard et al. 1983): jack pine (42%), red pine (16%), oak/pine (8%), oak/hardwood (12%), and aspen (14%). In the pre-European settlement period jack pine (52% i.v., Table 4), red pine (29% i.v.), and white pine (10% i.v.) were forest dominants with aspen (2% i.v.) and red oak (2% i.v.) associates. Hence, today's xeric forests are nearly a compositional facsimile of pre-European settlement conditions. Present-day jack pine stands on the Mack Lake outwash 95 plain range from 246-2962 trees/ha; mean density is 1975 trees/ha (Simard et al. 1983). It seems that.most present-day jack pine stands are far denser than pre-European settlement stands. On the Mack Lake outwash plain present-day mean d.b.h. ranges from 13 cm in jack pine stands to 20 cm in red pine stands. Basal areas likewise range from 25 nfi/ha to 33 mz/ha (Simard et al. 1983). Present-day basal areas are larger than in pre-European settlement time perhaps due to a decline in jack pine barrens. Jack pine's age-structure on xeric sites in Roscommon County, Michigan, is even-aged today (Larsen 1982) as it likely was 150 years ago. Structurally, these xeric sites seem. to have changed. Such changes undoubtedly have impacted populations of the endangered Kirtland's Warbler (Dendzgiga kirtlaggii) (Probst 1988). Whitney (1986) determined that the fire rotation period for the jack pine cover type in nearby Roscommon and Crawford CountieS‘wasiz 392 years during the fire protection era (1965- 1985). The fire rotation period for the Huron National Forest from 1950-1981 was s 398 years (Simard and Blank 1982). The average annual area burned in the Huron NF decreased from 5261 ha during the 19103 to 155 ha during the 19705 (USDA Forest Service 1980). Organized fire control beginning around 1930 diminished the frequency, severity, and extent of fires in the study area (Simard and Blank 1982). However, on May 5, 1980, a forest fire burned nearly 9713 ha on the Mack Lake outwash plain, crowning in several areas (Simard et al. 1983). This 96 was the sixth major fire (4047+ ha) on the Mack Lake plain since 1820 (Simard and Blank 1982) . Even today the study area’s xeric sites retain a pyric nature. This may have allowed the fire-dependent Kirtland's Warbler to survive 20th century fire management practices. Changes in the Dry-Mesic Forest Community This forest community has changed from a pre-European settlement red pine (41% i.v., Table 17, Appendix D: Figure 26), jack pine (24% i.v.), and white pine (15% i.v.) type to today’s red oak.complex (40% i.v., Table 17, Appendix C: Table 24), big-toothed aspen (23% i.v.), white oak (17% i.v.), and red maple (10% i.v.) type. Pre-European settlement forest associates included beech, sugar maple, and white oak (Table 5). Associates of the present-day forest are sugar maple, white ash, paper birch, basswood, and white pine (Table 17, Appendix C: Table 24). The tree species composition of this community has significantly changed (Table 18). 97 Table 17: Importance values and other attributes of tree species by community type in the pro-European settlement and present-day forest. Dry-Mesic Mesic Hat-Mesic Tree species PS1 90 PS 90 as 90 Abigg bglggggg 0.122 0.00 0.00 0.00 2.40 0.00 5535 Luggg! 0.51 10.11 0.79 4.39 1.03 9.70 Age; ggcchargg 1.86 5.10 15.10 41.31 4.42 25.56 figtglg gllgghaniensis 0.12 0.00 0.29 0.00 1.35 0.00 Qgtglg pggyrifera 0.29 1.12 0.83 1.21 4.78 20.18 Qgtglg species 0.00 0.00 0.46 0.00 3.53 0.00 gggy! species 0.00 0.00 0.00 0.45 0.00 0.00 {gags grandifglig 3.52 0.21 19.95 3.39 3.18 0.45 5535133; aggrigana 0.13 1.37 0.25 6.98 0.66 3.93 figggiggg Q1359 0.00 0.00 1.20 0.00 1.85 0.00 1ggipggg§ vigginiana 0.00 0.00 0.00 0.00 0.00 0.00 ngig larigina 0.00 0.00 0.00 0.00 1.55 0.00 Qgtgyg virginiggg 0.00 0.38 0.11 1.25 0.00 0.00 3195; species 0.32 0.00 0.00 0.00 0.35 0.00 giggg pgnkgigna 23.83 0.00 1.14 0.00 5.20 0.00 giggg gggigggg 41.49 0.16 9.65 0.00 11.08 0.00 glggg gtggggg 15.34 1.04 14.82 0.00 14.45 1.02 ggpglgg species 3.61 22.70 1.90 12.87 10.95 1.60 Eggggg sgrogina 0.00 0.43 0.00 0.11 0.00 0.00 ggggggg 319; 1.13 16.60 0.37 0.00 0.25 1.26 Qgggggg rggga-yglutina- 3.73 39.72 3.74 9.39 1.54 18.01 ellipsoidalis M! ocgidentglig 0.44 0.00 1.26 0.00 8.25 0.00 mg m 0.00 1.10 2.08 18.68 0.00 15.23 m Me 3.61 0.00 25.40 0.00 22.04 0.00 m species 0.00 0.00 0.70 0.00 1.17 3.08 Totals 100.00 100.00 100.00 100.00 100.00 100.00 Diversity um 0.7 0.7 0.9 0.8 1.1 0.8 Similarity uses 13 28 17 ; Where PS and PD indicate presettlement and present-day, respectively. Importance value a (relative density + relative dominance)/2 * 100 98 Table 18: Pair-wise comparisons of compositional differences between pre-European settlement and present-day forest communities on similar site types using the chi-squared goodness of fit test. Site Types Site Types Dry-mesic Mesic Wet-mesic Dry-mesic 946.6#* Mesic 528.4* Wet-mesic 503.7* # q-value (test statistic) * Present-day and pre-European settlement communities significantly different in composition, P50.05. The probable causes of these changes lie in the interaction between historical events of the past 150 years:available tree species pools: the adaptive strategies of the new dominants: and chance factors. Red and white pine have been nearly eliminated (Table 17) from this community. This occurred because of the destruction of their reproductive potential. Red and white pine were selectively logged from this community, removing most of the available seed trees. Kittredge and Chittenden (1929) determined that the fire rotation period was as nine years in cut-over mixed pine stands of northern lower Michigan in the logging era. The red and white pine advance regeneration provided by the few remaining seed trees was eliminated by these frequent and intense slash fires (Ahlgren and Ahlgren 1983). Red and white pine become reproductively mature between twenty and thirty years of age 99 (Burns and Honkala 1990). So, those juveniles which did escape one or two fires were most likely destroyed by a third before they could reproduce. The failure of the pines to reproduce in the logging era helps to explains the changes in this community. The introduction of white pine blister rust (grgpgrtium gibiggla) further hindered the regeneration of white pines. The present-day dominants in this community were either associates in the pre-European settlement forest (the red oaks, aspen, and white oak: Table 5) or at least present (red maple, Table 5) . The red oaks are vigorous sprouters and drought tolerant (Burns and Honkala 1990). Logging created drier soil conditions and frequent slash fires which gave the oaks a greater reproductive ability than the pines. The logging and repeated fires would have stimulated the suckering of aspen (Burns and Honkala 1991). New areas of mineral seedbed allowed for aspen's copious wind-dispersed seed to colonize large areas of this community. Red maple sprouts fairly well and produces a large seed rain (Burns and Honkala 1990). This may explain red maple’s present success in this community. The community-level overstory basal area (trees 2 10 cm d.b.h.) has decreased from 31 mZ/ha to 27 mZ/ha over the past 150 years (Table 19). The mean d.b.h. of this community has significantly decreased (PS0.05, paired t-test) from 30 cm to 20 cm (Table 19). This is not surprising, as the former 100 dominants of this community, white and red pine, were the largest in girth (Table 16) in the pre-European settlement forests Commensurate with this decrease in tree size has.been a three-fold increase in tree density in this community from 315 trees/ha to today’s 904 trees/ha (Table 19). The structural nature of this community has changed. Horizontally, the forest is less open and park-like. The canopy configuration has surely changed as the canopy is now composed of different tree species with different canopy architectures. The age-class diversity has likely decreased. Present-day stands do not exceed 100 years of age. In the pre-European settlement forest, judging by their size, some white and red pines may have been between 200 and 450 years old (Burns and Honkala 1990). Whitney (1987) determined that the fire rotation period for similar sites in nearby Roscommon and Crawford Counties was s 313 years for the fire protection era (1965-1985). The pre-European settlement fire rotation was s 100 years (Table 14). The advent of fire suppression in the study area may explain the presence of such mesic species as sugar maple, basswood, and white ash as associates in today’s overstory (Table 17) and maple’s (red and sugar) ubiquitous presence in the seedling and sapling stages (Padley 1989: Palik and Pregitzer 1992). Angiosperms have replaced gymnosperms as the dominants of this community. Such a change has likely had functional .ummuiu confine .mo.oum um usmquMflp wausmofiwflcmflm Anew. mumuuma ucmmmmmww sues memos 30m F .>Ho>wuooomou e>moiuoom0um com ucofioauumm smoaousmimmm oumofiocfi om poo mm muons m hm.oeme mm.ssom N.Pscm m.PFscm ~.ommsmomp e.ompeccm oemme l iumz ne.oeom was.osmm P.ss_m e.Psmm P.ms«mem m.mpssem names n~.oso~ me.oeom m.~«em m..epm m.cmseom o.msempm names on mm on we on we mass 4 seem _ 2mm 4 2mm 4 2mm 4 A50. .s.n.p com: Amc\~Ev mn\mn com: ms\moomu coo: A.:.n.p ESEHGAE so op. mwusnfiuuum Hmuouoouum umoHOm xmoiusmmomm pom usmsmauuom commousmimmm mo GOmHmeEOO < «mp manna 102 .ummuiu sweeps .mo.oum um uconMMHU >Hucmofiwflcmflm was “new. mouuoa ucouommep m ha ©030HHOH momma 30m 3.. now own poo3mmmm QNN amp mxmo pom wh— obs xmo ouflnz new mm. comma he. 6mm copes scape nap men magma Hmmmm moms mNN oases pom AEov .s.n.o coo: A80. .:.n.o com: >mpiusommum usmsmauumm commonsmioum mmfiommm some A.:.n.c EssacHE so op. ucosmauuom cmoeumaciousm women mmfluoem menu Sn .c.n.o CH momsmnu "om manna 103 repercussions in this ecosystem. Seral wildlife species haveundoubtedly benefited from the increase in palatable browse species here (Heinen and Sharik 1990). The C:N ratio of this community’s litter'has likely decreased, increasing N- mineralization rates (Pastor et al. 1984) and decreasing the residence time of litter on the forest flooru The presence of less pyric deciduous tree species has likely reinforced the human-imposed fire suppression program. Changes in the Mesic Forest Community This forest community was dominated by hemlock (25% i.v. , Table 6, Appendix D: Figure 27), beech (20% i.v.), sugar'maple (15% i.v.), and white pine (15% i.v.) in the pre-European settlement period. Present-day stand dominants are sugar maple (41% i.v., Table 17, Appendix D: Figure 27), basswood (19% i.v.), and big-toothed aspen (13% i.v.). Pre-European settlement associates included: red pine, red oaks, basswood, aspen, white cedar, black ash, and jack pine (Table 6). Red oaks, white ash, red maple, beech, ironwood, and paper birch are the present-day forest associates (Table 17, Appendix C: Table 25). The overstory composition of this community has changed significantly (Table 18). The tree species diversity (H’) of this community has declined (Table 17). Hemlock , white pine, and beech were eliminated from this community. Hemlock's demise was linked to its selective cutting for the leather industry as a source of tannin. This eliminated a great many seed trees. The increased exposure 104 and drying of the soil after logging prevented the establishment of hemlocks from the few seed trees left (Buttrick 1923). In recent times deer (Qggggilggs giggigiangg) browsing and the lack of coarse woody debris may have prevented regeneration of hemlock from the rare seed trees remaining (Frelich and Lorimer 1985) . While beech frequently root sprouts, it cannot regenerate under a regime of repeated cuttings: nor can it disperse its seeds by wind (Burns and Honkala 1990). Furthermore, beech is slow-growing and ineffectively responds to release (Burns and Honkala 1990). Beech's life history traits help explain its failure to reproduce after extensive logging. White pine's elimination here probably hinged on its subjection to exhaustive selective logging. Sugar maple had an extensive development of advance regeneration in old-growth northern hardwoods of northern Michigan (Frelich and Lorimer 1991). Sugar maple strongly responds to release (Burns and Honkala 1990). When the mesic community was logged, sugar maple seedlings and saplings probably rapidly established themselves as new canopy dominants. Sugar maple can also stump sprout and has a prolific seed rain (Burns and Honkala 1990). Taken together these factors may explain sugar maple's present-day dominance (41% i.v.) in this community. Basswood was an associate species in the pre-European settlement forest (Table 6). It achieved its present dominant status probably via sprouting; 105 wind dispersal of seeds: its rapid growth rate: and its ability to remain in the seed bank for up to three years (Burns and Honkala 1990) . Big-toothed aspen's present status depended on the logging and surface fires at the turn of the century. Aspen in pre-European settlement windfall gaps suckered into the newly opened spaces of this forest. Where surface fires exposed mineral soil on logged sites, aspen seeded in (Davis 1935). Overstory basal area increased from 25 mZ/ha to 31 mz/ha (Table 19) over the past 150 years. This may reflect a large ingrowth of pole-sized trees (Table 19) in recent time. As in the other forest communities, the mesic forest had less trees (241 trees/ha, Table 19) in pre-European settlement time than today (915 trees/ha, Table 19). Along with this increase in tree density, the mean d.b.h. has significantly (P50.05, paired t-test) decreased from 33 cm to 20 cm (Table 19) in this community. The significant (PS0.05, paired t-test) post- European settlement reduction in sugar maple’s mean d.b.h. (Table 20) may account for the decline in this community’s mean d.b.h. During the logging era (1870-1920) fires were less common here than in the previous communities (Roth 1905). Whitney (1987) calculated a fire rotation period of 276 years for the hemlock-white pine-northern hardwoods type of Roscommon and Crawford Counties during the fire suppression era (1965-1985) . I calculated a pre-European settlement fire rotation period of 106 5600 years. This indicates that fires may have been more frequent in this community since Euro-American settlement. This community has changed in composition and structure. Sugar maple has assumed dominance via release of advance regeneration and seeding in. Interestingly, basswood has also assumed a co-dominant position in this community. If gap- phase type disturbances are maintained in this community, sugar maple and basswood could likely perpetuate themselves here. Beech has low quality litter (Melillo et al. 1982). Sugar maple and basswood have high quality litter (Melillo et al. 1982: Pastor at al. 1984). The loss of beech and subsequent increase in sugar maple and basswood in this community has likely changed nutrient cycling. Present-day litter decomposition rates and N-mineralization rates may exceed pre-European settlement rates. Changes in the Wet-Mesic Forest Community This community has gone from a pre-European settlement conifer dominated system (65% i.v., Table 7) to today's angiosperm dominated forest (99% i.v., Table 17, Appendix D: Figure 28). Pre-European settlement dominants included hemlock, white pine, red pine, and aspen (Table 7). Today's dominants are sugar maple, paper birch, the red oaks, and basswood (Table 17). Pre-European settlement associate tree species included: white cedar, jack pine, sugar maple, birch species, beech, balsam fir, black ash, tamarack, red oaks, 107 yellow' birch, elms, and red. maple (Table 7). Today's associates are: red maple, white ash, elms, big-toothed aspen, white oak, and white pine (Table 17, Appendix C: Table 26). This community's tree species composition has significantly changed (Table 18) . Tree species diversity (H') has declined in this community (Table 17). The history of logging in this area resulted in today’s lack of hemlock, red pine, and white pine. But aspen’s move from a dominant to an associate appears counter-intuitive given the amount of disturbance during the logging era. Turn of the century slash fires likely fostered paper birch and red oaks: species jpresent. as associates in the jpre-European settlement forest. Sugar maple and basswood may have attained their present status here by sprouting: seed dispersal: and recruitment from the understory. While sugar maple was a pre- European settlement associate in this community, basswood was not recorded as a witness tree here (Table 7). Community-level overstory basal area (trees 2 10 cm d.b.h.) has remained nearly the same since Euro-American settlement (Table 19) . Yet, the mean d.b.h. has significantly declined from 30 cm to today's 18 cm (P50.05, paired t-test, Table 19) . Conversely, the mean tree density has tripled (Table 19) from 366 trees/ha to 1203 trees/ha. Such changes indicate the transition of this forest from one with a diversity of size classes to today’s more uniform structure. Like the dry-mesic forest community, this community has 108 changed in composition. All of the pre-European settlement dominants have been supplanted today by what were pre-European settlement associates. This change followed a pulse of disturbance at the turn of the century. This community is physiographically’ heterogeneous and actually consists of broad, wet, sandy flats interspersed with well-drained, narrow, linear ridges. Sugar maple and the red oaks do not tolerate soil wetness very well. Hence, they likely occupy today's ridge sites. Paper birch and basswood can tolerate more poorly drained areas (Burns and Honkala 1990) and may predominate on the wet flats. I propose that fires are much less frequent on this community's dry, sandy ridges than in pre-European settlement time. On the wet, sandy flats, the occurrence of windthrow probably exists as it did 150 years ago. Changes in the Hydrio Forest Community I lacked site specific present-day stand data for this community. However, some broad comparisons to observations in the literature were made. Veatch et al. (1931) reported that in Oscoda County the only major change within unlogged swamp conifer stands was the great decrease in tamarack due to larch sawfly (Lygaegnematus erichsonii) outbreaks. Where the swamp conifer type was logged and/or severly burned in the logging era, four new cover types came to dominate this community in different parts of Oscoda County: paper birch, alder (Linus ruggga) thicket, red maple, or aspen (Veatch et al. 1931). 109 Surveyor accounts of abundant yew (Taxus gangggnsis) populations on these sites constrast sharply with the lack of yews in today’s swamps (Frelich and Lorimer 1985). The modern irruption of deer populations in Michigan may account for this lack of yew regeneration. l2!"3 '1 i- _, oscao- '. t- , c- _u_o-:n- 5 .i W The study area's landscape patterns have definitely changed since 1840, albeit not nearly as drastically as in southern lower Michigan. Non-forested cover types have increased by 14% in the study area (excluding Alcona County, Table 21). Angiosperm cover types have increased by 22% in area while conifer cover types have declined by 35% in area (Table 21). 'These changes resulted from Ihuman. imposed disturbances. Pine cover types still cover the greatest area, but they have declined by 17% since the pre-European settlement period (Table 21). This loss approximates the pre-European settlement areal extent of the red pine-jack pine-white pine type (dry-mesic community, Table 11). 'Today's pine cover type is nearly all jack pine. Therefore, the pre-European settlement jack pine-dominated xeric community has not changed in area. The red pine-jack pine-white pine type has seemingly been replaced by the new aspen-birch cover type (Table 21). The northern hardwoods cover type has declined by 14% in areal extent (Table 21). In addition, it has lost hemlock, 110 Table 21: Pre-European settlement and present-day extent of various cover types in the study area (excluding Alcona County). Pre-European Settlement Present-day Area (ha) % of Area Area (ha) % of Area MIRIS cover type (code) Pine (421) 54,050 52.70 36,464 35.55 Northern Hardwoods (411) 22,542 21.98 7,851 7.67 Other upland conifers (422) 5,171 5.04 228 0.22 Lowland conifers (423) 15,649 15.26 3,998 3.89 Emergent marsh (622) 1,196 1.17 179 0.17 Lakes (52) 1,794 1.75 614 0.60 Pine barrens (333) 2,162 2.10 0 0.00 Central hardwoods (412) 0 0.00 16,466 16.05 Aspen-birch (413) 0 0.00 17,634 17.19 Lowland hardwoods (414) 0 0.00 2,944 2.87 X-mas trees (429) 0 0.00 5 0.00 Agriculture 0 0.00 2,586 2.52 Other wetlands 0 0.00 2,777 2.71 Other non-forested 0 0.00 6,255 6.10 Urban 0 0.00 4,573 4.46 Totals 102,564 100.00 102,564 100.00 111 beech, and white pine as integral components (Table 17).The 16% areal gain in the new central hardwoods cover type (sugar maple, basswood, red oak) has likely replaced the above loss in the northern hardwoods cover type. Wetland cover types (forested and non-forested) show a net loss of 41% in coverage (Table 21). The swamp conifer cover type shows a net area loss of 74% since the pre-European settlement period (Table 21) . The area lost by the swamp conifer type has likely been replaced by reservoirs (Alcona and Mio Dam Ponds) and three of the new cover types: lowland hardwoods (black ash, elm, red.maple, aspen, and.white birch), the aspen-birch type, and the "other"wetlands type (mainly shrub dominated wetlands). The areal decline in wetland cover types and conifer cover types (particularly swamp conifers and hemlock) since the pre-European settlement period.must have had and continue to have functional effects on the study area’s landscape ecosystems. The increase in non-pyric hardwoods on all but the xeric sites has certainly augmented modern fire control efforts. A decrease in wetland area has certainly changed landscape-level hyrologic processes. Surveyor accounts of Native American land-use in the study area only mention a major trading trail along the Au Sable River. Today, humans have fragmented the study area’s landscape to a much greater degree. 13% of the study area exists as direct human artifacts (i.e. , agriculture, right-of- 112 ways, roads, and urban areas: Table 21). These human artifacts increase the amount of forest edge area relative to interior forest conditions (Moss 1983). Thus, I suspect that edge species (e.g. deer, ruffed grouse [Bonasa umbellug], and cowbird [Molothzgs atg11) populations have increased in the study area as a result of such habitat modifications. The introduction of road systems, agriculture, and housing must have definite impacts on species dispersal patterns and forest fire behavior. 113 CONCLUSIONS EVIDENCE FOR/AGAINST HYPOTHESES Tree species were not randomly distributed across the pre-European settlement landscape. The correspondence analyses and nonmetric multidimensional analysis show that witness tree species were responding to an underlying gradients The. primary axis of this gradient ‘was soil moisture, but several other axes undoubtedly affected species distributions (i.e., fire regime, soil nutrients, light availability, and herbivory). These numerical analyses seem to substantiate the Gaussian model of community structure in the pre-European settlement forest. The binary discriminant analyses show that certain witness tree species occurred on certain landforms, soil types, and soil drainage conditions with statistically significant probabilities. An overlay of the two maps (see maps) shows that certain pre-European settlement forest cover types had definite affinities for certain landform positions. The fact that the grouping of the witness trees by site types explains much of the variance in the pre-European settlement forest demonstrates that tree species were probabilistically distributed. The cluster analyses interpreted against the composition of the site types and the indirect ordinations seem to indicate 'that loose associations did, occur' between. tree species in the pre-European settlement forest. However, each of the possible associations (beech-sugar maple-hemlock, white 114 cedar-tamarack, red pine-jack pine) were composed of species with similar responses to parts of the environmental gradient. Hence, these associations were not tightly linked and may merely indicate the response of physiologically similar species to their environment. The pre-European settlement forest communities all significantly differed from each other in composition save for the compositional similarity between the xeric and dry-mesic communitiese These two communities compositionally converged due to the overriding influence of the fire regime. Forest structure changed along the edaphic gradient. Disturbance regimes were correlated with the soil moisture gradient; fire being most common on the xeric sites and catastrophic windthrow most common on the hydric sites. Over the last 150 years, the composition and structure of the study area’s forests have changed. However, the magnitude of such changes are specific:to site types. lPresent-day xeric and hydric sites are compositionally similar to pre-European settlement conditions. Dry-mesic, mesic, and wet-mesic sites have changed in both composition and structure since Euro- American settlement. A VERBAL MODEL OF THE STUDY AREA' S VEGETATION .A complex interplay among substrate, disturbance, and biotic factors created the pre-European settlement forest patterns. Viewed as a whole, the study area in the pre- European settlement period was a non-equilibrium landscape as 115 defined by Shugart (1984) . Present-day forest communities show historic chance events often overriding competition- mediated environmental determinism in ‘their composition. Today’s communities are rapidly succeeding to other types (Palik and Pregitzer 1992). I suggest that the present-day landscape is still re-equilibrating to the turn of the century disturbance pulse. MANAGEMENT IMPLICATIONS Because forest communities can have multiple stable states (Shugart et al. 1981) and climate constantly changes (Brubaker 1988) defining the ’natural' vegetation for the study area is fraught with problems. Managers and the public must realize that we cannot return the landscape to 1840 for both practical and evolutionary-ecological reasons. The value of vegetation history lies in delimiting the range of vegetation types and disturbance regimes in which species have evolved (Sprugel 1991). The National Forest Management Act specifies that biological diversity be considered in national forest planning. What management practices might be considered in light of pre-European settlement biodiversity patterns? Today's forests lack: a mosaic of tree age-classes: coarse-woody debris; within-community tree species diversity: natural fire regimes: very old trees: and low forest edge:interior ratios. The following management proposals may help to restore the: dynamics of pre-European settlement 116 forests to the study area: 1. Stands should be managed with longer rotations. 2. Tree species such as red pine, white pine, beech, and hemlock might be artificially regenerated on the same sites they occupied in pre-European settlement time. 3. Dead trees should be left on the forest floor. 4. Some large "wolf" trees and dead and dying trees should be left in stands. 5. Some stands should be allowed to pass through several natural rotations. 6. Prescribed burning should be used more often in forest regeneration. 7. Clearcuts should be reduced in number and increased in mean size. APPENDICES APPENDIX A 3;. - 199‘ 0 . 15‘: e _:-‘C ‘- 110013 C ‘= .- : Figure 21: Importance values of witness tree species on xeric sites. 117 118 'vvvvvvvvvvvvvvvvvvvvv "' O...0.00.000000...0.0.0.0.0;0.000 ' ° ’ ’ ’e’e’e’e’e’ ’e ’e e e o e 0.0 9:0 O O O O O 0 L0 RT 0’) (\J *— eriieA eaueiipdwi P'me ‘09 SO- ‘5an ?\ OiCQGI 996 P\ \Nvfig 119 Figure 22: Importance values of witness tree species on dry- mesic sites. 120 @yxxyZ/ @rfi @zfi /O age 590 myoO new cocoa/c _ _ _ «flee/60%.... ca 02 O I Or r O N i Om i as _ Om enleA eaueiipdwi 121 Figure 23: Importance values of witness tree species on mesic sites. ® Q77 %9 O®/® ®¢® /O ®®/A/ fi®® @C/Qw ®// ©®i®C®/O O OmQ/®O eww @Wye/WOOO/Cv ®®®OOO 77m 122 _ _ _ _ _ _ _ O i O i C) (\i Om eri|e/\ eaueiipdw 123 Figure 24: Importance values of witness tree species on wet- mesic sites. 124 0492/5103; Romeo/e6 oo/ooioc/me ”ow/s1, ceases/«,0 one 09$. Wm. 00 09.0 «“4009sz 9m”, 09 c e as @0602 O 0 L0 ON mm enieA eaueiipdwl 125 Figure 25: Importance values of witness tree species on hydric sites. 04.6.7700 V 0 00w O 04 0 ® 0® A 00 0. 0 0 / ® 0/ 0 0 0900 v0.01”,V .00 v06,090 ”V00 900 COW/W, 6.000 C 0 126 _ r O 104 i O (\J 100 on enleA eaueiiodwi APPENDIX B tr r ri or -Eur an l t- ommuni ies Table 22: Density of tree species (trees 2 10 cm d.b.h.) by community type in the pre-European settlement and present-day forest. Dry-Mesic Mesic Wet-Mesic Tree Species rs1 PD PS P0 P8 P0 mm m 0.542 0.00 0.00 0.00 13.52 0.00 5gp; gghggg 2.62 113.40 2.49 46.11 5.08 127.92 M m 6.30 57.11 38.68 452.00 18.60 317.58 ggtglg gllgghggigggig 0.54 0.00 0.41 0.00 6.76 0.00 figtglg pggygjjggg 1.04 12.38 2.92 11.44 18.60 284.12 figtglg species 0.00 0.00 1.25 0.00 16.92 0.00 Qggyg species 0.00 0.00 0.00 4.39 0.00 0.00 figggg grgggifolig 15.22 1.45 60.37 29.64 13.52 4.93 figggings gaggjgggg 0.54 10.66 0.41 57.72 3.40 37.06 Eraxinus 91359 0.00 0.00 3.33 0.00 5.08 0.00 Jggjpgggg yicginigng 0.00 0.00 0.00 0.00 0.00 0.00 Lem mm 0.00 0.00 0.00 0.00 8.44 0.00 Qgtgyg viggigiang 0.00 2.98 0.41 16.56 0.00 0.00 flips; species 1.04 0.00 0.00 0.00 1.68 0.00 Pings panggigng 112.43 0.00 4.15 0.00 32.16 0.00 giggs rggjnogg 91.41 0.45 17.47 0.00 22.00 0.00 Eiggg gtggggs 32.06 13.37 23.72 0.00 33.85 14.80 ggpglgg species 16.80 184.61 7.07 93.58 57.53 13.60 Eggnus agggging 0.00 5.42 0.00 0.55 0.00 0.00 Qggggus alga 5.26 147.47 1.66 0.00 1.68 11.07 nggg g Lgpgg-yglggina- 16.80 344.45 9.15 59.83 5.08 193.39 g ipggigg 1g Ihgig occiggggglig 1.58 0.00 3.33 0.00 30.45 0.00 m m 0.00 9.85 4.99 142.98 0.00 143.93 Igggg ggggggggig 11.03 0.00 57.41 0.00 64.36 0.00 91099 species 0.00 0.00 2.08 0.00 6.76 55.00 Totals 315.20 903.60 241.30 914.80 365.50 1203.40 1 trees/ha 127 2 Where PS and PD indicate presettlement and present-day, respectively. 128 Table 23: Basal area of tree species (trees a 10 cm d.b.h.) by community type in the pre-European settlement and present-day forest. Dry-Mesic Mesic Bet-Mesic 1... Species 73—70 :77. H 591;; balsameg 0.022 0.00 0.00 0.00 0.39 0.00 Agg; ggggg! 0.06 2.05 0.14 1.16 0.24 3.83 55;; saggharum 0.53 1.03 3.60 10.33 1.34 9.50 gggglg allggganiggsig 0.02 0.00 0.10 0.00 0.30 0.00 figtglg pggygijggg 0.07 0.23 0.11 0.36 1.60 8.50 figtglg species 0.00 0.00 0.10 0.00 0.87 0.00 gggyg species 0.00 0.00 0.00 0.13 0.00 0.00 [gggg graggifolia 0.68 0.07 3.78 1.10 0.95 0.15 Fraxinus americana 0.03 0.41 0.08 2.38 0.14 1.11 Fraxiggg giggg 0.00 0.00 0.26 0.00 0.83 0.00 gunipgrgg virginiana 0.00 0.00 0.00 0.00 0.00 0.00 Lg; §,larigigg 0.00 0.00 0.00 0.00 0.28 0.00 ggggyg yiggflgiggg 0.00 0.11 0.01 0.21 0.00 0.00 315;; species 0.10 0.00 0.00 0.00 0.09 0.00 giggg gggggiggg 3.70 0.00 0.14 0.00 0.57 0.00 Eiggs resigggg 16.68 0.07 3.06 0.00 5.78 0.00 giggg ggggggg 6.34 0.16 5.03 0.00 7.03 0.44 ggpglgg species 0.58 6.66 0.22 4.82 2.20 0.41 gggggg serotina 0.00 0.07 0.00 0.05 0.00 0.00 Qgggggg 519; 0.18 4.50 0.01 0.00 0.01 0.33 Qgggggg.§g%gg-yglutina- 0.66 11.03 0.93 3.80 0.61 5.79 gllipggi 1; Ihgig gggiggggglig 0.11 0.00 0.29 0.00 2.92 0.00 11119 ggggjgggg 0.00 0.30 0.53 6.75 0.00 4.31 lgggg ggggggggig 1.15 0.00 6.86 0.00 9.48 0.00 giggg species 0.00 0.00 0.13 0.00 0.18 1.65 Totals 30.90 26.70 25.40 31.10 35.80 36.00 1 Hhere PS and PD indicate presettlement and present-day, respectively. 2 basal area (mg/ha) APPENDIX C Pres - ommunities: truc ure and i ion Table 24: Composition of present-day forests (trees - 10 cm d.b.h.) on dry-mesic sites. Percent Percent Relative Relative Tree species Density Dominance IV Density Dominance m m-velggina- 38.12 41.31 39.721 344.452 11.033 gllipgoidalig Eggglgg aggggigggtggg 20.43 24.96 22.70 184.61 6.66 ggggggs 519g 16.32 16.87 16.60 147.47 4.50 55;; ngggg 12.55 7.66 10.11 113.40 2.05 592; gacghargg 6.32 3.87 5.10 57.11 1.03 [gggiggg gaggigggg 1.18 1.55 1.37 10.66 0.41 gggglg pggygifggg 1.37 0.86 1.12 12.38 0.23 11119 americana 1.09 1.11 1.10 9.85 0.30 Eiggg ggggggg 1.48 0.60 1.04 13.37 0.16 Eggggg serotina 0.60 0.26 0.43 5.42 0.07 Qgggyg Virginians 0.33 0.43 0.38 2.98 0.11 figggg grandifolia 0.16 0.26 0.21 1.45 0.07 Eiggg resinosa 0.05 0.26 0.16 0.45 0.07 Totals 100.00 100.00 100.00 903.60 26.70 1 2 trees/ha 3 basal area (mg/ha) 129 Importance Value = (relative density + relative dominance)/2 * 100 130 Table 25: Composition of present-day forests (trees 2 10 cm d.b.h.) on mesic sites. Percent Percent Relative Relative Tree species Density’ Dominance 1V Density Dominance Agg; gggghgggg 49.41 33.21 41.31 452.00 10.33 1111; ggg§1gggg 15.63 21.72 18.68 142.98 6.75 ggpg1g§ grandigggtata 10.23 15.50 12.87 93.58 4.82 ngg§g§_rubga-velutina- 6.54 12.23 9.39 59.83 3.80 21.12s2192__a figg§1gg§ ggg§1§ggg 6.31 7.64 6.98 57.72 2.38 Agg; ngggg 5.04 3.74 4.39 46.11 1.16 figggg grgggifglia 3.24 3.53 3.39 29.64 1.10 Qgggyg virgigigna 1.81 0.69 1.25 16.56 0.21 figgg1g gggy£11ggg 1.25 1.16 1.21 11.44 0.36 gggyg species 0.48 0.42 0.45 4.39 0.13 Eggggg sergtina 0.06 0.16 0.11 0.55 0.05 Totals 100.00 100.00 100.00 914.80 31.10 1 Importance Value = (relative density + relative dominance)/2 * 100 2 trees/ha 3 basal area (mzlha) 131 Table 26: Composition of present-day forests (trees 2 10 cm d.b.h.) on wet-mesic sites. Percent Percent Relative Relative Tree species Density Dominance IV Density Dominance age; sgchgrun 26.39 24.72 25.561 317.582 9.503 §g1g1g pggyrifgra 23.61 16.75 20.18 284.12 8.50 Qgggggs ggggg-velutina- 16.07 19.94 18.01 193.39 5.79 gllipgoidalig Tilia americana 11.96 18.50 15.23 143.93 4.31 195; ngggg 10.63 8.77 9.70 127.92 3.83 Fraxinus americana 3.08 4.78 3.93 37.06 1.11 g1gg§ species 4.57 1.59 3.08 55.00 1.65 figgg1g§ graggidentata 1.13 2.07 1.60 13.60 0.41 Qgggggg 9193 0.92 1.59 1.26 11.07 0.33 fi1gg§ agggggg 1.23 0.80 1.02 14.80 0.44 figggg graggifolja 0.41 0.49 0.45 4.93 0.15 Totals 100.00 100.00 100.00 1203.40 36.00 1 Importance Value = (relative density + relative dominance)/2 * 100 2 trees/ha 3 basal area (mzlha) APPENDIX D :. _.. s .L 0 cos_ _- ._ 31°38 Tnc- u_--:m- 2 -n Figure 26: Importance values of pre-European settlement and present-day tree species on dry-mesic sites. 132 133 09.0 04.977 ,c, 90.0 000.00 77 a09% 4&0? 0mm. W0 ado/s. 000/00/70Wv 0060.0 «MUG/E0 EoEoEmmoea Q >inEomoi D EQ/L, C®Om4 0 04 O (\i O Of) 0s 00 mm BOUBIJOdwl 134 Figure 27: Importance values of pre-European settlement and present-day tree species on mesic sites. 135 096.770 ’9 06/0 6.0.40 77 8090 90 0:090 000 0,0 OWN/004400 o0 . few as 2953320 0 $015me D 000$ 00 00077 00 cocoa O or ON 00 0s 00 enleA eaueiiodwl 136 Figure 28: Importance values of pre-European settlement and present-day tree species on wet-mesic sites. 137 w/zfrvw/t Z/m Mix/07¢ 00000 0.90 £0,400 006,0 04/00 0A00<éry 000CWO0 $004019000$00 0000 MVW/Aww 0004 0469» on... 906, 006 z, 560 EmEofommea 0 3010535 B O O O (\J 00 enleA eaueiipdwl LITERATURE CITED LITERATURE CITED Aber, J.D. and J.M. Melillo. 1991. Terrestrial ecosystems. Saunders College Publishing. Ahlgren, C.E. and I.F. Ahlgren. 1983. The human impact on northern forest ecosystems pp.33-51. Ln S.L. Flader (ed.). The:great lakes forest.an.environmental and social history. University of Minnesota Press, Minneapolis. Albert, D.A., S.A. Denton, and BuV. Barnes. 1986. Regional landscape ecosystems of Michigan. School of Natural Resources, University of Michigan, Ann Arbor. Anderson, R.C. 1970. Prairies in the prairie state. Transactions of the Illinois State Academy of Sciences 63: 214-21. Austin, M.P. 1985. Continuum.concept, ordination.methods, and niche theory. Annual Review of Ecology and Systematics 16: 39-61. Barnes, B.V. 1989. Old-growth forests of the northern lake states: a landscape ecosystem perspective. Natural Areas Journal 9: 45-57. Bourdo, E.A., Jr. 1955. A validation of methods used in analyzing original forest cover. Dissertation. University of Michigan, Ann Arbor. ----. 1956. A review of the General Land Office Survey and of its use in quantitative studies of former forests. Ecology 37: 745-768. 138 139 Bray, J.R. and J.T. Curtis. 1957. An ordination of the upland forest communities of southern Wisconsin. Ecological Monographs 27: 325-349. Brubaker, L.B. 1988. Vegetation history and anticipating future vegetation change. pp.41-61. in J.K. Agee and D.R. Johnson (eds.). Ecosystem. management for parks and wilderness. Univ. of Washington Press, Seattle. Burgis, W.A. 1977. Late-Wisconsinan history of northeastern lower Michigan. Dissertation. University of Michigan, Ann Arbor. Burns, R.M. and B.H. Honkala (eds.). 1990. Silvics of North America. USDA Forest Service. Agricultural Handbook No. 654. Washington, DC. Buttrick, P.L. 1923. Second growth hardwood forests in Michigan. Special Bulletin 123, Michigan Agricultural Experiment Station. Canham, C.D. and 0.L. Loucks. 1984. Catastrophic windthrow in the presettlement forests of Wisconsin. Ecology 65: 803- 809. Clark, J.S. 1989. Effects of long-term water balances on fire regime, northwestern Minnesota. Journal of Ecology 77: 989-1004. Cottam, G. and.J.T. Curtis. 1956. The use of distance measures in phytosociological sampling. Ecology 37: 451-460. Crankshaw, W.B., S.A. Qadir, and A.A. Lindsey. 1965. Edaphic controls of tree species in presettlement Indiana. Ecology 46: 688-698. Curtis, J.T. 1959. The vegetation of Wisconsin. University of Wisconsin Press, Madison, Wisconsin. Davis, C.M. 1935. The High Plains of Michigan. Papers of the Michigan Academy of Science, Arts, and Letters 21: 303- 341. 140 Delcourt, H.R. and PuA. Delcourt. 1974. Primeval magnolia- holly-beech climax in Louisiana. Ecology 55: 638-644. Digby, P.G.N. and R.A. Kempton. 1987. Multivariate‘analysis of ecological communities. Chapman and Hall, New York. Finley, R.W. 1976. Original vegetation cover of Wisconsin from U.S. General Land Office Notes (map). U.S. Forest Service, NC Forest Experiment Station, St. Paul, Minnesota. Fralish, J.F., F.B. Crooks, J.L. Chambers, and F.M. Harty. 1991. Comparison of presettlement, second-growth and old- growth forest on six site types in the Illinois Shawnee Hills. American Midland Naturalist 125: 294-309. Franklin, J.F. 1989. Importance and justification of long-term studies in ecology. Pages 3-19 In G.E. Likens (ed.). Long-term studies in ecology: approaches and alternatives. Springer-Verlag, New York. Franklin, J.F., K. Cromack, W. Denison, A. McKee, C. Maser, J. Sedell, F. Swanson, and. G. [Juday. 1981. Ecological characteristics of old-growth Douglas-fir forests. Gen. Tech. Rep. PNW-118. Portland, Oregon: U:S.D.A. Forest Service, Pacific Northwest Forest and Range Experiment Station. Frelich, L.E. and C.G. Lorimer. 1985. Current and long-term effects of deer browsing in hemlock forests in Michigan, USA. Biological Conservation 34: 99-120. Frelich, L.E. and C.G. Lorimer. 1991. Natural disturbance regimes in hemlock-hardwood forests of the upper Great Lakes region. Ecological Monographs 61: 145-164. Fulling, E.H. 1956. Botanical aspects of the paper-pulp and tanning industries in the United States-an economic and historical survey. American Journal of Botany 43: 621- 634. 141 Gates, F.C. and G.E. Nichols. 1930. Relation between age and diameter in trees of the primeval northern hardwood forest. Journal of Forestry 28: 395. Gill, J.L. 1978. Design and analysis of experiments in the animal and medical sciences. Vol.1. Iowa State University Press, Ames, Iowa. Goff, F.G. and D. West. 1975. Canopy-understory interaction effects on forest population structure. Forest Science 21: 98-108. Gordon, R.B. 1969. The natural vegetation of Ohio in pioneer days. Bulletin of the Ohio Biological Survey 3: 10-22. Grime, J.P. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. American Naturalist 111: 1169-1194. Grimm, E.C. 1984. Fire and other factors controlling the Big Woods vegetation of Minnesota in the mid-nineteenth century. Ecological Monographs 54: 291-311. Haberman, SJ. 1973. The analysis of residuals in cross- classified tables. Biometrica 29: 205-220. Hansen, A.J., T.A. Spies, F.J. Swanson, and J.Lu Ohmann. 1991. Conserving biodiversity in managed forests: lessons from natural forests. Bioscience 41: 382-392. Harmon, M.E., J.F. Franklin, F.J. Swanson, P. Sollins, S.V. Gregory, J.D. Lattin, N.H. Anderson, S.P. Cline, N.G. Aumen,.J.R. Sedell, G.W. Lienkaemper, K. Cromack»Jr., and K.W. Cummins. 1986. Ecology of coarse woody debris in temperate ecosystems. Advances in Ecological Research 15: 133-302. Havens, R.R. 1915. History of the Federal surveys in Michigan. Thesis. Michigan State University, East Lansing. 142 Heinen, J.T. and T.L. Sharik. 1990. The influence of mammalian browsing on tree growth and mortality in the Pigeon River State Forest, Michigan. American Midland Naturalist 123: 202-206. Heinselman, M.L. 1973. Fire in the virgin forests of the Boundary Water Canoe Area, Minnesota. Quaternary Research 3: 329-382. Host, G.E. and K.S. Pregitzer. 1992. Geomorphic influences on ground-flora and overstory composition in upland forests of northwestern lower Michigan. Canadian Journal of Forest Research 22: 1547-1555. ----., K.S. Pregitzer, c.wu Ramm, J.B. Hart, and D.T. Cleland. 1987. Landform-mediated differences in successional pathways among upland forest ecosystems in northwestern lower Michigan. Forest Science 33: 445-457. ----., K.S. Pregitzer, C.W; Ramm, D.P. Lusch, and D.T. Cleland. 1988. Variation in overstory biomass among glacial landforms and ecological land units in northwestern lower Michigan. Canadian Journal of Forest Research 18: 659-668. Hutchinson, M. 1988. A guide to understanding, interpreting and using the Public Land Survey field notes in Illinois. Natural Areas Journal 8: 245-255. James, F.C. and C.E. McCulloch. 1990. Multivariate analysis in ecology and systematics: panacea or pandora’s box? Annual Review of Ecology and Systematics 21: 129-166. Jordan, W.R., M.E. Gilpin, and. J.D. .Aber (eds.). 1987. Restoration ecology. Cambridge University Press. Kenoyer, L.A. 1934. Forest distribution in southwestern Michigan as interpreted from the original land survey (1826-32). Papers of the Michigan Academy of Science, Arts, and Letters. 19: 107-111. 143 Kittredge, J. and A.K. Chittenden. 1929. Oak forests of northern Michigan. Special Bulletin 190, Michigan Agricultural Experiment Station. Kline, V.M. and G. Cottam. 1979. Vegetation response to climate and fire in the driftless area of Wisconsin. Ecology 60: 861-868. Kozlowski, T.T., P.J. Kramer, and S.G. Pallardy. 1991. The physiological ecology of woody plants. Academic Press, Inc. New York. Kruskal, J.B. 1964. Multidimensional scaling: a numerical method. Psychometrika 29: 115-129. Larsen, W.C. Structure, biomass, and net primary productivity for an age-sequence of jack pine ecosystems. Dissertation. Michigan State Univ., East Lansing, Michigan. Leitner,L.A., C.P. Dunn, G.R. Guntenspergen, F. Stearns, and D.M. Sharpe. 1991. Effects of site, landscape features, and fire regime on vegetation patterns in presettlement southern Wisconsin. Landscape Ecology 5: 203-217. Leopold, A. 1948. A Sand County Almanac. Oxford University Press: New York. Lindsey, A.A., W.B. Crankshaw, and S.A. Qadir. 1965. Soil relations and distribution map of the vegetation of presettlement Indiana. Botanical Gazette 126: 155-163. Lorimer, C.G. 1977. The presettlement forest and natural disturbance cycle of northeastern Maine. Ecology 58: 139- 148. Lorimer, C.G. 1980a. The use of land survey records in estimating presettlement fire frequency. Pages 57-62 In M.A. Stokes and.J.H. Dieterich (eds.). Proceedings.of the fire history workshop, 20-24 October 1980, Tucson, Arizona. U.S. Forest Service General Technical Report RM- 81. 144 ----. 1980b. Age structure and disturbance history of a southern Appalachian virgin forest. Ecology 61: 1169- 1184. Loucks, O.L. 1970. Evolution of diversity, efficiency, and community stability. American Zoologist 10: 17-25. Lovis, W.A., W.T. Langhorne, and S. Van Arsdel. 1978. A cultural resource overview of the Huron-Manistee National Forest, Michigan. Michigan State University‘ Museum, Archaeological Survey'iReport, No. 32, East. Lansing, Michigan. Marschner, F.J. 1974. The original vegetation of Minnesota (map). Redrafted from 1930. U.S. Forest. Service iNC Experiment Station, St. Paul, Minnesota. Maybee, R.H. 1976. Michigan’s white pine era: 1840-1900. Michigan Historical Commission. John M. Munson Michigan History Fund Publication Pamphlet 1. McCune, B. and G. Cottam. 1985. The successional status of a southern Wisconsin oak woods. Ecology 66: 1270-1278. McIntire, G.S. 1931. Theory and practice of forest typing, with special relation to the hardwood and hemlock associations of northern Michigan. Papers of the Michigan Academy of Science, Arts, and Letters 15: 239- 251. McIntosh, R.P. 1985. The background of ecology: concept and theory. Cambridge Univ. Press, New York. Melillo, J.M., J.D. Aber, and J.F. Muratore. 1982. Nitrogen and lignin control of hardwood leaf litter decomposition dynamics. Ecology 63: 621-626. Monk, C.D. 1966. An ecological significance of evergreen-ness. Ecology 47: 504-505. Mueller-Dombois, D. and H. Ellenberg. 1974. Aims and methods of vegetation ecology. John Wiley and Sons, Inc., New York. 145 Mustard, T.S. 1983. The vegetation of the Manistee National Forest, Oceana and Mason counties, Michigan. I. Physical, historical, and ecological aspects. Michigan Botanist 22: 111- 122. Noss, R.P. A regional landscape approach to maintain diversity. Bioscience 33: 700-706. Padley, E.A. 1989. Associations among glacial landforms, soils, and vegetation in northeastern lower Michigan. Ph.D. Dissertation, Michigan State University, East Lansing. Palik, B.J. and K.S. Pregitzer. 1992. A comparison of presettlement and present-day forests on two bigtooth aspen-dominated landscapes in northern lower Michigan. American Midland Naturalist 127: 327-338. Pastor, J., J.D. Aber, C.A. McClaugherty, and J.M. Melillo. 1984. Aboveground production and N and P cycling along a nitrogen mineralization gradient on Blackhawk Island, Wisconsin. Ecology 65: 256-268. Peet, R.K. 1974. The measurement of species diversity. Annual Review of Ecology and Systematics 5: 285-307. Peet, R.R. and O.L. Loucks. 1977. A gradient analysis of southern Wisconsin forests. Ecology 58: 485-499. Pickett, S.T.A., S.Iu Collins, ande.J..Arnesto. 1987. Models, mechanisms, and pathways of succession. Botanical Review 53: 335-371. Probst, J.R. 1988. Kirtland’s warbler breeding biology and habitat. management. pp.28-35. In Integrating Forest Management for Wildlife and Fish. U.S. Forest Service, NC-122, St. Paul, Minnesota. Probst, J.R. and T.R. Crow. 1991. Integrating biological diversity and resource management. Journal of Forestry 89: 12-17. 146 Romme, W.R. and D.H. Knight. 1981. Fire frequency and subalpine forest succession along a topographic gradient in Wyoming. Ecology 62: 319-326. Roth, F. 1905. Yearly report of the warden of the forest reserves. Report of the Michigan Forestry Commission for the Years 1903-1904, pp.22-41. Lansing, Michigan. Rowe, J.S. 1984. Forestland classification: limitations of the use of vegetation. Pages 132-147. In J.G. Bockheim (ed.) Proceedings of ‘the Symposium for Forest Land Classification: Experiences, Problems, and Perspectives, 18-20 Mar., Madison, Wisconsin. Runkle, J.R. 1982. Patterns of disturbance in some old-growth mesic forests of eastern north america. Ecology 63: 1533- 1546. Sandberg, L. 1983. The response of forest industries to a changing environment. pp.194-204. 1n S.L. Flader (ed.). The great lakes forest: an environmental and social history. University of Minnesota Press, Minneapolis. Sargent, G.S. 1894. Report on the Forests of North America. Census Office, U.S. Dept. of the Interior. SAS Institute Inc. 1985. SAS user’s.guide: statistics. Version 5 Edition. SAS Institute, Inc. Cary, North Carolina. Shugart, H.H. 1984. A theory of forest dynamics. Springer- Verlag, New York. Shugart, H.H., D.C. West, and W.R. Emanuel. 1981. Pattern and dynamics of forests: an application of simulation models. pp.74-94. 1n D.C. West, H.H. Shugart, and D.B. Botkin (eds.) . Forest succession: concepts and application. Springer-Verlag, New York. Simard, A.J. and R.W. Blank. 1982. Fire history of a Michigan jack pine forest. Michigan Academician 15: 59-71. 147 Simard, A.J., D.A. Haines, R.W. Blank, and J.S. Frost. 1983. The Mack Lake Fire. U.S.D.A. Forest Service. NC Experiment Station. General Technical Report NC-83. Smith, D.M. 1986. The practice of silviculture. 8th edition. John Wiley and Sons, New York. Sneath, P.H. and R.R. Sokal. 1973. Numerical taxonomy. W.H. Freeman and Co., San Francisco. Sokal, R.R. and F.J. Rohlf. 1981. Biometry. 2nd edition. W.H. Freeman and Co., San Francisco. Soule, M.E. 1986. Conservation biology: the science of scarcity and diversity. Sinauer .Assoc., Sunderland. Massachusetts. Sprugel, D.G. 1991. Disturbance, equilibrium, and environmental variability: what is 'natural' in a changing environment? Biological Conservation 58: 1-18. Spurr, S.H. and B.V. Barnes. 1980. Forest ecology. 3rd edition. John Wiley and Sons, New York. Stearns, F.W. 1949. Ninety-years change in a northern hardwood forest in Wisconsin. Ecology 30: 350-358. Strahler, A.H. 1978. Binary discriminant analysis: a new method for investigating species-environment relationships. Ecology 59: 108-116. Tilman, D. 1988. Plant strategies and the dynamics and structure of plant communities . Princeton: Princeton University Press. U.S.D.A. Forest Service. 1980. National Forest Fire Report. Washington, D.C. U.S.D.A. , Forest Service; annual report. U.S.D.A. Soil Conservation Service. 1990. Soil survey of Ogemaw County Michigan. 148 Veatch, J.O. 1959. Presettlement forest in Michigan (Map). Dept. of Resource Development, Michigan State University, East Lansing. Veatch, J.O., L.R. Schoenmann, C.E. Millar, and A.E. Shearin. 1931. Soil survey of Oscoda County, Michigan. U.S.D.A. Veatch, J.O., J.T. Stone, F.W. Suggitt, and L.S. Robertson. 1941. Soil associations, Alcona County, Michigan. U.S.D.A. Voss, E.G. 1985. Michigan flora. Part II. Dicots (Saururaceae- Cornaceae). Cranbrook Institute of Science Bulletin 59 and Univ. of Michigan Herbarium. Ward, J.H. 1963. Hierarchical grouping to optimize an objective function. Journal of the American Statistical Association 58: 236-244. Wartenberg, D., S. Ferson, and F.J. Rohlf. 1987. Putting things in order: a critique of detrended correspondence analysis. The American Naturalist 129: 434-448. White, C.A. 1984. A history of the rectangular survey system. Bureau of Land Management, U.S. Dept. of the Interior, Washington, DC. White, P.S. 1979. Pattern, process, and natural disturbance in vegetation. The Botanical Review 45: 229-299. White, P.S. and.S.T.A. Pickett. 1985. Natural disturbances and patch. dynamics: an introduction, pp.3-13. In, S.T.A. Pickett and P.S. White (eds.). The ecology of natural disturbance and patch dynamics. Academic Press, New York. Whitney, G.G. 1982. Vegetation-site, relationships in 'the presettlement forests of northeastern Ohio. Botanical Gazette 143: 225-237. ----. 1986. Relation of Michigan's presettlement pine forests to substrate and disturbance history. Ecology 67: 1548- 1559. 149 ----. 1987. An ecological history of the Great Lakes forest of Michigan. Journal of Ecology 75: 667-684. ----. 1990. The history and status of the hemlock-hardwood forests of the Allegheny Plateau. Journal of Ecology 78: 443-458. Whittaker, R.H. 1975. Communities and ecosystems. Macmillan Publ. Co., 2nd ed. Wilkinson, L. 1990. SYSTAT: The system for statistics. SYSTAT, Inc. Evanston, Illinois. Wilson, E.O. (ed.). 1988. Biodiversity. National Academy of Sciences. Zak, D.R., G.E. Host, and K.S. Pregitzer. 1989. Regional variability in nitrogen mineralization, nitrification, and overstory biomass in northern lower Michigan. Canadian Journal of Forest Research 19: 1521-1526. Glacial Landforms of the High Plains Region of the Huron National Forest (After Burgis1977) GLACIAL LANDFORM DELINEATION Glacial landforms were mapped on a base map constructed from 30 x 60 minute U.S. Geological Survey 0 topographic quadrangles using overlays of Burgis’s (1977)1 glacial landform maps for northeastern lower Michigan. Farrand’s (1982)2 map of the quaternary geology of Michigan was used as an independent test of Burgis’ maps. Soil series maps were also used to infer complex boundaries. 1 Burgis,W.A.1977. Late-Wisconsinan histow of north-eastern lower Michigan. Ph. D. Dissertation. University of Michigan, Ann Arbor, Michigan. 2Farrand, W.R.1982. Quaternary geology of Michigan. Geological Survey, Dept. of Natural Resources, State of Michigan, Lansing, Michigan. z' [x N ..." M . Z to N l.‘ 2' o 3 M l—' R. 5E. Scale 1: 100,000 metric z' 1 cm = 1 km a M.L. 3/3/94 I_- K.S. P. AuSable River 0 O | M ’M M O O Mack Lake ' GE ‘ R. 1 E. I R. 2E. M | Location of the Study Area M I! O (M R. 3E. I R. 4E. GLACIAL LANDFORM TYPES Type O: Glacial outwash sand and gravel and postglacial alluvium comprise this type. This type occurs as fluvial terraces along current and abandoned drainage- ways, as fans and sheets flanking end moraines, and as deltas along glacial lake margins. Glacial debris here is usually well sorted and well-stratified. Textures are predominantly sand, but significant areas of muck and peat and shallow organic soils occur. Topographically these areas are nearly level with only small kettle depressions interrupting the planar topography of this landscape. Soil drainage classes here tend to fall into either of the two extremes: excessively drained or poorly drained. Type M: End and ground moraines comprise this type. The bulk of these areas are end moraines which topographically occur as linear belts of hummocky relief marking former stillstands of the ice-sheet margin. Texturally these areas are heterogeneous with loamy sands, sandy Ioams, and loams predominating. Small areas of muck and peat occur in kettle features embedded in the morainal complex. Glacial debris here is usually non-sorted, but small areas of well- sorted fluvial deposits occur. Soils are mainly well drained to moderately well- drained. Type I: Ice-contact outwash sand and gravel comprise this type. Topographically this type is usually rugged, with short, steep hills, and abrupt depressions. The type can be subdivided into component landforms: kames, eskers, interlobate tracts, proglacial outwash, sandy till, and kettle holes. Textures here are predominantly sand and loamy sand with considerable cobbles and gravel. Glacial debris is often poorly sorted, non-stratified, and has slump and defor- mational features. Soil texture can be quite heterogeneous over short distances. Soils are somewhat excessively drained and well drained. Pre-European Settlement Forest Cover Types of the High Plains Region of the Huron National Forest FOREST COVER TYPE METHODS Each quarter-section and section corner within the study area was recorded as to its two General Land Office (GLO) survey witness tree species and soil drainage class. Witness tree data were grouped by soil drainage classes into five edaphic site types (xeric, dry-mesic, mesic, wet-mesic, and hydric). For each tree species in a site type the relative density, relative dominance (basal area), and an importance value (the average of relative density and dominance on a percentage basis) were calculated. Forest cover types were then defined by the dominant tree species occurring on each site type. Dominant tree species were those tree species with an importance value of Z 10% for a site type. A nonforested “marsh” cover type included wetland herbaceous emergents and was created from surveyor descriptions. The accuracy of cover types were checked against a correspondence analysis and a cluster analysis of the dominant forest species. The tree species patterns described by these numerical procedures fit well with the cover types as defined. The GLO survey section line descriptions were used to place every section mile in the study area within a cover type. Forested and non-forested wetlands were mapped from the surveyor’s plat maps. GLO survey section line descriptions listthe tree species encountered along the mile the surveyors traversed according to their order of importance. In this way, the section line descriptions were fitted to the quantitatively defined cover types. The map of the cover types was produced using 30 x 60 minute U.S. Geological Survey topographic quadrangles as a base map. When boundaries needed to be inferred, topography and soils data were used to define them. /VW T.27N. AuSable River 5 Location of the Study Area 3 J Mack Lake W @s >963? _ D Mi. .o/ii @ 5.. 43 5 233 21 T. 24N. 1: 100,000 metric Scale 1cm=1 km M. L. 3/3/94 K.S. P. FOREST COVER TYPE DESCRIPTIONS ‘ n _ Ulln‘, Type 1: Jack Pine-Redfine-White Pine 2:11:34. Jack pine, red pine, and white pine together accounted fér'Z 91% of the importance value on these sites. However, jack pine truly dominated this type with an importance value of 52%. Associates (those species which had an importance value between 1 and 9%) of this type included: species of the red oak complex, aspen species, and hemlock. This type occurred on xeric sites with low soil moisture and nutrient availability. These areas were dominated by sandy soils on topographically level glacial outwash plains and ”drainageways. Witness tree density and basal area were low on these sites, being approximately 214 trees/ha and 1 3.2 m2/ha, respectively. Average witness tree size was about 23.0-23.8 cm. Type 2: Red Pine-Jack Pine-White Pine - Red pine, jack pine, and white pine together accounted for Z 80% of the importance value on these sites. Associates ofthis type included: species of the red oak complex, hemlock, aspen species, beech, sugar maple, and white oak. This type occurred on dry-mesic sites with moderately low soil moisture and nutrient availability. Physiographically this type occurred on coarse-textured end-moraines; areas of broad rolling hills. Witness tree density and basal area were approximately31 5 trees/ha and 30.9 m2/ha, respectively. Average witness tree size was about 29.0-306 cm. ' ~ ' '- Type 3: Hemlock-Beech-Sugar Maple-White Pine Hemlock, beech, sugar maple, and white pine together accounted for Z 75% ofthe importance value on these sites. Associates ofthis type included: red pine, species of the red oak complex, basswood, aspen species, white cedar, black ash, and jack pine. This type occurred on mesic sites with moderate levels of soil moisture and nutrient availability. These sites were found on end moraines, ground moraines, and ice-disintegration features. Topographically this type occupied areas of short, steep hills an small holtows. Soii textures included loamy sandy, sandy Ioams, and loams. Witness tree density and basal area were approximately 241 trees/ha and 25.4 m2/ha, respectively. Average witness tree size was about 32.1-33.5 cm. Type 4: Hemlock-White Pine-Red Pine-Aspen Species Hemlock, white pine, red pine, and aspen species together accounted for_>_ 59% of the importance value on these sites. Associates of this type included: white cedar, jack pine, paper birch, sugar maple, birch species, beech, balsam fir, black ash, tamarack, species of the red oak complex, yellow birch, elm species, and red maple. These sites were wet-mesic. Physiographicallythis type occurred on small, wet, sandy flats where morainal compiexes and ice-contact features of substantial relief abutted outwash plains. Witness tree density and basal area were approximately 366 trees/ha and 35.8 mQ/ha, respectively. Average witness tree size was about 29.1-31.5 cm. Type 5: White Cedar-Tamarack White cedar and tamarack together accounted forZ 58% of the importance value on these sites. Associates ofthis type included: jack pine, spruce species, aspen species, balsam fir, black ash, birch species, paper birch, white pine, red maple, and hemlock. These sites were hydric. Topographically they occurred predominantly as poorly drained flats on outwash plains and as kettle features in all landforms. Muck and peat and thin organic deposits over sands dominated the soils of these sites. Nutrient availability varied. Witness tree density and basal area were approximately 366 trees/ha and 18.3 mQ/FIa,V’respectively. Average witness tree size was about 22.1 4-23.46 cm. ”’1“ Type 6: Marsh ',' ' ”3‘ This cover type was derived from surveyor descriptions o'f'non-forested wetlands dominated by herbaceous emergents. The type occurred sporadically in small kettle features and surrounding small lakes in all landform types. "it?liltiillliflill‘lillli