ll“\l‘lfifilfllllllllllllll u Milli ‘1‘2‘93 01091 1:551,— LififiéiRY Michigan State University This is to certify that the dissertation entitled Growth Dynamics and Successional Trends in an Old Growth, Cedar—Hardwood Dune Forest presented by Michael Lloyd Scott has been accepted towards fulfillment of the requirements for Ph.D. degreein Botany and Plant Pathology Mnjorprofessot Peter G. Murphy Date 7 January 1985 “(Iii-n-AMM A . '- lm' - y . . 0-1277' MSU LIBRARIES RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. GROWTH DYNAMICS AND SUCCESSIONAL TRENDS IN AN OLD GROWTH, CEDAR-HARDWOOD DUNE FOREST BY Michael Lloyd Scott A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Botany and Plant Pathology 1984 C>1985 MICHAEL LLOYD SCOTT All Rights Reserved ABSTRACT GROWTH DYNAMICS AND SUCCESSIONAL TRENDS IN AN OLD-GROWTH, CEDAR-HARDWOOD FOREST BY Michael Lloyd Scott A 20 hectare old-growth stand of northern white cedar (Thuja occidentalis L.) and associated hardwoods, including sugar maple (Acer saccharum Marsh.) and white ash (Fraxinus americana L.), exists on the southwestern corner of South Manitou Island, Michigan. Mature cedar within the stand were associated with a series of old-dune ridges and early growth rates for these trees were statistically equivalent to those for cedar currently establishing upon open dune slopes. It is suggested that the present forest association was initiated when cedar colonized the slopes of dunes which had advanced toward hardwood stands inland. The present stand has a total aboveground biomass of 148.5 metric tonnes/hectare and a total aboveground annual net primary productivity (NPP) of 7.1 mt/ha/yr. The dom- inance of cedar in the stand results from a collection of massive individuals which together account for 61% of total stand biomass. These large cedar are rather unproductive, however, and account for only 26% of total stand production. In contrast, numerous small. maple .represent. the largest percentage (42%) of stand NPP. Despite the current importance of cedar, this Species is not successfully replacing itself in smaller size- classes. Cedar seedling establishment has become increas- ingly restricted to log substrates and currently 99.7% of all first year seedling establishment occurs on logs. Logs in advanced stages of decay represent the best substrates for germination and early growth of seedlings. However, long-term survival on logs appears tenuous. Mortality for log-established cedar seedlings reaches 99.4% by year 12 and no cedar > 40 cm dbh showed evidence of having established on a log substrate. Old-growth cedar were logged from a 3.4 ha portion of mature forest in 1896. The stems removed represented a basal area of 49 mz/ha. Post-logging trends in stand com- position reveal that between 1920 and 1930 the basal area for second-growth sugar maple began to increase rapidly. Currently, maple accounts for 77% of total stand basal area (29 m2/ha). In the 85 years following logging, bole biomass and productivity in the second-growth stand have recovered to levels which exceed those found presently in the mature stand. Second-growth bole biomass (186.4 mt/ha) was 38% greater than mature bole biomass (allometrically estimated as 115.3 mt/ha) and bole NPP (3.9 mt/ha/yr) was nearly twice that of the mature stand (2.1 mt/ha/yr). To my wife Dr. Kathryn Vosburgh Scott and my parents Lloyd Allen Scott and Mary Margaret Scott ii ACKNOWLEDGEMENTS I would like to express my sincere appreciation to my major professor Dr. Pete Murphy who's articulate enthusiam for S. Manitou provided me with the incentive to initiate this project. His ideas, support, and willingness to listen helped also to bring the study to completion. Thanks are also extended to my committee members; Dr. Jay Harman, Dr. Donald Dickmann and Dr. Kurt Pregitzer. Their careful reading of this work did much to improve the end result. The National Park Service personnel were most helpful, particularly Steve Yancho, Max Holden and Don Hamilton. Several drafts were cycled past my office-mate and friend Gerry Donnally who was always willing to help with grammatical, conceptual and emotional rough spots. Valuable comments were also received from Bill Larsen, Lynn Robbins and Laura Huenneke who judiciously reviewed later drafts. I will never forget the many pleasant hours shared with Ray Sinclair on South Manitou. His unbounded energy and love of the outdoors significantly reduced the tedium of field work. Thanks also to my father, Bill Rose, Kurt and Keith Vosburgh and my wife who also provided welcomed help with data collection. Early versions of the dissertation were graciously typed by Keith Vosburgh and my wife. I would also like to iii acknowledge the support of my long time friends Mary Slocum Trauba, Jim Sluyter, Fred Hampel, and Cheryl Hannan Butman who have always been there with encouraging words. Finally, without the understanding and love of my wife and parents this would not have been possible. iv TABLE OF CONTENTS List of Tables ..................................... List of Figures .................................... INTRODUCTION ....................................... Cedar and Cedar Forests ....................... The South Manitou Forest ...................... Research Hypotheses and Objectives ............ METHODS ............................................ Forest Composition and Structure .............. Forest Soil ................................... Stand Origin .................................. Forest Succession ............................. Stand Mortality ............................... Biomass and Productivity ...................... Biomass .................................. Productivity ............................. Forest Recovery Following Logging ............. Stand Description ........................ Estimated Date of Logging ................ Forest Recovery .......................... RESULTS ............................................ Composition and Structure of the Old-Growth Stand .................................... Forest Soil ................................... 28 31 34 37 37 42 Stand Origin .................................. Successional Status of the Old-Growth Stand... Gap Reproduction ......................... Cedar Seedling Establishment ............. Stand Mortality ............................... Biomass and Productivity ...................... Biomass .................................. Productivity ............................. Forest Recovery ............................... Estimated Date of Logging ................ Forest Regeneration ...................... DISCUSSION ......................................... Composition, Structure, and Origin of the Old-Growth Stand .............................. Successional Status of the Old-Growth Stand... Gap Reproduction ......................... Regeneration and Growth of Cedar Under a Closed Canopy ............................ Patterns of Mortality for CanOpy and Sub-Canopy Trees ......................... Biomass and Productivity in the Old—Growth Stand ......................................... Forest Recovery Following Logging of the Old-Growth Stand .............................. Date of Logging .......................... Trends in Forest Structure and Function Following Logging ............................. vi 48 48 58 58 58 62 65 65 69 87 87 90 93 96 100 100 103 103 107 Page SUMMARY AND CONCLUSIONS ............................ 113 Al. Criteria used in assessing decay states of logs .......................................... 117 LITERATURE CITED ................................... 118 vii LIST OF TABLES Table 1. 10. 11. 12. 13. 14. Importance values for the old-growth forest ........................................ Average soil characteristics for the study area .......................................... A t-test comparison between the number of cedar stems per hectare, on dune ridges and between dune ridges ........................... A F-testl/ and Tukey's/honestly significant difference (HSD) test2 for comparison of mean annual growth rates for the first 50 years of growth for cedar in three locations ........... Summary data for standing dead cedar stems and the four nearest live trees ............... Seedling density (seedling/m2) for logs in four states of decay .......................... Life-table for white cedar seedlings in an old-growth stand .............................. Summary of mortality data from the mature stand for all trees (3 2.5 cm dbh) within twelve 250 m2 plots .......................... Estimated biomass values (mt/ha) .............. Estimated bole (wood and bark) annual net primary productivity .......................... Mean annual litter production ................. Estimated annual net primary productivity (mt/ha/yr) .................................... Summary of growth release in cedar in response to logging ........................... Summary information on trees which had established on logged cedar stumps within the second-growth stand ....................... viii 45 47 51 54 56 59 61 63 64 66 68 71 Table 15. 16. 17. 18. A1. Population age structure of second-growth sugar maple ................................... Statistical comparison (t-test) between density (A) and dominance (B), for cedar in the second-growth and old-growth stand prior to logging .............................. Future canopy composition estimated from relative density, relative dominance, and an average of these values for trees from four canopy gaps ................................... Comparison of biomass, production and biomass accumulation ratios for old-growth Michigan stands ........................................ Criteria used in assessing decay states of logs .......................................... ix Page 73 77 95 102 117 LIST OF FIGURES Fi re 1. Island archipelago of northwestern Lake Michigan, including South Manitou Island, Michigan ...................................... 2. Southwest corner of South Manitou Island, Michigan. Light stippling indicates approximate extent of old-growth cedar- hardwood forest, and heavy stippling de- limits the study area. Dashed lines define a two-track road. Contour lines in meters.... 3. Map of study area indicating locations of soil pits, sample plots, walking path and large cedar marked by the Michigan Botanical Club. Size of sample plots are drawn to scale, and all contour lines are in meters. The second-growth forest is indicated by stippling ..................................... 4. Location of profile diagram transects (solid bands) and transects assessing cedar density relative to dune topography (stippled bands). Study area delimited by dashed line. Contour lines in meters ............................... 5. Location of canopy gaps within the study area. Contour lines in meters ....................... 6. Regression of height on diameter for cedar. Y (height in meters) = 25.22 - 25.29 e -0.025 x (dbh in cm) (H = 23; r = 0.88) ................ 7. Regressions of height on diameter ratios for white cedar. Regressions for trees < 15.3 cm, dbh (A); X (height in meters) = —0.0655 Y (diameter to dbh, ratios) + 1.09 (n = 4, r = 0.9344). For trees 3 15.3 cm, dbh.(cn; X (height in meters) = -0.04414 Y (diameter to dbh, ratios) + 1.07 (n = 7, r = 0.9058) ....... 8. Location of litter plots (branches and leaves) within the study area. Contour lines in meters ............................... Figure 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Location of old-growth and second-growth study areas on South Manitou Island, Michigan. Old-growth stand indicated by tree symbols and second-growth stand by heavy stippling ........ Regression of diameter at breast height on diameter at stump height (diameter at 30 cm above ground surface), for old-growth cedar. Y (diameter at breast height) = 0.886 X (diameter at stump height - 0.659 (n = 29, r = 0.99) ..................................... Regression of age on diameter for second- growth sugar maple. Y (age in years) = 1.2 X (dbh cm) + 26.8 (n = 27, r = 0.85) .......... Size class distribution for all trees within the old-growth stand .......................... Size class distributions for maple, cedar and ash in the old-growth stand ................... Height class distribution for all trees (_>_ 2.5 cm dbh) ................................ Distribution of cedar in relation to dune topography within the mature stand ............ Comparison of cumulative growth for old- growth and open-dune cedar (A), and for old- growth and closed-forest cedar (B). Error bars reflect standard error of the mean ....... A) Species importance values for trees within four canopy gaps. B) Combined importance values from four canopy gaps. Tree species abbreviations: Acer saccharum (AS), Fraxinus americana (FA), Thuja occidentalis (TO), Prunus virginiana (PV), and Cornus alternifolia (CA). Importance values = (relative density) + (relative dominance) + (relative frequency) .......................... Importance values for four trees nearest dead standing cedar stems ..................... Occurrence of cedar stems by size classes relative to log and soil substrates ........... xi 29 3O 33 39 4O 41 44 46 49 50 53 Figgre 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. Population age structure for cedar seedling (5 1 m in height). Seedling numbers adjusted to a first year cohort of 1000 individuals ................................... Regression analysis of cedar seedling survivorship. A) Power function: Log Y (number of stems/ha) = 9.92 - 3.17 L08e X (age in years) (n = 12, r = -0.958). B) Negative exponential: Log Y (number of stems/ha) = 6.84 - 0.415 X (Sge in years) (n = 12, r = -0.924) .......................... Direction of tree falls within the study area. Chi-square goodness of fit test for direction of fall was not significant (x2 = 3.75; a > 0.05), indicating treefall was random .................................... Patterns of average annual growth (mm) for old-growth cedar adjacent to the logged stand. Dates indicate year of apparent growth release ................................ Patterns of average growth (mm) for old-growth cedar from within the mature stand ............ Age structure of second-growth sugar maple within logged study area. Negative exponen- tial model: Log Y (stems/ha) = 7.5 - 0.06 X (age in years a% breast height) (r = 0.957), fitted to age class distribution .............. Size range of cedar removed by logging adjacent to the old-growth stand. Dbh estimated from stumps based on a regression of dbh on diameter at stump height (stump height: defined as being 30 cm above surface of the ground) ................................ Comparisons of density and dominance for cedar in logged and old-growth locations. Density and dominance estimates reflect pre-logging values for both locations ..................... Importance values for the mature stand (A), and the logged stand (B) ...................... Basal area changes for cedar and maple in the second-growth stand since 1900 ................ xii Page 55 57 60 67 7O 72 74 75 78 80 Figure 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. Page Basal area changes for cedar and maple in the old-growth stand since 1900 ................... 81 Shifts in the relative contribution of cedar, maple, and other species to basal area, between the old-growth and second-growth stands. Circle size reflects relative basal area .......................................... 82 Stem biomass and net stem (bole) primary pro- ductivity for all trees in the second-growth stand ......................................... 83 Stem biomass and net stem (bole) primary pro- ductivity for trees in the old-growth stand... 84 Comparisons of total bole net primary produc- tivity (A) and total bole biomass (B) between old-growth and second-growth stands. Species categories include: cedar, maple, and other... 85 Photograph of cedar establishing on the open, lee slope of a dune which is encroaching upon hardwood forest of the island interior ........ 89 Profile diagram from within the mature forest. Cedar is indicated by heavy stippling ......... 91 Negative exponential model analysis of size distribution of maple, cedar, and ash in the old-growth forest ............................. 92 Aboveground biomass relative to net primary productivity (NPP). Closed circles represent old-growth Michigan stands; 1 (this study), 2 (Larsen 1982), 3 (Murphy and Kroh unpub. data), and 4 (Rose 1982). Dashed line is trend for climax biomass verses NPP. Solid lines include climax stands from woodlands through temperate—zone forests. Forest stands above biomass trend line were reported to be stable climax associations, while some climax or near-climax stands below this line had been affected by disturbance (Whittaker, Bormann, Likens and Siccama 1974). Graph modified from Whittaker et a1. 1974 ........... 104 Establishment of a second-growth birch on a cut cedar stump within the logged study area.. 106 xiii INTRODUCTION Cedar and Cedar Forests Northern white cedar (Thuja occidentalis L.) is a com- ponent of boreal and northern-mixed-hardwoods forest associ- ations (Braun 1950) and its range is limited to southeastern Canada and portions of the mid-western and northeastern U.S. (United States Forest Service 1965). The northern limit of the continous range for cedar extends from southeastern Manitoba to the southern end of James Bay and eastward to the Gulf of St. Lawrence. The southern limit stretches from east-central Minnesota through central Wisconsin and extends around the southern tip of Lake Michigan. From the lake eastward the range includes southern-lower Michigan, southern New York state through central Vermont and New Hampshire. Outside of this range cedar occurs in disjunct, localized populations through southern Ohio, western Pennsylvania and into the Appalachian Mountains from Virginia and West Virginia south to North Carolina and Tennessee. Cedar is common in areas characterized by low topo- graphy and high water tables sudh as sluggish streams and partially filled lake basins (Gates 1926). Most commer- cially harvested stands of cedar are from swampland, but the species is occasionally an old-field invader of moist, 1 2 well-drained soils of limestone origin. Old-field cedar typically show good fOrm on these sites, and exhibit rates of growth up to three times greater than trees from swamps (Curtis 1944). The occurrence of cedar in both poorly drained lowland habitats and on well-drained upland sites has led to suggestions that there are distinct uplands and lowland ecotypes (Potzger 1941 and 1944). Several studies indicate that there may be some morphological (Habeck 1959) and physiological (Caulkins 1967, Musselmann et a1. 1975) differences but the data are inconclusive. On certain swamp sites cedar is able to perpetuate itself in dense, almost pure stands but more often it is associated with a number of other tree species. On moist sites common associates include black ash (Fraxinus nigra), tamarack (Larix laricina), and black spruce (Picea mariana), while on better drained sites they would include eastern white pine (Pinus strobus), basswood (Tilia americana) and occasionally sugar maple (Acer saccharum). Cedar reproduces vegetatively by layering, and in some swamp stands this form of reproduction has been shown to account for as much as 73% of all new stems (Nelson 1951). In addition to layering, cedar reproduces effectively from seed. Germination and early growth of seedlings are reported to be best on moist, organic seedbeds including moss patches, rotting logs and burned-over soils (Johnston 1977). The seedlings are moisture sensitive and desiccation 3 in swamp sites appears to be a primary cause of mortality (Nelson 1951). In Michigan, the majority of cedar habitat is lowland swamp (Johnston 1977), but the species has been reported from a limited number of upland sites as well, including the dunes of Lake Michigan (Cowles 1901, Wolfe 1932). Cedar is rare in protected ravines along the southern dunes (Moran 1978) but becomes increasingly common northward (Waterman 1922). Olson (1958) characterizes cedar as an early suc- cessional dune species colonizing damp depressions and lee dune slopes. The association of cedar with dune slopes may be related in part to soil conditions. Specifically, cedar has been reported to have a very high calcium content in its foliage and wood (Dyer 1967, Young and Carpenter 1969) and the return of this element (and others) through litterfall may help to explain the rapid soil improvement and high exchangeable base values observed in these locations (Olson 1958). Additional nutrient inputs may result from the importation of carbonate rich sands and organic material, such as leaves, from more exposed dune slopes. Aside from soil improvement, these sites may also receive additional moisture along inequalities in the bedding planes of sand, as water moves from high. surrounding' dunes. Thus, the occurrence of cedar in such locations may relate to the unique soil conditions which develop on lee slopes. The South Manitou Forest The finest known example of upland cedar forest is found on South Manitou Island in Lake Michigan. South Manitou is the southernmost island of an archipelago located in northeastern Lake Michigan (Figure 1), and is situated 11 km to the northwest of Sleeping Bear Point and 24 km due west of the town of Leland, Michigan. On the southwest corner of the island is an old-growth forest of approxi- mately 20 ha (Figure 2), dominated by white cedar (Hatt 1948, Thompson 1963). Of particular note are the immense size and great age reported for cedar in the stand. Thompson reported tree heights in excess of 31 m (100 ft) and estimated the age of one tree to be more than 500 years. Because a portion of the forest is located between and on two large dune ridges, it was named "The Valley of the Giants". As the result of logging activities on South Manitou at the turn of the century (Rogers 1966), much of the timber on the island was cut, including a portion of this old-growth cedar stand as described by Hatt (1948). Many large cedar were removed during logging and presently the second-growth stand adjoins 10.9 ha of mature forest which was the focus of this investigation (Figure 3). At the time work was initiated, a footpath roughly paralled the east edge of the study area, extending from southeastern to northeastern corners. The defined eastern boundry of the stand follows the crest of an old dune ridge generally BEAVER I. c; N. MANITou I. . MANITou I. I \\ G TRAVERSE CITY FIGURE 1. Island archipelago of northern Lake Michigan, including South Manitou Island, Michigan. Z ,268 / 262 Marked cedar Soil pits 250 m2 sample plots 0 ‘.-----O. O ’ Footpath - -- ’0‘-‘ ' 0 100 200 meters FIGURE 3. Map of study area indicating locations of soil pits, sample plots, walking path and large cedar marked by the Michigan Botanical Club. Size of sample plots are drawn to scale and all contour lines are in meters. The second-growth forest is indicated by stippling. 8 outlined by the 244 m contour line. The western boundary is defined by a steep dune ridge corresponding with the 274 m contour line. Aside from several stumps along the southeastern boundary, the study area appears to be free from human-induced disturbance and is representative of this upland cedar forest. Research Hypotheses and Objectives Existing descriptions of the old-growth cedar forest on South Manitou Island are somewhat contradictory; Hatt (1948) described this stand as cedar—hardwood dominated while Thompson (1963) considered it to be an essentially pure cedar stand. Because of this, a major objective of this study was to provide a detailed description of stand com- position and structure. Based upon the observed association of cedar with the lee slopes of dunes, it is hypothesized that this forest community originated with the encroachment of dune ridggg into a hardwood forest, followed by establishment of cedar on the dune slopes. As a consequence, it is expected that old-growth cedar are associated with the old dune ridges and that early growth rates of these trees should be similar to those of cedar currently establishing on. open dune slopes. Thus, another objective of this research was to examine the occurrence of cedar in relation to dune topog- raphy and compare rates of growth for old—growth and open- dune cedar. 9 Although large cedar presently dominate the South Manitou forest community, there is reason to question whether cedar will maintain dominance or eventually’ be replaced by associated hardwoods. The limited infbrmation on cedar growth and regeneration suggests that with present conditions, mortality may exceed recruitment. Further, establishment and survival of small seeded species such as cedar may be inhibited by a continuous cover of leaf litter (Grime 1977), and the fact that cedar is rarely associated with hardwoods (Curtis 1946) may be a function of the lack of adequate seedbeds and the effects of hardwood leaf litter. It is, therefore! hypothesized that cedar repro- duction within the present old-growth stand is insufficient to maintain a stable age structure. Consequently, an eval- uation of the reproductive success and mortality patterns for cedar within the old-growth stand are important objec- tives of this study. Logging of a portion of the old-growth stand provided a unique opportunity to examine regeneration of cedar and other tree Species following this disturbance. Further, it provided valuable insight into how this regeneration has influenced long-term trends in stand composition and struc- ture, relative to the old-growth forest. A. number of factors influence the path of forest recovery following logging; these include successional status of the pre-cut stand, amount of disturbance to the litter and topsoil, availability of suitable seedbeds, species removed in 10 logging, species left intact, as well as others (Bormann and Likens 1979; Boring, Monk and Swank 1981). The factors or conditions which. have affected. the course of’ forest recovery following logging and how these have influenced the structure and composition of the second-growth stand rel- ative to the old-growth cedar stand are unknown. However, present composition of the old-growth stand suggests that it is a late successional forest, and at the time of logging likely had advanced regeneration of tolerant species such as sugar maple. In some late seral stands, removal of all or part of the canopy resulted in release of advance regenera- tion of tolerant or climax species, and thus accelerated the successional process in these stands (Spurr and Barnes 1973, Kapustka and Koch 1979). Removal of the overstory in the old-growth stand, without significant damage to the under- story or topsoil, likely did not create seedbeds favorable for cedar reproduction. Therefore, given advance regenera- tion of tolerant tree species like sugar ‘maple, it is hypothesized that such species will dominate the second- growth stand. An additional objective, then, was to gather information on compositional changes which have occurred in the second-growth stand as the result of logging. There are few data available on the biomass and produc- tivity of old-growth forest stands in the lake states (Crow 1978) and, therefore, a major objective was to document values of biomass and productivity in the mature cedar stand. Furthermore, the extent to which this old-growth 11 stand conformed to values presented for similar forest types (Whittaker and Likens 1975) was of interest. Closely linked to changes in structure and compostion following logging are trends in energy relations. The recovery of biomass and productivity in the second-growth stand, in comparison with the old-growth stand, was examined in light of recent theories concerning' patterns of’ forest. regeneration. fol- lowing logging (Bormann and Likens 1979, Peet and Christen- sen 1980, Feet 1981). Thus, a final major objective was to determine current levels of biomass and productivity in the second-growth stand, and contrast this with the uncut stand. METHODS Forest Compostion and Structure From an initial survey of' the southwest. corner‘ of South Manitou Island, a study area representative of the mature forest was selected and mapped (Figure 3). The site included a walking path and two exceptionally large white cedar which had been marked with aluminum plaques by the Michigan Botanical Club. Descriptive information on the forest was collected from twelve 250 m2 circular plots (Figure 3). A grid system was established for the site and plot selection was made with the use of a random numbers table (Steel and Torrie 1960). Within each plot the following data were collected on each tree 3 2.5 cm diameter at breast height (dbh): 1) tree species; 2) tree dbh; 3) live or dead; 4) height to top of crown; 5) crown width in two perpendicular direc- tions. Importance values for all tree species were calculated based upon relative dominance, relative density and relative frequency (Mueller-Dombois and Ellenberg 1974). Addition- ally, at least two stems per plot of each of the four dominant canopy trees were cored (north and south side of tree) with an increment borer. All stems were chosen so 12 13 that cores were obtained over the observed size range for each species. Cores were placed in straws and returned to the lab where they were air dried and mounted on grooved wooden blocks. Ring widths for each core were then enumerated using a dendrochronometer. Sampling adequacy was determined by graphing cumulative means of basal area and species richness. Finally, two 7 m wide belt transects (Figure 4) were extended in an east-west direction across the study area in order to construct profile diagrams of the forest. Information gathered within each transect included: 1) tree species and dbh; 2) height to top of crown; 3) length of crown along the transect; 4) crown depth (top to bottom of crown); and 5) tree position along the transect. Forest Soils Descriptions of soil within the study area was based on three randomly located soil pits (Figure 3) from which samples of the A1, B, and C horizons were collected. All samples were placed in polyethylene soil bags, sealed and returned to the lab for analysis. Percent organic matter for the A1 horizon was determined from ash-free dry weight measurements, and particle-size distribution for the profile was analyzed using the hydrometer method (Day 1965). Soil pH values were determined using a glass electrode pH meter 14 32 274 244 - 5 m wide belt transect — - 7 m wide belt transect 100 200 meters MICHIGAN FIGURE 4. Location of profile dia ‘ gram transects bands) and transects assessing cedar density relati(vse01t5 dune topography (stippled bands). Study area delimited by dashed line. Contour lines in meters. 15 with individual 10 9 samples of soil in a 50 ml solution of 0.01 M CaC12 . Stand Origin To examine the occurrence of old-growth cedar relative to dune tepography, three 5 m wide belt transects were established on a dune crest defined by the 274 m contour line. These transects were extended west to east from the crest into old-growth forest (Figure 4). Each transect was marked at 10 m intervals along its length creating a series of 5 m X 10 m plots. Within each plot the number of cedar stems 3 5.0 cm dbh were tallied. In addition, slope angle was measured within each plot with a clinometer to obtain a ground contour map along the length of each transect. Further information on stand origin was obtained from a number of canopy (n = 11) and sub-canopy (n = 10) cedar from within the study area as well as individuals (n = 5) which had established on open dune slopes adjacent to the study area. All trees were measured for dbh and increment cores were extracted at breast height, placed in straws and re- turned to the lab. Cores were mounted on grooved wooden blocks and growth over the first 50 years was measured with a dendrochronometer and recorded. Because of the size of old-growth trees, it ‘was difficult to consistently ob- tain cores which had passed through the center of these trees, however, a comparison of core length to tree diameter 16 indicated that all cores were within 8 cm of the center. The amount of wood missing from incomplete cores was determined from overall length of the core relative to stem radius. From this, the number of missing years was estimated by dividing the length of missing core by the mean annual increment from the five innermost rings. Lastly, average annual increments over the first 50 years of growth were estimated for old-growth, closed-forest and open dune cedar and were compared statistically using a F-test and Tukey's honestly significant difference (hsd) test (Sokal and Rolhf 1969). Forest Succession To examine possible successional trends within the mature stand, tree reproduction was examined within four canopy gaps (Figure 5). These gaps, each created from the fall of one to six trees, were of different sizes and gener- ally elliptical in shape. As such, gaps were measured along major and minor axes to determine the area of each canopy gap. Gap boundaries were defined by the edges of surround- ing canopy tree crowns. Within gaps, compositional features such as number of down trees were noted and all trees 3 2.5 cm dbh were identified to species and measured for dbh. From this information, importance values were calculated (Mueller-Dombois and Ellenberg 1974) for trees in each of the four gaps. 17 268 / L'h‘"'(.‘-f: - canopy gape 100 200 meters :— FIGURE 5. Location of canopy gaps within the study area. Contour lines in meters. 18 Additionally, replacement of old-growth cedar was examined by selecting standing dead. cedar (n = 10) and measuring dbh as well as distance to the four nearest live trees (3 2.5 cm dbh). Neighboring trees were identified by species and measured for dbh. From these data, calcula- tions similar to those for point-quarter sampling were used to establish importance values for trees nearest to standing dead cedar (Brower and Zar 1977). To determine if establishment of cedar seedlings was associated with a particular seedbed or substrate type, all seedlings (g 1.1n in height) in five, randomly located, 10 m X 10 m quadrats were identified and noted as to the type of substrate (log, stump, litter, other) upon which they had established. A chi-square goodness of fit test was used to examine whether seedling' number' was randomly associated with substrate type. From this information along with stand data (n1 trees (ferest structure and composition section), the number of cedar stems/ha which had established on logs or stumps was calculated for six size classes: 1) seedlings 5 1 cm tall (first year seedlings); 2) seedlings 3 1 cm tall and < 1 m tall; 3) stems 2.5 to 15 cm dbh; 4) 15.1 to 40 cm dbh; 5) 40.1 to 70 cm dbh; and 6) > 70 cm dbh. It was assumed that larger stems which had established on logs would appear to have aerial or stilt roots. Such roots result when seeds germinate on a substrate which eventually erodes away, exposing the roots (Kramer and Kozlowski 1979). 19 Preliminary observations on cedar seedlings within the study area suggested that recruitment was confined to emer- gent substrates such as logs and stumps. To evaluate the success of seedlings on such substrates three randomly located 10 m X 50 m transects were established in which the number of tree seedlings on logs and stumps was tallied. Seedling heights and surface area for all logs and stumps were also calculated. Cross-sectional stump area was measured as a circle and log area was approximated as a trapezoid: A = g (a+b)h (1) where: A = area a = diameter at one end of log b = diameter at Opposite end of log h = overall length of log State of decay for each log and stump sampled was also evaluated according to criteria presented in Appendix 1. As such, these different stages of decay represented a spectrum of substrate quality upon. which seedlings may establish. Success of seedling establishment with regard to substrate was assessed in terms of the number of seedlings per unit area of substrate type. A cumulative mean of seedling number per plot was calculated to ensure sampling adequacy. The population age structure of cedar seedlings was analyzed by harvesting seedlings (n = 65) throughout a height range (1 - 100 cm) from logs outside of the study 20 area. Harvested seedlings were placed in polyethylene bags, returned to lab, and. measured for total height, foliage height (bottom to top of foliage), foliage width (in two directions) and stem width at ground level. All seedlings were then sectioned at ground level with a razor blade and with a dissecting scope each stem was aged. The above data were entered into a multiple regression program (Stat 4 Computer Program, Michigan State University) to determine which variable or variables were the best pre- dictors of seedling age. From this, a population age structure was established. Survivorship was examined using the negative exponential and power function models as described by Hett (1971) and Hibbs (1979). Seedlings were grouped into two-year age classes and mid-points of each class were used in. the regression. analysis. From the survivorhsip data a life-table was also prepared. Stand Mortality Information on tree mortality was gathered from the sample quadrats on all fallen and standing dead stems. Where possible, dead stems were identified to species and measured for dbh. The direction of fall was recorded for all down stems and a chi-square goodness of fit test performed to determine if fallen trees were oriented randomly. Finally, the numbers of fallen and standing dead trees were tallied by Species. 21 Biomass and Productivity Biomass Because of the unique status of this old-growth stand, non-destructive techniques were employed to estimate the biomass components of the tree strata. Bole wood biomass was estimated using the equation for parabolic volume (Whittaker and Woodwell 1968, Newbould 1967). Wood volumes were multiplied by oven-dried wood densities for each species (Forbs 1955), summed over all sample trees and expressed on a metric tonne per hectare (mt/ha) basis by species. In addition to the parabolic estimate of hole wood biomass alternate estimates of cedar bole wood biomass were made from direct measurements of fallen cedar trees. Diameter measurements were taken at 2 m intervals along the bales of 23 down trees throughout the observed size range for cedar. From this information a regression of height on dbh was established (Figure 6), along with a regression of tree height on diameter ratios (or the diameter at height 2 divided by dbh) (Figure 7). The second distribution was best described by two separate curves: one for trees < 15.3 cm dbh and another for trees 3 15.3 cm dbh. These separate curves reflect a more rapid bole taper for the smaller trees. Next, a computer program was written to calculate individual tree heights and diameters at 2 m intervals from 22 D m ‘0‘ N E 8‘ p. 3: CD hd‘D‘ m C" I: -0.026X E 0 Y = 25.22-25.299 D 8‘ 0 2 :3 o R = 0.78 R = 0.887 m- c, I I I I 0 25 50 75 100 125 DBH (cm) FIGURE 6. Regression of height on diameter for cedar. Y (height in meters) = 25.22 - 25.29 e ‘0'025x (dbh 1“ cm). (n = 23, r = 0.887). 23 1.0 l 0 6 DIRHETER/DBH 0E4 0-2 1 0.0. I I I 0 5 10 15 20 25 CEDHR.HEIGHT (m) FIGURE 7. Regressions of height on diameter ratios for white cedar. Regressions for trees < 15.3 cm, dbh (A); X (height in meters) = -0.0655 Y (diameter to dbh, ratios) + 1.09 (R = 0.9344, n == 4). For trees 3 15.3 cm, dbh (G); X (height in meters) = -0.04414 Y (diameter to dbh, ratios) + 1.07 (r = 0.9058, n = 7). 24 the regression curves for a given dbh. Once diameters at each interval were calculated the program estimated volumes for each segment using the equation for the frustrum of a right circular cone, and the top-most segment was estimated as a right circular cone (Shelby 1967). Volumes of all stem segments were then summed to produce a value for total bole wood volume. Diameters (minus bark) of all standing cedar sampled were entered into the program and a second estimate of bole wood volume was obtained. Bole wood volumes were then converted to total bole wood biomass using oven-dried wood density and expressed on a mt/ha basis. Bole bark biomass was calculated for the three dominant species (cedar, maple and ash) by substracting bole wood volumes from total bole (wood and bark) volumes. Measure- ments of bark thickness from increment cores allowed for estimates of total bole volume (wood and bark) and bole wood alone fer each tree. Once bark volume was estimated, dry weight biomass was determined by multiplying volumes by bark specific gravity (dry) for' each. species. Only' specific gravity for sugar maple bark was available from the litera- ture (Whittaker et al. 1974). Values for cedar and ash were estimated from 10 samples each, using the immersion tech- nique (Forbs 1955). Bark biomass values were tallied and expressed as mt/ha for cedar, maple and ash. Estimates of bole bark biomass for minor species (importance values < 40) were not made. Branch biomass was not sampled directly but was estimated as a proportion of total bole biomass from the 25 literature. Cedar was estimated from Reiners (1972) while the remaining species were estimated from the data of Whit- taker et a1. (1974). Leaf biomass values for all deciduous species were estimated as litterfall collected over two growing seasons from fourteen 1 m X 1 m quadrats. All collected leaves were returned to the laboratory, separated by species and air dried. Foliage biomass was estimated as leaf dry weight in mt/ha. Cedar foliage biomass was estimated as a proportion of hole biomass (wood and bark) from existing data on cedar (Reiners 1972). Productivity As with biomass, net annual primary productivity (NPP) was estimated nondestructively. Because of the usually infrequent and often dispersed occurrence of tree falls, stem productivity, based on litter fall, is likely to be underestimated in short term studies which assume steady state conditions (Newbould 1967, Reiners 1972). For this reason a direct evaluation of stem wood production for the most important canopy tree species (cedar, ash, maple and basswood) was made. This involved calculation of the esti- mated volume increment (EVI) described by Whittaker and Woodwell (1968). Mean annual radial wood increments were obtained for these species by coring sample trees (n = 36) over their observed size ranges and averaging wood growth rates over the previous ten years (Whittaker and Marks 26 1975). Average annual increments were calculated by species and applied on a size class basis in each sample plot. Estimated volumes were then summed across plots, and with specific gravity values (dry’ weight) for each species, annual stem wood production was calculated for the stand. Since minor species (I.V. < 40) were not systematically cored, the ratio of bole wood productivity to bole wood biomass was determined for basswood (Tilia americana) and used to estimate bole wood production for all trees not cored ("other" category). Total bole wood. biomass for "others" was multiplied by this ratio to obtain bole wood productivity for this category. As with biomass, bark productivity was not estimated for minor tree species. For the dominant species, bole bark primary productivity was estimated as the ratio of bole wood productivity to bole wood biomass, multiplied by the biomass of bole bark (cf. Reiners 1972). Production of branches (3 2.5 cm basal diameter) was estimated over two growing seasons from litterfall collected in a 2 m X 350 m transect located within the study area (Figure 8). This assumes that branch production and branch fall are in steady state on an annual basis. All branches were collected, separated by species and weighed in the field. Subsamples obtained from each species were placed in polyethylene bags and returned to the lab. These branches were weighed within 24 hours and then oven dried to a con- stant weight to determine percent moisture. This value was 27 1:1:- - 2 m x 375 in branch transect withneatedmeZmand 1 m x l m quradats. 100 200 meters FIGURE 8. Location of litter plots (branches and leaves) within the study area. Contour lines in meters. 28 used to correct field weights to dry weight values. This method was also used for twigs (< 2.5 cm basal diameter) which were collected from fourteen 2 m X 2 m quadrats located at 25 m intervals along the branch transect. Net annual leaf production for deciduous trees was estimated from litter collections following leaf fall. The turnover of leaf biomass for cedar was not directly determined in this stand. For cedar, alternatively, leaf productivity was estimated as a percentage of total leaf biomass from a ratio presented by Reiners (1974) for cedar. All fruits and seeds for cedar, maple and ash from the litter collections were assumed to represent annual produc- tion. These fractions were separated by species, air dried and weighed. Fruits and seeds of the other tree species were included with leaf litter. Forest Recovery Following Logging Stand Description A logged-over area of 3.4 ha adjacent to the mature stand was selected to represent second-growth forest (Figure 9). The stand was characterized by a sugar maple canopy with many out cedar stumps in evidence throughout. To describe this stand, ten 10 m X 10 m quadrats were randomly located within the selected area and the following informa- tion collected for each tree 3 2.5 cm dbh: 1) tree species; 2) live or dead; 3) dbh; 4) tree height. Two increment 29 FIGURE 9. Location of old-growth and second-growth study areas on South Manitou Island, Michigan. Old-growth stand indicated by tree symbols and second-growth stand by heavy stippling. 30 120 l 100 Y = 0.886 X - 0.659 R = 0.99 80 60 40 DIAMETER AT BREAST HEIGHT (CM) v v V V V V 0 10 20 30 90 50 60 70 80 90 100 STUMP DIAMETER. AT A HEIGHT 0F 30 CM v v FIGURE 10. Regression of diameter at breast height on dia- meter at stump height (diameter at 30 cm above ground sur- face), for old-growth cedar. Y (diameter at breast height) - 0.886 X diameter at stump height - 0.659. (n= 29, r = 0.99). 31 cores were also extracted from several live trees per plot over the observed size ranges for each species. Additional- ly, all identifiable cedar stumps were measured for diameter at 30 cm above the surface of the ground. This was defined as diameter at stump height (dsh). An estimate of dbh for these stumps was obtained from a regression of dsh on dbh using measurements from 29 live cedar in the mature stand (Figure 10). From this, cumulative means of basal area for logged cedar along with second-growth maple were used to assess sampling adequacy. Based upon the above information importance values and total basal area for the present second-growth stand were calculated as in the mature stand. Estimated Date of Logging An attempt was made to pinpoint as accurately as pos- sible the date at which many large cedar were logged from the area adjacent to the mature stand. To do this the following approaches were taken: 1) old-growth cedar were examined for growth release; 2) trees which had established on cut cedar stumps were aged by ring count; 3) an age profile of second-growth maple was established. Four large cedar from the mature forest adjacent to the logged area were selected and cored with an increment borer. Two cores per tree were extracted and placed in straws. All cores were returned to the lab and measured for annual rates of growth. Since cedar is reported to respond well to release (united States Forest Service 1965, Johnston 1977) 32 it was expected that a disturbance such as logging would result in distinct, identifiable changes in growth of these trees at the time of‘ logging. To discriminate Ibetween changes in growth in response to logging as opposed to changes resulting from the effects of climate or small-scale disturbances (e.g. tree falls), criteria from Lorimer (1980) were applied to the data. The release criteria were: 1) the period of release had to occur abruptly and be sustained for at least 15 years; 2) release had to be preceded by at least 15 years of comparatively slow growth (< 1.0 mm/yr as mea- sured in the laboratory); 3) average growth rates for the 15 years following release had to be at least 100% greater than previous average growth rates; and 4) for trees with moder- ate rates of growth prior to release (i.e. 1.0 mm/yr) a 50% increase in average growth following release was required. By way of comparison, average rates of growth 15 years before and after the estimated cutting date were examined in four cedar from within the mature stand. Trees of various species which had established on cut cedar stumps were also cored as described above and aged by ring count. Further, age and diameter information gathered on second-growth maple were used to construct a linear regression of age on dbh (Figure 11) and from this an age profile for second-growth sugar maple was established. 33 701 e 60: 50: TB 0: 3; kg 404 E 5 Y = 1.2 X + 26.8 E w R = 0.85 E m 20« . 2 0 l0: 1'0 20 30 40 50 6'0 DIAMETER AT BREAST HEIGHT (04) FIGURE 11. Regression of age on diameter for second- growth sugar maple. Y (age in years) = 1.2 X (dbh cm) + 26.8 (n = 27, r = 0.85). 34 Forest Recovery To (contrast. structural and. functional changes ‘which have occurred in the second-growth and mature stands fol- lowing logging, it was important to first document how similar the original (pre-logging) stands were. This was accomplished by comparing pre-logging estimates of density (stems/ha) and basal area (mz/ha) for cedar in both stands. This information was reconstructed from stump data in the second-growth stand and compared with values from the mature stand, where mortality and growth were taken into account. Based upon the sound condition of cut cedar stumps in the second-growth stand, all sound, identifiable cedar stumps and fallen logs, as well as standing dead cedar, were assumed to be post-logging. All such mortality was included in estimates of density and basal area for cedar in both stands. Growth of cedar during this post-logging period was estimated from cored trees. Average annual increment over the last ten years was substracted from all live trees, on a per plot basis over the years following logging. Finally, a standard t-test was performed to test for differences in basal area and stem density for cedar in the second-growth and mature stands at the time of logging. To! examine forest. recovery, compositional shifts in basal area of the second-growth stand relative to the mature stand were evaluated. This parameter was chosen as it is a reflection of the dominance and physiological status of forest trees (Oliver and Stephens 1977). Basal area change 35 in ten-year intervals were calculated from stand data for cedar and maple for a period of 80 years. In each stand, maple and cedar were divided into 10 cm size classes and average growth increments over the last decade were esti- mated for both species, as above. The growth increments were subtracted from each tree in the appropriate size class on a per plot basis. Basal area was then calculated and summed across plots, providing an estimate of basal area (uncorrected for mortality) in each stand ten years earlier. This process was repeated for each successive decade from 1980 back to 1900. As discussed previously, a portion of cedar mortality was assumed to have occurred following logging. It was further assumed that mortality was constant over time and thus total basal area loss to mortality was factored out equally for each decade. Changes in biomass and productivity which resulted from logging were evaluated by contrasting bole biomass and net primary productivity in the second-growth stand with values from the mature stand. In the second-growth stand bole biomass for all trees was determined per plot by estimating total volume which was approximated from the equation for a paraboloid of rotation (Newbould 1967). Volumes were con- verted to biomass values by multiplying each tree by the appropriate specific gravity (dry weight) value. Biomass was summed over all plots and expressed as mt/ha. Similar- ly, net annual bole production for all trees was approxi- mated by calculating the estimated volume increment (EVI) 36 for the last ten year period (Whittaker and Marks 1975). These values were converted to dry weight values and ex- pressed as in the mature stand (mt/ha/yr). RESULTS Composition and Structure of the Old-Growth Stand Of the nine tree species encountered within the study area cedar was the stand dominant, accounting for 38% of the total importance value. Sugar maple and white ash were important co-dominant species, and together with cedar comprised 77% of total stand importance (Table 1). The importance of cedar in the stand derives primarily from the large basal area (dominance) of some individual cedar stems. One cedar tagged by the Michigan Botanical Club measured 162 cm dbh, and is listed as a national champion tree. There were a total of 884 stems (3 2.5 cm dbh)/ha in the stand, 61% of which were between 2.5 and 10 cm dbh (Figure 12). This reflects the large number of maple and ash iJI'this size range (Figure 13). By contrast, cedar is evenly distributed across a broad size range (Figure 13) and accounts for all stems 3 50 cm dbh. Despite the large diameters of some cedar within the stand, no trees measured were over 30 m in height and only 5% of all stems were between 20 and 25 m (Figure 14). For cedar specifically, 15% of all stems attain heights exceeding 20 meters. 37 38 .Aveefl muonsoasm use asonsoouuuaaosxe Amufimsoo o>fiumaouv + .oucmnwaoo o>fiumaouv + Amososvoum o>fiumaouv on~m> ooseuuonaa \s e.eeH e.ee~ masses m.H ¢.m H.H m.e a soasmaan «owns o.~ H.e a.H «.6 g asaamwmAa> assume H.~ e.o ~.¢ ~.e m aaaouasuouaa assuou m.~ m.e m.~ ~.H s successes «saws m.~ m.m m.e «.6 a. asassmuus> mousse e.HH H.mm ~.6H a.o es asuaosmu sous e.ma o.~¢ s.¢H m.m as assessosa assaxsus m.m~ s.es e.em s.mH aw asussoosm sous e.mm a.mHH g.sH m.~e a~ assassuesooo messy muz seamzma muzaszon scamsomms mmsummm arses so azuommm muzcamomzs m>Ha¢qmm m>HaHa¢qmm \ .umouom nusououoao on» How mosam> oosmuuooan .H mqmflumaosso mo confiumoaou .mH mmDUHm Amm>v mo< Amm>v mo< cm on em cm 2 o cm as om cm 3 o - . . . . x. «J «J . N m . N m .1. . a m .x . _N m x. .I . .I .s. M. a. . m A ~~\ s 0 DJ s an“. .x a a m . a m m. . 2 m. x . 3 ) X ) .. NH m . . NH w ESQ 128555 «5. mafia 52255.5 4---. i maewu Emmeom-smmonu oil. . SH e_e< :zsu Am muse“ rezoao m>.p5 0 l__ FA TO PV AS CA PA AS TO 250 - m gzoo 4.: L 150 GAP II GAP IV Ea Gap Area = 176 m2 Gap Area = 157 m2 £100 .J. 0i FA AS TO AS TO 1401 B) Combined Importance Values for all Gaps 100' g g; 50‘ 4.; 8 E 0. AS FA TO PV CA FIGURE 17. A) species importance values for trees within four canopy gaps, B) combined importance values from four canopy gaps. Tree species abbreviations: Acer saccharum (AS), Fraxinus americana (FA), Thuja occidentalis (TO), Prunus virginiana (PV), and Cornus alternafolia (CA). Importance values = (relative density) + (relative domi- nance) + (relative frequency). 50 150 ‘ 131.1 IMPORTANCE VALUE FIGURE 18. Importance values for four trees nearest standing dead cedar stems. 51 TABLE 5. Summary data for standing dead cedar stems and the four nearest live trees. A) MEAN DIAMETER OF DEAD STANDING TREES (CEDAR) n Mean dbh (cm), 1 SD 10 85.1 i 12.9 B) MEAN DIAMETERS OF ADJACENT LIVING TREES Species n Mean dbh (cm)y 1 SD Acer saccharum (AS) 27 13.7 i 8.9 Fraxinus americana (FA) 6 34.2 i 19.6 Thuja occidentalis (TO) 5 46.2 i 40.6 C) MEAN DISTANCE FROM LIVING TREES TO DEAD STANDING CEDAR Species n Mean distance (m), 1 SD AS 27 2.2 i 1.1 FA 6 2.9 i 1.2 T0 5 3.1 i 1.6 52 with increasing size there was a marked change in the occurrence of cedar, from log to soil substrates (Figure 19). While 99.7% (2382 stems/ha) of first season seedlings are found on logs, only 8.3% (or 3.3 stems/ha) of cedar be- tween 15 - 40 cm dbh were associated with logs. Although it may be difficult to accurately assess the original substrate of a large tree, no cedar > 40 cm dbh showed evidence of having established on a log substrate. Logs occupy an area of 680 m2 per hectare of forest floor and represent various stages of decay. Based on an index of decay (Appendix 1) 70.4% of all logs were found to be well or very well decom- posed. Additionally, the greatest seedling densities were found on logs in these more advanced states of decay (Table 6). Seedling height was the best predictor of seedling age for cedar (r = 0.84) and was used to derive a population age structure for seedlings g l m in height (Figure 20). Cedar seedling mortality was estimated to be 65% during the first growing season (Table 7). From years 4 through 12, mortal- ity averaged 24% of the previous age-class and 99.4% of the initial cohort was lost by year 13. Regression analysis of the seedling age structure (years 4 - 12) indicated that the power function model represented age distribution more accurately (r = -0.96) than did the negative exponential model (r = -0.92)(Figure 21). 53 .moumnumosm aflom one moa ou o>wumaon .mommmHo oNfim an mfioum Hmooo mo ousouusooo .mH mmDUHm .u: :w so H m N mmsflaooom .u: 5 E H «a 50 H A O mosfiaooom mm: are .5 3 I m.~ om m are so as I 3 m d om fig 80 2. I S o mm .56 .8 K o om coflml 91.3 8.3 6mm coo ooe oom cm as as cm 2.: cm; 6.3 mououumnsm mod oueuumnsm HHOM e co oumuoo: Rom mason mo Honfisz so vacuum: mom mfioum mo honEsz 54 TABLE 6. Seedling density (seedlings/m2) for logs in four states of decay. DECAY STATES 1/ SEEDLINGS/m2 OF LOG AREA (i SD), 1. Relatively Undecomposed 0.6 i 0.5 2. Moderately Decomposed 0.5 i 0.4 3. well Decomposed 1.7 i 0.3 4. Very well Decomposed 1.7 i 0.9 1/ Descriptive criteria for decay states given in Appendix 1. 55 .mamzow>flm:fl oooa mo unocoo Room women m on ooumsflom nudges: mafiaooom .Ausmfio: ca 8 H vV monfiaooom Hmooo Mom unoposhum mom sowumasoom .om mmDUHm Amm>v mmmm<4u ue< SNNN om ma 3 3 NH 3m 9 Fm . F P i . 5 y I ’ i Y H ’ o I 1 1 . com . co: coo . com 83 uaswnN SNIiaaag 56 TABLE 7. Life-table for white cedar seedlings in the old-growth stand. 1/ Proportion lost before next Age Numbers per (years) age classz/ Proportion of initial seedling population age class 0 - l 1000 1.00 0.65 1.1 - 2 348 0.38 0.57 2.1 - 3 150 0.15 0.00 3.1 - 4 282 0.28 0.51 4.1 - 5 138 0.14 0.26 5.1 - 6 102 0.10 0.76 6.1 - 7 24 0.02 0.00 7.1 - 8 30 0.03 0.60 8.1 - 9 12 0.01 0.00 9.1 - 10 12 0.01 0.00 10.1 - ll 12 0.01 0.00 11.1 - 12 12 0.01 0.00 12.1 - 13 0 0.00 1/ Life table extended to first unrepresented (no seedlings) age-class 2/ Numbers adjusted to an initial cohort of 1000. 57 <' if. 8 , A) POWER FUNCTION CURVE S a 3;; LL 6 * A LOGe Y=9.92-3.17L0G X o A e 35 g u 1 (R = -00958) D Z ’i 2 ‘ L9 3 0 . - I a l 2 3 ll LOGe X, AGE IN YEARS S A \ g 6 . ‘ ‘ B) NEGATIVE EXPONENTIAL CURVE l— ‘f ‘ L0Ge Y = 6.48 - 0.415 x o 4 g u (R = -0092”) g 2 J ‘ >: £3. 0 ‘ F . 2 LI 6 8 10 12 AGE IN YEARS Figure 21. Regression analysis of cedar seedling survivor- ship. A) Power Function: 9.92 - 3.17 Log B) Negative Eiponenti a1 : X (age in ygars) Log Y (number of stems/ha) = (n = 12, r = -0.958). Log Y(number of stems/ha) .-. 6.84 - 0.415 x (age in numbers) ‘in = 12, r = -0.924). 58 Stand Mortality Total stand tree mortality (Table 8) was comprised primarily of cedar (68%). The senescent status of cedar was further suggested by the fact that 31% of all cedar stems sampled were dead, compared with 3% and 7% for maple and ash, respectively. Cedar mortality consisted of six stand- ing dead stems (i dbh = 49.6 cm) and 15 fallen stems (i dbh = 46.9 cm). In contrast, sugar maple mortality was repre- sented by two standing dead saplings (i dbh = 6.9 cm) while mortality for white ash resulted from the death of two standing stems (3': dbh = 39.6 cm) and a single, large in- dividual which appeared to have been wind thrown (i dbh = 57.0 cm). Of particular note in this stand is balsam fir, a common associate of cedar on Open dunes (Olson 1958). Mortality data indicate that 50% of all balsam fir stems in this stand were dead. An analysis of all down trees indi- cated that direction of fall was not statistically different from random (Figure 22). Biomass and Productivity Biomass Distribution of aboveground biomass by component and species (Table 9) indicates the dominance of cedar in this stand. Cedar accounted for 61% of total bole biomass (wood and bark) in contrast to the codominants, maple and ash, £59 6.66 IIIII 666 an 66~ .s. 644966 6.6 6.6 «A 6.6 «A mmuwmmumww assess 6.6 6.6 «6 6.6 66 mmwaoamm Hone 6.6 . 6.6 «A 6.6 NA asaouasuouao assuou 6.6 6.6 6 6.6 6 mmmmmmummw mummmm 6H 66 6 n n averages avast 6 6H s A 6 asaoauoea cause ca 5 Ne . n on endowuele enswxeuh 6H m 66H m mes esuusooaa sou: 66 an 56 AN 66 asaausoeauuo «fimmw annotaao: macaw mmuummm e «on A6a06\o>A~. ¢:\m:uam wruam name mxuam m>na muuommm areas so szmuxmm name azuuamm so «more: areas we «were: ho mflmlbz I .euoHo cannon «I emu o>nosu canvas Anna eu em.N Av aoouu Ade new venue ouauel on» loan eueu mafiaeuuol uo hues-am .o munB~d¢fi¢02 Oldhm adfioa ho Hzmummm mmnummm < mom Gian quommm Aoaoexo>sa. ¢m\mzmam mzuam acme manna m>na G6 «66:62 aHozu swan“: Anna lo em.~ A. soon» «He now ocean ensue! ecu noun ever SufiueHROI no Swallow .m Manta 60 FIGURE 22. Direction of tree falls within the study area. Chi-square goodness of fit test for direction of fall was not significant (x2 = 3.75; a > 0.05), indicating tree fall was random . 61 TABLE 9. Estimated biomass values (mt/ha). BRANCHES TOTAL BOLE BOLE (wood + ABOVEGROUND SPECIES WOOD BARK bark) FOLIAGE BIOMASS Thuja occ1dentalis ‘66.8 3.6 15.0? 5.5? 90.9 Acer saccharum 25.9 1.7 6.2* 1.2* 35.1 Fraxinus americana 13.2 1.3 2.7* 0.3* 17.6 Other species 2.7 --- 1.6* 0.6*° 4.8 TOTAL 108.6 6.8 25.5 7.6 148.5 T Estimated from Reiners (1972) * Estimated from annual litterfall ° Includes fruits which were not separated by species 62 which comprise 24% and 13% of bole biomass, respectively. Total bole biomass alone represented 78% of the aboveground total for the stand (148.5 mt/ha). This pattern of cedar dominance is also reflected in leaf and branch biomass values where cedar represents 73% and 59% of stand totals for these components, respectively. A second, direct estimate of bole biomass for cedar, based on measurement of fallen stems, resulted in a value only 2% greater than the 69.8 mt/ha estimated from parabolic approximations of stem form. Both methods appeared to pro- vide acceptable estimates of bole volume and biomass: how- ever, for reasons discussed previously, all values reported here were based on parabolic estimates. Productivity Net annual production of bole wood and bark, by species, is presented in Table 10. Bole wood accounted for 94% of total bole production which equalled 2.1 mt/ha/yr. In contrast to the cedar-dominated pattern for biomass, maple and ash represented the greatest percentage of bole productivity (40% and 35%, respectively) while cedar ac- counted for 20% of the total. A summary of mean annual litter production (Table 11) indicates that leaves comprise the majority (66%) of litter production while branches were 33% of the total. As with bole wood and bark, maple was most productive, representing 58% of branch production and 38% of leaf production. 63 TABLE 10. Estimated bole (wood and bark) annual net primary productivity. SPECIES BOLE WOOD BOLE BARK TOTAL BOLE PERCENT OF PRODUCTION PRODUCTION PRODUCTION BOLE (mt/ha/yr) (mt/ha/yr) PRODUCTION Thu'a occ1dentalis 0.4 0.02 0.43 20.1% Acer saccharum 0.8 0.05 0.84 39.6% Fraxinus americana 0.7 0.05 0.75 35.4% Other species 0.1 0.01 0.11 5.4% TOTAL 2.0 0.13 2.13 100.5% 64 Hopped mood saw: ooosaosw o=Hm> Aehma .meav muocwom scum mmm30wn ummflaow no ooeusoouoo m we ooumawumm a “meow wmew eeeeeeeeeeeeeeeeeeeee eeeeeeeeeeeeeeeeeeeeeeeeinH-OH- Q o... oooooooooe mafiomgm “capo Ho.o .... ..... monomuoam assfixmuh No.0 0000 000...... gumnoomm H”U< ¢°o° .o.. no... mflHmucmvflouo ”flag“ NNeH geeo eeeeeeeeeeeeeeeeeeeee eeeeeeeeeeeeeeeeeeeeee mgfiah moo eeoe eeeeeeeeee mvflougm Hofio N.O .... ..... OgUMHOEM NSCflKflHh N.” 0... 00.0.0... gmnoomm “wot «N.H .... ..... wwamusoowooo chunk N0.m0 N.” ..................... .......................mQ~/fi0.H QN.O .... .......... QWHWOQW HOSUO oa.o .... ..... mamomuoam mzowxmum no.0 o... ooooooooo aghmnoomm “00¢ ¢~.o .... ..... mwamusocwooo mflsna NaeNm 00H eeeeeeeeeeeeeeeeeeeee eeeeeAgHmn ”a” voozv mmgucmum JCHOH A¢HHm6m< masses mochoE mmmuzwuoscouo mumafiun use deacon ovumawumm .NH Manda 67 CEDAR 1 2+ Vl‘ i| 3 Eli M W11. 3’ sq. . CEDAR 2 CEDAR A 31 2 . 1‘ ‘ f (1 J 132011860. 1900 19140 1980 1820 1860 1900 1990 1980 ‘DATE (YEARS) FIGURE 23. Patterns of average annual growth (in mm) for old-growth cedar adjacent to the logged stand. Dates indicate year of apparent growth release. 68 Aommav osnmuoa we Dunc coneswnmo onwsoaaow .An>\aav open nusono ommno>< % Aommnv unnamed mo once ouueawnmo on nonno .Anh\eav Dunn ausono ommnu>< # aukonm an omcmno unoonoo no woven nusono ummno>m .oumo unmoaon mo ommnu>e on oooaaosw no: use .mflnounno ommoaon nous no: one 0 R6 + mm.n ~¢.H IIII n x MN + «mm.n #Nm.n Q 6 nmoou Nee + «No.6 $66.0 @ m nmmou New + «oc.~ #mo.n Q N nmooo Nmn I «66.n #66.n e n Assoc Hmmmom ammoqo N66n + 66.~ 66.6 666s u m Q II @«m6.H onooon o: pnooon o: 6 neoou @Nwe + Qmm.m @~>.~ @momn m nmoou mean + 66.6 6~.n 666n ~ A6606 N6m~ + ~6.~ 6.6 666A n H6666 mwmm MOON Mmdmdmm mflhhd EH3OGU 2H moz< 6666666 mmosmm .An\ss. mesa 663666 664mm>< Aoocmno nnzono Deanna no obese mmqumm szmzdxm mammh .ocwomoH on uncommon on noooo on ommoaon nuzonu mo unmeasm .mH vanes 69 release. Based upon cedar 1 and 2 the average date of release was 1896 with an average growth increase of 156%. Old-growth cedar from within the mature stand did not exhibit pronounced increases in growth following 1896 (Fig- ure 24) and did not meet the established release criteria. These trees exhibited only a 9% increase in average growth in the period following 1896 (Table 13). The average age for trees which had established on cut cedar stumps was 55.2 yrs, SD :1: 7.5 (Table 14) and the oldest stump-established tree was a cedar estimated to be 64 years of age. The population-age structure of second-growth sugar maple (Figure 25) conformed to a negative exponential model (r = 0.985) in a slightly irregular form. There was a distinct peak in the 60 year age class while the 30 and 40 year age classes had fewer stems than predicted by the model. Two sugar maple individuals per hectare were esti- mated to be between 135 - 144 years of age, indicating that maple existed in the stand prior to logging in 1896 (Table 15). Forest Regeneration The strong correlation between dsh (diameter at stump height) and dbh for cedar (Figure 10) allowed for a direct estimate of the basal area of trees removed by logging. A relatively narrow size range of trees (Figure 26) was cut, representing a basal area of 48.9 mz/ha and a density of 220 stems/ha. A small number of cedar (10 stems/ha) with an 70 q CEDAR 1 ‘ CEDAR 2 3 2. ‘ 1Y1 E? 1 ‘ ‘5 . K “2’ . - :2 , - . :2 1% 4 ‘ CEDAR 3 ‘ CEDAR A 3: . 1 2‘ y 1 . . . . . l l. K II . 1820 1860 1900 1940 1980 1820 1860 1900 1940 1980 DATE (YEARS) FIGURE 24. Patterns of average annual growth (in mm), for old-growth cedar from within the mature stand. 71 TABLE 14. Summary information on trees which had established (n1 out cedar stumps, within the second-growth stand. DIAMETER AT TYPE OF ESTIMATED SPECIES BREAST HEIGHT REPRODUCTION AGE (in cm) (Years) White Birch 47.5 Seed 54 White Birch 43.0 Seed 49 Basswood 43.0 Seed 62 White Cedar 36.0 Sprout ? 64 White Cedar 36.8 Seed 47 i 41.3 cm (SD 1 4.8) 55.2 (so 1 7.5) years 72 220 < I \\ (I) 2 Lu )— OO LI. 0 (I L” m z D 2: 3o 40 50 60 70 80 90 100 110 120 130 140 AGE IN YEARS (AT BREAST HEIGHT) Figure 25. Age structure of second-growth sugar maple within logged study area. Negative exponential model: Log Y(stems/hectare) = 7.5 - 0.06 X(age in years at breast height) (I = 0.967), fitted to age class distribution. 73 Table 15. Population age structure of second-growth sugar maple1/. AGE CLASS NUMBER OF PERCENT OF PROPORTION LOST (years) TREES (in TOTAL BEFORE NEXT AGE 4500 m2) CLASS 25 - 34 95 41.3% 55.8% 35 - 44 42 18.3% 9.5% 45 - 54 38 16.5% 2.6% 55 — 64 37 16.1% 73.0% 65 - 74 10 4.3% 70.0% 75,— 84 3 1.3% 0.0% 85 - 94 3 1.3% 33.3% 95 2 0.9% 100.0% 1/ n = 230 70: 60: 501 404 20‘ 10I NUMBER OF TREES/HA 74 CEDAR [removed in logging) IF'” ,1, .1, FIGURE 26. ‘30 4050 6070 0'90 15 25 35 45 55 65 75 as 95 SIZE CLASS (CM) size range of cedar removed by logging adjacent to the old-growth stand. Dbh estimated from stumps based on a regression of dbh on diameter at stump height (stump height: defined as being 30 cm above surface of the ground). 75 .mGOHnmooa anon now monam> msnmmoHIono nooamon moumeumo 0056:6836 one Monacon .mcownmooa 530nmnoao one ooomoH on nmooo now 0056:6680 056 mnflmsoo mo msomnnmdeoo .SN mambo; none none absenmIpno Cowman zuaonmInHo commo~ . o 6n 0 6 m 0 S m . u . On N . can A V N \l .6. m . On . end 3 \I A w moon... 66>: I W A /z / H .66 H r . 66~ N. V ( 6.60am c023 I g .. . a...” . 66 E 62 . 66~ magnum nmooo I. I nmm ocuoceua poem I neooo usoIc: I \\ 76 estimated basal area of 0.33 mz/ha were present at the time of logging but were not harvested (Figure 27). In the oldrgrowth stand basal area for live cedar was 23.8 m2/ha with a density of 152 stems/ha. Cedar mortality, which had preseumably occurred since logging, contributed substantially to the decline in basal area and density (Figure 27). Regarding basal area, fallen stems accounted for 11.7 mz/ha while standing dead stems represented 4.5 mz/ha. Density values for logs and standing dead stems were 49.5 and 19.8 stems/ha, respectively. Stumps encountered within the study area had an estimated basal area of 11.8 mz/ha and a density of 29.7 stems/ha. A comparison of estimated basal area, at the time of logging, indicates that total basal area for cedar in the old-growth stand was higher (51.4 mz/ha) than in the second- growth stand (49.2 m2/ha). These differences ‘were also reflected in density values where, in the old-growth stand, there were 253 cedar stems/ha compared with 230 cedar stems/ ha in the second-growth stand. These differences, however, were not significant (Table 16). Relative to the mature (old-growth) stand, the logged stand is completely dominated by second-growth maple (Figure 28) which accounts for 72% of total importance, while cedar represented 12% of the total. Changes in species composi- tion following logging are suggested by shifts in basal area (Figure 29). Total basal area in the second-growth stand increased slowly over the first two decades after cutting, 77 Table 16. Statistical comparison (t-test) between density (A), and dominance (B), for' cedar' in. the second-growth stand, prior to logging. SECOND-GROWTH OLD-GROWTH STAND STAND A) DENSITY (stems/ha): n.s.1/ 2 230.0 253.2 SD 149.4 118.5 i: n 10 12 B) BASAL AREA(m2/ha): n.s.2/ i 49.2 52.9 SD 36.9 35.4 * n 10 12 Values on a per plot basis 1/ t = -o.50, « > 0.50 2/ t = -0.33, a > 0.50 78 A) IMPORTANCE VALUES. MATURE (Old-growth) 3mm A» i—‘ _>; “I an m l l l l I l 1‘"! [—1 r—L H I, 52' 0 e o A b b b A. 0% $9 0% Q @‘r‘ .300 . 0° .300 ‘9» S? 4 9°: q‘,’ o 9‘ o :0” 9° 2° 9’ e 7 I? 236 :0 B) IMPORTANCE VALUES, LOGG- (second-growth) STAND 1"- see .UI _>; 1001 seI ”1 [—1— l'_l r-—I r—'1 q, ‘5' ’Q 5 I to b .57 08) ~20 4°C ‘9? 4'4: goo 0 C’ ”O q," ‘90 O O 0 4 6 *9 Figure 28. Importance values for the mature stand (A), and the logged stand (B). 79 but by 1920 began to increase exponentially because of the rapid increase in basal area for maple. There was a gradual recovery of basal area for cedar until 1960, after which little further increase was observed (Figure 29). In contrast, the pattern of compositional change in the mature stand (Figure 30) was distinctly different. Since 1900 there has been a steady decline in cedar basal area. Overall stand basal area parallels the decline for cedar until 1940 and then increases coincidentally with maple until the present. Current basal area in the second-growth stand (29 mz/ha) is similar to that of the mature stand (34 mz/ha); however, there has been a dramatic reversal in the contribution of cedar and maple to total stand basal area (Figure 31). Whereas Old-growth cedar comprises 71% Of the basal area total in the mature stand, second—growth maple accounts for 77% of the basal area total in the logged stand. Maple now represents 87% of total bole biomass and 94% of annual bole net primary productivity in the second-growth stand (Figure 32). The distribution of bole biomass and productivity is quite different in the old-growth stand (Figure 33), where mature cedar dominates bole biomass (61% of total) whereas maple accounts for 40% of total bole net primary productivity. Comparisons between second-growth and mature stands (Figure 34) indicate that overall bole biomass in the second-growth stand (186.4 mt/ha) is currently 38% greater than that in the old-growth stand (115.3 mt/ha). 80 :etlmated date at logglno l | SECOND-GROWTH STAND 351 I B 30‘ ' STAND TOTAL + «g | (29 MZ/HA) . z 254 l *-—-* COMBINED BASAL AREA 9 | " —' MAPLE / < , . _’ 20( I “" "" CEDAR < (I) 6 I 15: I 104 I I / 51 I z ./ —a A d.— l _./ ’ IL: -—--3‘ "" 4r ’ P 0 3;.— =1: -l- "" 1890 1900 1910 1930 1930 1940 1950 1980 1970 1980 DATE (year) Figure 29. Basal area changes for cedar and maple in the second-growth stand, since 1900. 81 OLD-GROWTH STAND STAND TOTAL + estimated date of legglny (34 M2 /HA) 30: V A 25‘ s b —X _ fl .— flu— ._ x, 2 ~\ E 20 5 * ‘ COMBINED BASAL AREA £5 15+ x — " CEDAR 2' -- - ° MAPLE (D < m 101 5‘ ’ .e .0, u' .’ ’ _ 4' "’ 4" OL—H‘ ’ "- " "" 1890 1900 1910 I920 I990 I940 I950 1900 1970 1980 DATE (year) Figure 30. Basal area changes for cedar and maple in the old-growth stand, since 1900. 82 SECOND-GROWTH STAND OLD-GROWTH STAND OTHER OTHER 13.8% .. 14.12 CEDAR MAPLE 8.9% 14.7% MAPLE CEDAR 77.3% 71.3% 2 TOTAL BASAL AREA = 29 MZ/HA TOTAL BASAL AREA = SA M /HA Figure 31. Shifts in the relative contribution of cedar, maple and other species, to basal area, between the old- growth and second-growth stands. Circle size reflects relative basal area. £33 SECOND-GROWTH STAND STEM BIOMASS ANNUAL STEM PRODUCTIVITY CEDAR OTHER CEDAR OTHER 114. E7Mgégu 0 .2 MT/HA/Y 0' Gizlngg'A/YR 10.3 MT/HA ' (5.32) - (5.5%) MAPLE 3.7 MT/HA/YR (97.32) MAPLE 161.9 MT/HA (86.92) TOTAL = 186 MT/HA TOTAL = 3.9 MT/HA/YR Figure 32. Stem biomass and net stem (bole) primary pro- ductivity for all trees in the second-growth stand. £14 OLD-GROWTH STAND STEM BIOMASS ANNUAL STEM PRODUCTIVITY OTHER OTHER--\ C 17 MT/HA 0'87 "T/“A/YR OTEQRMT/HA/YR (14.72) (”0.72) (20.01) MmmE 28 MT/HA (24.32) CEDAR MAPLE 70 "T/HA 0.84 MT/HA/YR (60.81) (39.61) TOTAL = 115 MT/HA TOTAL = 2.1 MT/HA/YR Figure 33. Stem biomass and net stem (bole) primary pro- ductivity for trees in the old-growth stand. 85 12 >. \ < is A O I— . 1 5 I CEDAR )- .— Z 300 ‘ 3?}? E3 OTHER 3 D C d‘. 2.0 J ; MAPLE E g 100 ‘ ,— LU 2 Lu 3 an OLD-GRONTH SECOND-GRONTH STAND STAND 200 l 6 0 ' 31.131213: 2 T I I: 5; 120 U) U) < E 55 80 u‘ T .J O m “0 . M 1 OLD-GROWTH SECOND-GRONTH STAND STAND Figure 34. Comparisons of total bole net primary productiv- ity (A) and total bole biomass (B) between oldrgrowth and second—growth stands. Species categories include: cedar, maple and other. 86 Likewise, the second-growth stand is nearly twice as pro- ductive, with total bole production equalling 3.9 mt/ha/yr compared with 2.1 mt/ha/yr for the mature stand. DISCUSSION Composition, Structure and Origin of the Old-Growth Stand The large contribution of maple and ash to stand struc- ture, as described in this study, presents a different impression of stand composition than has been previously suggested by Thompson (1963) who described the stand as essentially pure cedar, based upon basal area values from two 25 m x 25 m quadrats. While specific locations were not given for these quadrats they may have been located in what is now the southern end of the study area and cannot be considered truly representative. Reports of cedar in excess of 31 m (100 ft) in height (Thompson 1963) were unsubstantiated. The tallest live cedar measured for this stand was 24 111 while the largest fallen cedar stem measured 26.1 m in length. These values agree closely with a reported height maximum of 25 m (80 ft) (United States Forest Service 1965). Growth curves for cedar presented by Gevorkiantz and Duerr (1939) suggest that cedar may have a.species-specific height maximum near 23 m (75 ft) on good sites (site index = 40). Overall stature of this forest is smaller than other mature forest stands in 87 88 Michigan (Rose 1982, Murphy and Kroh unpub. data) and re- flects the height maximum for cedar as well as the predom- inately small size (5 35 cm dbh) of the co-dominants. Regarding the origin of this old-growth stand, a close association was found between the occurrence of mature cedar and old-dune ridges. The hypothesis that these cedar had originally established on open dune slopes is supported by early growth rates of mature cedar, which were not statisti- cally different than growth of cedar currently colonizing the open lee slopes of dunes. I conclude that stand composition and the observed distribution of cedar resulted from encroachment of a series of dune ridges toward hardwood forests of the island inter- ior. Subsequent establishment of cedar on the dune ridges, along with hardwood regeneration, initiated development of the present old-growth stand. Such a process appears to be taking place in other island locations where cedar are currently establishing on arrested dune slopes adjacent to old-growth hardwood stands (Figure 35). Thompson's (1964) report of trees in this stand ex- ceeding 500 years of age was in accord with information obtained from dune soil profiles; textural analysis and pH values were similar to those reported by Olson (1958) for profiles between 500 and 1000 yr. It is suggested that this stand originated with dune migrations which occurred within the last several hundred years. 89 Figure 35. Photograph of cedar establishing on the open, lee slope of a dune which is encroaching upon hardwood forests of the island interior. 9O Successional Status of the Old-Growth Stand A profile diagram (Figure 36) suggests that old—growth cedar are being replaced by more tolerant maple and ash. Specifically, small numbers of mature cedar were observed in association with abundant maple and ash, largely g 35 cm dbh. Analysis of size-class distributions for the three dominants indicate that maple and ash are tolerant enough to maintain large numbers of stems in size classes 5 15 cm dbh while cedar shows correspondingly little recruitment. The negative exponential model provided a close fit (Figure 37) to distributions of maple and ash, which was in agreement with Lorimer (1980) who suggested that negative exponential distributions, in old-growth stands, should be expected for tolerant species only. In contrast, the negative exponen- tial model provided a poor fit with the size-class data of cedar (r == 0.37). The uniform size distributions of cedar was similar to species described by Hough (1932) and Christensen (1977), for which high rates of mortality in small size classes occurred in conjunction. with little population recruitment. Negative exponential distributions such as those ob- served for maple and ash do not necessarily imply the future importance of a species within a stand. A similar size class distribution was observed for Nyssa sylvatica over the course of 20 years in an oak-dominated forest (Christensen 1977), yet this species did not increase in importance. 91 .mcflaaaflum m>m0£ ma powwoflocfi wfl Amoco .umwuom ensues ecu canvas Scum EcumMAn waflmoum .om Ousmflm mmmhmz za muz 604 ,w””//’ \ \ 401 ' 201 <1: * ' ' T r T T ' T ‘ ' VT 2: \\ (I) LU Lu 1: A.— LL 0 60~ White Cedar a: 40 (Thuja occidentalis) LL] 4 g Y - 20.69 e "023" (r O 37) :3 . K ' - Z 20 .- ./ , -1 . 7 . . . , . . - é 80~ White Ash 60. (Fraxinus americana) 40< \ ~ Y - 58.0 e "065‘" (r .0,78) 201 q“””' 2. 10 20 30 40 50 T 60 ' TO T 3O ' 9O'T’1OO TITO 15 25 35 45 55 65 75 85 95 105 DIAMETER AT BREAST HEIGHT (CM) Figure 37. Negative' exponential. model analysis of' size distribution for maple, cedar and ash in the old-growth forest. 93 Christensen attributed this to a rapid turnover among indi- viduals in small size classes. Long-term patterns of re- cruitment and mortality for maple and ash were not examined but the dominance of mature individuals of these species in hardwood stands (Thompson 1974) adjacent to the study area suggests their future importance. Also, the absence of cedar in smaller size classes does not preclude the recruitment of this species at some future time. Recruitment of trees in some forest stands is asso- ciated with large disturbances such as fire or hurricanes (Henry and Swan 1974, Oliver and Stephens 1977). Although it is possible that the recruitment of cedar may be episodic and disturbance-dependent, the timing and direction of tree fall, and the lack of fire scars on old-growth trees sug- gests that the stand has not recently been subjected to large-scale disturbances. Future catastrophic disturbances, like fire or reinvasion of the forest by migrating dunes may, however, permit the establishment of young age classes of cedar. gap Reproduction Gap disturbance has been implicated in the maintenance of canopy diversity in old-growth or climax stands (Grubb 1977, White 1979 and Runkle 1981), and predictions of future canopy composition based upon species composition of sap- lings within forest gaps have been consistent with observed canopy importance values in several old—growth stands 94 (Runkle 1981). Canopy gaps in this study were similarly examined in an attempt to predict composition of the future canopy. Based upon relative density and dominance of tree seedlings (3 2.5 cm dbh), maple and ash are expected to be the major canopy dominants in this stand (Table 17). An analysis such as this is limited by the assumption that future canopy composition is accurately reflected in the densities and size of seedlings within gaps. However, factors such as tree position within the gap, tree life history strategies and gap size (Runkle 1982) are important determinants of successional change within gaps. The species composition of gaps in this investigation appears to be a function of gap size. Whereas ash was found to have higher stem densities (282 stems/ha) in gaps > 170 m2, maple had higher stem densities (1050 stems/ha) in gaps < 170 m2. These findings are consistent with Bormann and Likens (1979) who concluded that white ash is more likely to achieve importance in recently formed gaps 0f large size while maple appears to establish dominance more successfully in smaller gaps. Such patterns of reproduction and growth are attributed to differences in the reproductive strategies of these species and, more specifically, toler- ance. Maple is considered tolerant by Bormann and Likens (1979) and as such is expected to respond to small gap openings (> 50 m2) by establishment of seedlings or upgrowth of advance regeneration. Ash, of intermediate tolerance, requires larger openings (> 100 m2) to attain dominance with 95 Table 17. Anticipated canopy composition estimated from relative density, relative dominance and an average of these values for trees from four canopy gaps of various Sizes. PREDICTED CANOPY COMPOSITION SPECIES RELATIVE RELATIVE AVERAGEI/ DOMINANCE DENSITY Fraxinus americana 45 4O 42 Acer saccharum 34 45 40 Thuja occidentalis 16 9 13 TOTAL 95 1/ Most accurate predictor of canopy composition (Runkle 1981) 96 regeneration of seedlings and young, advance regeneration (generally < 10 years of age). Both species are considered able to respond to a spectrum of larger openings. Importance values for trees adjacent to or beneath the crowns of standing dead cedar were similar to those from gaps and suggest that maple and ash are most likely to succeed cedar in these locations. Together, these data imply an increasing importance of hardwoods in this stand and suggest that cedar is not successfully replacing itself. Regeneration and Growth of Cedar Under a Closed Canopy Following the orginal establishment of cedar in this stand, conditions for cedar reproduction and growth deter- iorated. This is made clear by the significanly slower growth rates of cedar establishing under closed-forest conditions relative to trees in the old-growth stand and on open-dunes, at comparable growth stages (Table 4). Addi- tionally, based upon published information on cedar regener- ation and growth, it was hypothesized that recruitment of cedar in this stand would be limited by hardwood leaf litter and the lack of adequate seedbeds. Presently, 99.7% of l-yr cedar seedlings are restricted to log substrates and it is suggested that this results from the inhibitory effects of leaf litter, primarily from maple and ash (Table 11). Such a pattern of seedling establishment was described for hem- lock in an old-growth, mixed-mesophytic forest in Pennsyl- vania. Keever (1973) observed that because of the annual 97 flush of leaf litter, small hemlock seedlings not on emer- gent substrates (stumps or logs) were physically destroyed by the litter cover. Successful cedar seedling establishment appears limited to logs and is greater on those logs in more advanced stages of decay (Table 6). Log substrates represented an area of 680 mZ/ha in the stand, with average seedling densities of 0.2 individuals/m2, compared to 0.006 individuals/m2 on leaf litter. This implies that logs serve as suitable seedbeds for cedar seedling establishment. For cedar seedlings in northern Michigan swamps, Nelson (1951) found that 87% had established on logs, and of these he noted that the highest root/shoot ratios were from seedlings growing on rotted wood. He speculated that intact (undecomposed) logs are less desirable growth sites because of resistance to root penetration. Similarly, in a spruce-fir forest, McCullough (1948) observed that the majority of seedlings were associ- ated with well-decayed logs. Initial establishment of cedar on logs was followed by high rates of mortality. The life-table and survivorship curve for cedar, based upon population age structure, demon- strate that there were two age-specific periods of mortal- ity: 1) an initial period of high mortality (years 0 - 3) followed by; 2) a longer period of decreasing mortality (years 4 - 12). Causes of mortality at each phase in the life cycle were not specifically identified but results from similar studies allow for speculation. High seedling 98 mortality, in some cases exceeding 95%, has been described for first year seedlings of other tree species (Hermann and Chricote 1965, Van Valen 1975, Hibbs 1979). Hibbs (1979) found that mortality of first year striped maple seedlings resulted from predation and unsuitability of the seedbed. Nelson (1951) found that the majority of young cedar seed- lings in a.1owland swamp died from desiccation as a result of having germinated on moss patches or logs with little decay; these sites tended to dry out as the growing season progressed, creating unfavorable moisture conditions. Other cause of mortality identified by Nelson (1951) included spring frosts, leaf-litter burial and competition from other plant species. Following an initial period of mortality (years 0 - 1) Hibbs (1979) noted a period (years 2 - 15) of no mortality for striped maple (Acer pennsylvanicum). This survival was thought to represent the ability of striped maple to persist in a suppressed state. In the present study, cedar exhibit- ed continuous mortality through year nine, followed by no apparent mortality through year 12 (Table 7). Survivorship during years 4 - 12 was analyzed using the negative exponen- tial and power function models. A power function curve more accurately described survivorship in this phase of estab- lishment (Figure 21), and such a model implies decreasing mortality with age (Hett 1971, Hibbs 1979). Inn this regard, Hett and Loucks (1968) related early differences in survivorship among white pine, balsam fir and 99 red maple to differences in shade tolerance for these three species. Cedar can survive in dense shade for several years and responds well to release at almost all ages (Johnston 1977). Thus, decreasing mortality for cedar in this stand may reflect increased shade tolerance with age. Neverthe- less, cedar seedlings in this investigation exhibited 99.4% mortality by age 12, suggesting that although logs serve as adequate seedbeds for establishment of cedar seedlings, long-term survival on these substrates is low. This is consistent with the observation that while seedlings are restricted to log substrates, larger stems (> 15 cm dbh) are associated with soil substrates (Figure 19). Aerial or stilt roots on mature cedar may become obscured with in- creasing stem size; however, Hough and Forbes (1943) found distinct aerial roots on white pine (Pinus strobus) in Pennsylvania, which had established on the root-plates of trees exposed by a wind storm in 1811. I assume, therefore, that the lack of aerial roots on mature cedar in this study reflects their establishment on a soil substrate. Patterns of mortality and survivorship data for seed- lings support the hypothesis that cedar is not replacing itself in smaller size classes in this stand. Furthermore, although logs in advanced stages of decay are adequate seed- beds for early cedar establishment, they may limit the long-term survival of these seedlings. 100 Patterns of Mortality for Canopy and Sub-canopy Trees Cedar in the old-growth stand appeared to represent a senescent population which accounted for 68% of total stand mortality (for all trees _>_ 2.5 cm dbh). Additionally, 28% of all cedar mortality was represented by standing dead stems, and it is likely that many of these individuals died as the result of physiological senescence or disease associated with reduced vigor. Such factors, along with lightning strikes and insects, have been cited as probable causes for the death of standing, old-growth trees in other mature stands (Lorimer 1980). In contrast to cedar, mor- tality for maple and ash represented a small percentage of stand mortality (9.6% each) and resulted from the death of small sub-canopy trees in the case of maple, along with wind-throw and death while standing for canopy-sized ash. Biomass and Productivity in the Old-Growth Stand Total aboveground biomass in the mature stand (176 Int/ha) is at the low end of the range for temperate decidu- ous forests (60 to 600 mt/ha; Whittaker and Likens 1975) and was less than the values reported for two old-growth Michi- gan stands. Murphy and Kroh (unpub. data) estimated total stand biomass for a mature stand of beech-maple to be 515 mt/ha, while Rose (1982) reported a value of 681 mt/ha for a relict stand of white pine. Biomass in this study was 101 comparable to regional second-growth forests and aboveground biomass (148 mt/ha) was bracketed by values presented by Crow (1978) for second-growth stands throughout the lake states. Because of the high stand mortality for massive, old-growth cedar and their apparent replacement by small maple and ash, biomass may be currently decreasing. Such a drop in stand biomass was noted by Larsen (1982) with canopy breakup in late successional jack-pine stands in Michigan. Presently, biomass is less than the 156.6 mt/ha reported for a 70 - 100 year-old stand of cedar in Minnesota (Reiners 1972). Maple and ash account for the majority of stand NPP (61%) which equalled 7.0 mt/ha/yr. This value was similar to second-growth stands in northern Wisconsin reported by Crow (1978) where aboveground NPP ranged between 7.1 and 10.4 mt/ha/yr, and was less than the second-growth stand of cedar in Minnesota which had a NPP of 10.1 mt/ha/yr (Reiners 1972). Old-growth stands are characterized by high standing I biomass and low rates of net primary productivity (Odum 1969). Correspondingly, biomass accumulation ratios (BARS, the aboveground biomass divided by aboveground net primary productivity) are high. Such a relationship is seen in the old-growth stands of white pine and beech-maple in Michigan (Table 18). In contrast, BARS in second-growth stands are lower because the trees are generally of smaller size and are more productive. The BAR in this study (22.2) was lower 102 manna away ~.- 4 A.A «AH nooscumm IHMUOU mafia: Amuse .n:a:=. nous . one means: m.Ho ¢.m mam madmzueooom .mmmflv omom ¢.m~H m.m Ham mafia mafia: Ammz\mmmsofinv Aum\ms\ua. oAumm Ammz. .mn\us. onus mousom cowumasaduud hufi>wuoscoum humawum mmmaowm Hmuoa Hmouom mmmaowm Hmsccd uoz H0009 .mccmpm somflnowz nuzouonpao How mowumu coaumasadooo mmmEOAQ use cofluoscoua .mmoao«n mo couwumaaou .ma manna 103 than other reported old-growth stands and was, in fact, comparable to second-growth stands in the northeastern U.S. (Whittaker et al. 1974). Additionally, Reiners (1972) presented a BAR value of 15.4 for a second-growth stand of cedar in Minnesota. That the BAR for this old-growth stand was similar to less massive, more productive second-growth stands appears to be a consequence of the shift in stand composition. With replacement of mature cedar by smaller, more vigorous maple and ash, a decrease in biomass is ex- pected to accompany an increase in productivity. Such a trend is suggested by Figure 38 where, compared to other old-growth stands, the present stand falls below the trend line for climax biomass relative to NPP. Stands below this line include immature and climax or near-climax stands which have been affected by disturbance (Whittaker et a1. 1974). Thus, with the apparent successional replacement of cedar, levels of biomass and productivity were not characteristic of an old-growth stand but were instead comparable to re- gional second-growth stands. Forest Recovery Following Logging of the Old-Growth Stand Date of Logging Changes in the annual growth rings of forest tree stems have provided accurate information for the reconstruction of disturbance history of various stands throughout the 104 7Ol ‘E5 60 + 9 :501 ‘2 Sun. .53 :30. Q 220. £3 E; 10 4 2 ' V I V f r fl (1 200 A00 600 800 1000 1200 1400 1600 1800 ABOVEGROUND NET PRIMARY PRODUCTIVITY (G/Mz/YR) Figure 38. Aboveground biomass relative to net primary pro- ductivity (NPP). Closed circles represent old-growth Michi- gan stands; 1 (this study). 2 (Larsen 1982), 3 (Murphy and Kroh unpub. data) and 4 (Rose 1982). Dashed line is trend for climax biomass verses NPP. Solid lines include climax stands from woodlands through temperate-zone forests. Forest stands above biomass trend line were reported to be stable climax associations, while some climax or near-climax stands below this line had been affected by disturbance (Whittaker, Bormann, Likens and Siccama 1974). Graph modified from Whittaker et al. (1974). 105 northeast (Henry and Swan 1974, Oliver and Stephens 1977, Lorimer 1980). In this study, old-growth cedar directly adjacent to the present second-growth stand displayed dra- matic and sustained increases in growth in approximately 1896. This growth release was thought to result from the catastrophic disturbance created by the removal of many large cedar during logging. Additionally, age data for trees which had established on cut cedar stumps and age- structure data for second-growth maple were used to cor- roborate growth release information in this investigation. The oldest stump-established tree examined (64 years) indicated that successful establishment occurred approxi- mately 20 years following the estimated logging date. Since seedling establishment for some tree species appears to be favored on logs in more advanced stages of decay (McCullough 1948, Nelson 1951) it may have taken up to 20 years for cut cedar stumps to decay sufficiently for successful seedling establishment to occur. This seems reasonable because even after 85 years these stumps were still recognizable (Figure 39). Age-structure analysis of maple in the second-growth stand indicates that although this species was present prior to logging the majority of stems are second-growth. A fit of the negative exponential model to these data showed a peak in the 60 year age-class followed by smaller than predicted numbers of individuals in the 30 to 40 year age- classes (Figure 25). Such a distribution suggests that 106 Figure 39. Establishment of a second-growth birch on a cut cedar stump within the logged study area. 107 although sugar maple existed in the stand as advance regen- eration, abundant recruitment occurred 20 - 25 years follow- ing logging. Bormann and Likens (1979) indicate that toler- ant species such as sugar maple are much less responsive to major disturbances such as logging than are intolerant, rapidly-growing species like pin cherry, trembling aspen and white birch. With a decline in the importance of these intolerant species they suggest that species such as maple may begin to increase in importance 20 to 30 years after logging. Following the eventual development of a second- growth maple canopy further recruitment may have been sup- pressed, accounting for lower than expected numbers of individuals in the 30 to 40 year age classes. Suppression of regeneration under dense, even-aged canopies has been previously described (Peet and Christensen 1980, Peet 1981). The popualtion age structure of second-growth maple and age analysis of stump-established trees support growth release information, which suggests that old-growth cedar were removed from a portion of the stand at the turn of the century. Trends in Forest Structure and Function Following Logging Pre-logging values of dominance and density were recon- structed for cedar in the old-growth and adjacent second- growth. stand. Originally, stands 'were :not. significantly 108 different (Table 16), and thus the second-growth stand is assumed to represent structural and functional changes which resulted from logging of the old-growth cedar stand. The harvesting of cedar from a portion of the old- growth stand at the turn of the century has given rise to a second-growth stand completely dominated by sugar maple. It was hypothesized that removal of old-growth cedar, without damage to the understory, would favor advance-regeneration of more tolerant species such as maple and, in effect, pro- mote succession. Basal area. changes subsequent; to 11900 demonstrate the suggested replacement of cedar by maple in both old-growth and second-growth stands. However, in contrast to the rapid recovery of second-growth maple fol- lowing logging is the gradual, tree-by-tree replacement of cedar by maple in the old-growth stand. Basal area of the mature stand declined over the period 1900 to 1930, then gradually increased to a current stand total of 34 mz/ha, as basal area increase for maple out-paced the loss for cedar (Figure 30). This successional process has been abbreviated in the second-growth stand, where a rather even-aged stand, dominated by maple, has recovered a total basal area of 29 mz/ha in the 85 years following logging. It is evident from this information that the logging of mature cedar acceler- ated a process which is presently occurring in the old- growth stand, that is, the successional replacement of cedar by maple and ash. 109 The cutting' of? cedar' also altered. energy' relations within the stand and as a consequence current levels of biomass and productivity in the second-growth stand exceed those of the mature forest (Figure 34). Based upon basal area change, bole biomass in the second-growth stand is assumed to have increased sharply since 1930 (Figure 29) and presently equals 186 mt/ha. This increase is expected to continue to an undetermined site maximum as the even-aged canopy of maple reaches maturity. Peak biomass values fol- lowing logging were observed within 70 years for Populus grandidentata stands in northern Michigan (Cooper 1981) and a subsequent biomass decline was anticipated with the mor- tality of mature aspen. Additionally, in disturbed Riggs contorta stands of Colorado, Peet (1981) noted coincident increases in basal area and biomass to maxima, followed by pronounced decreases. These decreases resulted from mortal- ity of mature canopy trees coupled with a subsequent lack of regeneration. In the old-growth stand, by comparison, a slow decline in basal area, and presumably biomass, occurred from 1900 until 1940 with the progressive loss of cedar (Figure 30). However, replacement of' basal area and. biomass through growth and reproduction of maple appears to have reversed this decline. Total bole biomass for the stand is 115 mt/ha and despite the apparent current increase, it is theoret- ically unlikely that this stand will ever approach a biomass maximum equal to that expected for the second-growth stand. 110 This follows from the fact that as cedar in the mature stand are replaced on a tree-by-tree basis by maple (or associated hardwoods), the resulting forest will have an all-aged population structure. Such a forest represents a spectrum of tree sizes which may not equal the biomass of an even- aged stand of mature trees (Peet 1981). In this regard, computer simulation of forest recovery following clear cuts of northeastern deciduous forest stands at Hubbard Brook, New Hampshire, indicate that biomass reaches a maximum in approximately 200 years with maturation of the even-aged canopy. After 350 years, biomass drops to an asymptote as the stand becomes more all-aged (Bormann and Likens 1979). Total bole productivity for the present second-growth stand is 3.9 mt/ha/yr with maple accounting for 94% of this total (Figure 32). Although trends for productivity in this stand are not easily discerned, it is suggested that produc- tivity in early successional communities may quickly reach levels approaching that for mature stands on the same or similar sites (Marks 1974, Bormann and Likens 1979). Boring et a1. (1981) noted that first-year net production on a clear-cut site equalled 22% that of adjacent, undisturbed hardwood forest in the southern Appalachians, suggesting very rapid recovery of NPP. Further, Bormann and Likens (1979) indicate that within the first one to two decades after logging, net primary productivity may initially peak and then decline. Cooper (1981) observed that trunk or bole productivity for aspen in northern Michigan reached a 111 maximum between 40 and 45 years after cutting while overall productivity for aspen peaked between years 45 - 50, and thereafter declined. Such declines in long-term net pro- duction for post—disturbance stands may be attributed to changes in stand density, nutrient availability and species composition (Peet 1981). Additionally, Kira and Shidei (1967) suggest that increased respiration associated with stand size and age may further limit net production. Wheth- er productivity in this second-growth stand is currently increasing, has reached a maximum, or is in decline is not known. However, with increasing tree size and decreasing density for this even-aged stand, an eventual drop in pro- ductivity from present levels is anticipated. Total bole productivity in the mature stand was 2.1 mt/ha/yr (Figure 33). Future trends in productivity for this stand will be strongly dependent upon tree population dynamics. For example, if the gradual attrition of old- growth cedar continues as indicated by basal area change (Figure 30) and replacement by maple is correspondingly gradual, net productivity may show little overall increase with time. If, on the other hand, loss of mature, unpro- ductive cedar occurs more rapidly than at present, rapid replacement of these trees by smaller, more productive maple and ash could result in a temporary production increase. Current levels of’ biomass and. productivity in the second-growth stand, relative to the old-growth stand, are summarized in Figure 34. The second-growth stand is nearly 112 twice as productive as is the mature stand, and has a higher total biomass. With synchronous growth and maturation of second-growth maple, biomass is predicted to reach an un- determined site maximum. The more gradual replacement of cedar by maple and ash in the old-growth stand has resulted in a more all-aged stand structure and as a consequence stand biomass, although increasing, is not anticipated to equal the peak biomass value expected for the second-growth stand. Barring future catastrophic disturbances, biomass and productivity in these two stands should converge with increased importance of maple in the mature stand, and development of a more all-aged population structure for the second-growth stand. SUMMARY AND CONCLUSIONS Results of this study provided interesting and sur- prising insights into the origin, structure and function of this unusual forest stand. Growth patterns exhibited by mature cedar on S. Manitou Island indicate that the trees established under open dune conditions. The present dis- tribution of cedar on dune ridges apparently resulted from the advancement of dunes toward hardwood forests inland, followed by colonization of the open dune slopes by cedar. Contrary to a report suggesting that the stand was essentially pure cedar, a total of nine dominant tree spe- cies were encountered ‘within the study area, including cedar, maple and ash. Furthermore, diameter distributions suggest that while maple and ash are tolerant enough to maintain large numbers of individuals in smaller size class- es (2.5 to 15 cm dbh), cedar is distributed throughout a broad size range with little population recruitment of stems 5 15 cm dbh. Prediction of future canopy composition im- plies the eventual dominance of maple and ash. White ash was more important in large canopy gaps (> 170 m2) while maple was more important in gaps < 170 m2. Presently, 99.7% of’ first; year' cedar‘ seedlings are restricted to log substrates, which appear to provide safe sites for establishment. Seedlings which establish on the 113 114 forest floor may, in large part, be destroyed by the annual fluSh of hardwood leaf litter. Logs in advanced stages of decay provide seedbeds suitable for germination and estab- lishment of cedar seedlings; however, long-term survival on these sites may be limited. Population age-structure infor- mation indicates that following germination 99.4% of all seedlings are lost by year 13. Further, no cedar stems > 40 cm dbh show evidence of having established on logs. Cedar stems represent 68% of total stand mortality for trees 3 2.5 cm dbh, and 31% of all cedar sampled were dead. In contrast, maple and ash together accounted for only 19% of total mortality and a much smaller percentage (3% and 7%, respectively) of stems for these species were dead. Despite the current dominance of cedar, patterns of tree mortality, in conjunction with diameter distributions, provide evidence that maple and ash are replacing mature cedar in this stand. The distribution of biomass and productivity in the mature stand was similar to second-growth stands throughout the lake states. Aboveground biomass was 148 mt/ha, with large, old-growth cedar comprising 61% of the aboveground total. Total annual aboveground net primary productivity (NPP) was 7.1 mt/ha/yr. While cedar represent 27% of above- ground NPP, small vigorous maple and ash account for 59% of the total. With the replacement of large, mature cedar by smaller, more productive maple and ash the biomass accumu- lation ratio (biomass/NPP) for the stand was lower than for 115 other old-growth Michigan stands, and comparable to a 70 - 100 year old stand of white cedar in Minnesota. A portion of the mature stand was logged of cedar in approximately 1896. By 1930, basal area for second-growth maple began to increase exponentially and presently accounts for 77% of total stand basal area. As a consequence of the removal of mature cedar 85 years previously, the second- growth stand has recovered levels of bole biomass and NPP which now exceed those found in the mature stand. Bole biomass for the logged stand (186.4 mt/ha) is 38% greater than that of the mature stand (115.3 mt/ha), while bole NPP (3.9 mt/ha/yr) is almost double (2.1 mt/ha/yr). Biomass and NPP in the two stands are expected to converge as the second-growth stand becomes more all—aged while maple and ash continue to replace cedar in the mature stand. APPENDIX 116 Appendix 1. Criteria used in assessing decay states of logs. DECAY STATE DESCRIPTION SUBJECTIVE CRITERIA 1. Relatively undecomposed Little or no moss; bark mostly intact 2. Moderately decomposed Scattered moss patches; little or no bark 3. Well decomposed Weed soft but not falling apart; at least a moss or herb covered; no bark 4. Very well decomposed W00d falling apart; outline of log not distinct LITERATURE CITED Boring, L. R., C. D. Monk, and W. T. Swank. 1981. Early regeneration of a clear-cut southern Appalachian forest. Ecology 62(5):1244—1253. Bormann, F. H., and G. E. Likens. 1979. Pattern and pro- cess in a forested ecosystem. Springer-Verlag, New York, New York, USA. Braun, E. L. 1950. Deciduous forests of eastern North America. Blakiston, Philadelphia, Pennsylvania, USA. Brower, J. E., and J. H. Zar. 1977. Field and laboratory methods for general ecology. William C. Brown Company, Dubuque, Iowa, USA. Caulkins, H. L., Jr. 1967. The ecology and reproduction of northern white-cedar. Thesis. School of Natural Re- sources, University of Michigan, Ann Arbor, Michigan, USA. Christensen, N. L. 1977. Changes in structure, pattern and diversity associated with climax fOrest maturation in Piedmont, North Carolina. American Midland Naturalist 97:176-188. Cooper, A. W. 1981. Above-ground biomass accumulation and net primary production during the first 70 years of succession in Populus grandidentata stands on. poor sites in northern lower Michigan. Pages 339—357. In D. C. West, H. H. Shugart, and D. B. Botkin, ed- Eors. Forest Succession: concepts and applications. Springer-Verlag, New York, New York, USA. Cowles, H. C. 1901. The physiographic ecology of Chicago and 'vicinity. Botanical Gazette 131:73-108, 145-182. Crow, T. R. 1978. Biomass and production in three contigu- ous forests in northern Wisconsin. Ecology 59:265-273. Curtis, J. D. 1944. N0rthern white cedar in upland soils in Maine. Journal of Forestry 42:756-759. 1946. Preliminary observations on northern white cedar in Maine. Ecology 27:23-26. 117 118 Day, P. R. 1965. Particle fractionation and particle size analysis. _I__n_ Black, C. A., editor. Methods of Soil Analysis, Part I. American Society of Agronomy, Inc., Madison, Wisconsin, USA. Dyer, R. G. 1967. Fresh and dry weight, nutrient elements and pulping characteristics of northern white cedar Thu'a occidentalis. Maine Agricultural Experiment Station Technical Bulletin 27:1-39. Forbs, R. D. 1955. Forestry Handbook. The Roland Press Company, New York, New York, USA. Gates, F2 (3. 1926. Plant succession about Douglas Lake, Cheboygan County, Michigan. Botanical Gazette 82: 170-182 . Gevorkiantz, S. R., and W. A. Duerr. 1939. Volume and yield of northern white cedar in the lake states. IL. S. Ferest Service. Lake States Forest Experiment Station, St. Paul, Minnesota, USA. Grime, J. P. 1977. Plant strategies and vegetation pro- cesses. J. Wiley and Sons, New Y0rk, New Y0rk, USA. Grubb, P. J. 1977. The maintenance of species-richness in plant communities: the importance of the regeneration niche. Biological Review 52:107-145. Habeck, J. 1959. Ecology of white cedar swamps in northern Wisconsin with special attention to their role as win- ter deer range. Dissertation Abstrasts 20(6):1191-1192. Hatt, R. T., J. Van Tyne, L. C. Stewart, C. H. Pope, and A. B. Grobman. 1948. Island Life: a study of the land vertebrates of the islands of eastern Lake Michigan. Cranbrook Institute of Science. Bulletin No. 27. Cranbrook Press, Bloomfield Hills, Michigan, USA. Henry, J. D., and J. M. A. Swan. 1974. Reconstructing forest history from live and dead plant material - an approach to the study of forest succession in south- west New Hampshire. Ecology 55:772-783. Hermann, R. K., and W. W. Chricote. 1965. Effects of seed- beds on germination and survival of douglas-fir. Research paper. (Forest Management Research) Oregon Forest Research Laboratory 4:1-28. 119 Hett, J. M. 1971. A dynamic analysis of age in sugar maple seedlings. Ecology 52:1071-1074. Hett, J. M., and O. L. Loucks. 1968. Application of life- table analyses to tree seedlings in Quetico Provincial Park, Ontario. The Forestry Chronicle 44(2):29-32. Hibbs, D. E. 1979. The age structure of a striped maple population. Canadian Journal of Forest Research 9: 504-508. Hough, A. F. 1932. Some diameter distributions in forest stands of northwestern Pennsylvania. Journal of Forestry 30:933-943. Hough, A. F., and R. D. Forbes. 1943. The ecology and silvics of Pennsylvania high-plateau forests. Ecolog- ical Monographs 13(3):301-320. Johnston, w. F. 1977. Managers' handbook for northern white cedar in the north central states. USDA Forest Service General Technical Report. NC-35. North Central Forest Experimental Station, St. Paul, Min- nesota, USA. Kapustka, L. A., and R. G. Koch. 1979. The acceleration of succession of a birch-maple forest due to high wind. The Michigan Botanist 18(2):83-87. Kira, T., and T. Shidei. 1967. Primary production and turnover of organic matter in different forest eco- systems of the western Pacific. Japanese Journal of Ecology 17:70-87. Keever, C. 1973. Distribution of major forest species in southeastern Pennsylvania. Ecological Monographs 43: 303-327. Kramer, P. J., and T. T. Kozlowski. 1979. Physiology of woody plants. Academic Press. New York, New York, USA. Larsen, W. M. 1982. Structure, biomass and net primary productivity for an age-sequence of jack pine eco- systems. Dissertation. Michigan State University, East Lansing, Michigan, USA. Lorimer, C. G. 1980. Age structure and disturbance history of a southern Appalachian virgin forest. Ecology 61: 1169-1184. 120 Marks, P. L. 1974. The role of pin cherry (Prunus pennsylvanica L.) in the maintenance of stability in northern hardwood ecosystems. Ecological Monographs 44:73-88. McCullough, H. A. 1948. Plant succession on fallen logs in a 'virgin spruce-fir forest. Ecology’ 29(4):508-513. Moran, R. C. 1978. Vascular flora of the ravines along Lake Michigan in Lake County, Illinois. Michigan Botanist 17(4):123-140. Mueller-Dombois, D., and H. Ellenberg. 1974. Aims and methods of vegetation ecology. John C. Wiley and Sons. New York, New York, USA. Musselman, R. C., D. T. Lester, and M. S. Adams. 1975. Localized ecotypes of Thuja occidentalis L. in Wiscon- sin. Ecology 56:647-655. Nelson, T. C. 1951. A reproductive study of northern white cedar . Game Division. Department of Conservation, Lansing, Michigan, USA. Newbould, P. :L. 1967. Methods for estimating the primary production of forests. IBP Handbook No. 2. Blackwell Scientific Publications, Oxford, England. Odum, E. P. 1969. The strategy of ecosystem development. Science 164:262-270. Oliver, C. D., and E. P. Stephens. 1977. Reconstruction of a mixed-species forest in central New England. Ecology 58:567-572. Olson, J. S. 1958. Rates of succession and soil changes on southern Lake Michigan dunes. Botanical Gazette 119 (3):125-170. Peet, R. K., and N. L. Christensen. 1980. Succession: a population process. Vegetatio 43:131-140. Peet, R. K. 1981. Changes in biomass and production during secondary forest succession. Pages 324-338. In D. C. west, H. H. Shugart, and D. B. Botkin, editors. Forest Succession: concepts and applications. Springer-Verlag, New York, New York, USA. Potzger, J. E. 1941. The vegetation of Mackinac Island, Michigan: an ecological survey. American Midland Naturalist 25:298-323. 121 Potzger, J. E. 1944. The vegetation of Round Island (Straits of’ Mackinac), Michigan. Butler ‘University Botanical Studies 6:116-121. Reiners, W. A. 1972. Structure and energetics of three Minnesota forests. Ecological Monographs 42:71-94. . 1974. Foliage production by Thuja occidentalis L. from biomass and litterfall estimates. American Midland Naturalist 92:340-345. Rogers, J. 1966. South Manitou Island. A field trip source book and guide. Northwestern Michigan College, Traverse City, Michigan, USA. Rose, W. M. 1982. Biomass, production and successional dynamics of a virgin white pine stand in northern Michigan . Di s sertati on . Michigan State University , East Lansing, Michigan, USA. Runkle, J. R. 1981. Gap regeneration in some old-growth forests of the eastern United States. Ecology 62(4): 1041-1051. . 1982. Patterns of disturbance in some old-growth mesic forests of eastern North America. Ecology 63(5): 1533-1546. Shelby, S. M. 1967. Standard Mathematical Tables. The Chemical Rubber Company, Cleveland, Ohio, USA. Sokal, R. R., and F. J. Rohlf. 1969. Biometry. W. H. Freeman and Company, San Francisco, California, USA. Spurr, S. H., and B. V. Barnes. 1973. Forest Ecology. Second Edition. Roland Press, New York, New York, USA. Steel, R. G. D., and J. H. Torrie. 1960. Principles and procedures of statistics. McGraw-Hill Book Company, Incorporated, New York, New York, USA. Thompson, P. W. 1963. The character of an ancient white cedar forest on South Manitou Island, Michigan. Papers of the Academy of Science, Arts, and Letters 47: 177-186. . 1974. Ecological survey of South Manitou Island. Prepared under contract with the National Park Service. United States Forest Service. 1965. Silvics of forest trees of the United States. Agricultural Handbook No. 271. United States Government Printing Office, Wash- ington, D.C., USA. 122 Van Valen, L. 1975. Life, death and energy of a tree. Biotropica 7(4):259-269. Waterman, W. G. 1922. Forests and dunes from Point Betsie to Sleeping Bear, Benzie, and Leelanau Counties, Michi- gan. Northwestern university, Evanston, Illinois, USA. Weber, H. L. 1973. Soil survey of Leelanau County, Michi- gan. United States Department of Agriculture, Soil Conservation Service. U. S. Government Printing Office, Washington, D.C., USA. White, P. S. 1979. Pattern, process, and natural disturb- ance in vegetation. The Botanical Review 45:229-299. Whittaker, R. H., and G. M. Woodwell. 1968. Dimension and production relations of trees and shrubs in the Brook- haven Forest, New York. Journal of Ecology 56:1-25. Whittaker, R. H., F. H. Bormann, G. E. Likens, and T. G. Siccama. 1974. The Hubbard Brook ecosystem study: forest biomass and production. Ecological Monographs 44:233-252. Whittaker, R. H., and G. E. Likens. 1975. The biosphere and man. Pages 305-328. 12. H. Leith and R. H. Whittaker, editors. Primary Productivity of the Biosphere. Springer-Verlag, New York, New York, USA. Whittaker, R. H., and P. L. Marks. 1975. Methods of asses- sing terrestrial producitivity. Pages 55-118. _I_r_1 H. Leith and R. H. Whittaker, editors. Primary Produc- tivity of the Biosphere. Springer-Verlag, New York, New York, USA. Wolfe, F. 1932. Annual rings of Thuja occidentajlis in relation to climatic conditions and movement of sand. The Botanical Gazette 93:328-335. Young, H. E., and P. M. Carpenter. 1967. weight, nutrient element and productivity studies of seedlings and saplings of eight tree species in natural ecosystems. Maine Agricultural Experiment Station Technical Bulletin 28: 1-39 . "11111111I“