Printer Friendly

CANOPY TREE TURNOVER IN OLD-GROWTH MESIC FORESTS OF EASTERN NORTH AMERICA.

JAMES R. RUNKLE [1]

Abstract. I studied the dynamic nature of old-growth, eastern U.S. forests by addressing the following questions: (1) How much do stand density, basal area, and size structure vary over time within several old-growth remnants? (2) How do mortality and growth rates vary with stem size? (3) How much does the importance of individual species vary over time and space? (4) At what rate do snags and stumps form and deteriorate?

In 1990-1991, I resampled canopy stems within several old-growth remnants in the southern Appalachians, in Hueston Woods State Park, Ohio, and in the Tionesta Scenic and Research Natural Areas, Pennsylvania. I had previously sampled those sites in 1976-1977 using the point-centered quarter method. I remeasured the same trees and measured new trees if the old ones had died or if a new stem closer to the point than the old stem for that quarter had grown to [greater than or equal to]25 cm in diameter at breast height.

Density and basal area changes were small. Density changes equaled -0.33%, -0.52%, and 0.00%/yr for the southern Appalachians, Hueston Woods, and Tionesta sites, respectively. Corresponding basal-area changes were 0.03%, -0.22%, and 0.45%/yr. Mortality increased consistently with stem size in all three areas, However, growth rates of smaller stems more than compensated for the higher mortality rates of larger stems, so that overall stem size increased between samples. Most species changed little in relative density or basal area. Overall Fagus showed the largest changes between samples, with a small decrease in the southern Appalachians, a larger decrease in Hueston Woods, and an increase in Tionesta. Trees usually died standing, breaking off at a variety of heights. For example, only 6-27% of trees that died were uprooted, depending on region, whereas 16-31% broke at a height of [greater than or equal to]16 m. Total snag densities were 15-18 snags/ha.

In general these stands were marked by slow changes toward fewer, larger stems, even after centuries in which no major disturbances had occurred. However, changes in pathogens, climate, and atmospheric chemistry could change these trends in the future.

Key words: Acer; Appalachians, southern; Fagus; forest, eastern USA, old-growth; Hueston Woods (Ohio, USA); snags, formation and deterioration; species composition, temporal and regional variation; stem size, effects on tree growth and mortality; Tionesta (Pennsylvania, USA); tree demography; tree mortality, function of stem size, Tsuga.

INTRODUCTION

Following a major disturbance natural communities undergo succession, resulting, hypothetically, in a steady state, or climax, in which no more net changes in species composition or ecosystem properties such as biomass occur (Whittaker 1975, Connell and Slatyer 1977, Bormann and Likens 1979). The process of forest development has been divided into four stages: stand initiation, stem exclusion, understory reinitiation, and old growth (Oliver 1981, Oliver and Larson 1990). The old-growth stage occurs when enough time has elapsed since a major disturbance that the species composition of the understory and overstory have converged, and the stand as a whole is in equilibrium with regard to species composition and stand structure.

The properties of old-growth forests, particularly in the eastern United States, have recently been the focus of much attention (Parker 1989, Martin 1992, Davis 1996). Many aspects of old-growth communities still are not clear (Davis 1996). It is unlikely that even the oldest and least disturbed forests are in the perfect equilibrium hypothesized above. Studies therefore are necessary not to document a theoretical state of no change but to quantify the normal rates of change that do occur. Without knowing normal rates of variation we cannot evaluate the significance of changes ostensibly caused by such anthropogenic factors as climate change or introduced pathogens (Millers et al. 1989, MeClenahen and Long 1993, Houston 1994, Vitousek 1994, Hoffard et al. 1995, Hom et al. 1996, SAMAB 1996).

Here I document changes in structure and species composition of canopy trees in several forests of eastern North America that may not have experienced large disturbances for centuries to millennia to see how nearly they approach a steady-state condition. Sites were sampled using the point-centered quarter method in 1976-1977 and again in 1990-1991. These results are compared to other studies done on old-growth forests primarily in the eastern United States. In particular I address four major questions.

1. How much do stand characteristics (density, basal area, stem size distribution) vary over time within several old-growth remnants?

2. How do mortality and growth rates vary with stem size? To understand community changes over time and space it is necessary to know how the patterns of mortality and growth vary with stem size, species, and geographic location. Previous literature on this relationship is sparse and ambiguous (Harcombe 1987, Sheil and May 1996).

3. How much does the importance of individual species vary over time and space? Species respond differently to the environment. Therefore, at any given time some species will be increasing and some decreasing in importance due to recent conditions, such as changes in climate. Quantifying those changes now is important because a variety of changes are about to take place in the areas I studied that will influence species success. Examples include global warming and introduced pathogens (McClenahen and Dochinger 1985, Houston 1994, Hom et al. 1996, SAMAB 1996).

4. At what rates do snags and stumps form and deteriorate? An important, defining characteristic of old-growth forests is the presence of dead, standing stems (snags) and logs (Parker 1989, Martin 1992, Runkle 1996). In an equilibrium forest, stems die to form snags and logs of different heights and lengths. Snags gradually deteriorate to form smaller snags and, eventually, logs. The dynamics of snag formation and deterioration affect ecosystem dynamics (biomass accumulation and decay) and wildlife habitat (Harmon et al. 1986). However, few data are available to quantify those dynamics.

METHODS

Study areas

I sampled old-growth, mesophytic stands from three regions in eastern North America: the southern Appalachians (in Tennessee and North Carolina), Ohio, and Pennsylvania. Within the southern Appalachians I sampled stands from the Great Smoky Mountains National Park in Tennessee and North Carolina, the Joyce Kilmer Wilderness Area in the Nantahala National Forest, North Carolina, and the Walker Cove Research Natural Area in the Pisgah National Forest, North Carolina. Study areas in the southern Appalachians occupied mid-elevations (766-1236 m) and latitudes 35-36[degrees] N. They were chosen to avoid obvious human impacts (Pyle 1988). Dominant species included sugar maple (Acer saccharum), eastern hemlock (Tsuga canadensis), American beech (Fagus grandifolia), mountain silverbell (Halesia carolina), white basswood (Tilia heterophylla), yellow buckeye (Aesculus octandra), and yellow birch (Betula allegheniensis). Fires were rare to absent. Few stems were affected by any single windstorm. I avoided areas in whic h rhododendron (Rhododendron maximum) formed a dense shrub layer or in which American chestnut (Castanea dentata) had been abundant before the chestnut blight. Several pathogens are likely to increase tree mortality in the future: dogwood anthracnose was first reported in the region in 1988, and beech bark disease was first reported in the Great Smoky Mountains National Park in 1993; the European gypsy moth and hemlock woolly adelgid have not yet arrived in my study areas (SAMAB 1996). Further site details are given in Runkle (1981, 1982, 1985a, b, 1998) and Runkle and Yetter (1987).

For some analyses I subdivided the southern Appalachian data into the same three relatively homogeneous groups as in Runkle (1981). Roaring Fork is a mid-elevation cove (mean 929 m) in the Great Smoky Mountains with much hemlock. Albright Grove-Kalanu Prong are mid-elevation coves (mean 985 m) in the Great Smoky Mountains with substantially less hemlock. Walker Cove is a high-elevation cove (mean 1213 m) with no hemlock or silverbell.

The Ohio site is a 67-ha old-growth stand in Hueston Woods State Park near Oxford, southwest Ohio (39[degrees]34' N, 84[degrees]45' W). It has remained relatively undisturbed since its purchase in 1797. The forest overstory is dominated by beech and sugar maple. It has been the subject of several studies on forest dynamics (Vankat et al. 1975, Runkle 1981, 1982, 1990a, Moore and Vankat 1986, Peters and Poulson 1994, Hicks 1996, Kupfer and Runkle 1996, Fore et al. 1997, Kupfer et al. 1997).

The Pennsylvania site is part of the Tionesta Scenic and Research Natural Areas in the Allegheny National Forest (41[degrees]39' N, 78[degrees]57' W) and was a large area dominated by beech and hemlock. Within the site were some areas recovering from large disturbances (e.g., windstorms in 1808 and 1870) and some areas in which small-scale disturbance ("gaps") were most important (Bjorkbom and Larson 1977). In 1977 I sampled only within the latter areas (Runkle 1981, 1982, 1985a). In 1985 a large section was affected by another large blowdown (Petersen and Pickett 1995). In 1991 I re-sampled only areas not obviously affected by that blowdown. Beech bark disease reached these stands in the late 1980s (Millers et al. 1989, Houston 1994).

Field methods

In 1976-1977 I chose suitable study areas within each of the sites described above. Beginning at randomly chosen points, transects were walked along compass lines parallel to the long axis of each suitable study area. At random places along those transects the point-centered quarter method (Cottam and Curtis 1956) was used to record the canopy vegetation: the first point was located 0-25 paces from the beginning of the transect and subsequent points were located 25-75 paces ([tilde]17-50 m) apart. At each point I measured distances to and diameters of the nearest trees [greater than or equal to]25 cm in diameter at breast height (dbh; diameter at breast height = 137 cm) in each quarter. In each study area (seven locations within Great Smoky Mountains National Park plus Joyce Kilmer, Walker Cove, Hueston Woods, and Tionesta) I sampled 30-200 points along transects totaling 800-6500 m (Runkle 1985b). In 1990-1991 I relocated the transects using my original field notes, as described in Runkle and Yetter (1987). Some transects were not relocated, usually because the small coves in which they were located were not adequately described in the field notes.

I located the position of each point using my records of the species, sizes, distances, and quarters of trees from the previous sample and the distances between points. Configurations of these parameters for points were unique, so ambiguity was rare even though the trees had not been marked. After relocating a point I remeasured all the original trees. I indicated whether or not they had died since the last sample. If they were alive, I measured their dbh. If they had died, I estimated their stump height and found a replacement tree for their quarter, recording its species, dbh, and distance from the original point. In some quarters the original tree was still alive, but a small tree closer to the point had grown to be [greater than or equal to]25 cm dbh. In that situation I recorded the species, dbh, and distance of the new replacement tree.

Species nomenclature follows Little (1979).

Analytical methods

I calculated the mortality and growth rates of the original trees as well as overall changes in the species and size class distributions, density, and basal area of each stand. Tables were constructed of initial characteristics; decreases by death, replacement, and outgrowth to larger size classes; and increases by new trees or ingrowth from smaller size classes.

I used t tests to judge the significance (P [less than or equal to] 0.05) of changes in stand density and basal area from 1976-1977 to 1990-1991. I based these t tests on the raw data collected using the point-centered quarter method: the distances from point to tree, tree dbh, and tree basal area.

Mortality rates were calculated as exponential decay rates:

Average annual mortality rate = 1 - [(S/[N.sub.0]).sup.(l/y)]

where S = number of survivors, [N.sub.0] original number of stems, and y = number of years between samples (Runkle 1990b, Sheil et al. 1995).

Growth rates were compared among size classes and among regions using an analysis of variance followed by Tukey's Studentized range test to look for significant (P [less than or equal to] 0.05) differences between specific size classes (GLM procedure with TUKEY option [SAS 1985]).

Compositions were compared using percentage similarity, defined as follows:

[PS.sub.ab] = [[[sigma].sup.n].sub.i=1] min([p.sub.[i.sub.a]], [P.sub.[i.sub.b]])

where [PS.sub.ab] is the percentage similarity value, [p.sub.[i.sub.a]] and [p.sub.[i.sub.b]] are the relative importance values (as percentages) for a given species i in the two samples a and b, and the summation is over n species (Whittaker 1975). The importance value used here is the average of relative density and relative dominance (basal area).

Snag dynamics were modeled assuming that snags form at a constant rate (number per hectare per year) from live trees and that a constant fraction of snags deteriorates to non-snags each year (Gore et al. 1985, Runkle 1991), although Harmon et al. (1986) found some populations to have an initial lag period in which little fragmentation occurred. For this study I define snags to be stems [greater than or equal to]25 cm dbh and [greater than]2.5 m high. A snag becomes a non-snag when it has deteriorated to a height [less than or equal to]2.5 m high. Given these assumptions, snag numbers follow the formula

[N.sub.t] = [N.sub.0][e.sup.-rt] (1)

where [N.sub.t] = number of snags at time t, [N.sub.0] = number of snags at the beginning of the time interval, and r = a constant indicating the rate of fragmentation. The value for r can be obtained if a set of snags is followed through time, recording the number of snags present in the initial sample ([N.sub.0]), the number still occurring as snags in the second sample ([N.sub.t]) and the number of years between samples (t). I record these values in Runkle (1998). The initial set of snags consisted of gapmakers measured as part of a broad overview of gap processes in the three sites (Runkle 1982). On return visits to the sites I recorded the numbers of snags that were still snags and the number that had broken off near the ground. Using these data I calculated values for the decay rate r as follows. For the southern Appalachians I obtained values of [N.sub.0] = 149, [N.sub.t] = 52, and t = 14 (Runkle 1998). Solving gives r = 0.075. For Hueston Woods 5 out of 15 snags remained as snags after 12 yr (Runkle 199 1), resulting in r = 0.092. For Tionesta 4 out of 14 snags remained as snags after 13 yr (unpublished data), resulting in r = 0.0964.

This r value was then used to calculate the survival and deterioration rates of snags. The survival rate is the fraction of snags that remain as snags after 1 yr, i.e., ([N.sub.t+1]/[N.sub.t]) = [e.sup.-r]. The deterioration rate is the fraction that becomes non-snags (broken to a height [less than or equal to]2.5 m) each year, i.e., 1 - [e.sup.-r].

I did not measure the density of snags directly. However, given a set of assumptions, I can estimate snag density from my data. I first assume that snag density is a function of snag formation from living trees and snag deterioration to stumps and fallen logs. This assumption leads to the equation

dN/dt = mT - rN (2)

where N = snag density, t = time, T = live-tree density, m = fraction (assumed constant) of live trees becoming snags per time, and r is the snag deterioration constant calculated above. Integrating and rearranging terms yields

[N.sub.t] = mT/r[1 - (1 - r[N.sub.0]/mT) [e.sup.-rt]] (3)

where [N.sub.t] = number of snags at time t, [N.sub.0] = number of snags at time 0, and other terms are defined as above. The number of snags aged [less than or equal to]t can be estimated by setting [N.sub.0] = 0, resulting in

[N.sub.t] = (mT/r)*(l - [e.sup.-rt]). (4)

Two particular values of interest are [N.sub.1], the number of snags formed per year, and [N.sub.[infinity]], the total number of snags per area of all ages.

To solve for these values I assumed that T, the number of live trees, was constant and equal to my calculated values. The value for r came from Eq. 1. I then estimated [N.sub.t], the density of snags [less than or equal to]t yr old, and solved for m. I estimated [N.sub.t] as the product of the ratio of snags! dead trees, the ratio of dead trees/total trees, and the calculated value of trees/ha. Using the calculated value of m I solved Eq. 4 for [N.sub.1] and [N.sub.[infinity]].

RESULTS

The changes in total stand density and basal area for the study areas were slight (Table 1). Density tended to decrease (southern Appalachians, Hueston Woods) and basal area to increase (southern Appalachians, Tionesta). The same trends were generally true of more homogeneous areas (Roaring Fork, Ablight Grove-Kalanu Prong, and Walker Cove) within the southern Appalachians (Table 1). For the three main study regions the largest relative changes were a 6% increase in basal area for Tionesta and a 7% decrease in density for Hueston Woods. Despite these changes, no significant (P [less than or equal to] 0.05) changes in distance from point to plant, in tree dbh (diameter at breast height), or in tree basal area occurred in any region or in the three more homogeneous areas except that in Walker Cove both dbh and basal area were significantly higher in the 1990--1991 sample.

Both mortality rates and growth rates varied with stem size (Figs. 1 and 2). In all three study areas mortality rates were higher for large stems. Hueston Woods had the highest mortality rates of all three areas for stems [greater than]40 cm dbh (Fig. 1).

The relationship between diameter growth and initial (1976-1977) diameter was not consistent for the three study areas (Fig. 2). In the southern Appalachians growth was significantly lower for stems [less than]40 cm dbh than for larger sizes (except for stems [greater than]100 cm dbh, for which a smaller sample size led to increased standard error and decreased significance). In Hueston Woods growth decreased with increased stem size but the relationship was not significant. In Tionesta growth varied significantly overall with initial diameter primarily because of high growth rates of stems 61-80 cm dbh. Overall, trees in Hueston Woods grew significantly (P [less than or equal to] 0.05) faster than trees in the other two sites.

Although large stems had the highest mortality rates in all three regions, the fraction of large stems increased (Table 2). The number in each size class was decreased by the death of stems, the replacement of living stems by small trees growing closer to the point-centered quarter points, and the growth of stems to larger size classes. The number in a given size class was increased by growth of stems from smaller size classes and by new stems that replaced either more distant living stems or dead stems. Overall, the growth of smaller stems more than compensated for the death of larger stems so that the fraction of the largest size classes increased by 16% in the southern Appalachians, 11% in Hueston Woods, and 38% in Tionesta.

Mortality rates varied by species and region (Fig. 3). Beech mortality was highest at Hueston Woods and lowest at Tionesta. Sugar maple had the opposite pattern. Other high mortality rates were for basswood and yellow birch. Low mortality rates were observed for yellow buckeye, red maple (Acer rubrum), and tulip poplar (Liriodendron tulipifera).

The average basal area of individual trees increased 3-5% for each region (Table 3). However, stand basal area remained nearly constant (southern Appalachians) or decreased (Hueston Woods) because density decreased (Table 1). Growth and new stems produced about the same basal area as was lost by death and replacement of living trees. Most individual species remained constant or increased in basal area (Table 3). Beech had the largest species declines in the southern Appalachians and Hueston Woods, largely due to high mortality rates. In Hueston Woods beech was less often a replacement tree (small value for "new" stems), especially compared to sugar maple. Most other species gained basal area. In Hueston Woods, ash and tulip tree increased in relative basal areas, primarily by rapid growth and low mortality rates. In Tionesta beech increased due to low mortality rates and high inputs of new stems.

Some species varied in their relative sizes (based on the ratio of relative dominance to relative density) among regions (Table 4). Beech was one of the smaller species in the southern Appalachians but was average-sized in Tionesta and large in Hueston Woods. Sugar maple was large in the southern Appalachians, medium in Tionesta, and small in Hueston Woods.

The three regions changed in species composition (Table 4). In the southern Appalachians beech decreased both in relative density and in relative dominance. Hueston Woods showed a larger decrease in beech and an increase in other species. Tionesta showed an increase in beech and a decrease in the other major species, hemlock. Percentage similarity values for each location, comparing the two sampling dates, were 98.4% for the southern Appalachians, 94.9% for Tionesta, and 91.0% for Hueston Woods.

Within the southern Appalachians the patterns varied among sites (Table 5). The relative sizes of the species were consistent: hemlock and sugar maple were larger than average (based on the ratio of relative dominance to relative density), silverbell and beech were smaller, buckeye and basswood were intermediate. Beech consistently decreased in importance, although less in the highest elevation site (Walker Cove). Hemlock decreased in relative dominance but increased in relative density in both sites where it was present. All other species increased.

Because Hueston Woods showed the largest changes I examined the dynamics of beech and sugar maple there in more detail (Table 6). Beech dominated the larger size classes and sugar maple the smaller size classes. The main change over time was in the midsize class (41-60 cm dbh). The death and outgrowth of beech of those sizes was higher than the input of new or ingrowth stems. However, sugar maple increased due to new stems (typically stems growing past the 25 cm dbh cut-off) and growth of the smallest size class. Beech showed a pattern suggestive of a species that had become dominant in a situation that has changed, resulting in a decrease back to a level appropriate to the new environment. Perhaps the natural or human-related environment in the 1800s was more favorable for beech than the more recent environment.

Trees in these stands usually died standing, breaking off at a variety of heights (Fig. 4). Uprooting percentage increased from Tionesta to the southern Appalachians to Hueston Woods but at most only 27% of the trees uprooted. In contrast, 16-31% of the trees produced stumps [degrees]16 m high, with Tionesta possessing the highest percentage.

Additional snag characteristics come from the snag decay equations given earlier. The fraction of snags observed in the southern Appalachians is 39% (number of snags per dead tree), using the percentage of dead point-centered quarter trees [degrees]2.5 m high (Fig. 4). The fraction of dead trees is 10.2%, calculated by dividing the total number of dead trees (138) by the total number of trees initially (1350, Table 2). Stem density is 244 trees/ha (Table 1). The product of these terms (0.39 X 0.102 X 244) = 9.7 snags/ha, formed over a 14-yr period. Thus, from Eq. 4, using the previously calculated value of the decay rate, r (=0.75; see Methods: Analytical methods, above),

[N.sub.14] = 9.7 = [m X (244)/0.0751(1 - [e.sup.-(0.0745)X14]).

Solving for m, the fraction (assumed constant) of live trees becoming snags per time, gives m = 0.00459. Using that m value gives [N.sub.1] = 1.08 snags/ha for new ([less than or equal to]1 yr) snags and [N.sub.[infinity]] = 14.9 snags/ha for snags of all ages. Further, annual transformation of snags to non-snags occurs as

1 - ([N.sub.t-1]/[N.sub.t]) = 1 - ([e.sup.-0.075]) = 1 - 0.928 = 7.2%/yr.

For Hueston Woods, r = 0.092 and the observed number of snags after 12 yr = 11.3 snags/ha (=0.56 snags/dead tree X 48 dead trees/408 trees X 171 trees/ ha). From Eq. 4, m = 0.00909, [N.sub.1] = 1.49 new snags/ha/yr, and total number of snags = 16.9 snags/ha. Also, annual snag loss = 8.8%/yr. For Tionesta, r = 0.0964 and the observed number of snags after 13 yr = 13.0 snags/ha (=0.71 snags/dead tree X 32 dead trees/386 trees X 221 trees/ha). From Eq. 4, m = 0.00794, [N.sub.1] = 1.67 new snags/ha/yr, and total number of snags = 18.2 snags/ha. Annual snag loss = 9.2%/yr.

DISCUSSION

Changes in stand structure over time

Net changes over time in the number and basal area of stems and in mortality rates, as calculated from the present study, were compared to rates for other old-growth eastern forests sampled at least twice (Table 7). I took initial and final values given in the literature for these stands and used the stem-mortality-rate formula assuming exponential decay (see Methods: Analytical methods) to calculate comparable values for the stands. Stands were judged to be old growth based on descriptions in the articles considering characteristics such as tree sizes and stand histories (Davis 1996). Also, most stands had basal areas [greater than]30 [m.sup.2]/ha, which sometimes is used as a criterion for old-growth stands in mesic habitats (Held and Winstead 1975). If possible, annual mortality was calculated based on basal-area losses: differences in lower size limits for stems sampled influence density values more than basal-area values.

Mortality rates calculated in the present study were similar to rates calculated for other old-growth sites, averaging about 1%/yr. An exception is the Hueston Woods site, whose average mortality of 1.85%/yr was high. This result supports my earlier conclusion that the Hueston Woods site is showing a readjustment of stand and species characteristics. High values also were found in two other studies of beech-sugar maple woods in Ohio, one also from Hueston Woods (Fore et al. 1997, Forrester 1998).

For almost all sites, stem density decreased and basal area increased. However, most changes in basal area were under [tilde]0.5%/yr (Fig. 5). Losses in basal area, such as occurred in Hueston Woods, or increases much greater than 0.5%/yr indicate stands not likely to be near equilibrium. These patterns showed no obvious relationship to geography or lower stem-size limits.

In my study, the net increase in basal area for the southern Appalachians of only 0.03%/yr was remarkably small. Individual stands in that region varied more but still showed less change than most other locations. The smallness of the change masked differences in stand structure, though, to fewer, larger stems (Table 1).

Despite these low average values some stands showed fairly large changes between the dates studied, because the small changes were sustained for long time periods. For example, a net increase of 0.5%/yr sustained over 50 yr produces a total increase of 28% (Parker et al. 1985). The few stands studied more than twice demonstrated that growth rates did vary from time to time. Not enough such studies were found to examine that variation, however.

Mortality and growth rates as functions of stem size

The relationship between stem size (dbh) and probability of mortality is not clear in the literature (Sheil and May 1996). Goff and West (1975) hypothesized that diameter distributions (which also depend on the relationship between growth rate and size) of old-growth stands could be generated by high mortality rates in small stems, low mortality rates in stems in the smaller canopy sizes, and then intermediate to high mortality rates in the largest stems. Several widely used models of forest dynamics (JABOWA, Botkin et al. 1972; FORET, Shugart 1984; general review by Urban and Shugart [1992]) assume that the probability of mortality increases monotonically with tree size. Harcombe (1987) found evidence that mortality decreases from small to intermediate sizes, i.e., from subcanopy to small canopy stems. He stated, however, that the data are weak and equivocal in showing a possible increase in mortality in the larger sizes.

The present study (Fig. 1) and some recent literature (Fig. 6) agree with the hypothesis of Gaff and West (1975). Mortality rates are highest in the smallest, sub-canopy size classes. Rates decrease and reach minimum values in stems of [tilde]20-50 cm dbh. Mortality then almost always increases in the larger size classes.

Fewer studies of old-growth forests have examined the relationship between stem diameter and growth rates measured as changes in dbh. Harcombe (1987) suggested that growth normally increases with stem size to intermediate sizes and then decreases in larger sizes. I found no obvious relationship between diameter and growth in the present study (Fig. 2). Runkle (1990b) and Parker et al. (1985) both found growth rates to continue increasing with stem size.

Changes in species importance over time

I compared changes in species importance calculated from my values with changes calculated from other studies involving repeat samples of old-growth eastern forests (Table 8). I used relative basal area as a measure of importance, because that value is less influenced by differences in sampling procedure than is relative density. I compared the three species most important in my studies, because they are widely distributed throughout the eastern deciduous forest: eastern hem-lock, sugar maple, and American beech.

Eastern hemlock showed no clear pattern of change in importance for the southern Appalachians, decreasing in some areas and increasing in others. Hemlock decreased in northwest Pennsylvania, but increased in New York. The hemlock woolly adelgid had not yet influenced the stands in Table 8 (SAMAB 1996).

Sugar maple increased over most of its range, sometimes rapidly, such as in Missouri, the southern Appalachians, Illinois, Indiana, Ohio, and New York. However, it decreased in northwest Pennsylvania and Vermont, as has been noted by other studies (Millers et al. 1989, McWilliams et al. 1996).

American beech, in contrast, decreased over most of its range, including the southern Appalachians, Ohio, New York, and Vermont. It increased in northwest Pennsylvania and showed various patterns in Indiana. Beech bark disease can account for the high mortality rates in New York and New England, but this disease has reached northwest Pennsylvania too recently to have affected mortality rates in the literature or present study (Millers et al. 1989, Houston 1994). The decreases in Ohio are puzzling but, as stated earlier, may represent a change in stand environment (past disturbances? more recent pollution?) from 1800 until the present. Decreases in the southern Appalachians have not been reported previously (SAMAB 1996); the significance of that decrease is unknown.

Snag formation rates

The snag densities caculated in this study (14.9, 16.9, and 18.2 snags/ha for the southern Appalachians, Hueston Woods, and Tionesta, respectively) are similar to those suggested for old-growth eastern forests in general (Parker 1989, Martin 1992, Runkle 1996) and obtained for other old-growth sites with a similar dbh cutoff: 20.8 snags/ha for a New York hemlock-beech stand (Runkle 1991), 12.2 snags/ha for a Kentucky mixed mesophytic stand (Muller and Liu 1991), and 20.4 snags/ha for a Virginia oak stand (Rosenberg et al. 1988). Studies with lower size limits for snags (e.g., [greater than or equal to]10 cm dbh) have found much higher snag densities (Spies et al. 1988, Tritton and Siccama 1990, Tyrrell and Crow 1994a). My values may be high, because I assumed that snag deterioration rates do not increase with age despite some evidence to the contrary (Harmon et al. 1986). However, my model predicts that few older snags occur, so their deterioration rates will not have a great impact on my estimates for total snag densities.

Tyrrell and Crow (1994b) studied the formation and deterioration of snags. For stems [greater than or equal to]10 cm dbh they found annual snag formation rates of 3.0 snags/ha standing dead, 1.2 snags/ha broken off, and 0.6 snags/ ha uprooted. Their relative proportions with injuries were similar for gapmakers ([greater than or equal to]25 cm dbh): 53% standing dead, 30% broken off, and 16% uprooted. My injury distribution was similar, though with more broken off and fewer standing dead. Our definitions are not exactly the same, however, so the figures cannot be compared directly. Tyrrell and Crow (1994b) also found that 1.3 out of 14.3 snags/ha collapsed at the base annually. This rate (9.1%/yr) is similar to those calculated from my data (7.2% for the southern Appalachians, 8.8% for Hueston Woods, and 9.2% for Tionesta).

Because some mortality occurs in these stands most years, snags continue to be present despite their tendency to deteriorate to smaller sizes. They thus are available, dependably, for the species that require them (Harmon et al. 1986).

(1.) E-mail: JAMES.RUNKLE@WRIGHT.EDU

ACKNOWLEDGMENTS

This research was supported by the National Science Foundation (DEB-8214774, BSR-9006779). I thank my field assistants for 1990/1991: Emily Maimon, Jovanka Kink, Debbie Premus, Amy Levi, Charles Powers, Ben Runkle, Matt Runkle, Brian Current, and, especially, Holly Proudfoot. The staffs of the field sites were very helpful. Special thanks go to GSMNP ranger Jack Campbell. I benefitted from the comments of the reviewers and the editor, James S. Clark, who also helped me develop the mathematics of snag deterioration.

LITERATURE CITED

Abrell, D. B., and M. T. Jackson. 1977. A decade of change in an old-growth beech-maple forest in Indiana. American Midland Naturalist 98:22-32.

Bell, D. T. 1997. Eighteen years of change in an Illinois streamside deciduous forest. Journal of the Torrey Botanical Society 124:174-188.

Bjorkbom, J. C., and R. G. Larson. 1977. The Tionesta scenic and research natural areas. USDA Forest Service General Technical Report NE-31.

Bormann, F. H., and G. E. Likens. 1979. Pattern and process in a forested ecosystem. Springer-Verlag, New York, New York, USA.

Botkin, D. B., J. F. Janak, and J. R. Wallis. 1972. Some ecological consequences of a computer model of forest growth. Journal of Ecology 60:849-872.

Busing, R. T. 1989. A half century of change in a Great Smoky Mountains cove forest. Bulletin of the Torrey Botanical Club 116:283-288.

Busing, R. T. 1993. Three decades of change at Albright Grove, Tennessee. Castanea 58:231-242.

Campbell, T. 1995. The community composition and mortality of woody plants in the Wright State University woods, Ohio. Thesis. Wright State University, Dayton, Ohio, USA.

Chiles, L. 1985. Woody species distribution and soil properties for the Wright State Biology Preserve. Thesis. Wright State University, Dayton, Ohio, USA.

Connell, J. H., and R. 0. Slatyer. 1977. Mechanisms of succession in natural communities and their role in community stability and organization. American Naturalist 111:1119-1144.

Cottam, G., and J. T. Curtis. 1956. The use of distance measures in phytosociological sampling. Ecology 37:451-460.

Davis, M. B. 1996. Eastern old-growth forests. Island Press, Washington, D.C., USA.

Eickmeier, W. G. 1988. Ten years of forest dynamics at Radnor Lake, Tennessee. Bulletin of the Torrey Botanical Club 115:100-107.

Fore, S. A., J. L. Vankat, and R. L. Schaefer. 1997. Temporal variation in the woody understory of an old-growth Fagus-Acer forest and implications for overstory recruitment. Journal of Vegetation Science 8:607-614.

Forrester, J. A. 1998. The structure and dynamics of an old-growth Acer-Fagus woods in the Holden Arboretum, northeastern Ohio. Thesis. Wright State University, Dayton, Ohio, USA.

Goff, F. G., and D. West. 1975. Canopy-understory interaction effects on forest population structure. Forest Science 21:98-108.

Gore, A. P., E. A. Johnson, and H. P. Lo. 1985. Estimating the time a dead tree has been on the ground. Ecology 66: 1981-1983.

Gunasekaran, M., D. J. Weber, S. Sanderson, and M. M. Devall. 1992. Reanalysis of the vegetation of Bee Branch Gorge Research Natural Area, a hemlock-beech community on the Warrior River Basin of Alabama. Castanea 57:34-45.

Harcombe, P. A. 1987. Tree life tables. BioScience 37:557-568.

Hardin, E. D., and K. P. Lewis. 1980. Vegetation analysis of Bee Branch Gorge, a hemlock-beech community on the Warrior River Basin of Alabama. Castanea 45:248-256.

Harmon, M. E., J. F. Franklin, F. J. Swanson, P. Sollins, S. V. Gregory, J. D. Lattin, N. H. Anderson, S. P. Cline, N. G. Aumen, J. R. Sedell, G. W. Lienkaemper, K. Cromack, and K. W. Cummins. 1986. Ecology of coarse woody debris in temperate ecosystems. Advances in Ecological Research 15:133-302.

Held, M. E., and J. F. Winstead. 1975. Basal area and climax status in mesic forest systems. Annuals of Botany 39:1147-1148.

Hicks, W. T. 1996. Tree mortality, coarse woody debris, and the equilibrium status of an old-growth Fagus--Acer forest in southwestern Ohio. Thesis. Miami University, Oxford, Ohio, USA.

Hoffard, W. H., D. H. Marx, and H. D. Brown. 1995. The health of southern forests. USDA Forest Service Southern Region Publication R-8 PR 27.

Hom, J., R. Birdsey, and K. O'Brian, editors. 1996. Proceedings, 1995 meeting of the Northern Global Change Program, 14--16 March 1995, Pittsburgh, Pennsylvania. USDA Forest Service General Technical Report NE-214.

Houston, D. R. 1994. Major new tree disease epidemics--beech bark disease. Annual Review of Phytopathology 32:75-87.

Kupfer, J. A., G. P. Malanson, and J. R. Runkle. 1997. Factors influencing species composition in canopy gaps: the importance of edge proximity in Hueston Woods, Ohio. Professional Geographer 49:165-178.

Kupfer, J. A., and J. R. Runkle. 1996. Early gap successional pathways in a Fagus-Acer forest preserve: pattern and determinants. Journal of Vegetation Science 7:247-256.

Little, E. L. 1979. Checklist of United States trees (native and naturalized). United States Agriculture Handbook 541.

Martin, W. H. 1992. Characteristics of old-growth mixed mesophytic forests. Natural Areas Journal 12:127-135.

McClenahen, J. R., and L. S. Dochinger. 1985. Tree ring response of white oak to climate and air pollution near the Ohio River valley. Journal of Environmental Quality 14:274-280.

McClenahen, J. R., and R. P. Long. 1993. Height growth of northern red oak in relation to site and atmospheric deposition in Pennsylvania. Environmental Pollution 80:105-114.

McWilliams, W. H., R. White, S. L. Arner, C. A. Nowak, and S. L. Stout. 1996. Characteristics of declining forest stands on the Allegheny National Forest. USDA Forest Service Northeastern Forest Experiment Station Research Note NE- 360.

Miceli, J. C., G. L. Rolfe, D. R. Pelz, and J. M. Edgington. 1977. Brownfield Woods, Illinois: woody vegetation and changes since 1960. American Midland Naturalist 98:469-476.

Millers, I., D. S. Shriner, and D. Rizzo. 1989. History of hardwood decline in the Eastern United States. USDA Forest Service General Technical Report NE-126.

Moore, M. R., and J. L. Vankat. 1986. Responses of the herb layer to the gap dynamics of a mature beech-maple forest. American Midland Naturalist 115:336-347.

Muller, R. N., and Y. Liu. 1991. Coarse woody debris in an old-growth deciduous forest on the Cumberland Plateau, southeastern Kentucky. Canadian Journal of Forest Research 21:1567-1572.

Nakashizuka, T., S. Iida, H. Tanaka, M. Shibata, S. Abe, T. Masaki, and K. Niijama. 1992. Community dynamics of Ogawa Forest Reserve, a species rich deciduous forest, central Japan. Vegetatio 103:105-112.

Oliver, C. D. 1981. Forest development in North America following major disturbances. Forest Ecology and Management 3:153-168.

Oliver, C. D., and B. C. Larson. 1990. Forest stand dynamics. McGraw-Hill, New York, New York, USA.

Pallardy, S. G., T. A. Nigh, and H. E. Garrett. 1988. Changes in forest composition in Central Missouri: 1968-1982. American Midland Naturalist 120:380-390.

Parker, G. R. 1989. Old-growth forests of the central hardwood region. Natural Areas Journal 9:5-11.

Parker, G. R., D. J. Leopold, and J. K. Eichenberger. 1985. Tree dynamics in an old-growth deciduous forest. Forest Ecology and Management 11:31-57.

Peterken, G. F, and E. W. Jones. 1987. Forty years change in Lady Park Wood: the old-growth stands. Journal of Ecology 75:477-512.

Peters, R., and T. L. Poulson. 1994. Stem growth and canopy dynamics in a worldwide range of Fagus forests. Journal of Vegetation Science 5:421-432.

Peterson, C. J., and S. T A. Pickett. 1995. Forest reorganization: a case study in an old-growth forest catastrophic blowdown. Ecology 76:763-774.

Phillips, O. L., and A. H. Gentry. 1994. Increasing turnover through time in tropical forests. Science 263:954-958.

Platt, W. J., G. W. Evans, and S. L. Rathbun. 1988. The population dynamics of a long-lived conifer (Pinus palustris). American Naturalist 131:491-525.

Pyle, C. 1988. The type and extent of anthropogenic vegetation disturbance in the Great Smoky Mountains before National Park Service acquisition. Castanea 53:225-235.

Rosenberg, D. K., J. D. Fraser, and D. F Stauffer. 1988. Use and characteristics of snags in young and old forest stands in southwest Virginia. Forest Science 34:224-228.

Runkle, J. R. 1981. Gap regeneration in some old-growth forests of the eastern United States. Ecology 62:1041-1051.

Runkle, J. R. 1982. Patterns of disturbance in some old-growth mesic forests of eastern North America. Ecology 63:1533-1546.

Runkle, J. R. 1985a. Disturbance regimes in temperate forests. Pages 17-33 in S. T. A. Pickett and P. S. White, editors. The ecology of natural disturbance and patch dynamics. Academic Press, New York, New York, USA.

Runkle, J. R. 1985b. Comparison of methods for determining fraction of land area in treefall gaps. Forest Science 13:15-19.

Runkle, J. R. 1990a. Gap dynamics in an Ohio Acer-Fagus forest and speculations on the geography of disturbance. Canadian Journal of Forest Research 20:632-641.

Runkle, J. R. 1990b. Eight years change in an old Tsuga canadensis woods affected by beech bark disease. Bulletin of the Torrey Botanical Club 117:409-419.

Runkle, J. R. 1991. Gap dynamics of old-growth eastern forests: management implications. Natural Areas Journal 11:19-25.

Runkle, J. R. 1996. Central mesophytic forests. Pages 161-177 in Mary B. Davis, editor. Eastern old-growth forests. Island Press, Washington, D.C., USA.

Runkle, J. R. 1998. Changes in southern Appalachian canopy tree gaps sampled thrice. Ecology 79:1768-1780.

Runkle, J. R., and T. C. Yetter. 1987. Treefalls revisited: gap dynamics in the southern Appalachians. Ecology 68:417-424.

SAMAB [Southern Appalachian Man and the Biosphere]. 1996. The Southern Appalachian assessment terrestrial technical report. Report 5 of 5. USDA Forest Service Southern Region, Atlanta, Georgia, USA.

SAS Institute staff. 1985. SAS user's guide: statistics, version 5 edition. SAS Institute, Cary, North Carolina, USA.

Schmelz, D. V., J. D. Barton, and A. A. Lindsey. 1975. Donaldson's Woods: two decades of change. Proceedings of the Indiana Academy of Science 84:234-243.

Sheil, D., D. F. R. P. Burslem, and D. Alder. 1995. The interpretation and misinterpretation of mortality rate measures. Journal of Ecology 83:331-333.

Sheil, D., and R. M. May. 1996. Mortality and recruitment rate evaluations in heterogeneous tropical forests. Journal of Ecology 84:91-100.

Shugart, H. H. 1984. A theory of forest dynamics. Springer-Verlag. New York, New York, USA.

Spies, T. A., J. E Franklin, and T. B. Thomas. 1988. Coarse woody debris in Douglas-fir forests of western Oregon and Washington. Ecology 69:1689-1702.

Tritton, L. M., and T. G. Siccama. 1990. What proportion of standing trees in forests of the Northeast are dead? Bulletin of the Torrey Botanical Club 117:163-166.

Tyrrell, L. E., and T. R. Crow. 1994a. Structural characteristics of old-growth hemlock-hardwood forests in relation to age. Ecology 75:370-386.

Tyrrell, L. E., and T. R. Crow. 1994b. Dynamics of dead wood in old-growth hemlock-hardwood forests of northern Wisconsin and northern Michigan. Canadian Journal of Forest Research 24:1672-1683.

Urban, D. L., and H. H. Shugart. 1992. Individual-based models of forest succession. Pages 249-292 in D. C. GlennLewin, R. K. Peet, and T. T. Veblen, editors. Plant succession: theory and prediction. Chapman & Hall, London, UK.

Vankat, J. L., W. H. Blackwell, and W. E. Hopkins. 1975. The dynamics of Hueston Woods and a review of the question of the successional status of the southern beech-maple forest. Castanea 40:290-308.

Vitousek, P. M. 1994. Beyond global warming: ecology and global change. Ecology 75:1861-1876.

Vogelmann, H. W., G. J. Badger, M. Bliss, and R. M. Klein. 1985. Forest decline on Camels Hump, Vermont. Bulletin of the Torrey Botanical Club 112:274-287.

Whitney, G. G. 1984. Fifty years of change in the arboreal vegetation of Heart's Content, an old-growth hemlock-white pine-northern hardwood stand. Ecology 65:403-408.

Whittaker, R. H. 1975. Communities and ecosystems. Second edition. Macmillan, New York, New York, USA.
               Overall basal area and density, by region and
              subregion studied in the eastern United States.
                                             Basal area
                              No. of trees ([m.sup.2]/ha)
Region and subregion           1990-1991     1976-1977    1990-1991
Southern Appalachians, SA [+]    1350           64.3        64.6
 RF                               264           71.3        72.9
 AG-KP                            365           77.1        74.3
 WC                               299           44.3        47.3
Hueston Woods (Ohio), HW          408           36.7        35.6
Tionesta (Pennsylvania), TA       386           41.4        43.9
                                      Density
                              (trees [greater than or
                              equal to]25 cm dbh/ha)
Region and subregion                 1976-1977        1990-1991
Southern Appalachians, SA [+]           244              233
 RF                                     261              271
 AG-KP                                  279              261
 WC                                     201              182
Hueston Woods (Ohio), HW                171              159
Tionesta (Pennsylvania), TA             221              221
(+.)Subregion code: RF = Roaring Fork (Tennessee),
AG-KP = Albright Grove and Kalanu Prong
(Tennessee), WC = Walker's Cove (North Carolina).
             Changes in stem size distribution from 1976-1977
                 stand samples to 1990-1991 stand samples.
              dbh        Initial [+] Died Replaced [++] Outgrowth [ss]
Region       (cm)          (n)       (%)    (%)            (%)
SA           25-40         501        8      4             21
             41-60         428        7      4             21
             61-80         248       15      4             14
            81-100         115       15      3             21
       [greater than]100    58       26      7              0
HW           25-40         154        5      6             27
             41-60         140       14      6             28
             61-80          95       31      4              9
       [greater than]80     19       42      5              0
TA           25-40         174        3      6             17
             41-60         144        8      5             15
             61-80          53       21      0             15
       [greater than]80     16       25      0              0
       New [II] Ingrow [n] Final [#]
Region (%)        (%)       (%)
SA     23          0         89
        8         25        100
       11         38        116
       11         30        103
        7         41        116
HW     35          0         97
       11         29         92
       17         41        114
       11         47        111
TA     16          0         90
       11         22        105
        2         40        106
       12         50        138


Notes: All percentages are based on the initial number. Regions are abbreviated as in Table 1. Sampling was done by the point-centered quarter method (Cottam and Curtis 1956); see Methods: Field methods for further explanation. For region code, see Table 1.

(+.)Initial number alive in 1976-1977.

(++.)Replaced = living trees no longer included in estimating density because another living tree closer to the point-centered quarter point had reached [greater than or equal to]25 cm dbh by 1990-1991.

(ss.)Outgrowth = trees leaving the given size class because they had grown into a larger size class.

(II.)New = trees that have replaced initial trees, either because the initial tree died or because the new tree is closer to the point-centered quarter point than an initial, still-living (replaced) tree.

(n.)Ingrowth = trees entering the given size class because they had grown out of a smaller size class.

(#.)Final = trees in 1990-1991 as percentage of the number of trees in the same size class in 1976-1977.
               Changes in basal area, by region and species.
                          Basal area
                            Initial      Initial     Died Replaced Growth New
Region and species        ([m.sup.2]) ([m.sup.2]/ha) (%)    (%)     (%)   (%)
Southern Appalachians       355.6         64.3        14     4       12   11
  Tsuga canadensis          105.7         19.1        15     4       10   10
  Acer saccharum             72.3         13.1        10     6       16   10
  Fagus grandifolia          40.8          7.4        23     4       12   10
  Halesia carolina           39.7          7.2        12     2       14   13
  Aesculus octandra          29.6          5.4         7     5       11    6
  Tilia heterophylla         26.0          4.7        18     2       12   22
Hueston Woods                93.3         36.7        23     5       16   15
  Fagus grandifolia          54.1         21.3        33     2        8   12
  Acer saccharum             21.7          8.5        10     5       22   21
  Fraxinus americana          8.0          3.1         3    20       30   21
  Liriodendron tulipifera     4.8          1.9         0     9       40   12
Tionesta                     72.5         41.4        14     3       13    9
  Fagus grandifolia          36.8         21.0        10     3       15   11
  Tsuga canadensis           21.3         12.2        17     2        9    8
  Acer rubrum                 5.3          3.0         4     4       20    8
  Betula allegheniensis       3.9          2.2        24     2        7    5
  Acer saccharum              3.2          1.8        62     6        1    0
                            Tree      Stand
                          final [+] final [++]
Region and species           (%)       (%)
Southern Appalachians        105       100
  Tsuga canadensis           100        95
  Acer saccharum             110       105
  Fagus grandifolia           95        91
  Halesia carolina           112       107
  Aesculus octandra          104        99
  Tilia heterophylla         114       109
Hueston Woods                103        96
  Fagus grandifolia           85        79
  Acer saccharum             128       119
  Fraxinus americana         128       119
  Liriodendron tulipifera    144       134
Tionesta                     105       105
  Fagus grandifolia          113       113
  Tsuga canadensis            98        98
  Acer rubrum                122       122
  Betula allegheniensis       87        87
  Acer saccharum              34        34
Notes: Values for "Died," "Replaced," "Growth," "New," and
"Tree final" are percentages of "Initial"; "Stand Final" is
percentage of "Initial." Terms are defined as in Table 2,
except that Growth = dbh increases of trees alive both in
1976-1977 and 1990-1991.
(+.)Based on the average basal area per tree.
(++.)Corrected for density changes and represents [m.sup.2]/ha.
              Changes in importance values of tree species in
         the three regions, based on point-centered quarter data.
                               1976-1977
                                                             RDOM/
Region     Species                 N [+] RDEN [++] RDOM [ss] RDEN
  SA   Tsuga canadensis           310    23.0      29.7      1.3
       Acer saccharum             230    17.0      20.3      1.2
       Fagus grandifolia          243    18.0      11.5      0.6
       Halesia carolina           228    16.9      11.2      0.7
       Aesculus octandra          108     8.0       8.3      1.0
       Tilia heterophylla          96     7.1       7.3      1.0
  HW   Fagus grandifolia          168    41.2      57.5      1.4
       Acer saccharum             159    39.0      24.2      0.6
       Fraxinus americana          37     9.1       7.9      0.9
       Liriodendron tulipifera     20     4.9       5.2      1.1
  TA   Fagus grandifolia          215    55.6      20.8      0.9
       Tsuga canadensis            75    19.4      29.3      1.5
       Acer rubrum                 44    11.4       7.3      0.6
       Betula allegheniensis       25     6.5       5.3      0.8
       Acer saccharum              16     4.1       4.5      1.1
       RDEN 1990-1991/ RDOM 1990-1991/
Region RDEN 1976-1977  RDOM 1976-1977
  SA        1.07           0.96
            1.00           1.04
            0.95           0.90
            1.00           1.06
            1.02           1.00
            0.99           1.08
  HW        0.86           0.82
            1.16           1.22
            1.02           1.23
            1.00           1.38
  TA        1.04           1.07
            0.94           0.94
            1.05           1.15
            0.85           0.83
            0.51           0.31
Note: Regions are defined in Table 1.
(+.)No. of individual trees.
(++.)RDEN, relative density = percentage of stand density.
(ss.)RDOM, relative dominance = percentage of stand basal area.
            Changes in importancevalues of tree species based on
              point-centered quarter data  for three selected
              subregions within the southern Appalachians, USA.
                                         1976-1977
                                                             RDOM/
Subregion [+] PS [++]      Species           N     RDEN RDOM RDEN
RF            95.7    Tsuga canadensis      108    40.9 48.8 1.2
                      Halesia carolina       62    23.5 13.5 0.6
                      Acer saccharum         32    12.1 16.0 1.3
                      Fagus grandifolia      40    15.2 11.7 0.8
AG-KP         96.4    Tsuga canadensis       66    18.1 22.9 1.3
                      Halesia carolina       73    20.0 14.2 0.7
                      Fagus grandifolia      74    20.3 13.4 0.7
                      Acer saccharum         51    14.0 16.2 1.2
                      Aesculus octandra      32     8.8 10.8 1.2
                      Tilia heterophylla     31     8.5  9.5 1.1
WC            97.8    Acer saccharum        102    34.1 47.6 1.4
                      Aesculus octandra      59    19.7 17.3 0.9
                      Fagus grandifolia      68    22.7 11.1 0.5
                      Tilia heterophylla     32    10.7 11.6 1.1
              RDEN 1990-1991/ RDOM 1990-1991/
Subregion [+] RDEN 1976-1977  RDOM 1976-1977
RF                 1.03           0.95
                   1.05           1.19
                   1.00           1.11
                   0.85           0.74
AG-KP              1.17           0.96
                   1.02           1.13
                   0.89           0.86
                   1.00           0.99
                   1.06           1.03
                   0.96           1.18
WC                 1.03           1.03
                   1.02           0.99
                   0.96           0.88
                   1.09           1.03
Note: See Table 4 forcolumn-head explanations.
(+.)RF = Roaring Fork(Tennessee), AG-KP = Albright Grove and
Kalanu Prong (Tennessee), WC = Walker's Cove (North Carolina).
(++.)PS = Percent similarity between importance values (average
of relative density and relative dominance) for the two times.
               Changes in size-class distribution for beech
               and sugar maple in Hueston Woods, Ohio, USA.
                        dbh        Initial Died Replaced Outgrowth New
Species                 (cm)         (n)   (n)    (n)       (n)    (n)
Fagus grandifolia      25-40          27    3      1         5     15
                       41-60          53   14      3        17      5
                       61-80          74   24      2         6     10
                  [greater than]80    14    8      0         0      1
                        all          168   49      6        28     31
Acer saccharum         25-40          99    1      7        26     35
                       41-60          50    3      3         6      7
                       61-80           8    4      0         0      2
                  [greater than]80     2    0      0         0      0
                        all          159    8     10        32     44
                  Ingrowth Final
Species             (n)     (n)
Fagus grandifolia    0       33
                     5       29
                    17       69
                     6       13
                    28      144
Acer saccharum       0      100
                    26       71
                     6       12
                     0        2
                    32      185
Note: Column headings are defined in Table 2.
              Net annual changes in relative basal area (RBA)
               for three widely distributed tree species in
          eastern U.S. stands sampled twice (all dates in 1990s).
                                Net annual
                                   RBA     Minimum
                                  change    (dbh)
       Site             Years      (%)      (cm)    N
Tsuga canadensis
  Alabama             1977-1989    2.34       4     149
  Tennessee and North 1976-1990   -0.84      25     310
    Carolina
  Tennessee           1962-1992    0.66       2     106
                      1935-1987    2.62       1      90
                      1935-1988   -0.50       1     125
  Pennsylvania        1977-1990   -0.49      25      75
                      1929-1978   -0.25      24     217
                      1929-1978   -0.31      24     197
  New York            1978-1986    0.21       0    1889
Acer saccharum
  Missouri            1968-1882    3.24       9      95
  Alabama             1977-1989    2.84       4      26
  Tennessee and North 1976-1990    0.31      25     230
    Carolina
  Tennessee           1935-1987    0.34       1      35
                      1935-1988    1.05       1     115
                      1976-1986    2.60      10     103
  Illinois            1977-1995    1.31       4     196
                      1925-1975    1.03      10    5217
  Indiana             1926-1976    1.68      10     481
                      1954-1974    2.31      10     572
                      1965-1975    0.96      10     278
  Ohio                1982-1993    0.38      25      47
                      1981-1988    0.79      11     196
                      1977-1991    1.42      25     159
                      1991-1997    3.11      10     214
  Pennsylvania        1977-1990   -8.59      25      16
  New York            1978-1986    0.88       0     139
  Vermont             1965-1983   -0.35       2     225
Fagus grandifolia
  Alabama             1977-1989   -1.40       4     103
  Tennessee and North 1976-1990   -0.72      25     243
    Carolina
  Tennessee           1962-1992   -1.55       2       9
  Indiana             1954-1974    1.07      10     508
                      1965-1975   -0.74      10     373
  Ohio                1981-1988   -2.12      11      57
                      1977-1991   -1.37      25     168
                      1991-1997   -3.46      10     141
  Pennsylvania        1977-1990    0.54      25     215
                      1929-1978    0.64      24     221
                      1929-1978    1.84      24      24
  New York            1978-1986   -4.08       0     327
  Vermont             1965-1983   -0.39       2     104
       Site           Reference
Tsuga canadensis
  Alabama             Hardlin and Lewis (1980), Gunasekaran et al. (1992)
  Tennessee and North present study
    Carolina
  Tennessee           Busing (1993)
                      Busing (1989): lower stand
                      Busing (1989): upper stand
  Pennsylvania        present study
                      Whitney (1984): association
                      Whitney (1984): consociation
  New York            Runkle (1990b)
Acer saccharum
  Missouri            Pallardy et al. (1988)
  Alabama             Hardin and Lewis (1980), Gunasekaran et al. (1992)
  Tennessee and North present study
    Carolina
  Tennessee           Busing (1989): lower stand
                      Busing (1989): upper stand
                      Eickmeier (1988)
  Illinois            Bell (1997)
                      Miceli et al. (1977)
  Indiana             Parker et al. (1985)
                      Schmelz et al. (1975)
                      Abrell and Jackson (1977)
  Ohio                Chiles (1985), Campbell (1995)
                      Fore et al. (1997)
                      present study
                      Forrester (1998)
  Pennsylvania        present study
  New York            Runkle (1990)
  Vermont             Vogelmann et al. (1985)
Fagus grandifolia
  Alabama             Hardin and Lewis (1980), Gunasekeran et al. (1992)
  Tennessee and North present study
    Carolina
  Tennessee           Busing (1993)
  Indiana             Schmelz et al. (1975)
                      Abrell and Jackson (1977)
  Ohio                Fore et al. (1997)
                      present study
                      Forrester (1998)
  Pennsylvania        present study
                      Whitney (1984): association
                      Whitney (1984): consociation
  New York            Runkle (1990)
  Vermont             Vogelmann et al. (1985)
Note: Sample size (N) is the number of stems in the latest sample.
COPYRIGHT 2000 Ecological Society of America
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2000 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:RUNKLE, JAMES R.
Publication:Ecology
Article Type:Statistical Data Included
Geographic Code:1U2PA
Date:Feb 1, 2000
Words:9420
Previous Article:GEOSTATISTICAL ESTIMATION VARIANCE FOR THE SPATIAL MEAN IN TWO-DIMENSIONAL SYSTEMATIC SAMPLING.
Next Article:SEEDLING DYNAMICS OVER THIRTY-TWO YEARS IN A TROPICAL RAIN FOREST TREE.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters