Printer Friendly

Analysis of forest community structure at the Margaret & Luke Pettit environmental preserve, Bartow County, Georgia.


A community analysis was performed on a 0.75 ha forested area in order to gather baseline data for a newly formed environmental preserve in Bartow County, Georgia. A total of 17 different woody species were observed. Dominant woody species were Oxydendrum arboreum (L.) DC., Quercus alba L., and Pinus echinata Miller. Two pairs of species with similar ranges and ecological characteristics were compared. The first pair of species were the broadleaf species Nyssa sylvatica Marshall and Acer rubrum L., and the second pair were the pine species Pinus taeda L. and P. echinata. Within each pair species showed opposite trends in importance values with increasing distance from a reservoir. For N. sylvatica and P. taeda, importance values increased, whereas for A. rubrum and P. echinata importance values decreased with distance from a reservoir. Pinus echinata, P. taeda, and A. rubrum had even distribution patterns, whereas N. sylvatica was randomly distributed.

Keywords: Pinus taeda, Pinus echinata, Acer rubrum, Nyssa sylvatica, community structure.



In order to monitor long-term changes in a community, it is first necessary to obtain baseline data on the structure of that community. Community structure can be examined through the determination of various attributes of its component species and their relationships to each other. Examples of such attributes include species composition, density, dominance, and spatial distribution patterns. The purpose of this study was to perform a community analysis to provide baseline data to aid future research at the Margaret and Luke Pettit Environmental Preserve, Bartow County, Georgia.

The community studied was a second growth mixed pine-hardwood forest. Attributes investigated within the sampling area were species composition, abundance, dominance, and frequency. For four of the species, we also determined importance values (as determined by relative density, dominance, and frequency), and spatial distribution patterns. The four species included two broadleaf species: Acer rubrum L. (red maple) and Nyssa sylvatica Marshall (blackgum), and two pine species: Pinus taeda L. (loblolly pine) and Pinus echinata Miller (shortleaf pine).

Previous studies have demonstrated changes in community structure in relation to environmental gradients such as changes in soil water content, nutrient levels, elevation, climate, and light. For example, Montague and Givnish (1) established that the distribution of Picea mariana (black spruce) and Larix laricina (eastern larch) followed a fertility/moisture gradient where L. laricina dominated wetter sites and P. mariana dominated drier sites. A study by Vazquez-G. and Givnish (2) found a decrease in diversity of species with increasing elevation in the Sierra de Manantlan, Mexico, however at lower elevations this diversity was spatially more patchy. More locally, a study on forest composition in the Georgia Piedmont (3) across various topographical sites (floodplain, transitional, and upland) indicated that within each site there were differences in the composition of the community in response to environmental gradients. Furthermore, results indicated that species composition of the upland sites was distinct from the other two types of sites, and that stand maturity (successional status) and nutrient status were also factors affecting community structure.

Within our study area there were several environmental gradients that may have affected community structure. For example, a change in elevation as a function of distance from a reservoir (and thus presumably a change in soil moisture) may be a factor influencing the spatial arrangement of species. In this study we used random stratified sampling to analyze change in community structure with increasing distance from a reservoir in a 0.75 ha area of the forest community.


The study took place at the Margaret and Luke Pettit Environmental Preserve located in Bartow County, GA, at 34[degrees] 05' 14" N and 84[degrees] 48' 55" W. The Preserve consists of an 18 ha second growth forest adjoining a 6.5 ha reservoir. The reservoir was created in 1976 and its long axis runs generally north to south. The specific area studied was a 0.75 ha forest located at the north end of the reservoir where the terrestrial elevation ranged from approximately 220 m to 250 m above sea level. A 150-m baseline was established on the west shore of the lake, perpendicular to the shore of the reservoir. Every 20 m along the baseline 50-m transect lines were laid down. Three 10 m X 10 m quadrats were randomly located along each transect (Fig. 1). Total area within quadrats was 0.24 ha. Within each quadrat, we identified, mapped, and measured the diameter at breast height (dbh) of all trees with a dbh greater than or equal to 2 cm.


In addition, for comparison purposes, we chose two pairs of species with similar ecological characteristics. One pair consisted of the pine species P. taeda and P. echinata, and the other pair consisted of the broadleaf species A. rubrum, and N. sylvatica. For these focus species, relative density, relative dominance, relative frequency, importance values, and spatial distribution patterns were determined. Relative density and relative dominance were calculated in each quadrat using the following formulas (4):

Relative density = [# of individuals of species A/total # of individuals of all species] X 100

Relative dominance = [total basal area of species A/total basal area of all trees] X 100

Frequency values were calculated as follows:

Frequency = [# of quadrats where species A occurs]/[total number of quadrats per transect]

Relative frequency was then calculated as follows:

Relative frequency = [frequency value of species A/total frequency value of all species] X 100

Importance values were calculated based on the following formula:

Importance value = relative density + relative dominance + relative frequency

Spatial distribution patterns were determined by nearest neighbor analysis using the T-square distance sampling method (5). The basic procedure was as follows:

1. Five points (O) along each transect were randomly identified.

2. From each point O the distance to the nearest individual (P) of the species of interest was measured (x).

3. From individual P, a line perpendicular to x was drawn.

4. The distance (y) from P to the nearest individual of the same species (Q) on the opposite side of the perpendicular line from O was measured (see Fig. 2).


In order to establish the spatial distribution of our four focus species, the T-square index of spatial pattern was calculated using the following formula, where N = sample size:

C = [[[summation].sub.i=1.sup.N] [[x.sub.i.sup.2]/([x.sub.i.sup.2] + [1/2][y.sub.i.sup.2])]]/N

Species with a random distribution show C-values close to 0.5. Values significantly lower than 0.5 indicate an even distribution, whereas values significantly greater than 0.5 are indicative of a clumped distribution. To test the significance of a departure of C from 0.5, a z-value was calculated, using (from 5):

z = [C - 0.5]/[square root of (1/(12N))]

where N = 40.


In the 50 m X 150 m area sampled, we identified a total of 17 woody species. The most dominant species in our forest were Oxydendrum arboreum (L.) DC. (15.9%), Quercus alba L. (10.5%), P. echinata (10.2%), and A. rubrum (10.2%) (Table I). The three species with the greatest densities were N. sylvatica, A. rubrum, and O. arboreum (Table II). Oxydendrum arboreum and A. rubrum also showed high relative dominance values, however N. sylvatica had a much lower relative dominance (Table I).

For comparison purposes, with the four focus species, we chose to pair the two pine species and the two broadleaf species. The importance values for both pine species increased with distance from the reservoir (Fig. 3). For P. echinata however, importance values increased from 0 m up to 100 m away from the reservoir at which point they reached their maximum. Pinus taeda on the other hand, showed importance values close to zero from 0-40 m away from the reservoir. As distance from the reservoir increased beyond 100 m, P. taeda surpassed P. echinata in importance value (Fig. 3).


Both midstory broadleaf species (A. rubrum and N. sylvatica) displayed opposite trends in relative density, dominance, and importance values. Nyssa sylvatica was absent from the transect closest to the reservoir (transect 0), but then increased in relative density (Fig. 4), and relative dominance (Fig. 5), with increasing distance from the reservoir. Acer rubrum showed an opposite trend, with relative density and dominance initially high at transect 0, then decreasing with increasing distance from the reservoir (Figs. 4 and 5). These trends are reflected in the importance values with A. rubrum having the higher importance value at transect 0, then decreasing; whereas N. sylvatica had the lowest importance value at transect 0, then increased (Fig. 6). The T-square index of spatial pattern C-values indicated that P. echinata, P. taeda, and A. rubrum had an even distribution, whereas N. sylvatica had a random distribution (Table III).





As with most Piedmont forests of the Southeastern USA, historical human disturbance has almost certainly played a role in species distributions in the forest at the Preserve. The Georgia Piedmont was rapidly converted from virgin timber stands to agricultural fields in the late 18th century (6). Since conversion, soils in the Piedmont have been subject to considerable erosion from destructive land use practices (7). As agricultural land was abandoned, it reverted through natural succession to pine forest, dominated primarily by P. taeda and P. echinata (6), and subsequently to a pine-hardwood mixed forest. Previous studies of community structure in Piedmont forests of nearby states such as Alabama have found that disturbance history, and hence successional status, is a primary determinant of forest composition (8). Selective logging and fire suppression are other factors that have undoubtedly influenced forest composition in this area over the last century (3).

In comparing community attributes of the forest at the Preserve with those of other Piedmont forests of comparable age, some interesting similarities and differences were found. The most dominant species in our forest were O. arboreum (15.9%), Q. alba (10.5%), P. echinata (10.2%), and A. rubrum (10.2%). In contrast, in upland forests of the Georgia Piedmont studied by Cowell (3), which represented a variety of ages, topographical positions, and exposures, the most dominant species was P. taeda. At the Preserve, P. echinata was the dominant pine. In both studies, Q. alba was the second most dominant species. It seems somewhat surprising that we should have found O. arboreum to be the most dominant species in our forest, given that in the forests studied by Cowell (3) it had a very low relative dominance. This species was historically used in Southern Appalachia for honey production (9), however, at this time, we are unsure of the factors contributing to its high dominance at the Preserve.

When relative dominance was compared at the genus level, the three most dominant genera in our forest were Quercus (43.7%), Pinus (19.2%), and Oxydendrum (15.9%), while in Cowell's study (3) they were Pinus (20.8%), Quercus (34.3%), and Liriodendron (9.7%). The relative dominance of Pinus was very similar in both studies, however Quercus was more dominant at the Preserve.

Nyssa sylvatica had by far the greatest relative density at the Preserve, probably because the N. sylvatica population consisted of many young individuals. When abundances of genera in the forest at the Preserve were compared to historical abundances in Georgia based on original survey maps (10), we found them to be relatively similar. Plummer (10) found Quercus, Pinus, and Carya to occur in the proportions 53:23:8 in early Piedmont forests. The proportions we found for Quercus, Pinus, and Carya in the Preserve were similar (42:20:5).

The two species within each pair of our four focus species had very strong similarities to each other in terms of range, habitat, and shade tolerance. Both of the Pinus species range across the southeastern USA from Texas up to Maryland. They are both shade intolerant, and invaders of abandoned fields. The two broadleaf species (A. rubrum and N. sylvatica) have even more extensive ranges, covering most of the eastern USA (9). They are both midstory trees, shade tolerant and can be found on dry upland sites as well as moist sites. Therefore, the fact that within each pair there are opposing patterns in importance with distance from the reservoir, as well as different patterns of distribution within the broadleaf pair, does not appear to be easily explained by habitat preferences. Given the similarity of the two species in each pair, the opposing patterns may be the result of resource partitioning among these species. It is also possible that the soil is less homogenous than we assumed, and that differences in pH and nutrients may help explain the distributions observed. In addition to successional status, Cowell found that the composition of upland Piedmont forest sites was strongly related to nutrients (3). While species composition of our site was generally consistent with a nutrient poor status (3), soil variables were not analyzed. Future forest community studies at the Preserve should examine nutrients and other soil variables to further address this issue.

In addition to the results presented, we also mapped the location of each tree with a dbh [greater than or equal to]2 cm (data not shown). This information, in addition to the information presented on overall community structure will be useful to future studies at the Preserve, as well as to other studies of forest communities. Because the Preserve is located in a county currently undergoing rapid development, there is significant need for research on change in forest community attributes relative to increasing development in the surrounding area.
Table I. Relative dominance for 17 species sampled at the Margaret &
Luke Pettit Environmental Preserve. Species are ranked from highest to
lowest values of relative dominance.

Species Relative Dominance (%)

Oxydendrum arboreum (L.) DC. 15.9
Quercus alba L. 10.5
Acer rubrum L. 10.2
Pinus echinata Miller 10.2
Quercus marilandica Muenchh. 9.2
Pinus taeda L. 9.0
Quercus falcata Michaux 8.7
Quercus stellata Wangenh. 7.0
Quercus rubra L. 7.0
Carya glabra (Miller) Sweet 5.6
Nyssa sylvatica Marshall 3.9
Quercus sp.* 1.4
Carya tomentosa (Poiret) Nutt. 0.6
Liriodendron tulipifera L. 0.2
Prunus serotina Ehrh. 0.1
Vaccinium arboreum Marshall 0.02
Fagus grandifolia Ehrh. <0.01

(*) unidentified oaks

Table II. Relative density for 17 species sampled at the Margaret &
Luke Pettit Environmental Preserve. Species are ranked from highest to
lowest values of relative density.

Species Relative Density (%)

Nyssa sylvatica Marshall 43.1
Acer rubrum L. 12.4
Oxydendrum arboreum (L.) DC. 9.0
Carya glabra (Miller) Sweet 6.4
Quercus marilandica Muenchh 4.4
Pinus echinata Miller 4.1
Quercus alba L. 3.5
Quercus stellata Wangenh 3.5
Pinus taeda L. 2.4
Quercus falcata Michaux 2.2
Quercus sp. * 1.5
Carya tomentosa (Poiret) Nutt. 1.4
Quercus rubra L. 1.3
Prunus serotina Ehrh. 0.7
Liriodendron tulipifera L. 0.3
Fagus grandifolia Ehrh. 0.3
Vaccinium arboreum Marshall 0.2

(*) unidentified oaks

Table III. Results from the T-square index of spatial pattern test for
four focus species. C-values significantly less than 0.5 indicate an
even pattern of distribution; values ~0.5 indicate a random pattern;
whereas values significantly greater than 0.5 indicate a clumped pattern

Species Common Name C-Value Pattern p-value

Pinus echinata Miller Shortleaf Pine 0.36* Even 0.002
Pinus taeda L. Loblolly Pine 0.41* Even 0.049
Acer rubrum L. Red Maple 0.38* Even 0.009
Nyssa sylvatica Marshall Blackgum 0.44 Random 0.19

(*) indicate values significantly different from 0.5 (random); N = 40.


1. Montague TG and Givnish TJ: Distribution of black spruce versus eastern larch along peatland gradients: relationship to relative stature, growth rate, and shade tolerance. Can J Bot 74: 1514-1532, 1996.

2. Vazquez-G. JA and Givnish TJ: Altitudinal gradients in tropical forest composition, structure, and diversity in the Sierra de Manantlan. J Ecol 6: 999-1020, 1988.

3. Cowell CM: Environmental gradients in secondary forests of the Georgia Piedmont, U.S.A. J Biogeog 20: 199-207, 1993.

4. Cox GW: "Laboratory Manual of General Ecology." Boston: W C Brown Publishers, p90, 1996.

5. Ludwig JA and Reynolds JF: "Statistical Ecology--A Primer on Methods and Computing." New York: J Wiley & Sons, p55-58, 1988.

6. Odum EP and Turner MG: "The Georgia Landscape: A Changing Resource." Institute of Ecology, U of Georgia, Athens, GA, 1987.

7. Trimble SW: "Man-induced soil erosion on the Southern Piedmont, 1700-1970." Soil Conserv Soc Am, Ankeny, IA, 1974.

8. Golden MS: Forest vegetation of the lower Alabama Piedmont. Ecol 60: 770-782. 1979.

9. Brown CL and Kirkman LK: "Trees of Georgia and Adjacent States." Portland, Oregon, Timber Press, 1990.

10. Plummer GL: 18th century forests in Georgia. Bull Ga Acad Sci 33: 1-9, 1975.

Shannon B. Cutler

Eric M. Johnson

Heather D. Sutton

Paula C. Jackson*

Kennesaw State University, College of Science and Mathematics, Department of Biological and Physical Sciences, 1000 Chastain Rd., Kennesaw, GA, 30144-5591

* corresponding author and faculty supervisor e-mail:
COPYRIGHT 2003 Georgia Academy of Science
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2003 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Student Paper
Author:Cutler, Shannon B.; Johnson, Eric M.; Sutton, Heather D.; Jackson, Paula C.
Publication:Georgia Journal of Science
Date:Dec 22, 2003
Previous Article:Saturn in 2002-03.
Next Article:Size and age class estimates of North American Eocene palaeopheid snakes.

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