Relationships between plant biomass and cover in the ground cover layer of longleaf 
pine forests at Fort Benning, GA 
 
by 
 
Lorenzo Ferrari 
 
 
 
 
A thesis submitted to the Graduate Faculty of 
Auburn University 
in partial fulfillment of the 
requirements for the Degree of 
Master of Science 
 
Auburn, Alabama 
December 14, 2013 
 
 
 
 
Keywords: Pinus palustris, stand structure, basal area, growth form 
 
 
Copyright 2013 by Lorenzo Ferrari 
 
 
Approved by 
 
Lisa J. Samuelson, Chair, Professor, School of Forestry and Wildlife Sciences 
Edward F. Loewenstein, Associate Dean, School of Forestry and Wildlife Sciences 
Robert S. Boyd, Professor, Biological Sciences 
Dale G. Brockway, Affiliate Associate Professor, USDA Forest Service  
Greg L. Somers, Professor Emeritus, School of Forestry and Wildlife Sciences 
 
 
 
 
 
 
ii 
 
Abstract 
 
 
 Interest in restoring longleaf pine (Pinus palustris Mill.) ecosystems has recently 
grown but little information is available for estimating biomass in the highly diverse 
ground cover layer of longleaf pine forests. Aboveground biomass-cover relationships in 
the ground cover were examined by growth form (shrubs/tree seedlings, vines, 
graminoids, legumes, forbs, ferns) in five longleaf pine stands ranging in age (5, 12, 21, 
64, and 87 years) and forest structure at Fort Benning, GA. Cover was visually estimated 
and live biomass was determined through destructive harvests. Total live biomass in the 
ground cover layer ranged from 28 to 171 g m-2. Linear relationships were observed but 
different models were required for different growth forms and stands. The increase in 
biomass with increasing cover was greatest in the youngest stand for all growth forms 
except shrubs/tree seedlings, where the slope coefficient was highest for the 12-year-old 
stand. For forbs, the two oldest stands demonstrated a greater increase in biomass with 
increasing cover than the 12- and 21-year-old plantations. Live, herbaceous, graminoid, 
and dead biomass decreased linearly with increasing basal area. Results suggest that 
percent cover can be used to obtain a rapid estimate of biomass in the ground cover layer 
of longleaf pine forests. 
 
 
 
 
iii 
 
Acknowledgments 
 
 
 The author would first like to thank my advisor Dr. Lisa Samuelson for her 
guidance and instruction throughout this process. Thanks also to my committee members 
Dr. Robert Boyd, Dr. Dale Brockway, Dr. Ed Loewenstein, and Dr. Greg Somers for 
their time and assistance. The author would also like to thank Tom Stokes and Althea 
ArchMiller for their advice along the way. Thanks to all people who helped in the field 
and laboratory. Special thanks to my beautiful wife Katy, my daughter Evie, and to my 
fellow graduate students for their support and motivation.  
 
  
iv 
 
Computer Software Used: 
 SAS version 9.2/9.3 
Microsoft Word 2010 
Microsoft Excel 2010 
Sigma Plot 10.0 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
v 
 
Table of Contents 
 
 
Abstract ............................................................................................................................... ii 
Acknowledgments.............................................................................................................. iii  
List of Software Used ........................................................................................................ iv  
List of Tables ..................................................................................................................... vi  
List of Figures .................................................................................................................. viii 
Chapter 1 Introduction   
 1.1 Background ........................................................................................................1 
 1.1 Objectives ..........................................................................................................3 
Chapter 2 Relationships between plant biomass and cover in the ground cover layer of 
longleaf pine forests at Fort Benning, GA. 
 
 2.1 Introduction ........................................................................................................4 
 2.2 Materials and Methods .......................................................................................9 
 2.3 Results ..............................................................................................................19 
 2.4 Discussion ........................................................................................................30 
Chapter 3 Conclusions .......................................................................................................36 
References ..........................................................................................................................37 
 
 
 
 
 
 
 
vi 
 
List of Tables 
 
 
Table 2.2.1. Structure of longleaf pine (LLP) stands sampled at Fort Benning, GA. 
Overstory1 includes all woody stems ?10 cm dbh and Overstory2 is woody 
stems > 2 m height and < 10 cm dbh. The 5-year-old stand understory 
includes all planted longleaf < 2 m height and is included only for the 5-year-
old stand to account for all planted trees. .......................................................15 
 
Table 2.2.2. Ground cover plant species in longleaf pine stands at Fort Benning, GA .....16 
 
Table 2.3.1. Observed probability values for the effect of stand age on groundcover layer 
variables in longleaf pine stands at Fort Benning, GA. Total live and dead 
biomass is the sum of biomass of all growth forms within each 1-m2 plot ....19 
 
Table 2.3.2. Slopes of the relationship (y= ?1x+?) between biomass (g m-2) and ground 
cover (%) by growth form in longleaf pine stands varying in age at Fort 
Benning, GA ...................................................................................................23 
 
Table 2.3.3. Relationship between biomass and percent cover in the ground cover layer in 
longleaf pine stands at Fort Benning, GA. Likelihood ratio tests were used to 
test reduced models against the full five-parameter model by different growth 
forms. G is the test statistic for the likelihood ratio test. Non-significant p-
values at ?=0.05 indicate that the reduced model is as appropriate as the full 
model. .............................................................................................................24 
 
Table 2.3.4. Regression functions for relationships between ground cover biomass (g m-2) 
variables and total basal area (m2 ha-1) of woody stems > 2m in height in 
longleaf pine stands at Fort Benning, GA (?=0.05). Herbaceous biomass is 
total live biomass without shrub/tree seedling biomass and vine biomass. 
Regression coefficients are indicated by a and b. The number of observations 
is 20 for all models .........................................................................................27 
 
 
 
 
 
 
 
 
 
vii 
 
List of Figures 
 
 
Figure 2.2.1. Representative schematic of the 1-ha field plot sampling structure for each 
longleaf pine stand at Fort Benning, GA. Within each subplot five circular 
1-m2 plots were randomly selected. ..............................................................14 
 
Figure 2.3.1. Mean (SE) cover by growth form in the ground cover layer in longleaf pine 
stands at Fort Benning, GA. Standard errors represent variability among 
subplots. Within each growth form, different letters indicate significant 
differences in least square means among stands at ?=0.05...........................20 
 
Figure 2.3.2. Mean (SE) total live biomass (sum of all growth forms) and dead biomass in 
the ground cover layer in longleaf pine stands at Fort Benning, GA. 
Standard errors represent variability among subplots. Different capital 
letters, for live biomass, and different lowercase letters, for dead biomass, 
indicate significant differences in least square means between stands at 
?=0.05 ...........................................................................................................21 
  
Figure 2.3.3. Live biomass by growth form in the ground cover layer in longleaf pine 
stands at Fort Benning, GA. Standard errors represent variability among 
subplots. Within each growth form, different letters indicate significant 
differences in least square means of biomass between stands at ?=0.05 ......22 
 
Figure 2.3.4. Biomass of shrubs/tree seedlings, vines, and graminoids in relation to cover 
in the ground cover layer in longleaf pine stands at Fort Benning, GA .......25 
 
Figure 2.3.5. Biomass of legumes and forbs in relation to cover in the ground cover layer 
in longleaf pine stands at Fort Benning, GA.................................................26 
 
Figure 2.3.6. Total live biomass (sum of all growth forms) and dead biomass in the 
ground cover layer as a function of basal area of all woody stems ? 2 m in 
height in longleaf pine stands at Fort Benning, GA ......................................28 
  
Figure 2.3.7. Herbaceous biomass (total of graminoids, legumes, forbs, and ferns) and 
graminoid biomass in the ground cover layer as a function of basal area of 
all woody stems ? 2 m in height in longleaf pine stands at Fort Benning, GA 
  ......................................................................................................................29 
 
  
1 
CHAPTER 1. INTRODUCTION 
 
1.1. Background 
Longleaf pine occurs over the large physiographic region of the Atlantic and Gulf 
Coastal Plain from southeastern Virginia to central Florida and west to east Texas (Ware 
et al., 1993). Woodlands, savannas and other ecosystems dominated by longleaf pine 
occupied approximately 37 million ha in the southeastern United States prior to European 
settlement (Frost, 1993). Since then, these communities have undergone a reduction in 
occurrence of longleaf pine to less than 3% of their overall original extent (Frost, 2006) 
due to logging, production of turpentine, interruption of natural fire regimes, and 
conversion of longleaf pine forests to crops, pasture or other southern pine species (Noss, 
1989; Frost, 1993; Van Lear et al., 2005). 
Longleaf pine forests are important ecosystems in terms of biological diversity 
(Sorrie and Weakley, 2001) and provide an essential habitat for sensitive animal and 
plant species (Van Lear et al., 2005). Longleaf pine ecosystems constitute a critical 
environment for federally listed animal species such as the red-cockaded woodpecker 
(Picoides borealis Vieillot). On frequently burned sites, plant species richness in the 
understory is among the highest in the world (Peet and Allard, 1993). Moreover, nearly 
60% of the vascular plant taxa endemic to the entire Coastal Plain are obligate associates 
of the longleaf pine systems and occur almost only within the range of longleaf pine 
(Sorrie and Weakley, 2006). The loss of longleaf pine forests and the decline of many 
2 
species that are found primarily in these habitats have recently led to increasing 
interest in restoring longleaf pine ecosystems. Furthermore, longleaf pine forests are 
considered ecosystems that can contribute to minimizing the effects of climate change 
and increasing carbon sequestration in the southern U.S. because of longleaf pine?s 
higher longevity and longer rotations, wider ecological amplitude, and greater resistance 
to diseases, insects, and wind damage compared to other southern pine species (Stainback 
and Alavalapati, 2004; Stanturf et al., 2007; Johnsen et al., 2009).  
The largest blocks of longleaf pine communities in the Coastal Plain remain on 
federal lands, principally National Forests, military installations, and National Wildlife 
Refuges (Sorrie and Weakley 2006). The Department of Defense is one of the largest 
federal land managers in the U.S. and is in charge of managing longleaf pine forests and 
protecting listed species. The Strategic Environmental Research and Development 
Program (SERDP) is the means through which the Department of Defense addresses 
environmental concerns. SERDP?s objective is to improve military readiness and to 
manage the environment of military installations by supporting the development of new 
scientific knowledge and cost-effective technologies in areas such as resource 
conservation and climate change. This study was part of a project funded by the 
Department of Defense through the SERDP program. The goal of the overall project, 
called ?Developing Tools for Ecological Forestry and Carbon Management in Longleaf 
Pine?, was to create two carbon management models (an even-aged management model 
and an uneven-aged management model) and provide managers with the necessary 
knowledge to balance military training objectives with the maintenance of native 
3 
biodiversity, sustainable yield of forest products and enhancement of forest carbon 
sequestration. 
1.2 Objectives  
The primary objective of this study was to investigate relationships between 
biomass, percent cover and forest structure in the ground cover layer of longleaf pine 
forests varying in age and stand structure in order to better understand forest carbon 
balance and provide land managers in military installations with tools to efficiently 
estimate carbon pools in the ground cover layer of longleaf pine ecosystems. Biomass 
and carbon pool data will be used to develop even-aged and uneven-aged longleaf pine 
ecosystem models and information about the relationships between ground cover 
biomass, percent cover, and stand structure will be incorporated into the models to better 
describe important ecosystem processes. Five longleaf pine stands of the ages of 5, 12, 
21, 64, and 87 years with different stand structures were selected for this study. The three 
youngest stands were plantations while the two oldest stands were naturally regenerated. 
Specific objectives were to:  
1. Examine the relationship between percent cover and biomass in the ground cover 
layer by growth form (shrubs/tree seedlings, vines, graminoids, legumes, forbs, and 
ferns) in order to determine whether the relationship varied among stands and 
whether percent cover can be used to estimate ground cover biomass in longleaf pine 
forests, and   
2. Develop relationships between ground cover biomass and forest structure in order to 
assess whether biomass in the ground cover can be estimated by basic forest indices, 
such as basal area. 
4 
CHAPTER 2. RELATIONSHIPS BETWEEN BIOMASS AND COVER IN THE 
GROUND COVER LAYER OF LONGEAF PINE FORESTS AT FORT 
BENNING, GA 
 
2.1 - Introduction 
Though the ground cover normally constitutes less than 1% of the total aboveground 
forest biomass in the Northern Hemisphere (Gilliam, 2007; Muller, 2003), this layer is an 
ecologically important component in forest ecosystems (Augusto et al., 2003; Gilliam, 
2007). In most temperate forests, the ground cover is a species-rich stratum (Whigham, 
2004) and it is two to 10 times more diverse than the overstory (Gilliam, 2007). In 
southern temperate zone forests, while understory plants and herbs are on average 4% of 
the forest community biomass, the herbaceous litter can account for approximately 9% of 
the annual foliar litter fall (DeAngelis et al., 1981; Muller, 2003). Furthermore, the high 
proportion of nutrient uptake and recycling in this layer is important to overall forest 
productivity (Yarie, 1980; Moore et al., 2007). The ground cover layer is generally a 
small fraction of the overall forest biomass in many forest ecosystems (Peichl and Arain 
2006; Cao et al., 2012) and therefore it has received little attention in carbon (Nabuurs et 
al., 2003) and forest process modeling (Loudermilk et al., 2011). However, values of 
ground cover biomass reported by some studies in temperate forests suggest that the 
aboveground portion in the ground cover may not be negligible during early successional 
5 
stages when stands still lack a well-developed tree canopy and in low-density forests, 
such as longleaf pine (Pinus palustris Mill.) forests (Litton et al., 2004; Lavoie et al., 
2010).  
Longleaf pine forests, which were once an important ecosystem in the Southeast, are 
now the focus of restoration (Brockway et al., 1998, 2005; Harrington et al., 2003; 
Aschenbach et al., 2010. These provide ecosystem services (e.g. biodiversity) and refugia 
for rare species, and also there is the opportunity to use longleaf pine which is a long-
lived species (up to 400 years) in carbon sequestration projects. Naturally developed 
longleaf pine ecosystems are characterized by a dense and diverse ground cover layer 
(Kirkman et al., 2001; Peet, 2006). Mitchell et al. (1999) determined that across a soil 
drainage gradient of longleaf pine-wiregrass sites, net primary productivity of the 
understory (all herbaceous vegetation and woody plants with basal diameter of 1 cm or 
smaller) was on average 68% of total aboveground productivity. Plant species richness in 
longleaf pine forests can be considerable, especially at the small scale, with up to 42 
species per 0.25 m2 (Walker and Peet, 1984; Drew et al., 1998), and species associated 
with longleaf pine forests make up the majority of the endemic plant taxa of the Coastal 
Plain (Sorrie and Weakley, 2006). The ground layer of a longleaf pine forest is 
particularly well-developed in open woodlands, where longleaf pine tree crowns in the 
overstory typically do not overlap (Brockway and Outcalt, 1998). Frequent burning 
increases cover of grasses and forbs (Glitzenstein et al., 2003; Brockway et al., 2009; 
Outcalt and Brockway, 2010) and keeps shrubs and woody understory species from 
increasing in dominance (Outcalt and Brockway, 2010). In the absence of a hardwood 
midstory, the open canopy of longleaf pine forests allows a substantial amount of light to 
6 
reach the ground cover and therefore vegetation abundance is higher in this layer than in 
closed canopy forests (Harrington and Edwards, 1999; McGuire et al., 2001; Pecot et al., 
2007). Given the renewed interest in longleaf pine restoration by land managers and the 
opportunities that these forests may offer to sequester carbon, more information on the 
proportion of forest biomass stored in the ground cover of these ecosystems is needed. 
Biomass is used to quantitatively describe ground cover vegetation and is considered 
the species variable that best reflects differences in community dominance and diversity 
(Wilson, 1991; Guo and Rundel, 1997; Chiarucci et al., 1999). However, collecting 
biomass data through direct destructive methods is labor intensive and time-consuming; 
therefore, indirect methods have been sought to estimate biomass in ground cover. Cover 
is suitable to predict biomass because cover is repeatable, can assess different plant life 
forms in comparable terms, and is more closely related to biomass than are other 
measures of abundance, such as density and frequency (Muller-Dombois and Ellemberg, 
1974). A visual estimate of cover is a popular method to predict biomass since it only 
requires a small set of field tools (Daubenmire, 1959; Peet et al., 1998). In addition, the 
visual estimate is considered reliable (Br?kenhielm and Qinghong, 1995; Vanha-
Majamaa et al., 2000), is precise in estimating cover of growth forms in the understory, 
and requires reasonable sample sizes (3-36 replicates for growth form groups) 
(Abrahamnson et al., 2011). Significant linear and nonlinear relationships between 
biomass and cover have been demonstrated in different ecosystems at the level of species 
(Ohmann et al., 1981; Alaback, 1986; Halpern et al., 1996; Halpern and Lutz, 2013), life 
form or growth form (Mitchell et al., 1987), and understory layer (Joyce and Mitchell, 
1989; Gilliam and Turrill, 1993). For example, allometric regressions for the prediction 
7 
of ground cover biomass based on percent cover have been developed for understory 
growth forms in boreal forests (MacDonald et al., 2012; Muukkonen et al., 2006), 
southern Appalachian spruce-fir forests (Moore et al., 2007), and ponderosa pine forests 
(Pinus ponderosa Douglas ex Lawson & C. Lawson) (Rose and Eddleman, 1994; 
Mitchell et al., 1987). In the southern U.S., relationships between biomass and cover have 
been studied for some broad forest types, such as planted mixed pine, natural mixed pine, 
oak-pine, and upland hardwoods (Joyce and Mitchell, 1989). Linear relationships 
between cover and biomass were reported for three-year-old longleaf pine plantations on 
hydric soils in North Carolina (Walker and Cohen, 2009) and were described by 
logarithmic functions. However, there is no study we are aware of that has investigated 
biomass-cover relationships in the ground cover layer in longleaf pine dominated forests 
varying in age. As a faster and non-destructive method of estimating ground cover 
biomass, cover analysis may be better suited to land managers who are interested in 
managing for biodiversity and carbon in longleaf pine forests.  
The biomass versus cover relationship is influenced by plant traits and characteristics 
such as plant physiognomy, carbon content, and biomass distribution, therefore it varies 
between growth forms but it is likely to be comparable among species of the same growth 
form (R?ttgermann et al., 2000; Diaz and Cabido, 1997; Chapin, 1993; Frank and 
McNaughton, 1990; Hermy, 1988). For example, biomass per unit of cover is higher in 
woody than in herbaceous growth forms. The biomass-cover relationship in the ground 
layer is expected to be linear whenever ground cover height and composition are similar 
(Hermy, 1988; Muukkonen et al., 2006). In general, growth forms with multilayered 
canopies accumulate more biomass per unit of cover than prostrate monolayer growth 
8 
forms (Alaback, 1986). The ratio of biomass to cover can decrease with decreasing light 
availability, because plants living under dense overstory canopies show higher specific 
leaf area, higher leaf area ratio, lower leaf thickness, and reduced branching compared to 
plants growing in open vegetation (Hutchings and de Kroon, 1994; Pearcy, 1999). 
Moreover, where light is a limiting factor and ground cover vegetation is sparse or less 
stratified, such as in high-density forests, plants may tend to explore space more with 
plagiotropic (horizontal) stems rather than by vertical growth, thus decreasing the 
biomass-cover ratio (Hutchings and de Kroon, 1994; Liira et al., 2002). 
The objective of this study was to examine the relationship between biomass and 
percent cover in the ground cover layer of longleaf pine forests varying in age and stand 
structure. To achieve this objective, percent cover was estimated and biomass was 
harvested by growth form (shrubs and tree seedlings, vines, graminoids, legumes, forbs, 
and ferns) in five stands dominated by longleaf pine at Fort Benning, GA. Stands varied 
in age from 5 to 87 years. We tested the hypothesis that ground cover biomass will be 
linearly related to cover in each stand and for each growth form. We expected to find 
linearity between biomass and cover because: i) plant height and composition of each 
growth form were probably homogeneous within stands; and ii) data were collected in 
Spring and the prediction of biomass by means of cover has been shown to be more 
accurate when not at the peak of standing crop (Hermy, 1988). We tested the null 
hypothesis that the biomass-cover relationship of each growth form will be the same for 
all stands in each growth form. Lastly, we expected biomass will decline with increasing 
basal area. 
 
9 
2.2 ? Materials and Methods 
Study Site 
The study was conducted at the Fort Benning Army Base, on the Coastal Plain-
Piedmont Fall-Line of west-central Georgia and eastern Alabama. Fort Benning is a 
United States military installation serving as a basic training site for infantry. The 
majority of the soil on the installation is represented by highly weathered Ultisols, 
originated by Coastal Plain material and few alluvial deposits from the Piedmont (Garten, 
2006). Average annual precipitation is 1180 mm, and temperatures range from an average 
minimum of 12.8?C to an average maximum of 24.6?C, with an average annual 
temperature of 18.7?C (data from the Columbus Airport weather station, for the years 
1982-2011; http://www.ncdc.noaa.gov). Soil series were: Nankin sandy clay loam (5-
year-old stand), Nankin sandy loam (12-year-old), Troup loamy sand (21- and 87-year-
old), and Troup Springhill Luverne Complex (64-year-old) (USDA NRCS, 1997, 1999). 
Fort Benning is located in the central-interior portion of the range of longleaf pine, 
but is outside the range of wiregrass. Five different age longleaf pine stands were 
selected. Stand ages were: 5, 12, 21, 64, and 87 years. The three youngest stands were 
plantations, whereas the 87-year-old and the 64-year-old stands were natural. Planting 
density was 1494 trees ha-1 in the 5-year-old stand, 1494 trees ha-1 in the 12-year-old 
stand, and 2240 trees ha-1 in the 21-year-old stand. Basal area of all overstory trees ranged 
from 0.5 m? ha-1 in the 5-year-old stand to 22.4 m? ha-1 in the 21-year-old stand. Total 
overstory density also was highest in the 21-year-old stand at 1982 trees ha-1 and lowest 
in the 5-year-old stand at 214 trees ha-1. Understory density varied between 500 trees ha-1 
in the 21-year-old stand to 2500 trees ha-1 in the 12-year-old stand. All stands were last 
burned in winter or early spring 2010. The actual burning history is unknown therefore, 
10 
variation in burn history may confound the results, in addition to variation in stand 
structure among stands. 
Study Design 
Research plots 1-ha in size were selected. In each research plot, four 0.04-ha circular 
subplots were established (Figure 2.2.1). Overstory 1 was defined as all woody stems ?10 
cm dbh (main trees) and Overstory 2 was woody stems > 2 m height and < 10 cm dbh. 
Understory was all stems between 1 and 2 m height, except in the 5-year-old stand where 
understory included all planted longleaf pine < 2 m height. Ground cover was defined as 
trees and shrubs < 1 m in height and all herbaceous species. Five 1-m2 circular plots were 
located within each 0.04-ha subplot for a total of 20 1-m2 sample plots per stand. In order 
to avoid over-sampling the inner one-half of the plot area, the stratified-random polar 
coordinates method described by Gaiser (1951) was used. Cover was defined as the 
percentage of ground surface obscured by the vertical projection of all live aboveground 
parts of a given growth form onto that surface. Relatively small gaps between plant live 
material were ignored. Cover classes followed the ten-point scale of Peet et al. (1998): 1 
= trace (< 0.1%), 2 = 0-1%, 3 = 1-2%, 4 = 2-5%, 5 = 5-10%, 6 = 10-25%, 7 = 25-50%, 8 
= 50-75%, 9 = 75-95%, 10 = > 95%. 
A stand inventory was conducted in February 2012 (Table 2.2.1). Cover and biomass 
data were collected during May and June 2012. Cover was estimated for six plant growth 
forms: (1) tree seedlings and shrubs < 1 meter in height, (2) vines, (3) graminoids, (4) 
legumes, (5) forbs, and (6) ferns. Although no growth form category may exceed 100% 
cover, total percent cover estimates across all categories exceeded 100% when multiple 
overlapping plants occurred. After visual estimation of cover, plant biomass was 
11 
harvested by hand clipping all plants at the root collar. Plant biomass was sorted by 
growth form and live biomass was separated from dead biomass. Plant material was then 
dried at 70?C for 72 h and weighed. Plants were identified at the species level when 
possible. Some species could only be identified at the genus level. A total of 73 plant taxa 
was encountered, 27 woody and 46 herbaceous (Table 2.2.2). Sixty three taxa were 
identified at the species level and 10 taxa only at the genus level.  
Statistical methods 
Effects of stand age on cover and biomass 
Average cover and biomass of each growth form, total biomass, and dead standing 
biomass were determined for each stand age. Data from 1-m2 plots were averaged by 
subplot and used to calculate an overall mean and standard error by stand age. 
The influence of stand age on mean cover and biomass was examined using a mixed 
linear model. The model was in the following form: 
y = X? + Z? + ?         (3) 
where ? is the fixed-effect parameter and ? the random parameter. Stand age was the 
fixed-effect parameter and subplot was the random parameter. Total biomass was defined 
as the sum of each growth form?s biomass in the 1-m2 plots. The mean cover for each 
growth form was the mid-point of each cover class (Peet et al., 1998). Midpoints of cover 
classes were 0.05, 0.5, 1.5, 3.5, 7.5, 17.5, 37.5, 62.5, 85, and 97.5. The Bonferroni 
adjustment was applied since after fitting the model residuals showed heteroscedasticity.  
Cover-biomass relationships by growth form  
The relationship between biomass and cover was determined by pooling all stands by 
each growth form using data from the 1-m2 plots. Cover values were the same cover class 
12 
midpoints that were used to test the effect of stand age on cover. Regressions were forced 
through 0 (model without intercept) because this regression function is known to pass 
through it (biomass = 0 g m-2 will always correspond to cover = 0%). Simple linear 
regressions run separately for each stand and that included an intercept showed that the 
intercept did not differ from zero in almost all cases. Data points with very unusual 
relative proportion of growth forms and that produced a change of more than 20% on the 
regression slope were considered outliers, therefore those points were taken out and the 
models were re-fitted. These outliers were excluded from the other analyses as well. 
Since dead biomass was not sorted by growth forms and outliers could not be excluded, 
we used dead biomass data with outliers for analyses. 
To determine effect of the different stands on the relationship between biomass and 
cover for each growth form, the likelihood-ratio test was used. Under this test procedure, 
a full model and a reduced model were compared, as in the general linear test (Kutner et 
al., 2004). The full model represents the scenario in which all the slopes of the 
relationship between biomass and cover are different from each other and therefore every 
stand age has a different model. The full model was compared to the reduced model that 
describes the relationship between biomass and cover as being the same for all stand 
ages, in order to determine if there were any differences between any of the stand ages. 
The full model was:  
Yi = ?1 Xi1 + ?2 Xi2 + ?3 Xi3 + ?4 Xi4 + ?5 Xi5 + ?     (1)  
where Yi is the aboveground biomass measurements, Xi1, Xi2, Xi3, Xi4, and Xi5 are the 
dummy variables that represent five different stands, and ?1, ?2, ?3, ?4, and ?5 the slope 
terms of the dummy variables. The hypothesis tested was: 
13 
H0: ?1 = ?2 = ?3 = ?4 = ?5  Ha: at least one ?k is different from the others 
The null hypothesis was the reduced model and the alternative hypothesis the full 
model. The test statistic for the likelihood-ratio test was: 
G = -2logeL(R) ? (-2logeL(F))       (2) 
where -2loge(F) is the log-likelihood for the full model and -2logeL(R) is the log-
likelihood for the reduced model. If H0 holds, G is approximately distributed as ?2. The 
loge(F) and the logeL(R) were calculated using the maximum likelihood (ML) estimation 
method [Proc Mixed (SAS version 9.2; SAS Institute, Inc., Cary, NC)]. The level of 
significance was set at 0.05. If G (d.f.=4) ? ?2 was not significant then H0 was not 
rejected. Whenever H0 was rejected, combinations of stands were tested against the full 
model using the appropriate degrees of freedom in order to determine whether a coarser 
group of stands could share the same slope. Smallest number of groups in each growth 
form was the criterion through which the most satisfactory combinations were selected. 
Biomass - basal area relationships  
The relationship between ground cover biomass and total basal area of trees > 2 m in 
height was determined by pooling subplots and stands (n=20). Both linear and non-linear 
models were tested (SAS version 9.2; SAS Institute, Inc., Cary, NC). Live total biomass, 
herbaceous biomass, biomass by growth form, and dead biomass were averaged by 
subplot. Basal area data from the forest inventory were also averaged by subplot. Linear 
functions fit the data. The linear function was in the form of Y = a + bX where y is 
biomass, x is total basal area of trees > 2 m in height, and a and b estimated regression 
coefficients.   
14 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1 ha  plot  
r = 56 . 4  m  
3 5 m  at 0 ?  
35m  at 1 2 0 ?   
3 5 m  at 2 4 0 ?  
 
 
r  = 1 1 . 3 m  
Sub pl o t 2  
Sub pl o t 3  
Subpl o t 1  
Sub pl o t 4  
 
 
  
Fig. 2.2.1. Representative schematic of the 1-ha field plot sampling structure for each 
longleaf pine stand at Fort Benning, GA. Within each subplot five circular 1-m2 plots 
were randomly selected. 
 
15 
 
  
Age Stand Type Canopy Layer Species Basal Area Density Dbh H 
        (m2 ha-1) (trees ha-1) (cm) (m) 
5 Planted Overstory1 LLP 0.0 0.0 - - 
   
Other 0.0 0.0 - - 
  
Overstory2 LLP 0.4 150 6.6 2.4 
   
Other 0.04 64 3.2 2.7 
Understory LLP - 625 - <2 
   
Other - 1,000 - <2 
12 Planted Overstory1 LLP 3.7 300 12.2 9.2 
   
Other 4.4 219 14.6 8.6 
  
Overstory2 LLP 2.7 619 7.2 6.2 
   
Other 0.6 159 6.6 5.8 
21 Planted Overstory1 LLP 18.6 1,331 13.2 11.8 
   
Other 0.9 75 12.0 9.3 
  
Overstory2 LLP 2.3 481 7.7 8.7 
   
Other 0.6 95 9.3 8.0 
64 Natural Overstory1 LLP 7.5 94 29.7 18.3 
   
Other 2.4 37 29.5 18.3 
  
Overstory2 LLP 0.3 159 4.5 3.1 
   
Other 0.1 127 2.3 2.3 
87 Natural Overstory1 LLP 13.4 87 42.5 28.9 
   
Other 0.0 0.0 - - 
  
Overstory2 LLP 1.1 765 4.1 3.7 
    Other 0.0 0.0 - <2 
Table 2.2.1. Structure of longleaf pine (LLP) stands sampled at Fort Benning, GA. 
Overstory1 includes all woody stems ?10 cm dbh and Overstory2 is woody stems > 2 
m height and < 10 cm dbh. The 5-year-old stand understory includes all planted 
longleaf < 2 m height and is included only for the 5-year-old stand to account for all 
planted trees. 
 
 
16 
 
Scientific Name Family Common Name 
Andropogon/Schizachyrium Poaceae bluestem 
Asimina parviflora (Michx.) Dunal Annonaceae smallflower pawpaw 
Callicarpa americana L. Verbenaceae American beautyberry 
Carya alba (L.) Nutt. Juglandaceae mockernut hickory 
Centrosema virginianum (L.) 
Benth 
Fabaceae spurred butterfly pea 
Chamaecrista spp. Fabaceae partridge pea 
Chrysopsis mariana (L.) Elliott Asteraceae Maryland goldenaster 
Cnidoscolus urens (L.) Arthur var. 
stimulosus (Michx.) Govaerts 
Euphorbiaceae finger rot 
Coreopsis major Walter Asteraceae greater tickseed 
Danthonia sericea Nutt. Poaceae downy danthonia 
Desmodium lineatum DC. Fabaceae sand ticktrefoil 
Desmodium obtusum Muhl. ex 
Willd.) DC 
Fabaceae stiff ticktrefoil 
Desmodium paniculatum (L.) DC. Fabaceae panicledleaf ticktrefoil 
Desmodium rotundifolium DC. Fabaceae prostrate ticktrefoil 
Dichanthelium aciculare (Desv. ex 
Poir.) Gould & C.A. Clark 
Poaceae needleleaf rosette grass 
Dichanthelium acuminatum (Sw.) 
Gould & C.A. Clark 
Poaceae tapered rosette grass 
Dichanthelium commutatum 
(Schult.) Gould 
Poaceae variable panicgrass 
Diospyros virginiana L. Ebenaceae common persimmon 
Eupatorium compositifolium 
Walter 
Asteraceae yankeeweed 
Eupatorium hyssopifolium L. Asteraceae hyssopleaf thoroughwort 
Euphorbia pubentissima Michx. Euphorbiaceae false flowering spurge 
Galium hispidulum Michx. Rubiaceae coastal bedstraw 
Galium pilosum Aiton Rubiaceae hairy bedstraw 
Gelsemium sempervirens (L.) 
W.T. Aiton 
Loganiaceae evening trumpetflower 
Hypericum gentianoides (L.) 
Britton, Sterns & Poggenb. 
Clusiaceae orangegrass 
Hypericum hypericoides (L.) 
Crantz 
Clusiaceae St. Andrew?s cross 
Ipomoea pandurata (L.) G. Mey. Convolvulaceae man of the earth 
Lespedeza cuneata (Dum. Cours.) 
G. Don 
Fabaceae sericea lespedeza 
Lespedeza procumbens Michx. Fabaceae trailing lespedeza 
Lespedeza repens (L.) W.P.C. 
Barton 
Fabaceae creeping lespedeza 
Table 2.2.2. Ground cover plant species in longleaf pine stands at Fort Benning, GA. 
 
17 
Lespedeza stuevei Nutt. Fabaceae tall lespedeza 
Liquidambar styraciflua L. Hamamelidaceae sweetgum 
Lygodium japonicum (Thunb.) Sw. Lygodiaceae Japanese climbing fern 
Mimosa quadrivalvis L. Fabaceae fourvalve mimosa 
Morella cerifera (L.) Small Myricaceae wax myrtle 
Opuntia sp. Cactaceae  
Oxalis spp. Oxalidaceae woodsorrel 
Packera anonyma (Alph. Wood) 
W.A. Weber & ?. L?ve 
Asteraceae Small?s ragwort 
Parthenocissus quinquefolia (L.) 
Planch. 
Vitaceae Virginia creeper 
Physalis virginiana Mill. Solanaceae Virginia groundcherry 
Pinus palustris Mill. Pinaceae longleaf pine 
Pinus spp. Pinaceae pine 
Pityopsis spp. Asteraceae silkgrass 
Pseudognaphalium spp. Asteraceae cudweed 
Pteridium aquilinum (L.) Kuhn Dennstaedtiaceae western brackenfern 
Quercus falcata Michx. Fagaceae southern red oak 
Quercus hemisphaerica W. 
Bartram ex Willd. 
Fagaceae Darlington oak 
Quercus incana W. Bartram Fagaceae bluejack oak 
Quercus marilandica M?nchh. Fagaceae blackjack oak 
Quercus nigra L. Fagaceae water oak 
Rhus copallinum L. Anacardiaceae winged sumac 
Rhynchosia reniformis DC. Fabaceae dollarleaf 
Rubus cuneifolius Pursh Rosaceae sand blackberry 
Rubus sp.  Rosaceae blackberry 
Scleria ciliata Michx. var. ciliata Cyperaceae fringed nutrush 
Sericocarpus tortifolius (Michx.) 
Nees 
Asteraceae Dixie whitetop aster 
Smilax bona-nox L. Smilacaceae saw greenbrier 
Smilax glauca Walter Smilacaceae cat greenbrier 
Solidago odora Aiton Asteraceae anisescented goldenrod 
Stylosanthes biflora (L.) Britton, 
Sterns & Poggenb. 
Fabaceae sidebeak pencilflower 
Symphyotrichum patens (Aiton) 
G.L. Nesom 
Asteraceae late purple aster 
Sisyrinchium sp. Iridaceae blue-eyed grass 
Tephrosia spicata (Walter) Torr. 
& A. Gray 
Fabaceae spiked hoarypea 
Tephrosia virginiana (L.) Pers. Fabaceae Virginia tephrosia 
Toxicodendron pubescens Mill. Anacardiaceae Atlantic poison oak 
Tragia urens L. Euphorbiaceae wavyleaf noseburn 
Triodanis perfoliata (L.) Nieuwl. Campanulaceae clasping Venus? looking-glass 
Ulmus alata Michx. Ulmaceae winged elm 
18 
Vernonia angustifolia Michx. Asteraceae tall ironweed 
Vaccinium arboreum Marshall Ericaceae farkleberry 
Vaccinium myrsinites Lam. Ericaceae shiny blueberry 
Vaccinium stamineum L. Ericaceae deerberry 
  
19 
2.3 - Results 
Cover of shrubs/tree seedlings, vines, graminoids, and forbs varied significantly 
among stands (Table 2.3.1). Stand age had no significant effect on cover of legumes.  
Ferns occurred only in the 64-year-old stand. Within the shrub/tree seedling growth form, 
cover ranged from 5% in the 5-year-old stand to 42% in the 12-year-old stand (Figure 
2.3.1). Shrub/tree seedling cover in the 12-year-old stand was significantly higher than in 
the 5-year-old stand but neither stand was significantly different from the other stands 
(Figure 2.3.1). Vine cover varied between 7% in the 21-year-old stand to 29% in the 5-
year-old stand and was higher in the 5-year-old stand than in the 12-, 21-, and 87-year-
old stands (Figure 2.3.1).  Cover of graminoids was lowest in the 21-year-old stand (4%) 
but was similar among the other stands and was on average 42% (Figure 2.3.1). Average 
legume cover was 8%. Forb cover was lower in the 12-year-old stand compared to the 5, 
21, and 87-year-old stands and forb cover was similar among the 5, 21, 64, and 87-year-
old stands (Figure 2.3.1). 
 
 
 
 
 
 
 
 
 
Variable Stand age (P>F) 
Shrub/tree seedling cover 0.009 
Vine cover 0.001 
Graminoid cover < 0.001 
Legume cover 0.080 
Forb cover 0.002 
Shrub/tree seedling biomass 0.003 
Vine biomass < 0.001 
Graminoid biomass < 0.001 
Legume biomass 0.187 
Forb biomass < 0.001 
Total live biomass 0.001 
Dead biomass  < 0.001 
 
Table 2.3.1. Observed probability values for the effect of stand age on groundcover 
layer variables in longleaf pine stands at Fort Benning, GA. Total live and dead 
biomass is the sum of biomass of all growth forms within each 1-m2 plot.  
 
20 
Figure 2.3.1. Mean (SE) cover by growth form in the ground cover layer in longleaf 
pine stands at Fort Benning, GA. Standard errors represent variability among subplots. 
Within each growth form, different letters indicate significant differences in least 
square means among stands at ?=0.05.  
G r o w t h  f o r mShr
u
b
/
t
r
e
e
 se
e
d
l
i
n
g
V
i
n
e
G
r
a
mi
n
o
i
d
L
e
g
u
me
F
o
r
b
C
o
v
e
r
 (
%
)
0
20
40
60
80
100
5 - y e a r - o l d  st a n d   
1 2 - y e a r - o l d  st a n d  
2 1 - y e a r - o l d  st a n d  
6 4 - y e a r - o l d  st a n d  
8 7 - y e a r - o l d  st a n d  
d
e
e
de
e
g
g
h
g
g
m
n
m
mn
m
a
b
ab
ab
ab
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Total live biomass in the ground cover layer varied among stand ages and values 
ranged from 28.3 g m-2 in the 21-year-old stand to 171.5 g m-2 in the 12-year-old stand 
(Table 2.3.1; Figure 2.3.2). Total live biomass was significantly lower in the 21-year-old 
stand than in the 5- and 12-year-old stands but similar to live biomass in the two oldest 
stands (Figure 2.3.2). Dead biomass also varied among stands (Table 2.4.1). Dead 
biomass was highest in the 5-year-old stand and higher in the 12-year-old stand than in 
the 21-year-old stand (Figure 2.3.2).  
21 
Figure 2.3.2. Mean (SE) total live biomass (sum of all growth forms) and dead 
biomass in the ground cover layer in longleaf pine stands at Fort Benning, GA. 
Standard errors represent variability among subplots. Different capital letters, for live 
biomass, and different lowercase letters, for dead biomass, indicate significant 
differences in least square means between stands at ?=0.05. 
Sta n d  a g e  (y e a r s)
5 12 21 64 87
B
ioma
ss (
g
 m
-2
)
0
50
100
150
200
250
To ta l li v e  bio ma ss
D e a d  bio ma ss 
a
b
bc
bc
A
A
AB
AB
B
c
 
 
 
 
 
 
 
 
 
 
 
 
 
Live biomass by growth form differed among stand ages except for legumes (Table 
2.3.1). Shrub/tree seedling biomass was highest in the 12-year-old stand, 132.3 g m-2, 
compared to an average of 24.8 g m-2 in the other stands, which were similar (Figure 
2.3.3). Biomass of vines ranged between 6.6 g m-2 in the 12-year-old stand to 58.9 g m-2 
in the 5-year-old stand, which had the highest vine biomass (Figure 2.3.3).  Graminoid 
biomass in the 21-year-old stand was 1.5 g m-2 and lower than graminoid biomass in the 
two youngest stands, which had an average graminoid biomass of 35.1 g m-2 (Figure 
2.3.3). Average legume biomass across all stands was 5.3 g m-2. Biomass of forbs varied 
22 
Figure 2.3.3. Live biomass by growth form in the ground cover layer in longleaf pine 
stands at Fort Benning, GA. Standard errors represent variability among subplots. 
Within each growth form, different letters indicate significant differences in least 
square means of biomass between stands at ?=0.05.  
G r o w t h  f o r mShr
u
b
/ t
r
e
e
 se
e
d
l i
n
g
V
i n
e
G
r
a
mi
n
o
i d
L
e
g
u
me
F
o
r
b
B
i
o
ma
ss (g
 m
-2
)
0
20
40
60
80
100
120
140
160
180
200
5 - y e a r - o l d  st a n d
1 2 - y e a r - o l d  st a n d  
2 1 - y e a r - o l d  st a n d  
6 4 - y e a r - o l d  st a n d  
8 7 - y e a r - o l d  st a n d  
b
a
b
b
b
d
e
e
e
e
g
gh
i
hi
hi
m
o
no
no
n
between 2.2 g m-2 in the 12-year-old stand and 34.2 g m-2 in the 5-year-old stand, and was 
highest in the 5-year-old stand (Figure 2.3.3). The 12-year-old stand had lower forb 
biomass than the 87-year-old stand (Figure 2.3.3). 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
At the stand level, the linear relationship between biomass and cover in the form of y = ?x 
+ ? was significant for almost all growth forms and stands. The regression for ferns in the 
64-year-old stand was also significant. For shrubs, vines and graminoid growth forms a 
reduced model with two equations was as good as the full model in estimating biomass 
(Tables 2.3.2 and 2.3.3). For legume and forb growth forms the relationship between 
23 
biomass and cover was explained by three different functions (Table 2.3.2 and 2.3.3). For 
shrub/tree seedlings, the 12-year-old stand demonstrated a greater slope compared to all 
the other stands combined (Table 2.3.2; Figure 2.3.4). The slope of the 5-year-old stand 
was greater compared to all other stands combined for vine and graminoid growth forms 
(Table 2.3.2; Figure 2.3.4). For legumes, slopes of the 12-, 21-, and 87-year-old stands 
were similar and smaller than the 64-year-old stand and the 5-year-old stand which were 
grouped separately (Table 2.3.2; Figure 2.3.5). In the forb growth form, three equations 
were necessary to describe the relationship between cover and biomass with the slope of 
the two oldest stands combined being higher than the 12- and the 21-year-old stands 
combined slope and smaller than 5-year-old stand (Table 2.3.2; Figure 2.3.5). 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Growth form  ?1 S.E. of  ?1 
Shrubs/tree seedlings 
       12-year-old stand 
       All other stands 
 
3.180 
2.021 
 
0.112 
0.176 
Vines 
       5-year-old stand 
       All other stands 
 
1.932 
0.877 
 
0.101 
0.110 
Graminoids 
       5-year-old stand 
       All other stands 
 
0.829 
0.391 
 
0.048 
0.027 
Legumes 
       5-year-old stand 
       64-year-old stand 
       12-, 21-, and 87-year-old stands 
 
1.344 
1.014 
0.509 
 
0.082 
0.087 
0.067 
Forbs 
        5-year-old stand 
        64-and 87-year-old stands                                                                                              
        12-and 21-year-old stands 
 
1.163 
0.663 
0.241 
 
0.072 
0.078 
0.078 
Table 2.3.2. Slopes of the relationship (y= ?1x+?) between biomass (g m-2) and 
ground cover (%) by growth form in longleaf pine stands varying in age at Fort 
Benning, GA. 
 
24 
 
 
 
 
 
 
Growth form  Reduced model  G         df P>G 
Shrubs/tree 
seedlings 
 1 function for all stands 
 
1) 12-year-old stand                       
2) All the other stands 
    32.7  
 
6.8 
4 
 
3 
< 0.001 
 
0.079 
  
 
Vines  1 function for all stands 
 
1) 5-year-old stand                         
2) All the other stands 
 39.9 
 
0.0 
4 
 
3 
< 0.001 
 
0.873   
Graminoids  1 function for all stands 
 
1) 5-year-old stand                         
2) All the other stands 
 56.2 
 
7.3 
4 
 
3 
< 0.001 
 
0.06   
Legumes  1 function for all stands 
 
1) 12, 21, and 87-year-old stands        
2) 5 and 64-year-old stands 
 
1) 12, 21, and 87-year-old stands                 
2) 5-year-old stand                             
3) 64-year-old stand 
 55.5 
 
11.4 
 
 
 
3.7 
4 
 
3 
 
 
 
2 
< 0.001 
 
0.001 
 
 
 
0.157 
  
  
Forbs  1 function for all stands 
 
1) 5-year-old stand                         
2) All the other stands 
 
1) 5-year-old stand                                                                                            
2) 12 and 21-year-old stands                   
3)  64 and 87-year-old stands 
 64.2 
 
18.8 
 
 
1.5 
4 
 
3 
 
 
2 
< 0.001 
 
< 0.001 
 
 
0.472 
  
Table 2.3.3. Relationship between biomass and percent cover in the ground cover 
layer in longleaf pine stands at Fort Benning, GA. Likelihood ratio tests were used to 
test reduced models against the full five-parameter model by different growth forms. 
G is the test statistic for the likelihood ratio test. Non-significant p-values at ?=0.05 
indicate that the reduced model is as appropriate as the full model. 
 
25 
Fig. 2.3.4. Biomass of shrubs/tree seedlings, vines, and graminoids in relation to cover 
in the ground cover layer in longleaf pine stands at Fort Benning, GA.  
0
40
80
120
160
200
240
280
320
360
400
0 20 40 60 80 100
Bi
o
m
a
s
s
 
(g
 
m
?
?)
C o v er  (% )
12-ye a r- ol d st a nd
All t he  other  sta nds
Shrubs /tree seedling s
0
20
40
60
80
100
120
140
160
0 10 20 30 40 50 60 70
Bi
o
m
a
s
s
 
(g
 
m
?
?)
C o v er  (% )
5- ye a r- ol d st a nd
All t he  other  sta nds
Vines
0
20
40
60
80
100
120
140
0 20 40 60 80 100
Bi
o
m
a
s
s
 
(g
 
m
?
?)
C o v er  (% )
5- ye a r- ol d st a nd
All t he  other  sta nds
G ra m ino ids
 
 
 
 
  
 
 
 
 
 
 
 
 
 
 
 
 
 
  
 
 
 
 
26 
Fig. 2.3.5. Biomass of legumes and forbs in relation to cover in the ground cover layer 
in longleaf pine stands at Fort Benning, GA.  
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30 35 40
Bi
o
m
a
s
s
 
(g
 
m
?
?)
C o v er  (% )
5- ye a r- ol d st a nd
64-ye a r- ol d st a nd
12-, 2 1- , &  87- ye a r -old stands
Leg um es
0
20
40
60
80
100
120
0 10 20 30 40 50 60 70
Bi
o
m
a
s
s
 
(g
 
m
?
?)
C o v er  (% )
5- ye a r- ol d st a nd
64- & 87- ye a r- old s ta nds
12- & 21- ye a r- old s ta nds
Forbs
 
 
 
 
 
 
 
  
 
 
  
 
 
 
 
 
 
 
 
 
 
 
27 
Total live, dead, herbaceous, and graminoid biomass at the 1-m2 plot level were 
averaged by subplot and relationships between biomass and stand basal area were 
explored. Total live and dead biomass were linearly and negatively related to basal area 
and basal area explained 46 and 65% of the variability in live and dead biomass, 
respectively (Table 2.3.4, Fig. 2.4.6). Herbaceous biomass, which included graminoids, 
legumes, forbs, and ferns, declined linearly with increasing basal area (Fig. 2.3.7). 
Graminoid biomass was also significantly and negatively related to basal area (Fig. 
2.3.7).  
 
 
 
 
 
 
 
 
  
Biomass Model equation a b R? P>F 
Total live y =  a + bx 170.44 -5.61 0.46 0.001 
Dead y =  a + bx 88.95 -3.44 0.65 < 0.001 
Herbaceous y =  a + bx 79.88 -3.01 0.68 < 0.001 
Graminoid y =  a + bx 39.03 -1.60 0.58 < 0.001 
Table 2.3.4. Regression functions for relationships between ground cover biomass (g 
m-2) variables and total basal area (m2 ha-1) of woody stems > 2m in height in longleaf 
pine stands at Fort Benning, GA (?=0.05). Herbaceous biomass is total live biomass 
without shrub/tree seedling biomass and vine biomass. Regression coefficients are 
indicated by a and b. The number of observations is 20 for all models. 
28 
 
 
 
 
  
Figure 2.3.6. Total live biomass (sum of all growth forms) and dead biomass in the 
ground cover layer as a function of basal area of all woody stems ? 2 m in height in 
longleaf pine stands at Fort Benning, GA.  
Ba sal a r e a  ( m
2
 ha
-1
)
0 5 10 15 20 25 30
L
iv
e
 bio
m
a
ss
 (
g
 m
-2
)
0
50
100
150
200
250
300
5 - y e a r - o l d  sta n d
1 2 - y e a r - o l d  sta n d
2 1 - y e a r - o l d  sta n d
6 4 - y e a r - o l d  sta n d
8 7 - y e a r - o l d  sta n d
Ba sal a r e a  ( m
2
 ha
-1
)
0 5 10 15 20 25 30
De
a
d
 bio
m
a
ss
 (
g
 m
-2
)
0
20
40
60
80
100
120
140
5 - y e a r - o l d  st a n d
1 2 - y e a r - o l d  st a n d
2 1 - y e a r - o l d  st a n d
6 4 - y e a r - o l d  st a n d
8 7 - y e a r - o l d  st a n d
29 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 2.3.7. Herbaceous biomass (total of graminoids, legumes, forbs, and ferns) and 
graminoid biomass in the ground cover layer as a function of basal area of all woody 
stems ? 2 m in height in longleaf pine stands at Fort Benning, GA.  
Ba sal a r e a  ( m
2
 ha
-1
)
0 5 10 15 20 25 30
He
r
b
a
c
e
o
u
s bi
o
m
a
ss
 (
g
 m
-2
)
0
20
40
60
80
100
120
140
5 - y e a r - o l d  st a n d
1 2 - y e a r - o l d  st a n d
2 1 - y e a r - o l d  st a n d
6 4 - y e a r - o l d  st a n d
8 7 - y e a r - o l d  st a n d
Ba sal a r e a  ( m
2
 ha
-1
)
0 5 10 15 20 25 30
Gr
a
m
in
o
id
 bio
m
a
ss
 (
g
 m
-2
)
0
20
40
60
80
5 - y e a r - o l d  st a n d
1 2 - y e a r - o l d  st a n d
2 1 - y e a r - o l d  st a n d
6 4 - y e a r - o l d  st a n d
8 7 - y e a r - o l d  st a n d
30 
2.4 - Discussion 
Biomass Production 
Aboveground biomass in the ground cover of forests dominated by loblolly pine 
(Pinus taeda L.), longleaf pine (Pinus palustris Miller), and shortleaf pine (Pinus 
echinata Miller) was studied at Fort Benning by Lajeunesse et al. (2006). Mean total live 
biomass in the third season following fire was 88 g m-2, which was lower than the 
average (104 g m-2) across all stands observed in this study. Herbaceous, graminoid, and 
legume biomass was on average higher and forb biomass lower than in Lajeunesse et al. 
(2006). Total live biomass in this study (104 g m-2) was within the range of biomass 
values found in a longleaf pine-wiregrass woodland of southern Georgia, where total live 
biomass in ground cover varied from approximately 81 g m-2 in xeric sites to 157 g m-2 in 
wet-mesic sites and basal area of overstory trees was lower in xeric sites than in wet-
mesic sites (Kirkman et al., 2001). However, total live and herbaceous biomass by stand 
at Fort Benning, GA, was generally lower than biomass values reported in the literature 
for longleaf pine forests, especially in the 21-year-old plantation. For example, a 21-year-
old stand in Louisiana had higher herbaceous biomass (41.8 g m-2) than the stand of same 
age at Fort Benning, GA (10 g m-2), even though basal area was similar between the two 
studies (Wolters, 1982). In the West Gulf region, herbaceous biomass in a 37-year-old 
longleaf pine stand with a basal area of 22 m2 ha-1 was 94 g m-2 (Haywood et al., 2001) 
and higher than in the 21-year-old plantation at Fort Benning, GA, which had similar 
basal area but higher density. Differences between studies in the literature and this study 
in graminoid, total live, and herbaceous biomass can be attributed not only to differences 
in stand structure but also to  the contribution of wiregrass, which normally constitutes a 
31 
high proportion of ground cover biomass in longleaf pine-wiregrass woodlands (Kirkman 
et al., 2001). Moreover, ground cover biomass was measured in May and June and was 
likely lower than values at the peak biomass of herbaceous plants, which comes in 
July/August at Fort Benning, GA (Lajeunesse et al., 2006). In the 21-year-old stand, 
specifically, high planting density has led to a closed canopy structure and to less 
available space for plants to grow in the ground cover compared to the stands in 
Louisiana and West Gulf Region. 
Basal Area Influence 
In longleaf pine forests, herbaceous plant production has been shown to decrease 
linearly with increasing overstory basal area and to increase with increasing seasonal 
precipitation (Grelen and Lohrey, 1978; Wolters, 1973, 1982). At Fort Benning, 
herbaceous biomass and graminoid biomass of all stands pooled together decreased 
linearly with increasing basal area. The regression coefficient of 3 g m-2 for herbaceous 
biomass was substantially lower than the slope obtained in longleaf pine plantations in 
Louisiana varying in age from 8 to 23 years, where biomass decreased 10 g m-2 with each 
m2 ha-1 increase in tree basal area (Wolters, 1982). Precipitation at Fort Benning in 
Spring 2012 (12 mm in April and 85 mm in May) was below average (101 mm in April 
and 95 mm in May), particularly in April, and was lower compared to the average rainfall 
at the Louisiana site (195 mm in April and 175 mm in May between 1969 and 1975) 
(http://www.ncdc.noaa.gov; accessed October 2013). Differences in precipitation 
between the two sites therefore may have accounted for the lower decrease in biomass 
with increasing basal area in this study.  The relationship between live biomass 
(herbaceous plus woody) and tree basal area was significant but basal area explained less 
32 
variation in biomass compared to the relationship with herbaceous vegetation. Pecot et al. 
(2007), for a longleaf pine woodland in west central Georgia, showed that growth of 
herbaceous vegetation increased with overstory tree removal while woody understory did 
not. The authors suggest that herbaceous plants may be competing primarily for light 
while shade-tolerant understory woody plants and canopy trees competed mostly for soil 
nitrogen. The equations to estimate biomass from basal area can be particularly useful in 
fuel and carbon models where separate quantifications of herbaceous and dead biomass in 
the ground cover are needed. 
Plant Cover-Biomass relationships 
As hypothesized, and consistent with similar studies in the literature, biomass and 
cover at Fort Benning, GA were positively and linearly related for all growth forms and 
stands. For example, in the understory of Canadian riparian forests, MacDonald et al. 
(2012) found that percent cover varied linearly with biomass in dwarf shrubs and 
graminoids. Similarly, Muukkonen et al. (2006) reported significant linear biomass-cover 
relationships for dwarf shrubs and herbaceous plants (forbs and grasses) in boreal 
coniferous upland forests. According to Muukkonen et al. (2006), a linear relationship 
was suitable because species composition in the ground cover did not vary with total 
percent cover. In southern forests, understory herbaceous biomass has been reported to 
increase linearly with percent cover. For example, in Alabama, slopes of the regressions 
for relationships between cover and biomass were 1.9 g m-2, 1.7 g m-2, and 1.1 g m-2 
respectively for planted pine, natural pine, and oak-pine forest types (Joyce and Mitchell, 
1989). Significant linear relationships between biomass and cover were reported for herbs 
33 
and shrubs in a 2-year-old longleaf pine plantation on hydric soils at Camp Lejeune, NC 
(Walker and Cohen, 2009) and were best described by a log-transformed equation. 
Models and Prediction 
A single model was not sufficient to describe the relationship between biomass and 
cover in the ground cover layer. For shrubs/tree seedlings, vines, and graminoids two 
models were needed while for legumes and forbs three different models were necessary. 
The slope of the relationship of biomass versus cover was highest in the 5-year-old stand 
for vines, graminoids, legumes, and forbs. In the ground cover layer, the ratio of biomass 
to cover within a growth form can vary with stand age or seral stage (Alaback, 1986), 
because of different light conditions determined by stand structure. Higher light 
availability can allow plants to produce more biomass per unit of cover and also can 
influence plant architecture and species composition. In the youngest stand, there was no 
overstory canopy. In order to limit the effect of stand age on understory biomass-cover 
relationships, Muukkonen et al. (2006) did not include very young stands in their 
analyses. In contrast to our results, the increase in biomass with increasing percent cover 
did not differ between early and late seral stages for graminoids and dwarf shrubs in 
riparian boreal forests (MacDonald et al., 2012). In riparian forests, the relationship 
between biomass and cover for such growth forms may be driven more by hydrologic 
processes.  
While differences between the 5-year-old stand and the other stands can be explained 
by light limitation under tree canopies compared to an early seral stage, further 
separations into three models for legumes may reflect a difference in species composition 
between the 64-year-old stand and the younger stands. The legume species Tephrosia 
34 
virginiana (L.) Pers., which is denser and taller than most legume species we 
encountered, occurred and was abundant only in the 64-year-old stand. In forbs, higher 
slope for the natural stands compared to the 12- and 21-year-old stands suggest an 
influence of stand origin, and possibly past land use, on the biomass cover ratio. Within 
the shrub/tree seedling growth form, the slope of the biomass versus cover relationship 
was higher in the 12-year-old stand than in all other stands. The slope for the 12-year-old 
stand was similar to that of dwarf shrubs in Muukkonen et al. (2006) and MacDonald et 
al. (2012) (both coefficients were 2.1 g m-2). A separate model for the 12-year-old stand 
was the result of generally higher biomass but not higher cover in the 12-year-old stand 
compared to the other stands. The two different models may be a result of differences in 
species composition within the growth form, because dense and many branched shrub 
species with high biomass per unit cover, such as Vaccinium myrsinites Lam. and other 
species in the Ericaceae family, were more common in the 12-year-old stand compared to 
all other stands. These results suggest that a single linear relationship in the graminoid 
and vine growth forms could be used to estimate biomass from cover for longleaf pine 
forests similar in ground cover species composition and growth forms to the stands at 
Fort Benning, GA, regardless of stand age and structure. However, graminoid and vine 
biomass will be probably underestimated in very young plantations without a tree 
canopy. 
Results from this study provide a better understanding of the relationship between 
aboveground biomass and cover of plant growth forms in the ground cover of longleaf 
pine stands. In shrubs/tree seedlings, vines, graminoids, legumes, and forbs, biomass was 
linearly related to percent cover. Two different equations for shrubs/tree seedlings, vines, 
35 
and graminoids and three equations for legumes and forbs were developed. In the 5-year-
old stand, which lacked overstory trees, the slopes of the regressions were generally 
higher than in the other stands, suggesting that a lower slope under tree canopy was 
determined by light limitation. Other changes in biomass-cover relationships between 
some stands for shrubs/tree seedlings, legumes, and forbs may be related to stand-specific 
differences in species composition and dominant species and to stand origin (planted 
versus naturally regenerated stands). Through its influence on species composition, a 
different fire history among stands may have also indirectly affected the relationships 
between biomass and cover. Live, herbaceous, graminoid, and dead biomass were 
linearly and inversely related to basal area. Biomass-basal area equations developed in 
this study can be used to obtain a rapid estimate of herbaceous, dead, and graminoid 
biomass when stand structure data are available but ground cover data are not. 
  
36 
CHAPTER 3. CONCLUSIONS 
Results from this study provide information useful for estimating ground cover 
biomass in longleaf pine ecosystems. Evidence supports use of the fast and non-
destructive method of visual cover estimate to predict biomass of growth forms in the 
ground cover layer. Relationships between biomass and cover were linear but different 
models were needed for different growth forms and stands.  Regression slopes were 
highest in the youngest stand, which did not have a tree canopy, for all three herbaceous 
growth forms and for vines. Other slope differences in biomass-cover relationships 
seemed to be linked to differences in ground cover composition, as in the case of 
shrubs/tree seedlings and legumes, or to stand origin (planted versus natural) as in the 
case of forbs. Live, herbaceous, graminoid, and dead biomass in the ground cover layer 
can be estimated by using linear relationships with basal area. Further research should 
investigate the relationship between biomass and cover in other areas of the longleaf pine 
range, especially in areas of bluestem and wiregrass, and the influence of plant height, 
which is probably more variable within each growth form in other seasons. Our results 
also support the findings of other studies indicating that herbaceous biomass can be 
efficiently estimated by basal area. 
37 
REFERENCES 
 
Abrahamson, I.L.; Nelson, C.R.; Affleck, D.L. Assessing the performance of sampling 
designs for measuring the abundance of understory plants. Ecol. Appl. 2011, 21, 
452-464. 
 
Alaback, P.B. Biomass regression equations for understory plants in coastal Alaska: 
effects of species and sampling design on estimates. Northwest Sci. 1986, 60, 90-
103. 
 
Andrew Stainback, G.; Alavalapati, J.R. Restoring longleaf pine through silvopasture 
practices: an economic analysis. Forest Policy Econ. 2004, 6, 371-378. 
 
Aschenbach, T.A.; Foster, B.L.; Imm, D.W. The Initial Phase of a Longleaf Pine?
Wiregrass Savanna Restoration: Species Establishment and Community 
Responses. Restor. Ecol. 2010, 18, 762-771. 
 
Augusto, L.; Dupouey, J.L.; Ranger, J. Effects of tree species on understory vegetation 
and environmental conditions in temperate forests. Ann. For. Sci. 2003, 60, 823-
831. 
 
Br?kenhielm, S.; Quinghong, L. Comparison of field methods in vegetation monitoring. 
In Biogeochemical Monitoring in Small Catchments: Refereed Papers from 
BIOGEOMON, the Symposium on Ecosystem Behaviour: Evaluation of 
Integrated Monitoring in Small Catchments, Held in Prague, Czech Republic, 
September 18-20, 1993; ?ern?, J., Novak, M., Pa?es, T., Wieder K., Eds.; 
Springer: Dordrecht, Netherlands, 1995; 79, pp. 75-87. 
 
Brockway, D.G.; Outcalt K.W. Gap-phase regeneration in longleaf pine wiregrass 
ecosystems. Forest Ecol. Manag. 1998, 106, 125-139. 
 
Brockway, D.G.; Outcalt, K.W.; Wilkins, R.N. Restoring longleaf pine wiregrass 
ecosystems: plant cover, diversity and biomass following low-rate hexazinone 
application on Florida sandhills. Forest Ecol. Manag. 1998, 103, 159-175. 
 
Brockway, D.G.; Outcalt, K.W.; Tomczak, D.J.; Johnson, E.E. Restoring longleaf pine 
forest ecosystems in the southern US. In Restoration of boreal and temperate 
forests; Stanturf, J.A., Madsen, P., Eds.; CRC Press: Boca Raton, FL, USA, 2005; 
pp. 501-519. 
 
Brockway, D.G.; Outcalt, K.W.; Estes, B.L.; Rummer, R.B. Vegetation response to 
midstorey mulching and prescribed burning for wildfire hazard reduction and 
longleaf pine (Pinus palustris Mill.) ecosystem restoration. Forestry 2009, 82, 
299-314.  
 
38 
Cao, J.; Wang, X.; Tian, Y.; Wen, Z.; Zha, T. Pattern of carbon allocation across three 
different stages of stand development of a Chinese pine (Pinus tabulaeformis) 
forest. Ecol. Res. 2012, 27, 883-892. 
 
Chapin, F.S. Functional role of growth forms in ecosystems and global processes. In 
Scaling physiological processes: Leaf to globe; Ehleringer, J.R., Field, C.B., Eds.; 
Academic Press: San Diego, CA, USA, 1993; pp. 287-312.  
 
Chiarucci, A.; Wilson, J.B.; Anderson, B.J.; De Dominicis, V. Cover versus biomass as 
an estimate of species abundance: does it make a difference to conclusions? J. 
Veg. Sci. 1999, 10, 35-42.  
 
Daubenmire, R. A canopy-coverage method of vegetational analysis. Northwest Sci. 
1959, 33, 43-64. 
 
DeAngelis, D.L.; Gardner, R.H.; Shugart, H.H. Productivity of forest ecosystems studied 
during the IBP: the woodlands data set. In Dynamic properties of forest 
ecosystems; Reichle, D.E., Ed.; Cambridge University Press: Cambridge, UK, 
1981; pp. 567-672. 
 
D?az, S.; Cabido, M. Plant functional types and ecosystem function in relation to global 
change. J. Veg. Sci. 1997, 8, 463-474. 
 
Drew, M.B.; Kirkman, L.K.; Gholson, A.K. The vascular flora of Ichauway, Baker 
County, Georgia: a remnant longleaf pine/wiregrass ecosystem. Castanea 1998, 
63, 1-24.  
 
Frank, D.A.; McNaughton, S.J. Aboveground biomass estimations with the canopy 
intercept method: a plant growth form caveat. Oikos 1990, 57, 57-60. 
 
Frost, C.C. Four centuries of changing landscape patterns in the longleaf pine ecosystem. 
In The longleaf pine ecosystem: ecology, restoration and management: 
proceedings, 18th Tall Timbers Fire Ecology Conference, May 30-June 2, 1991, 
Tallahassee, Florida.; Hermann, S.M., Ed.; Tall Timbers Research Station: 
Tallahassee, FL, USA, 1993; 18, pp. 17-43. 
 
Gaiser, R.N. Random sampling within circular plots by means of polar coordinates. J. 
Forest. 1951, 49, 916-917.   
 
Garten, C.T. Predicted effects of prescribed burning and harvesting on forest recovery 
and sustainability in southwest Georgia, USA. J. Environ. Manage. 2006, 81, 
323-332. 
Gilliam, F.S. The ecological significance of the herbaceous layer in temperate forest 
ecosystems. BioScience 2007, 57, 845-858.  
 
39 
Gilliam, F.S.; Turrill, N.L. Herbaceous layer cover and biomass in a young versus a 
mature stand of a central Appalachian hardwood forest. B. Torrey Bot. Club 1993, 
445-450. 
 
Glitzenstein, J.S.; Streng, D.R.; Wade, D.D. Fire frequency effects on longleaf pine 
(Pinus palustris P. Miller) vegetation in South Carolina and northeast Florida, 
USA. Nat. Area. J. 2003, 23, 22-37.  
 
Grelen, H.E., Lohrey, R.E. Herbage yield related to basal area and rainfall in a thinned 
longleaf plantation. In Res. Note SO-232; Dept. of Agriculture, Forest Service, 
Southern Forest Experiment Station: New Orleans, LA, USA, 1978. 
 
Guo, Q.; Rundel, P.W. Measuring dominance and diversity in ecological communities: 
choosing the right variables. J. Veg. Sci., 1997, 8, 405-408. 
 
Halpern, C.B.; Lutz, J.A. Canopy closure exerts weak controls on understory dynamics: a 
30-year study of overstory-understory interactions. Ecol. Monogr., 2013, 83, 221-
237. 
 
Halpern, C.B.; Miller, E.A.; Geyer, M.A. Equations for predicting above-ground biomass 
of plant species in early successional forests of the western Cascade Range, 
Oregon. Northwest Sci. 1996, 70, 306-320. 
 
Harrington, T.B.; Dagley, C.M.; Edwards, M.B. Above-and belowground competition 
from longleaf pine plantations limits performance of reintroduced herbaceous 
species. Forest Sci. 2003, 49, 681-695. 
 
Harrington, T.B.; Edwards, M.B. Understory vegetation, resource availability, and 
litterfall responses to pine thinning and wood vegetation control in longleaf pine 
plantations. Can. J. Forest Res. 1999, 29, 1055-1064.  
 
Haywood, J.D.; Harris, F.L.; Grelen, H.E.; Pearson, H.A. Vegetative response to 37 years 
of seasonal burning on a Louisiana longleaf pine site. South. J. Appl. For. 2001, 
25, 122-130. 
 
Hermy, M. Accuracy of visual cover assessments in predieting standing crop and 
environmental correlation in deciduous forests. Vegetatio 1988, 75, 57-64. 
 
Hutchings, M.J.; de Kroon, H. Foraging in plants: the role of morphological plasticity in 
resource acquisition. In Advances in ecological research; Begon, M., Fitter, A.H., 
Eds.; Academic Press: London, UK, 1994; 25, 159-238. 
 
Johnsen, K.H.; Butnor, J.R.; Kush, J.S.; Schmidtling, R.C.; Nelson, C.D. Hurricane 
Katrina winds damaged longleaf pine less than loblolly pine. South. J. Appl. For. 
2009, 33, 178-181. 
 
40 
Joyce, L.A.; Mitchell, J.E. Understory cover/biomass relationships in Alabama forest 
types. Agrof. Syst. 1989, 9, 205-210. 
 
Kirkman, L.K.; Mitchell, R.J.; Helton, R.C.; Drew, M.B. Productivity and species 
richness across an environmental gradient in a fire-dependent ecosystem. Am. J. 
Bot. 2001, 88, 2119-2128.  
 
Kutner, M.; Nachtsheim, C.; Neter, J. Applied Linear Regression Models, 4th ed.; 
McGraw Hill: New York, NY, USA, 2004. 
 
Lajeunesse, S.D.; Dilustro, J.J.; Sharitz, R.R., Collins, B.S. Ground layer carbon and 
nitrogen cycling and legume nitrogen inputs following fire in mixed pine forests. 
Am. J.  Bot. 2006, 93, 84-93. 
 
Lavoie, M.; Starr, G.; Mack, M.C.; Martin, T.A.; Gholz, H.L. Effects of a prescribed fire 
on understory vegetation, carbon pools, and soil nutrients in a longleaf pine-slash 
pine forest in Florida. Nat. Area. J. 2010, 30, 82-94. 
 
Liira, J.; Zobel, K.; M?gi, R.; Molenberghs, G. Vertical structure of herbaceous canopies: 
the importance of plant growth-form and species-specific traits. Plant Ecol. 2002, 
163, 123-134. 
 
Litton, C.M.; Ryan, M.G.; Knight, D.H. Effects of tree density and stand age on carbon 
allocation patterns in postfire lodgepole pine. Ecol. Appl. 2004, 14, 460-475. 
 
Loudermilk, E.L.; Cropper Jr, W.P.; Mitchell, R.J.; Lee, H. Longleaf pine (Pinus 
palustris) and hardwood dynamics in a fire-maintained ecosystem: A simulation 
approach. Ecol. Model., 2011, 222, 2733-2750. 
 
MacDonald, R.L.; Burke, J.M.; Chen, H.Y.; Prepas, E.E. Relationship between 
aboveground biomass and percent cover of ground vegetation in Canadian Boreal 
Plain riparian forests. Forest Sci. 2012, 58, 47-53. 
 
McGuire, J.P.; Mitchell, R.J.; Moser, E.B.; Pecot, S.D.; Gjerstad, D.H.; Hedman, C.W. 
Gaps in a gappy forest: plant resources, longleaf pine regeneration, and 
understory response to tree removal in longleaf pine savannas. Can. J. Forest Res. 
2001, 31, 765-778. 
Mitchell, J.E.; Bartling, P.N.S.; O'Brien, R. Understory cover-biomass relationships in 
the Front Range ponderosa pine zone. In Research note RM; USDA Forest 
Service, Rocky Mountain Forest and Range Experiment Station: Fort Collins, CO, 
USA, 1987. 
 
Mitchell, R.J.; Kirkman, L.K.; Pecot, S.D.; Wilson, C.A.; Palik, B.J.; Boring, L.R. 
Patterns and controls of ecosystem functions in longleaf pine ? wiregrass 
savannas. I. Aboveground net primary productivity. Can. J.  Forest Res. 1999, 29, 
743-751.  
41 
 
Moore, P.T.; Van Miegroet, H.; Nicholas, N.S. Relative role of understory and overstory 
in carbon and nitrogen cycling in a southern Appalachian spruce-fir forest AES 
Publication 7863. Utah Agricultural Experiment Station, Utah State University, 
Logan, Utah. Can. J. Forest Res. 2007, 37, 2689-2700. 
 
Muller, R.N. Nutrient relations of the herbaceous layer in deciduous forest ecosystems. In 
The herbaceous layer in forests of eastern North America; Gilliam, F.S., Roberts, 
M.R., Eds.; Oxford University Press: New York, NY, USA, 2003; pp. 15?37. 
 
Muller-Dumbois, D.; Ellenberg, H. Aims and Methods of Vegetation Ecology; Wiley: 
New York, NY, USA, 1974. 
 
Muukkonen, P.; M?kip??, R.; Laiho, R.; Minkkinen, K.; Vasander H., Fin?r, L. 
Relationship between biomass and percentage cover in understorey vegetation of 
boreal coniferous forests. Silva Fenn. 2006, 40, 231-245. 
 
Nabuurs, G.J.; Schelhaas, M.J.; Field, C.B. Temporal evolution of the European forest 
sector carbon sink from 1950 to 1999. Glob. Change Biol. 2003, 9, 152-160. 
 
Noss, R.F. Longleaf pine and wiregrass: keystone components of an endangered 
ecosystem. Nat. Area. J. 1989, 9, 211-213. 
 
NRCS, USDA. Soil Survey of Chattahoochee and Marion Counties, Georgia; U.S. 
Government Printing Office: Washington, DC, USA, 1997. 
 
NRCS, USDA. Soil Survey of Russell County, Alabama; U.S. Government Printing 
Office: Washington, DC, USA, 1999. 
 
Ohmann, L.F.; Grigal, D.F.; Rogers, L.L. Estimating plant biomass for undergrowth 
species of northeastern Minnesota forest communities; General Technical Report 
NC-61; U.S. Dept. of Agriculture, Forest Service, North Central Forest 
Experiment Station: St. Paul, MN, USA, 1981.  
 
Outcalt, K.W.; D.G. Brockway. Structure and composition changes following restoration 
treatments of longleaf pine forests on the Gulf Coastal Plain of Alabama. Forest 
Ecol. Manag. 2010, 259, 1615-1623.  
 
Pearcy, R.W. Responses of plants to heterogeneous light environments. In Handbook of 
Functional Plant Ecology; Pugnaire, F., Valladares, F., Eds.; CRC Press: Boca 
Raton, FL, USA, 1999. 
 
Pecot, S.D.; Mitchell, R.J.; Palik, B.J.; Moser, E.B.; Hiers, J.K. Competitive responses of 
seedlings and understory plants in longleaf pine woodlands: separating canopy 
influences above and below ground. Can. J. Forest Res. 2007, 37, 634-648. 
 
42 
Peet, R.K. Ecological classification of longleaf pine woodlands. In The Longleaf Pine 
Ecosystem; Jose, S., Jokela, E.J., Miller, D.L., Eds.; Springer New York: New 
York, NY, USA, 2006; pp 51-93. 
 
Peet, R.K.; Allard, D.J. Longleaf pine vegetation of the southern Atlantic and eastern 
Gulf Coast regions: a preliminary classification. In The longleaf pine ecosystem: 
ecology, restoration and management: proceedings, 18th Tall Timbers Fire 
Ecology Conference, May 30-June 2, 1991, Tallahassee, Florida.; Hermann, 
S.M., Ed.; Tall Timbers Research Station: Tallahassee, FL, USA, 1993; 18, pp. 
45-81. 
 
Peet, R.K.; Wentworth, T.R.; White, P. A flexible, multipurpose method for recording 
vegetation composition and structure. Castanea 1998, 63, 262-274.  
 
Peichl, M.; Arain, M.A. Above- and belowground ecosystem biomass and carbon pools 
in an age-sequence of temperate pine plantation forests. Agr. Forest Meteorol. 
2006, 140, 51-63. 
 
Rose, J.A.; Eddleman, L.E. Ponderosa pine and understory growth following western 
juniper removal. Northwest Sci. 1994, 68, 79-85. 
 
R?ttgermann, M.; Stenlein, T.; Beyschlag, W.; Dietz, H. Linear relationship between 
aboveground biomass and plant cover in low open herbaceous vegetation. J. Veg. 
Sci. 2000, 11, 145-148.  
 
Sorrie, B.A.; Weakley, A.S. Coastal Plain vascular plant endemics: phytogeographic 
patterns. Castanea 2001, 50-82. 
 
Sorrie, B.A.; Weakley, A.S. Conservation of the endangered Pinus palustris ecosystem 
based on Coastal Plain centres of plant endemism. Appl. Veg. Sci. 2006, 9, 59-66. 
 
Stanturf, J.A.; Goodrick, S.L.; Outcalt, K.W. Disturbance and coastal forests: a strategic 
approach to forest management in hurricane impact zones. Forest Ecol. Manag. 
2007, 250, 119-135. 
 
Vanha?Majamaa, I.; Salemaa, M.; Tuominen, S.; Mikkola, K. Digitized photographs in 
vegetation analysis?a comparison of cover estimates. Appl. Veg. Sci. 2000, 3, 89-
94. 
 
Van Lear, D.H.; Carroll, W.D.; Kapeluck, P.R.; Johnson, R. History and restoration of 
the longleaf pine-grassland ecosystem: implications for species at risk. Forest 
Ecol. Manag. 2005, 211, 150-165. 
 
Walker, J.L.; Cohen, S. Regenerating Longleaf Pine on Hydric Soils: Short-and Long-
term Effects on Native Ground-Layer Vegetation; Final Report - SERDP Project 
43 
SI-1303, 2009; http://prod.serdp-estcp.org/content/download/7241/92972/file/SI-
1303-FR.pdf (accessed Sept 20 2013). 
 
Walker, J.; Peet, R.K. Composition and species diversity of pine-wiregrass savannas of 
the Green Swamp, North Carolina. Vegetatio 1984, 55, 163-179. 
 
Ware, S.; Frost, C.; Doerr, P.D. Southern mixed hardwood forest: the former longleaf 
pine forest. Biodiversity of the southeastern United States: Lowland terrestrial 
communities; Martin, W.H., Boyce, S.G., Echternacht, A.C., Eds.; John Wiley 
and Sons, Inc.: Hoboken, NJ, USA, 1993; pp. 447-493. 
 
Whigham, D.F. Ecology of woodland herbs in temperate deciduous forests. Annu. Rev. 
Ecol., Evol. S. 2004, 35, 583-621. 
 
Wilson, J.B. Methods for fitting dominance/diversity curves. J. Veg. Sci. 1991, 2, 35-46. 
 
Wolters, G.L. Southern pine overstories influence herbage quality. J. Range Manage. 
1973, 423-426. 
 
Wolters, G.L. Longleaf and slash pine decreases herbage production and alters herbage 
composition. J. Range Manage. 1982, 761-763. 
 
Yarie, J. The role of understory vegetation in the nutrient cycle of forested ecosystems in 
the mountain hemlock biogeoclimatic zone. Ecology 1980, 1498-1514.