\documentclass[twocolumn,letterpaper,dvips]{article} \pagestyle{myheadings} \input{seteps}
\usepackage{mcfns} %using mcfns.sty version 9 3Feb09 -- NOTE the MCFNS.STY variables that have to be updated below
\usepackage{amsmath,bm,url} \usepackage[dvips]{graphicx} \usepackage{setspace,lineno} \usepackage[sort]{natbib}
\usepackage[breaklinks,pdfstartview={FitH -32768},pdfborder={0 0 0},bookmarksopen,bookmarksnumbered]{hyperref} %\usepackage{bibtexlogo}
%______________________________________________________________________
% Define MCFNS variables: 
% Moved to mcfns.sty like the \issueyear \publish	{Feb.~00} 		%Still not sure if Issue Manuscript publication date
\setcounter{page}{52}
\def\editors	{\href{mailto://mike@mcfns.com} {Editor:~Mike~Strub}}
\def\submit 	{Apr.~4,~2009} 	%Submission date can be different than the issue year \issueyear
\def\accept 	{Aug.~24,~2009} 		%The works should be Accepted & Published in the year of the Current_Issue \issueyear
\def\lasterrata	{Aug.~28,~2009} 	%Last Errata date can be different than the Issue-Year \issueyear
\def\citename	{Liu} 		%"Author" or "FirstAuthor et al."
\def\citeemail	{shangbinl@gmail.com} 	% Use later: {\href{mailto://\citeemail}{\citename}}
\def\citeetal	{~et~al.} 		% or {} %for a single author; or 
\author{	{\href{mailto://\citeemail}{Shangbin \citename}}$^1$, 		% Change only the first name of the first author
		{\href{mailto://c@cjci.net}{Chris J.\ Cieszewski}}$^2$ 		% Complete for other
}\affiliation{	\small\it{$^1$Biostatistician, {{}{Premier Research Group Limited, Ph./FAX: (678)279-4848/4880}}} \\ 
		\small\it{$^2$Professor, University of Georgia, Ph./FAX: (706)542-8169/8356}
}\def\yourtitle	{{Impacts of Management Intensity and Harvesting Practices on Long-term Forest Resource Sustainability in Georgia}} 				%need double {{ for \\ e.g.: {{Title \\ Subtitle}}
\def\yourkwords	{Intensive management practices; forestland; harvest level; rotation age; sustainability; simulation; OPTIONS; SRTS}
\def\yourabstract{ 
Using a spatially explicit forest management model called 
OPTIONS simulation analyses are conducted to investigate the impact of 
intensive management practices, rotation age, and harvest level on 
long-term wood production, harvest opportunities, and resource 
sustainability. The initial forest inventory is compiled from datasets of 
the USDA Forest Service Forest Inventory Analysis Unit, various GIS data, 
Landsat thematic mapper imagery, and simplified assumptions about the 
spatial distribution of different forest cover types. The parameters of the 
model are determined from published and unpublished literature, and from 
interviews with experts in the area of forest management in the Southeastern 
US. The sensitivity analyses reveal the impacts of various combinations of 
rotation age, harvest level and percentage of land put into intensive 
management and of the interaction of these factors on the sustainability of 
the forest resource production under the condition of a 4{\%} net reduction 
in the forestland area. The results of the analyses suggest that IMP acreage 
and rotation length are key factors in sustaining an increased harvest 
level. The volume available for harvest increases with an increasing rate of 
transition to intensively managed pine plantations (IMP rate) for each 
harvest level and rotation age. Even a reduced forest land base (4{\%} net 
reduction) in Georgia can easily sustain the current level of harvest with 
the current level of intensive pine plantation management for short and 
medium rotation ages. Increased pine plantation management intensity could 
lead to sustainable or even increased future wood production despite a 
decline in the forest land base and an increased wood demand. Timber growth 
would exceed removals in most of each of the projection periods. Throughout 
the projections the distribution of the harvestable volume by species group 
shows that the traditionally managed pine plantations (PSOF) contribute to 
the largest share of the total harvestable volume. The distribution of the 
harvest by species group indicates that the harvest come mainly from PSOF 
and IMP. The merits of definitions of the scenarios in this study are 
discussed and compared with those used in the subregional timber supply 
(SRTS) modeling. 
}%----------------------------------------------------------------------

% put any of your personal LaTeX definitions etc here.
\newcommand{\der}[2]{\frac{{\mathrm d}#1}{{\mathrm d}#2}}
\newcommand{\dr}[2]{{\mathrm d}#1/{\mathrm d}#2}
\newcommand{\dd}{\,{\mathrm d}}
\bibliographystyle{SAF}
\bibpunct{(}{)}{,}{a}{}{,}

% THE REST SHOULD BE AUTHOMATIC ... Go To the first Section ... 

\title{\Large\bf\uppercase\yourtitle} 
\begin{document} \markright{\hfil{{{\href{mailto://\citeemail}{\citename}\citeetal}}~(\issueyear)/\mcfnshead}} \twocolumn[ \begin{@twocolumnfalse} 
\maketitle \hypersetup{pdftitle={\mcfnshead}, pdfauthor={\citename~(\issueyear)}, pdfsubject={\yourtitle}, pdfkeywords={\yourkwords}}  \hrule 
\begin{abstract}\yourabstract\\\\{\bf Keywords:}~\yourkwords \end{abstract}\hrule\vspace{.3truein}\end{@twocolumnfalse}] 
%\numberwithin{figure}{section}


% Continue with the first Section: 

\section{Introduction} 

The forest products industry in Georgia is one of the most important 
contributors to the state's economy with an estimated total annual impact of 
{\$}28.5 billion in output, 141,155 jobs, and {\$}6.7 billion in 
compensation to employees and proprietors (Riall 2008). In addition to the 
direct economic benefit, Georgia's forests provide other indirect benefits, 
such as hunting, fishing, hiking, and other outdoor recreational 
opportunities. The forests help to maintain a clean water supply, conserve 
soil and provide habitat for many fish and wildlife species, some of which 
are presently endangered.

Current projections show an increasing demand on wood and wood products, 
while the areas of commercial forests are decreasing due to fast population 
growth and subsequent urban expansion and other development uses. While some 
agricultural areas have been converted to forestry use, as the state becomes 
more populated, urban/suburban expansion will result in fewer acres 
available for forestry production. At the same time, as the population of 
the state increases the demand for clean water and other non-timber 
forest uses such as hiking and camping will be increasing. This will further 
reduce the number of acres available for production forestry. Various 
possible regulatory constraints, such as mandated streamside management 
zones and road beautifying buffers may contribute to reduction of the 
commercial land base. Most of the projected net reduction is in the 
Southeast region, especially around fast-growing areas such as the 
Atlanta (Ahn et al. 2001, Dangerfield and Hubbard 2001, Ahn et al. 2002, 
Alig et al. 2002, Prestemon and Abt 2002, Wear and Greis 2002a,b, Alig et 
al. 2003, Alig and Butler 2004).

Forest management in the South has been intensifying over the past two 
decades, setting a trend that is expected to continue (Siry et al. 2001, 
Siry 2002). In the face of urban expansion and environmental pressure to 
reduce the numbers of acres dedicated to plantation forestry, the 
implementation of intensive management practices in pine plantations 
provides more production due to significantly increased growth rates (Sedjo 
and Botkin 1997, Daniels 1999, Borders and Bailey 2001, Alig et al. 2002, 
Martin and Shiver 2002).

Thus, a compelling question in this changing environment is whether 
Georgia's forests can sustain the needed production of forest products in 
the state while excluding a substantial part of the resource from non-forest 
industry related uses. Towards that end Cieszewski et al. (2004) proposed a 
simulation-based approach for analysis of the impact of various 
management practices and regulatory constraints on the resource and 
production sustainability. Based on limited data, simplified species groups 
and site index classification, arbitrarily selected harvest levels and basic 
assumptions of management regimes, Zasada et al. (2002) performed a study on 
the impacts of IMP on long-term sustainability of forest resources in 
Georgia. Liu et al. (2009) conducted a simulation study on long-term 
fiber supply assessment, which was based on detailed species group, site 
index, and rotation age classification and comprehensive harvest levels and 
management regimes, while the forestland base was kept unchanged and at the 
current level throughout the entire simulation period. In this study, the 
sensitivity analysis is performed with respect to impacts of rotation age, 
transition rate to IMPs, and harvesting limits on long-term forest 
resource sustainability of Georgia under the condition of a reduction in the 
forest land base, along with the application of updated yield tables of IMPs 
and different harvesting priorities. 

\section{Data}

The initial forest inventory is compiled from datasets of the USDA Forest 
Service Forest Inventory Analysis Unit, various GIS data, Landsat thematic 
mapper imagery, and simplified assumptions about the spatial distribution of 
different forest cover types. The database creation process and advantages 
of using Landsat imagery for landscape analysis were described in Liu et al. 
(2009).

\section{Methods}

A conceptual framework was proposed for analysis of various management 
practices and regulatory constraints and their impacts on resource and 
harvest sustainability (Cieszewski et al. 2004). Using the principles of 
this methodology, and applying them with comprehensive information on the 
related factors, such as impacts of the rotation ages, IMP rates, and 
harvest levels on the long-term fiber supply in Georgia we continue the 
investigation under the reduced forestland base assumption.

\subsection{Simulator, Species Group, Site Index Class, and Management 
Regime}

The simulator, definitions of species group and site index class, and 
management regimes for each species group are the same as in Liu et al. 
(2009). In brief, a spatially explicit forest estate model called OPTIONS 
from DR Systems Inc. is used to determine impacts of various forest 
management activities at the forest and stand levels. The simulation lengths 
are 200-years. The six broad species groups are natural pine stands, 
traditionally managed pine stands, intensively managed pine stands, upland 
and bottomland hardwoods, and oak-pine forests, abbreviated as NSOF, 
PSOF, IMP, UWDS, BHWD, and OAKP respectively. The seven site index classes 
are extremely low, very low, low, medium, high, very high, and extremely 
high, abbreviated as ELOW, VLOW, LOW, MED, HIGH, VHIG, EHIG respectively. A 
management regime is composed of combinations of individual silvicultural 
treatments, such as regeneration, fertilization, thinning, genetically 
improved stock, etc.

\subsection{Yield Table for IMP and Management Regime after Harvest}

The yield tables for NSOF, PSOF, UWDS, BHWD, and OAKP at each site index 
class are the same as in Liu et al. (2009). The yield tables for IMP are 
significantly different from in Liu et al. (2009), in which the IMP growth 
is assumed 2 times higher than in unmanaged stands. In this study, a 
state-of-the-art individual stand growth simulator called SiMS from 
ForesTech International, LLC is used to generate the yield tables of IMP 
(SiMS 2003). These tables consider the responses from genetically improved 
seedlings and fertilization, which are defined for management regimes of 
IMP. The responses from thinning are considered in OPTIONS as a part of the 
definition of scenarios. 

The definition of the management regimes after harvest is almost the same as 
in Cieszewski et al. (2004), i.e. 90{\%} of NSOF, 50{\%} of OAKP, 10{\%} or 
40{\%} UHWD, and 10{\%} of BHWD are converted to PSOF after harvesting, with 
a revision to manage as BHWD for BHWD stands with a site index class below 
MED, as per an unpublished report (Goetzl 1998). 

\subsection{Harvest Priorities}

Harvest priorities are set for wood types and species group. For priorities 
of wood types, the following assumptions are used: thinning should be 
performed as the first priority, the over-mature stands should be cut 
next, and the mature stands should be cut last. The definition of the 
harvest priority by species group is the same as in Cieszewski et al. 
(2004), i.e. in the order of IMP, PSOF, NSOF, OAKP, UHWD, and BHWD.

A minimum volume from thinning, which is equal to the initially qualified 
thinning volume from IMP, is set to ensure the qualified IMP stands will be 
thinned. Consequently the minimum harvest from IMP is the same as what we 
wished to thin from the IMP. The minimum harvesting volumes are set to 
ensure certain minimum wood supply for species groups with a lower priority 
in harvest (e.g., BHWD, UHWD, NSOF, and OAKP). The minimum volume that has 
to be cut is based on results of the recent FIA reports for Georgia 
(Thompson, 1998).

\subsection{Land Base Losses}

The land base loss is defined as the percentage of the projected net 
forestland reduction to all forestland acreage in a given region. The 
estimates of an average of 4{\%}~forestland loss for Georgia (see the 
discussion section in this paper for the details) are based on the recent 
national Renewable Resources Planning Act~[RPA] (Haynes et al. 2001, Haynes 
2003) and the Southern Forest Resource Assessment Consortium~(SOFAC) 
findings (Prestemon and Abt 2002). In this study, simulations are run based 
on 96{\%}~of the area of current forestland in Georgia.

\section{Investigated Factors}

\subsection{Rate of Transition to IMP}

The rate of transition to IMP is the proportion of stands to be artificially 
regenerated that will fall into intensive management. To test the potential 
impact of increasing the scale of intensive management on Georgia forestry 
and wood supply, 11 rates of transition to IMP, are considered~(0 to 100{\%} 
increments of 10{\%}). Current IMP acreage is about 30{\%} of all 
plantations. P0 means maintaining the current status of all plantations. P30 
means that 30{\%} regenerated plantation is managed as IMP, which is a 
scenario that is most likely (Goetzl 1998). As an extreme case, P100 assumes 
that all newly planted pine stands would be managed as IMP. P10, P20, P40, 
P50, P60, P70, P80, and P90 mean that 10{\%}, 20{\%}, 40{\%}, 50{\%}, 
60{\%}, 70{\%}, 80{\%}, and 90{\%} of regenerated plantation is managed as 
IMP, respectively.

\subsection{Harvest Level}

For the projections six harvest levels are considered, which is based on 0 
to 50{\%} harvest increases over limits at current level in increment of 
10{\%}. The current harvest level of removals in Georgia is approximately 
42.5~mm$^{3}$/yr [million m$^{3}$ per year] (Thompson 1998). H0 means 
keeping the current annual harvest during the entire simulation period. On 
the other hand, H40, the most likely case in practice (Haynes et al. 2001, 
Prestemon and Abt 2002, Haynes 2003), means harvesting 40{\%} more than the 
current harvest by 2050, i.e., 59.5~mm$^{3}$/yr, and then holding it 
constant. H10, H20, H30, and H50 mean harvesting 10{\%}, 20{\%}, 30{\%}, and 
50{\%} more than the current harvest by 2050 and then holding it constant, 
respectively. The period between 2000 and 2050 is evenly divided into 
10~intervals. The increase in the harvest between two adjacent intervals is 
one-tenth of the total harvest increase.

\subsection{Rotation Age}

Three types of rotation ages, short (S), medium (M), and long (L), are 
defined for each site/species group combination based on the similar 
research as what are used to define the yield tables. Values of rotation age 
are the same as in Liu et al. (2009). With respect to the short rotation 
age, the medium and long rotation ages are 5 and 10 years longer for NSOF, 
OAKP, UHWD, BHWD, and 2 and 5 years longer for PSOF and IMP.

\subsection{Execution of the Simulations}

The 198 simulations are run for all combinations (scenarios) of different 
levels of three factors: the IMP rate, rotation age, and harvest level. The 
model generates outputs that contain the volume inventory, forested area, 
volume available for harvest, and harvest by species group and by wood type 
at each year of the period of projection. The OPTIONS outputs are presented 
in figures illustrating potential trends in long-term fiber supply in 
Georgia.

\section{Results}

For convenience a scenario is denoted by the combination of any different 
levels (abbreviated) of the three investigated factors: harvest level, IMP 
rate, and rotation age. For example, H0-P0-S means a scenario of 
keeping the current harvest level (42.5~mm$^{3}$/yr), keeping current IMP 
levels (30{\%} of all pine plantation areas in Georgia), and using a short 
rotation age class during the entire simulation period. H40-P30-S 
means a scenario of 40{\%} more than the current harvest level by 2050 
(59.5~mm$^{3}$/yr) and then holding it constant, 30{\%} of stands to be 
artificially regenerated falling into intensive management, and use of a 
short rotation age class. H40-S means a scenario of 40{\%} more than the 
current harvest level by 2050 and then holding it constant, and the use of a 
short rotation age class.

\subsection{Impacts from Three Factors and Their Interactions on the Forest 
Resource Sustainability}

Volume available for harvest, or harvestable volume, which is the volume 
represented by stands at ages equal to or older than the harvest ages, shows 
similar trends over time for all scenarios. That is, the harvestable volume 
would initially decrease during the first two to three decades followed with 
an increase. After reaching a maximum, over the next century the volume 
available for harvest would remain steady at a level slightly lower than its 
peak (Fig.~1 and 2). Although 198 scenarios are modeled in this analysis, 
only parts of them are shown in the paper. However the trends shown in the 
paper are common to all scenarios.

To investigate the impacts on the sustainability of the fiber supply from 
the IMP rate and harvest level, the results from the short rotation age are 
used. Under the short rotation age the harvestable volume increases with an 
increasing IMP rate for different harvest levels. On average over the entire 
projection period, with H0-S applied, the volume available for harvest 
would vary from 952 to 1740~mm$^{3}$ with an increment of about 6{\%} for 
each IMP increase as IMP rate changes from P0 to P100 (Fig.~1a). Under 
H40-S it would vary from 560 to 1274~mm$^{3}$ with a 9{\%} increment 
(Fig.~1b). Under certain harvest levels the minimum harvestable volume comes 
early with the increase of the IMP rate. For example, under H40-S the 
years to reach the minimum harvestable volume change from 21 to 31~years as 
IMP rate changes from P100 to P0.

As expected, the total harvestable volume decreases with the increase of the 
harvest level for any given level of IMP rate (Fig.~1a,b,c). For example, 
with P30-S applied, the volumes available for harvest are 1116, 729, and 
629~mm$^{3}$, on average, for H0, H40, and H50 respectively. With the 
increase of the harvest level, it takes much more time to reach the minimum 
harvestable volume for any given level of IMP rate. For example, with P0 
applied, the years to reach the minimum harvestable volume are 23 and 
31~years for H0 and H40 respectively.

\begin{figure*}[htbp]
\centerline{\includegraphics[width=6.5in,height=2.4in]{291.eps}}
\centerline{\includegraphics[width=6.5in,height=2.4in]{292.eps}}
\centerline{\includegraphics[width=6.5in,height=2.4in]{293.eps}}
\caption{Volumes available for harvest of the different IMP rates over 
years under short rotation age for harvest levels of (a)~keeping the current 
harvest level (42.5~mm$^{3}$/yr) during the entire simulation period; 
(b)~40{\%} more than the current harvest level by 2050 (59.5~mm$^{3}$/yr) 
and then holding it constant; and (c)~50{\%} more than the current harvest 
level by 2050 (63.7~mm$^{3}$/yr) and then holding it constant.}
\label{fig1}
\end{figure*}

\begin{figure*}[htbp]
\centerline{\includegraphics[width=6.5in,height=2.4in]{294.eps}}
\centerline{\includegraphics[width=6.5in,height=2.4in]{295.eps}}
\centerline{\includegraphics[width=6.5in,height=2.4in]{296.eps}}
\caption{Volumes available for harvest and standing total inventory over 
time for scenarios of (a)~keeping the current harvest level 
(42.5~mm$^{3}$/yr) and current level of IMP (30{\%} of all pine plantation 
areas in Georgia) during the entire simulation period (H0-P0); (b)~40{\%} 
more than the current harvest level by 2050 (59.5~mm$^{3}$/yr) and then 
holding it constant and 30{\%} of stands that are to be artificially 
regenerated falling into intensive management (H40-P30); (c)~50{\%} more 
than the current harvest level by 2050 (63.7~mm$^{3}$/yr) and then holding 
it constant and 30{\%} of stands that are to be artificially regenerated 
falling into intensive management (H50-P30). Mature-S/M/L represents 
the volume available for harvest under short/medium/long rotation age. 
Inventory-S/M/L represents the inventory under short/medium/long rotation 
age.}
\label{fig2}
\end{figure*}

Even in the presence of a declining forestland (e.g. 4{\%} net reduction) 
Georgia can easily sustain the current harvest (Fig.~1a) and would sustain 
the increasing harvests (Fig.~1b,c) with the different levels of intensive 
management. With H0-P0-S applied, the harvestable volume is larger 
than its initial level, 621~mm$^{3}$, for most of the simulation period 
except for the first three decades when it initially decreases to 
199~mm$^{3}$ and then increases to reach its initial level. Beyond 2060 it 
would remain steady at twice its initial level (Fig.~1a, the square labeled 
line). Under H40-P0-S the harvestable volume decreases to 132~mm$^{3}$ 
in the first three decades, and then takes two decades to come back to its 
initial level. After that the volume would be stable at its initial level 
(Fig.~1b, the square labeled line). At H50-P0-S harvestable volume is 
lower than its initial level throughout the projections, at a level of about 
454~mm$^{3}$ for most of the simulation period, which is about 30{\%} lower 
than its initial level (Fig.~1c, the line labeled with squares). Thus, to 
keep the initial level of the harvestable volume beyond year~2050, at least 
30{\%}~(Fig.~1c, the line labeled with short dashes) of new plantations 
would have to be converted to intensive management.

Changes in rotation ages impact the volume available for harvest in a 
similar way, but with different magnitudes (Fig.~2, the lines labeled with 
hollow symbols). Throughout the simulation, the harvestable volume of the 
short rotation is larger than those of the medium and long rotations since 
the volume available for harvest is defined as the volume represented by 
stands at ages equal to or older than the rotation ages. Under the long 
rotation the forestland in Georgia cannot sustain the current harvest with 
the current IMP level~(Fig.~2a, the line labeled with hollow triangles). 
With an increasing harvest level applied, even with part of the stands to be 
artificially regenerated being converted to intensive management, 
sustainability in the fiber supply cannot be assured~(Fig.~2b,c, the 
hollow-triangle labeled lines). Even though the harvestable volumes are 
different among three rotation age classes, the total volume 
inventory~(Fig.~2, lines labeled with solid symbols) shows no significant 
differences in magnitude among them. At the beginning of the projections the 
total volume of the short rotation is equal to or slightly larger than those 
of the medium and long rotations, while beyond a certain year, which depends 
on the harvest level and IMP rate, the total volume of the long rotation 
would be the largest and that of the short rotation would be the smallest. 
Despite the initial decrease in volume available for harvest, the total 
volume shows a very slight decrease in the first decade and follows a steady 
increase for three to four decades. After that it would slowly level off.

\subsection{Two Investigated Cases}

Two scenarios, abbreviated as H0-P0-S and H40-P30-S, are 
designated as the ``investigated cases'' for two reasons. First, the 
H40-P30-S would occur most likely. Its assumption is consistent with 
the findings of the SOFAC and RPA projections on harvest (H40), the findings 
of the AF{\&}PA survey on southern forest management intensity (P30), and 
the empirical findings of the rotation ages in practice (S); Second, to 
investigate the impacts on the fiber supply from IMP rate and harvest level, 
H0-P0-S can be considered as a ``control'' in the experiment.

The increased pine plantation management intensity could lead to sustainable 
or even increased future wood production despite a decline in the forest 
land base and an increased wood demand (Fig.~3, 4, 5 and 6). Such 
productivity increases could prevent timber shortages, at least for 
pulpwood.

With H40-P30-S, a scenario that most likely will occur, in the first 
several years of the projection the removals slightly exceed growth. Then 
the timber growth is projected to exceed removals till 2050. After that they 
would be close to each other (Fig.~3). The volume available for harvest 
would quickly decrease from 621 to 185~mm$^{3}$ during the first two 
decades, and then rapidly increase in the next 15 years to reach its initial 
level. After this quick increase, it would gradually reach the maximum of 
910~mm$^{3}$ around 2070. After reaching the maximum, it would remain steady 
at a level of about 820~mm$^{3}$, which is about 30{\%} larger than its 
initial level (Fig. 5a,b). Correspondingly, under H0-P0-S the 
harvestable volume shows an almost identical pattern as H40-P30-S in 
the first three decades. After that it reaches a higher maximum of 
1196~mm$^{3}$ by the end of this century. Over the next century it would 
remain steady at a level of about 1107~mm$^{3}$, which is about twice its 
initial level (Fig. 4a,b).

The portions of the volume available for harvest by wood type vary between 
different simulations. With H40-P30-S applied the portions of mature 
and over-mature volume start out initially at about 325 and 268~mm$^{3}$ 
respectively. After dropping in the first three decades, the mature volume 
becomes 5~times that of the over-mature volume, on average 511 and 
102~mm$^{3}$ respectively, in the following two decades. After that the two 
volumes would remain steady at on average 412~mm$^{3}$, moderately larger 
than their initials (Fig. 5a). With H0-P0-S applied the mature volumes 
are larger than the over-mature volumes throughout the projections except 
for the first several years. Overall, the mature volume is nearly twice that 
of the over-mature volume, 599 and 349~mm$^{3}$ respectively, for the 
entire projection period (Fig. 4a).

On the other hand, the harvestable volume by species group shows a similar 
pattern for two investigated scenarios, with differences in magnitude (Fig. 
5b and 4b). Under H0-P0-S the harvestable volume of PSOF, which is the 
largest share of the total harvestable volume, shows a typical 
decrease-increase-constant pattern throughout the projections (Fig. 5b 
and 4b, thick diamond labeled lines). However, with H40-P30-S applied 
it is much lower than that under H0-P0-S beyond 2035. The shares of 
PSOF are 29{\%} and 51{\%}~of the total harvestable volume for 
H40-P30-S and H0-P0-S respectively. Throughout the projection 
the harvestable volume of IMP fluctuates at a low level compared with those 
of other species groups. Since the IMP has the first priority in harvest, 
any mature volume in this type of stand is always harvested once it becomes 
available (Fig. 5c and 4c), which is quite common in real forest management 
practices. The harvestable volume of IMP under H40-P30-S is much 
higher than that under H0-P0-S. Over the entire projection period, 
66{\%}~of the years have harvestable volumes of IMP between 20 and 
40~mm$^{3}$ under H40-P30-S (Fig. 5b). While under H0-P0-S 
82{\%}~of the years have harvestable volumes of IMP less than 20~mm$^{3}$ 
(Fig. 4b). Since 30{\%}~of stands to be artificially regenerated will fall 
into intensive management, there is an increase in the harvestable volume in 
IMP under H40-P30-S. With H40-P30-S applied, the harvestable 
volumes of NSOF, UHWD, and OAKP show a decrease-increase-level off 
pattern. The availability of harvestable volume in BHWD remains unchanged 
during the whole simulation period (Fig. 5b).

The distribution of the harvest (cut volume) by species group indicates that 
the harvest comes mainly from the pines, which includes IMP, PSOF, and NSOF. 
Overall 79{\%}~(45{\%} from IMP, 31{\%} from PSOF, and 3{\%} from NSOF) and 
71{\%}~(24{\%} from IMP, 44{\%} from PSOF, and 3{\%} from NSOF) of the 
harvest would be harvested from the pine stands for H40-P30-S (Fig. 
5c) and H0-P0-S (Fig. 4c) respectively. The harvests from hardwoods 
(UHWD and BHWD) and OAKP account for 20{\%} and 8{\%} of the total harvest 
in H0-P0-S respectively. With H40-P30-S hardwoods contribute 
15{\%} of the total harvest and OAKP make up the rest of 6{\%}.

\begin{figure*}[htbp]
\centerline{\includegraphics[width=7.00in,height=2.4in]{297.eps}}
\caption{OPTIONS model projections of timber growth and removals volumes in 
the state of Georgia, 1997 to 2200, under H40-P30-S assumptions of 
40{\%} more than the current harvest level by 2050 (59.5~mm$^{3}$/yr) and 
then holding it constant, 30{\%} of stands to be artificially regenerated 
falling into intensive management, and use of a short rotation age.}
\label{fig3}
\end{figure*}

\begin{figure*}[htbp]
\centerline{\includegraphics[width=6.5in,height=2.4in]{298.eps}}
\centerline{\includegraphics[width=6.5in,height=2.4in]{299.eps}}
\centerline{\includegraphics[width=6.5in,height=2.4in]{2910.eps}}
\caption{Harvest sources for a scenario of keeping the current harvest 
level (42.5~mm$^{3}$/yr), keeping current IMP levels (30{\%} of all pine 
plantation areas in Georgia), and using a short rotation age (H0-P0-S) 
during the entire simulation period. (a)~the volume available for harvest by 
wood type; (b)~the volume available for harvest by species group; and 
(c)~the harvested volume by species group. The pine includes IMP, PSOF, and 
NSOF.}
\label{fig4}
\end{figure*}

\begin{figure*}[htbp]
\centerline{\includegraphics[width=6.5in,height=2.4in]{2911.eps}}
\centerline{\includegraphics[width=6.5in,height=2.4in]{2912.eps}}
\centerline{\includegraphics[width=6.5in,height=2.4in]{2913.eps}}
\caption{Harvest sources for a scenario of 40{\%} more than current harvest 
level by 2050 (59.5~mm$^{3}$/yr) and then holding it constant, 30{\%} of 
stands to be artificially regenerated falling into intensive management, and 
use of a short rotation age (H40-P30-S). (a)~the volume available for 
harvest by wood type; (b)~the volume available for harvest by species group; 
and (c)~the harvested volume by species group. The pine includes IMP, PSOF, 
and NSOF.}
\label{fig5}
\end{figure*}

\begin{figure*}[htbp]
\centerline{\includegraphics[width=6.5in,height=2.4in]{2914.eps}}
\centerline{\includegraphics[width=6.5in,height=2.4in]{2915.eps}}
\centerline{\includegraphics[width=6.5in,height=2.4in]{2916.eps}}
\caption{Changes in the state of forest over years for a scenario of 40{\%} 
more than current harvest level by 2050 (59.5~mm$^{3}$/yr) and then holding 
it constant, 30{\%} of stands to be artificially regenerated falling into 
intensive management, and use of a short rotation age (H40-P30-S). 
(a)~The volume inventory (mm$^{3})$ by maturity; (b)~the volume inventory by 
species group; and (c)~the forested area (1000ha) by species group.}
\label{fig6}
\end{figure*}

Different from the harvestable volume, which is a part of total volume, the 
total volume by maturity indicates that with H40-P30-S the mature 
(including the mature and over-mature mentioned in the analysis of the 
volume available for harvest by wood type) initially is about twice of the 
immature, 593~vs. 308~mm$^{3}$. During the first four decades they go in the 
opposite directions, i.e. with a decrease-increase pattern in the mature 
and an increase-decrease pattern in the immature. At about the fourth 
decade of this century, the mature again exceeds the immature. After that, 
the two volumes would remain steady at levels moderately larger than their 
corresponding initial values (Fig. 6a). The distribution of the total volume 
by species groups (Fig. 6b) changes the most in the planted (PSOF and IMP) 
and natural pines (NSOF). With H40-P30-S applied the proportions of 
PSOF and IMP increase from about 14{\%} and 2{\%} of the total volume at the 
beginning of the simulation to about 27{\%} and 20{\%} of the total volume 
at the end of the simulation. The large increase in the volume of planted 
pines comes mostly at the expense of the NSOF (decreases from 25{\%} to 
9{\%}), UHWD (decreases from 22{\%} to 14{\%}), and BHWD (decreases from 
24{\%} to 19{\%}). Similar to the volume changes by species group, the area 
(Fig. 6c) also changes the most in the planted and natural pines. The areas 
of PSOF and IMP increase from about 18{\%} and 8.5{\%} to 30{\%} and 
21.5{\%} of the total area respectively. The area of NSOF drops from 20{\%} 
to 6{\%}. The area of OAKP decreases from 15{\%} to 8{\%} and UHWD decreases 
from 21{\%} to 17{\%}. The area of BHWD remains unchanged throughout the 
simulation.

\section{Discussion}

The goal of this study is to investigate the potential impacts of the 
individual factors, intensive pine plantation management, rotation age, and 
harvest level, and the interaction of these factors on the production and 
sustainability of forest resources in Georgia. There are certain necessary 
simplifications and uncertainties in the assumptions that could affect the 
results. The reliability of such projections becomes progressively lower as 
the time projected into the future increases. Yet the general trends 
revealed in this study are based on state-of-the-art analysis and 
are correct and conclusive.

The assumption that the intensive management plantations are introduced in 
the mid-eighties (see Liu et al. 2009 for the definition of IMP) implies 
that these plantations are still not available for harvest. Since the 
intensive plantations produce most of the harvested volume in the study 
area, all simulations are characterized by an initial reduction of available 
volumes.

\subsection{Increased Wood Demand and Declining Forestland Base}

Current projections show an increasing demand on wood and wood products from 
decreasing areas of commercial forests due to fast population growth and 
subsequent urban expansion and development uses. The most recent RPA 
predicted that under assumptions regarding economic trends through 2050, 
U.S. timber harvest would increase 38{\%} (Haynes et al. 2001, Haynes 2003). 
Meanwhile the timberland area is projected to decline by 3 percent. The 
SOFAC made projections for the South independently from RPA projections by 
using the subregional timber supply (SRTS) model (Adams and Haynes 1996, Abt 
et al. 2000). In aggregate, harvests from private lands were projected to 
increase by 53{\%} (56{\%} and 49{\%} for softwoods and hardwoods 
respectively) and private timberland will decrease by 1{\%} between 1995 and 
2040 under the base scenario, abbreviated as IH for inelastic demand-high 
plantation growth rate increase~(Prestemon and Abt 2002). Harvests from 
private lands and land losses were projected to change unevenly across the 
South. State-level projected harvests and land losses in private 
timberland under the~IH showed that the harvests would increase from 38.6 to 
51.0~mm$^{3}$ or by 32{\%} (33{\%} and 30{\%} for softwoods and hardwoods 
respectively) between 1995 and 2040 for the state of Georgia. Meanwhile, the 
projected net forestland base reduction was about 4{\%} for the state of 
Georgia; 3.1{\%}, 19.9{\%}, 3.9{\%}, -39.8{\%}, and -19.9{\%} 
(``-`` means converting to non-forest uses; data from a support table 
in Prestemon and Abt 2002) for southeast, southwest, central, north central, 
and north of Georgia respectively. The distribution of the land loss by 
analysis survey unit is consistent with other available reports~(Ahn et al. 
2001, Dangerfield and Hubbard 2001, Ahn et al. 2002, Wear and Greis 2002, a, 
b, Alig et al. 2003, Alig and Butler 2004), which project that most of the 
projected net reduction was in the Southeast region, especially around 
fast-growing areas such as Atlanta.

The assumption of the harvest and net reduction on the forest land base for 
the state of Georgia is consistent with the above projections done for this 
study, especially for the investigated case H40-P30-S. The 4{\%} net 
reduction in the forest land base is set by considering the variation in 
five Forest Inventory Analysis survey units. Compared with the projection of 
the SOFAC, which projected a 32{\%} increase on harvest between 1995 and 
2040 by using the SRTS model, a scenario is investigated with a 40{\%} 
increase to the current harvest by 2050. The harvests are compiled in the 
simulator, OPTIONS, and are very close to the modified SRTS values (Fig.~7, 
SRTS-MOD). Since the differences between the projected harvest at 1997 by 
the SRTS model and those estimated at 1997 by FIA (Thompson 1998) are 38.9 
and 42.5~mm$^{3}$ respectively, the harvest is modified to keep the same 
rate as SRTS. If the harvest projected by SRTS occurs, the conclusions on 
sustainability of forest resources are still valid, or even stronger, since 
a higher level of harvest is utilized in the OPTIONS than in SRTS.

\begin{figure*}[htbp]
\centerline{\includegraphics[width=6.5in,height=2.4in]{2917.eps}}
\caption{Harvests used in OPTIONS and projected by SRTS model.}
\label{fig7}
\end{figure*}


\subsection{Pine Plantation Management Intensity}

Forest management in the South has been intensifying over the past two 
decades, setting a trend that is expected to continue (Siry et al. 2001, 
Siry 2002). Recent declines in harvest on public lands in the West have 
significantly deviated from historic growth/removal patterns and have placed 
more pressure on eastern forests that are predominantly in private ownership 
(Smith et al. 2004). The Southeast, with its shorter rotation and a greater 
possibility to augment growth and yield through intensive management 
practices, represents a region with the potential to retain a portion of the 
market share lost by timber producers in the West (Alig 1990, Garcia and Abt 
1997, Cubbage et al. 1998, Joyce and Birdsey 2000, USDA 2001, Adams 2002, 
Haynes 2002, Mills and Zhou 2003).

The simulation results in this study suggest that the reduced (4{\%}) 
timberland in Georgia can easily sustain the current level of harvest (H0) 
with the current level of intensive management (P0) for different choices of 
the rotation age classes. While with an increased harvest, the 
sustainability of the wood supply would depend on the rotation ages and also 
on an increase in IMP. The results of the analysis also indicate that the 
minimum harvestable volume is closer to the annual harvest with the harvest 
level increase. It would imply that part of new plantations would have to be 
converted to intensive management to assure sustainability of forest 
resources under an increasing harvest scenario. The higher the harvest 
level, the more plantations would have to be converted.

The results of the analysis show that the increased harvest could be 
supplied by increasing intensive pine plantation management. By comparison 
of the cases of keeping the current harvest level and IMP rate 
(H0-P0-S), and of applying an increasing harvest level and an 
increasing IMP rate (H40-P30-S), the majority of harvest comes from 
PSOF and IMP. The different patterns between them should be that the 
proportions of the harvest from PSOF and IMP are reversed, that is, the main 
part of the harvest comes from PSOF~(44{\%}) under H0-P0-S (Fig. 4c). 
While with H40-P30-S applied IMPs~(45{\%}) account for the major part 
of the harvest (Fig. 5c). In both cases of H40-P30-S and H0-P0-S 
the harvests from PSOF are similar, mainly varying from 5 to 30~mm$^{3}$. 
While under H40-P30-S the harvest from IMP is much higher than that 
under H0-P0-S. Compared with the harvest from IMP in H0-P0-S, 
which changes from 0 to 31~mm$^{3}$, the harvest from IMP in H40-P30-S 
changes from 20 to 47~mm$^{3}$. For the total 204~years, the percentages of 
the years with a harvest larger than 20~mm$^{3}$ are 79{\%} and 25{\%} for 
H40-P30-S and H0-P0-S respectively. According to Goetzl (1998) 
the current average IMP rate should be between 25{\%} and 50{\%}, which 
would assure sustainable wood supply in Georgia even under an increased 
harvest scenario~(Fig.~1). The fact that 66{\%}~of the years had a 
harvestable volume of IMP between 20 and 40~mm$^{3}$ under H40-P30-S 
suggests that with an assumption that 30{\%} of newly regenerated stands 
would convert to IMP, the sustainable harvest level could rise beyond 
59.5~mm$^{3}$/yr, i.e., 40{\%} more than the current harvest level by 2050, 
and then hold constant (Fig. 5b).

\subsection{Advantages of this Study}

Most of values of simulation parameters used in the scenario H40-P30-S 
are compiled from outputs of SOFAC's projection using the SRTS model. These 
values include changes in harvest level, forestland, forest cover type, etc. 
As expected, general trends in growth and removal, volume inventory and 
forested area by species group from H40-P30-S are consistent with 
trends from the base scenario (IH) in SRTS. Advantages of this study over 
SRTS modeling are as follows: the planted pine is split into two parts, PSOF 
and IMP. The proportion of IMP is consistent with the literature (Zasada et 
al. 2002). All factors compiled in the OPTIONS model are different for PSOF 
and IMP. For example, the yield tables are based on the most reliable model 
system in this area: the PMRC (Harrison and Borders 1996, Borders et al. 
2004) for PSOF and SiMS for IMP. The significantly different management 
regimes and the management regimes after harvest are applied for PSOF and 
IMP. Correspondingly, there is no such split in SRTS modeling. Throughout 
the projections an increase in the pine plantation growth rate by ownership, 
75{\%} by 2040 for the forest industry and 37.5{\%} for the non-industry 
private forest (NIPF), was considered in SRTS.

\subsection{Future Directions and Related Studies}

A meaningful sustainability analysis is an ongoing process. In fact, most 
analyses show that forest management intensities on pine plantations will be 
the key in determining inventory, harvest and price levels for softwoods in 
the South in the future. To obtain better and more reliable results the 
method of assigning acres of intensively managed plantations to records of 
the database need to be revised and possibly some differences in management 
as relates to physiographic regions should be introduced.

So far, the intensively managed pine plantations have been mainly 
concentrated on the forestland owned by forest industries. The NIPF owner 
does not yet apply the intensive management practices but may in the future 
(Goetzl 1998, Moffat et. al 1998). On the other hand, the dramatic land 
ownership change is currently in progress, i.e. Timberland Investment 
Management Organizations and Real Estate Investment Trust ownerships are 
taking industry stands. Will they maintain/decrease/increase acres of IMP? 
To obtain better and more reliable results the ownership of the forested 
land needs to be considered since it causes significant differences in 
decisions on the future management and intensity of management.

Since pine pulpwood has such low value and paper-making capacity has been 
dramatically reduced in the South over the past few years, more emphasis is 
now placed on sawtimber production. For a complete understanding of the 
impacts of rotation ages on forest resources sustainability, a type of 
rotation age much longer than those used here needs to be investigated.

\section*{Acknowledgements}
{We are grateful for helpful comments provided by the editor and three anonymous referees.}

\section*{Literature Cited}
\begin{description}

\item Abt, R. C., F. W. Cubbage, and G. Pacheco. 2000. Southern forest resource 
assessment using the subregional timber supply (SRTS) model. For. Prod. J. 
50 (4), 25--33.

\item Adams, D. M. and R. W. Haynes. 1996. The 1993 timber assessment market 
model: structure, projections, and policy simulations. Gen. Tech. Rep. 
PNW--GTR--368. Portland, OR: USDA Forest Service Pacific Northwest Research 
Station. 58 p.

\item Adams, D. M. 2002. Solid wood products: rising consumption and imports, 
modest price growth. J. For., 100(2):14-19.

\item Ahn, S., A. J. Plantinga, and R. J. Alig. 2001. Historical trends and 
projections of land use for the South-Central United States. Res. Pap. 
PNW-RP-530. Portland, OR: USDA Forest Service Pacific Northwest 
Research Station. 20 p.

\item Ahn, S., A. J. Plantinga, and R. J. Alig. 2002. Determinants and projections 
of land use for the South-Central United States. South. J. Appl. For.. 
26(2):78--84.

\item Alig, R. J. 1990. Changes in area of timberland in the United States, 
1952-2040, by ownership, forest type, region and state. Publication 
SE-64. Asheville, NC: USDA Forest Service Southeastern Forest Experiment 
Station. 34p.

\item Alig, R. J., J. R. Mills, and B. J. Butler. 2002. Private timberlands: 
growing demands, shrinking land base. Journal of Forestry. 100(2): 32--37.

\item Alig, R. J., A. J. Plantinga, S. Ahn, and J. D. Kline. 2003. Land use 
changes involving forestry in the United States: 1952 to 1997, with 
projections to 2050. Gen. Tech. Rep. PNW-GTR-587. Portland, OR: USDA 
Forest Service Pacific Northwest Research Station. 92 p.

\item Alig, R. J. and B. J. Butler. 2004. Area changes for forest cover types in 
the United States, 1952 to 1997, with projections to 2050. Gen. Tech. Rep. 
PNW-GTR-613. Portland, OR: USDA Forest Service Pacific Northwest 
Research Station. 106 p.

\item Borders, B. E. and R. L. Bailey. 2001. Loblolly pine - pushing the limits 
of growth. South. J. Appl. For.25(2): 1-6.

\item Borders, B. E., W. M. Harrison, Y. Zhang, B. D. Shiver, M. L. Clutter, C. J. 
Cieszewski, and R. F. Daniels. 2004. Growth and yield models for second 
rotation loblolly pine plantations in the piedmont/upper coastal plain and 
lower coastal plain of the Southeastern U.S. PMRC Tech. Rep. 2004. 
Plantation Management Research Cooperative. Daniel B. Warnell School of 
Forest Resources. University of Georgia. Athens, GA. 63p.

\item Cieszewski, C. J., M. Zasada, B. E. Borders, R. C. Lowe, J. Zawadzki, M. L. 
Clutter and R. F. Daniels. 2004. Spatially Explicit Sustainability Analysis 
of Long-term Fiber Supply in Georgia, USA. Forest Ecology and Management. 
187(2-3): 349-359.

\item Cubbage F. W., J. P. Siry, R. C. Abt, D. N. Wear, and S. Moffat. 1998. 
Forest productivity and timber supply modeling in the South. In: Proceedings 
of conference ``Improving Forest Productivity for Timber{\ldots} A Key to 
Sustainability'' held in Duluth, MN, December 1-3, 1998.

\item Dangerfield, C. W. and W. H. Hubbard. 2001. Forestry Situation and Outlook 
In the U.S. South, 2001. Center For Forest Business. Warnell School of 
Forest Resources. Athens, GA. Research Note No. 7.

\item Daniels, R. F. 1999. Sustainability and forest industry: growing more wood 
than ever before. In: Proc. 1998 SAF Natl. Conv. Society of American 
Foresters, Bethesda, MD, USA. 75-80.

\item Goetzl, A. 1998. AF{\&}PA southern forest management intensity survey: Data 
summary and survey results. Unpublished report. Washington, DC: American 
Forest and Paper Association.

\item Harrison W. M., and B. E. Borders. 1996. Yield Prediction and Growth 
Projection For Site-prepared Loblolly Pine Plantations in the Carolinas, 
Georgia, Alabama and Florida. PMRC Tech. Rep. 1996-1 Plantation 
Management Research Cooperative. Daniel B. Warnell School of Forest 
Resources. University of Georgia. Athens, GA. 64p.

\item Haynes, R. W., D. M. Adams, R. J. Alig, and others. 2001. Projections of the 
U.S. timber supply and demand situation to 2050: draft findings from the 
USDA Forest Service 2000 RPA Timber Assessment. Proceedings of the 2000 
Society of American Foresters national convention, Washington, DC, November 
16-20, 2000. Bethesda, MD: Society of American Foresters, 2001: Pages 
451-453.

\item Haynes, R. W. 2003. An analysis of the timber situation in the United 
States: 1952 to 2050. Gen. Tech. Rep. PNW-GTR-560. Portland, OR: USDA 
Forest Service Pacific Northwest Research Station. 254p.

\item Joyce, L. A., and R. Birdsey. 2000. The impact of climate change on 
America's forests. Gen. Tech. Rep. RMRS-GTR-59. Fort Collins, CO: USDA 
Forest Service Rocky Mountain Research Station. 133p.

\item Liu, S., C. J. Cieszewski, R. C. Lowe, and M. Zasada. 2009. Sensitivity 
analysis on long-term fiber supply simulations in Georgia. South. J. Appl. 
For. 33:81-90.

\item Martin, S. W. and B. D. Shiver. 2002. Impacts of vegetation control, genetic 
improvement and their interaction on loblolly pine growth in the southern 
United States-age 12 results. South. J. Appl. For. 26:37-42.

\item Mills, J. R. and X. Zhou. 2003. Projecting national forest inventories for 
the 2000 RPA timber assessment. Gen. Tech. Rep. PNW-GTR-568. Portland, 
OR: USDA Forest Service Pacific Northwest Research Station. 58p.

\item Moffat S., F. W. Cubbage, A. Cascio, and R. Sheffield. 1998. Estimations of 
future forest management intensity on NIPF lands in the South: Results of 
the Southern State Foresters' Survey. Working Paper Series, SOFAC Report 14. 
Research Triangle Park, NC: Southern Forest Resource Assessment Consortium.

\addtolength{\textheight}{-4.25truein}

\item Perez-Garcia, J. and R. C. Abt. 1997. An assessment of southern timber 
markets in a global context: Scarcity of timber resources in the South. 
SOFAC Report No. 2. Southern Forest Assessment Consortium, Working Paper 
Series. 23 p.

\item Prestemon, J. P. and R. C. Abt. 2002. Chapter 13 (TIMBR-1): Timber 
Products Supply and Demand. In: The Southern Forest Assessment Final 
Technical Report. Gen. Tech. Rep. SRS-53. USDA Forest Service, Southern 
Research Station, Asheville, NC.

\item Riall, W. B. 2008. Economic Benefits of the Forestry Industry in Georgia: 
2007. Report of the Georgia Tech Research Corporation, Atlanta, GA, prepared 
for the Georgia Forestry Commission.

\item Sedjo R. A. and D. Botkin. 1997. Using Forest Plantations to Spare Natural 
Forests. Environment. Vol. 39, Issue 10: 15-20, 30.

\item Smith, W. B., P. D. Miles, J. S. Vissage, and S. A. Pugh. 2004. Forest 
Resources of the United States, 2002. General Technical Report NC-241. 
St. Paul, MN: U.S. Dept. of Agriculture, Forest Service, North Central 
Forest Experiment Station.

\item SiMS. 2003. ForesTech International, LLC. \underline 
{http://www.forestech.us}

\item Siry, J. P., F. W. Cubbage, and A. J. Malmquist. 2001. Potential impacts of 
increased management intensities on planted pine growth and yield and timber 
supply modeling in the South. Forest Prod. J. 51 (3): 42-48.

\item Siry, J. P. 2002. Chapter 14 (TIMBR-2): Intensive Timber Management 
Practices. In: The Southern Forest Assessment Final Technical Report. Gen. 
Tech. Rep. SRS-53. USDA Forest Service, Southern Research Station, 
Asheville, NC.

\item Thompson M. T. 1998. Forest Statistics for Georgia, 1997. Resource Bulletin 
SRS-36. Asheville, NC: USDA Forest Service Southern Research Station. 
92p.

\item USDA Forest Service. 2001. 2000 RPA Assessment of forest and range lands. 
Publication FS-687. Washington, DC: U.S. Department of Agriculture, 
Forest Service. 78p.

\item Wear, D. N. and J. G. Greis. 2002a. Southern forest resource assessment: 
summary of findings. Journal of Forestry. 100(7): 6--15.

\item Wear, D. N. and J. G. Greis. 2002b. Southern forest resource assessment: 
summary report. Gen. Tech. Rep. SRS-54. Asheville, NC: USDA Forest 
Service Southern Research Station. 103p.

\label{docend}
\item Zasada M., C. J. Cieszewski, and B. E. Borders. 2002. Impact of Intensive 
Management Practices on Long-Term Simulations of Forest Resources in 
Georgia. In: Proceedings of the 2002 SAF National Convention. Society of 
American Foresters, Winston-Salem, NC, USA. October 5-9, 2002. 64p.
\end{description}
\end{document}