The Southern Timber Market to 2040

Jeffrey P. Prestemon and Robert C. Abt

Timber market analysis of the South’s predominantly private timberland finds that the 13southern states produce nearly 60 percent of the nation’s timber, an increase from the midmid-1900s. Projections with the Subregional Timber Supply model show that, despite a 67 percentincrease in the area of pine plantations, the South will experience a 2 percent decline in privatetimberland area as other forest types shrink. Because of expected productivity gains for plan-tation forests and conversions of some agricultural lands to natural forests, the South’s indus-trial wood output is projected to increase by more than 50 percent between 1995 and 2040.

Keywords: economics; harvesting; plantations

S outhern timber markets have un-dergone a steady transformationover the past 40 years. Invest-

ments in plantation and intensive forestmanagement, population growth, andthe decline of the southern agriculturalsector have moved the timber sectorfrom an equal player with western mar-kets to one dominant in North Amer-ica and the world. Southern timber isdemanded and supplied at increasinglyhigher levels. The South is the world’slargest single industrial wood producer,and its influence is felt in internationalmarkets for many kinds of timber prod-ucts-from kraft paper to hardwoodlumber to fine furniture. Changing de-mographic characteristics and growingdemand for nontimber values obtainedfrom forests have given rise to concernsabout increasing harvests, intensive for-est management, and the future of nat-ural forests.

The Southern Forest Resource As-sessment (SFRA) addressed questionsregarding the history, status and pro-jected future demands for and supplies

of wood products in the South (Preste-mon and Abt, in press).

Histor ica l Background

The South produced 41 percent ofthe country’s wood fiber output in

Timber production in the UnitedStates has shifted among regions since

1952 and 58 percent in 1997. Over the

the 18OOs, and today finds the South

same period, the South’s share of the

the primary producing region in thecountry (Haynes et al. 2002). Because

world’s industrial wood production

the South produces most of its timberproducts from private forests, the in-crease in output observed in the South

rose from 6.3 to 15.8 percent. Today,

implies that investment opportunitiesfor intensive forest management on

fully 18 percent of the world’s indus-

these lands and subsequent product

trial timber products comes from

manufacture have improved in theSouth relative to other regions (Murrayand Wear 1998; Guan and Munn2000).

southern states. Although the UnitedStates has remained the world’s largestproducer of industrial timber for thepast 40 years, producing a stable 25percent of total world supply, risingsouthern output means that the South’sshare of that production has grownsubstantially.

Model ing Approach

SRTS has several exogenous inputsand assumptions-prespecified vari-ables that set the context of the modelsolution. One variable that affects fu-ture harvest behavior is the level of na-

Trends in the southern timber sectorwere projected with the Subregional

tional timber demand, which had been

Timber Supply (SRTS) model (Abt etal. 2000). The SRTS model projects

continually rising, at least until the lat-

private timber inventories, growth, re-movals, prices, land use, and timber-

est economic recession. However, de-

land area by five broad forest manage-ment types at substate and ecoregion(Bailey 1995) levels. Results are brokenout by broad ownership categories:nonindustrial private forestry (NII’F)and industry (which includes nonforestindustry corporate owners, such as tim-ber investment management organiza-tions). The SRTS projections are basedon the results of empirical models oftimber and land supply and demand inrelation to prices, income, and othervariables. The projection period forSFRA was 1995 to 2040.

1 6 Reprinted from Journal Forestry, Vol. 100, N O. 7, October/November 2002. Not for further reproduction.

mand is inherently hard to predict: Itdepends on the status of the nationaland global economy, interest rates,housing markets, and a variety of otherhighly volatile factors. Our SRTS pro-jections examined the detailed supplyresponse to overall increasing demandto the year 2040.

We ran the model under four projec-tion scenarios: a base case and three al-ternative scenarios using different as-sumptions about how pine plantationsgrow on industrial and nonindustrialprivate lands and how timber demandresponds to prices. Scenario-based mod-eling permits one to examine the conse-quences of modeling assumptions.Studying the differences in results acrossscenarios can be more informative thanthe specific results of any one scenario.Hence, we projected the future based ontwo demand price-sensitivity assump-tions and two pine plantation produc- plantation growth rate increase (75tivity assumptions. The combinations percent increase in growth rate on in-of these yielded four scenarios: dustrial and corporate-owned planta-

l IH: the base case of tions between 1995 and 2040, and(price-insensitive) demand and high 37.5 percent on NIPF plantations).

M a n a g e m e n t t y p e a r e a Total timberland area(million acres) (million acres)

6 0

5 0

4 0

+ Mixed pine- Total timberland

3 0

2 0 5 0

1 9 9 5 2000 2005 2010 2 0 1 5 2020 2025 2 0 3 0 2 0 3 5 2 0 4 0

ections of private timberland area by management type, 1995-2040, under the basemptions of inelastic timber demand and high plantation volume growth rate increases.

l IL: inelastic demand and low plan-tation growth rate increase (50 percentincrease in growth rate on industrialand corporate-owned plantations be-tween 1995 and 2040, and 25 percenton NIPF plantations).

l EH: elastic (price-sensitive) de-mand and high growth rate increase.

l EL: elastic demand and lowgrowth rate increase.

The 2000 Resource Planning Act(RPA) assessment projections (Hayneset al. 2002) provided national andglobal context, but we used SRTS toproject southern forest area, harvests(removals), growth, and inventory. Ini-tial inventory, net growth, and re-movals used in the SRTS projectionswere obtained from USDA Forest Ser-vice Forest Inventory and Analysis(FIA) data, reported periodically by thestates. Some of these states had rela-t ively old surveys; the older the survey,the less confidence we have in specificprojections. These concerns arise espe-cially in South Carolina (because ofHurricane Hugo’s effects on growthrates of trees in natural stands) and inAlabama, Kentucky, Louisiana, NorthCarolina, Texas, and Virginia.

October/November 2002 - Journal of Forestry 17

..-

‘L

I) 10% to 27%

timberland area, 1995-2040, under assumptions of inelastic timber demand and high plantationvolume growth rate increases.

Southern pine plantation area(million acres)

7 0

6 0

- * Projected, IH

- A Projected, IL

- A Projected, EH

- [ 3 Projected, EL

- Historical

Figure 3 . Pine plantation area projections by scenario, and historical pine plantation area onprivate land in the South, 1952-2040.

Project ion Resul tsProjections for private timberland

cent from 1995 to 2040, is net of an ag-

under the base case (II-I) scenario forgregate increase in the area of pine plan-tations and an aggregate decrease in the

FL4 survey uni ts show the South los ing area of other forest types (fig. 1). A de-private timberland over the comingdecades.This loss, amounting to 2 per-

tailed map of forest area changes (fig. 2)shows that private timberland area is

projected to increase in the westernparts of the South, with losses in s tatesalong the southern Atlantic seaboard.The projected gains in private timber-land area, facilitated by rising timberprices relative to agricultural rents, areconcentrated in Alabama, Arkansas,Louisiana, and Mississ ippi . Signif icantpercentage losses are projected forFlorida, North Carolina, South Car-olina, Tennessee, Texas, and Virginiaand will likely be concentrated nearurban areas; some rural locations maygain forest area. But all of Florida andSouth Carolina’s FIA survey units areprojected to lose private timberland.

Softwoo& Pine plantation areas areprojected to increase by 21 millionacres (67 percent) under the IH sce-nario between 1995 and 2040 (f;. 3).This increase mirrors the losses of nat-ural forest management types underprivate ownership, but private timber-land area remains largely unchangedfrom 1995 to 2040. For the EH andEL scenarios-in which demand is

’igh ly price sensitive-pine plantationarea is projected to increase by onlyabout 25 percent , result ing in a net lossin private timberland area of just over27 million acres (15 percent) between1995 and 2040. This slower increase inplantation acres occurs because prices,to which pine planting responds posi-tively, do not increase as much underthe elast ic demand scenarios.

Under the IH and IL scenarios, pineplantation area increases at the expenseof private timberland in other foresttypes, but other land-use changes alsooccur. During the 1980s and 199Os,about 30 percent of new pine planta-tion acres in the South derived fromagricultural land, and the rest camefrom conversion of natural forest man-agement types. Natural forests havealso been converted to urban uses-atrend that is projected to continue. Weanticipate that increased private pineplantation acreage will come from agri-cultural land as well as from naturalforest management types. Gulf Coaststates and the coastal and Piedmont re-gions of Atlantic Coast states will gainthe most pine plantations; northernand interior regions will gain the least.

Projections of pine plantation areaare sensitive to assumptions regarding

18 Journal of Forestry * October/November 2002

forest productivity. Comparing scenar-ios IH and IL shows that each percent-age point increase in growth rate abovea 50 percent increase for industry (andeach 0.5 percentage point increaseabove 25 percent for NIPF) results inabout 170,000 fewer acres of projectedpine plantations by 2040. Similarly,each percentage point increase in thepine plantation growth rate for indus-try (0.5 percent for NIPF) is projectedto “save” about 50,000 acres of naturalforest. If demand is very responsive totimber prices, as in the EH and EL sce-narios, however, the importance of thistradeoff is diminished.

Figure 4 detai ls the changes by s tatein pine plantation area projected in theIH scenario. Pine plantation areachanges vary among southern statesmostly because of differences in thearea of industrial forests, the propor-tion of natural pine forests to othertypes (natural pine stands are more fre-quently converted to plantations), andland-use changes to and from nonfor-est. In all states except Kentucky, pineplantation area is projected to grow atleast 45 percent from 1995 to 2040,with the largest percentage gains inTennessee (120 percent), Arkansas(117 percent), and Alabama (89 per-cent). Georgia, the state with the mostpine plantations in 1995 (6.4 millionacres), is projected to have 9.3 millionacres in 2040. Alabama, with the sec-ond most in 1995 (4 million acres), isprojected to have 7.5 million acres.

Increased area in planted pine underthe IH and the IL scenarios would alsolead to a rise in timber inventories.Under the base case (IH) scenario, soft-wood growth is projected to exceed re-movals during the entire 40-year pe-riod (fig. 5). This finding holds for theother scenarios as well. Softwood har-vests are projected to increase most inpercentage terms in the northernreaches of the South (Kentucky, Ten-nessee, Arkansas, and Oklahoma) andleast in southeastern parts (fig. 6a,p. 20). In absolute terms (volume peryear), the results are more complicated(fig. 6b, p. 20). Large volume increasesare projected in some places that havealways been major producing regions(Georgia, Alabama, and Louisiana) andin some that have not (parts of the

Pine plantation area( t h o u s a n d a c r e s )

1 0 , 0 0 0

8 , 0 0 0

6,000

4,000

1 9 9 5

II 2020

a 2040

Figure 4. Pine plantation area by state on private land in the South for 1995, 2020, and 2040,as projected by SRTS, under the base case (IH) scenario, with inelastic demand and high pineplantation growth rate increases.

I n v e n t o r y(bcf)

Growth orr e m o v a l s

(bcf)

1 4 0

1 2 0

1 0 0

8 0

6 0

4 0

2 0

0

I n v e n t o r y- Growth------- Removals

Figure 5. Subregional Timber Supply model projections of softwood timber growth and removalsvolumes (bcf) on private timberland in the South, 1995-2040, under the base case (IH) assumptionsof inelastic timber demand and high plantation volume growth rate increases.

October/November 2002 - Journal of Forestry 19

P e r c e n t c h a n g e ( 1 9 9 5 - 2 0 4 0 )cy--J -50% to 4%

- 6 % t o 6 6 %6 6 % t o 1 4 6 %1 4 6 % t o 2 5 8 %

. 258% to 453%

Change in removal (million cubic feet) \&

- 8 6 , 0 6 9 t o 01 to 75,827

I, 75,826 to 165.318r) 165,319 to 306,493

Figure 6. Percentage (a) and absolute (b) changes in annual softwood harvest levels, 19952040,as projected by the Subregional Timber Supply model projections, by FIA survey, under the basecase (IH) assumptions of inelastic timber demand and high plantation volume growth rate increases.

Piedmont and mountains of NorthCarolina and Virginia, central Ten-nessee, and the Ozarks of Arkansas).

Even parts of the South projected tolose forest area would experience in-creased softwood harvests. This increasecomes not only from rising growth rateson plantations but also from timber en-tering the market as forests are con-verted to nonforest uses; the Piedmont

and Florida are examples of this. Otherplaces (parts of Mississippi, Arkansas,and Louisiana) are projected to have de-creased harvests even if forest area mightbe stable or rising, because many of theexpected new acres of pine plantationswill not be harvested until after 2040.

In aggregate, softwood harvestsfrom private lands are projected to in-crease by 56 percent between 1995 and

2040 under the IH scenario. This re-sults from both the increase in the areaof pine plantations and the projectedrise in productivity of those planta-tions. Pine plantations, which yieldwood at least 50 percent faster thannatural pine stands, now account fornearly half of southern timber volumegrowth. Rising productivity over timemeans that more wood can be pro-duced on a smaller land base. In otherwords, the projection shows a largeshift in the share of harvests from nat-ural forest management types to pineplantations. The effect then is to raisethe harvest intensity (harvest volumeper acre) on plantation pine stands andlower it on natural types over time.

Natural forests. Projected changes innatural forest management types underthe IH scenario also vary by state (fig.7). All states are projected to loseacreage in natural forest types underthis scenario. The states with the great-est loss in natural forest types areFlorida (58 percent), South Carolina(35 percent), and North Carolina (30percent)-the result of pine plantationexpansion plus a loss of forests to resi-dential and urban uses. In other sce-narios, the losses projected for naturalforest management types in those statesare similar, and the same states are pro-jected to lose most.

Hardwoods. Intensive silviculture hasnot been widely applied in hardwoods.Assuming no major change in manage-ment practices, hardwood growth isprojected to stay ahead of removals onlythrough the mid-2020s, after whichhardwood inventory is projected to de-cline. This finding is common to allscenarios and is displayed graphically infigure 8 in the base case (1H). Here,growth is projected to exceed removalsuntil about 2025, when removals over-take growth. Much of the increasingrate of removals can be ascribed to agrowing demand for hardwood fiber forengineered wood products, especiallystructural and nonstructural wood pan-els (Haynes et al. 2002).

Hardwood harvests from privatelands are projected to change unevenlyacross the South. In percentage terms,projected increases are largest for thenorthern and western parts of theSouth (e.g., Kentucky, Tennessee,

20 J o u r n a l of Forestry * October/November 2002

northern Alabama, northern Arkan-sas), where these harvests will bemostly from areas not projected to loseforests. In Florida, however, the pro-jected harvests will often be associatedwith conversion from forest to urbanuses (fig. 9a, p. 22). Total volume willlikely reflect a combination of harvestsentering the market as hardwoodstands are converted to nonforest usesand to pine plantations, and higherharvesting rates in remaining hard-wood forests (fig. 9b, p. 22).

Inventories. Changes in inventoryresulting from private timberland areafluctuations, management type areashifts, and plantation growth vary con-siderably across subregions of theSouth. For most states, inventories ofboth hardwood and softwood are pro-jected to exceed those present in 1995.Also for most states, growth and re-movals of both hardwood and soft-wood species are projected to increasethrough 2040, but there are notableexceptions, including Mississippi andSouth Carolina, where hardwood re-movals outpace growth during the en-tire projection. The falling hardwoodinventories in many areas can be as-cribed primarily to vigorous conversionof natural forest management types topine plantations. Softwood inventoriesin both Mississippi and South Carolinaare projected to rise through 2040.Kentucky and Oklahoma, with largeinventories relative to local demand,are projected to have steadily rising in-ventories of both hardwood and soft-wood throughout the period.

Discussion and ConclusionsProjections suggest that the southern

United States will remain the largestsingle producer of timber products inthe world. The South has become in-creasingly prominent in domestic tim-ber product markets because of rapidlyincreasing productivity on private land,improved product manufacturing tech-nology, and the shrinking timber har-vests in other parts of the country. Soft-wood growth will easily accommodatehigher harvest volumes because of con-tinual expansion of pine plantation areaand faster growth on those acres, buthardwood resources will see increasedharvests as well.

Natural forest type area( t h o u s a n d a c r e s )

20,000

1 6 , 0 0 0

1 2 , 0 0 0

8,000

4 , 0 0 0

0

1 9 9 5

I 2020

2040

Figure 7. Natural forest management type (natural pine, oak-pine, upland hardwood, bottomlandhardwood) area on private timberland by state in the South for 1995, 2020, and 2040, as projectedby SRTS, under the base case (IH) scenario, with inelastic demand and high pine plantation growthrate increases.

I n v e n t o r y(bcf)

200 ’

G r o w t h oremovals

(bcf)

; 5

160 1

80 y

2 ’

I n v e n t o r y

- Growth------- Removals

Figures. Subregional Timber Supply model projections of hardwood timber growth and removalsvolumes (bcf) on private timberland in the South, 1995-2040, under the base case (IH) scenarioassumptions of inelastic demand and high plantation volume growth rate increases.

Decreases in the area of private tim-berland in natural forest managementtypes will come from projected in-creases in pine plantations and the liq-uidation of forests to accommodateurban expansion. Land-use pressuresare projected to depress the total area oftimberland in some parts of the South,

especially in the heavily populated At-lantic Coast states. The loss of timber-land in these areas is projected to beoffset by gains in some parts of theSouth’s northern and western regions.

Projected increases in acreage andgrowth rates of southern pine planta-tions imply that forest product manu-

October/November 2002 * Journal of’ Forestry 21

H a r d w o o d

Percent change (1995-2040)-27% to 0%

95% to 198%

1 9 8 % t o 2 9 9 %

299% to 2,123%

Change in removal (million cubic feet)

*c:; -38,773 to 0

15,002 to 28,690

28,691 to 46,314

46,315 to 76,973

Figure 9. Percentage (a) and absolute (b) changes in annual hardwood harvest levels, 1995-2040, -240,as projected by the Subregional Timber Supply model projections, by FIA survey unit under basecase (IH) assumptions of inelastic demand and high plantation volume growth rate increases.

facturing opportunities will improve.Investment opportunities will exist fordeveloping capacity and technology toutilize small-diameter logs comingfrom pine plantations. But such eco-nomic opportunities may have to besquared with, or be limited by, issuessurrounding the losses of natural for-est types and their associated ecologicalvalues.

Our projections were based onsteadily increasing national income anddemand, consistent with historical evi-dence and published RPA projections.To the extent that such income and de-mand growth do not materialize, re-sults would differ: Harvest pressureswould be lower, and plantation areasmight not increase as much. The issueis complex, however, because rising

timber demand also translates intolower pressure to convert forest toother uses and fewer incentives to con-vert agricultural land to forest. Morerobust projections will require addi-tional research that fills several infor-mation gaps, including a more refinedunderstanding of how land uses can beprojected at fine scales, how changes inlandowner demographics will affecttimber supply, and how urbanizationand demographic changes are likely tofragment both forests and their owner-ship. Projections would be improvedwith better data on forest conditionsand uses and land-use trends compiledat finer scales of resolution.

literature CitedAHI; R.C., F . W . CURIMX, and G. PACHECO. 2000.

Southern forest resource assessment using the Subre-gional Timber Supply (SRTS) model. Forest ProductsJournal50(4):25-33.

B A I I F Y , R.G. 1995. Description of the ecoregions of theUnited States. Miscellaneous Publication 1391. Wash-ington, DC: US Department of Agriculture.

GUAN, H., and LA. MUNN. 2000. Harvest restrictions:An analysis of new capital expenditures in the PacificNorthwest and the South. Journal of Forestry98(4):11-17.

HARES, R., D. A L ~ A M S , R. ALIG D. B ROOKS, B. BUTLER,1 . D U R R A K , J. HOWARD , I? INCE, D. MCKEEVER, J.Mi1.1 S, K. S K < ) < ; , and X. ZHOU. 2002. The 2000 RPAtimber assessment: An analysis of the timber situation inthe United Starer, 1952 to 2050. Technical documentsupporting the 2000 USDA Forest Service RPA as-sessment. Available online at www.fs.fed.us/pnw/sev/rpa/; accessed by authors May 2002.

M C R K A Y , B.C., and D.N. WEAR. 1998. Federal timberrestrictions and interregional arbitrage in US lumber.Land Economics 74( 1):76-9 1.

&WrEMON, J.P, and R.C. ABT. In press. Timber prod-ucts supply and demand. In Southern forest resource as-sessment, eds. D.N. Wear and J.G. Greis. GeneralTechnical Report. Asheville, NC: USDA Forest Ser-vice, Southern Research Station.

Jeffrey P. Prestemon (jprestemon@fs.fed.us) is research forester, Southern Re-search Station, USDA Forest Service,PO Box 12254, Research Triangle Park,NC 27709; Robert C. Abt is professor,Department o f Fores try , North Carol inaState Universi ty , Raleigh.

22 Journal of Forestry * October/November 2002