document Historic, archived Do not assume content reflects current scientific knowledge, policies, or practices. A I A99.9 F7625U USDA United States Economic Feasibility of Longer Department of Agriculture Management Regimes the in Forest Service Douglas-Fir Region Pacific Northwest Research Station —3> Richard Haynes7 Research Note PNW-RN-547 February 2005 1 Abstract D The financial returns associated with extended management regimes have been the subject ofrecurring debate in the Pacific Northwest. Proponents argue that the amount and value ofhigher quality timber associated with older trees will offset the costs associated with longer management regimes. Land managers and owners express concerns about diminished financial returns depending on the expected costs ofholding timber for long periods. The increase in average lumber prices for high-quality timber is insufficient, on average, to offset the costs oflonger manage- ment regimes. On public land, where ownership continuity is assured and the requirement forpositive rates ofreturn is less, longer management regimes may be attractive when they involve thejoint production ofvarious public goods such as wildlife habitat and scenery. Keywords: Forest management, economic feasibility, management regimes. Introduction The Pacific Northwest (PNW) is considered one ofthe premier regions for forest management in the United States. For almost a century, intensive forest manage- ment has supported a forest sectorthat has grown in both size and complexity. Various forest management regimes have evolved in response to changes in land- owner objectives, the development ofsilvicultural information (including growth and yield information), and changes in the utilization patterns ofharvested timber. Debates about appropriate management regimes have been frequent and often contentious. Management regimes2 that have longer rotations3 have been the focus ofrecurring debate. Proponents argue that the amount and value ofhigher quality 1Richard W. Haynes isaresearch forester, Forestry Sciences Laboratory, P.O. Box OR 3890, Portland, 97208. 2Mostfrequentlytheseregimesaredescribedas includingplanting, stockingcontrol (often achievedby usingprecommercial thinning), multiplethinnings, andfinal harvest ofarelativelyhigherqualitymix ofsawlogsassociatedwith oldertrees. *In some applications thesearereferredtoasextendedrotations. RESEARCH NOTE PNW-RN-547 timber associated with oldertrees will offset the costs associated with longer rotations. The revealed choices by landowners and managers (discussed in the next section) suggest a steady decline in rotation lengths forprivate timberlands. This suggests that economic benefits associatedwith longerrotations have failedto materialize. These arguments have reappearedrecently as part ofdeveloping a new land managementplan forthe Tillamook State Forest.4 There the Oregon Department of Forestry is considering alternative forest management prescriptions to protect key habitats andprocesses in selected areas. Among the alternatives are several that involve longerrotations that provide greatertimberrevenues over a longertime- frame than canbe obtained by the short-term high-volume harvest regime currently in operation. This paper describes a study done forthe Salmon Anchor Habitat Working 5 Group as backgroundto theirreview ofalternative management strategies forthe Tillamook State Forest. I review the evolution ofmanagement regimes in the Pacific Northwest andthen develop three specific management regimes that encompass the types ofalternatives being considered. I also review the management methods and some ofthe critical assumptions behind the regimes. Finally, I discuss the implica- tions for land management and managers. The interest in growing timber for high-quality purposes by using longerrota- tions is not unique to the Pacific Northwest. It has been discussed in the context of plantation management in South Africa (Craib 1948), New Zealand (Sutton 1984, Whiteside and others 1975), British Columbia (Kellogg 1989), the Pacific Northwest (Barbour and others 2003), and southeast Alaska (Barbour and others, in press). The general conclusion ofthis work is that the tradeoffs among rotation length, timber quality, and financial returns depend on the way in which costs are discount- ed and whetherthe computations include the value ofamenities associated with longerrotations. Evolution of Management Regimes Speculation continues about the appropriate management regimes in the PNW, but there has been tentative consensus on the basic set ofpractices that compose the core ofthe management regimes. Some ofthis discussion dates from the early TheTillamook StateForest is inthenorthwestcornerofOregon and includesmuch ofthe areaplantedafteraseriesoffiresbetween 1930and 1945. 5TheSalmonAnchorHabitatWorkingGroupwasconvenedbytheOregon Departmentof Forestryatthedirection ofthe 2003 Oregon Legislature. Itwasacitizenworkgroupwhose purpose was toreview how foreststandswith salmon habitatwerebeingmanaged in the Tillamook State Forest. 2 EconomicFeasibilityofLongerManagementRegimesintheDouglas-FirRegion 1960s when the Timber Trends study (USDA FS 1963) identifiedthree management regimes differentiatedby the array ofpractices in each. The first regime, which we would now call custodial, represented those owners who operate expecting high rates ofreturn6 on theirtimber investment andwho prefer little active timber management. Instead, these are owners who harvest frequently anduse low-cost regeneration regimes. The secondmanagement regime was called intermediate and assumedthat landowners operated expecting moderate (6 percent in nominal terms) rates ofreturn on their investments. Rotation lengths in this regime were assumed to be shorter (60 years) than with the third management regime. The intermediate regime was aplant, thin, and harvestregime differing (based on site and owner) in regeneration methods and the number ofthinnings. The third managementregime (which we would call intensive) assumed landowners who operated expecting lower rates ofreturn on their investments andwho would adopt longerrotations (120 years) and more thinnings and other cultural treatments. It included site prepara- tion, planting, fertilization, precommercial thinning (PCT), commercial thinning (CT), andfinal harvests. In all three managementregimes, final harvests were as- sumed to be clearcuts. As private forests in the PNW transitioned from old growth to second growth, the Timber Trends study assumedthat 16, 22, and 62 percent of the private land wouldbe assignedto each ofthese management regimes, respec- tively. Overthe last 30 years an array ofstudies have examined the timber supply PNWW prospects forprivate forest lands in the (Pacific Northwest West, western Oregon and Washington). In each case, following the lead ofthe Timber Trends study, efforts were made to identify the range ofmanagement practices appliedto the currentprivate forest inventory. These studies provide differentperspectives of managementpractices (ormanagement intentions) at three times: the early 1970s, the late 1980s, and the late 1990s. A summary ofthese perspectives is shown in table 1 for seven types ofmanagementregimes (see Haynes and others 2003 for more details). In general, industrial owners' management intentions have consistently in- volvedpractices beyond simply securing regeneration, and usually some form ofcommercial thinning. The proportion ofthinning has differed overtime, with some ofthe more recent surveys showing a decline. Practices otherthan commer- cial thinning also appearto have gained in importance as we move from the earliest to the most recent study, including precommercial thinning, fertilization, partial cutting regimes, anduse ofgenetically improvedplanting stock. There has also Landownersdemandahighrateofreturntocapitalwheninterestratesarehigh. Typically these landownersdonotholdtimberforlongperiodswithoutfinancialreturns. 3 RESEARCH NOTE PNW-RN-547 — Table 1 Summary ofprivate timber land base allocation to management regimes from six studies ofthe Pacific Northwest region WOR WWA WOR WWA WOR PNWW Early 1970s Early 1970s Late 1980s Late 1980s Late 1990s Late 1990s Beuter WaFPS£ Sessions Adams Adams RPA/ Industrial managementregime: Age (years) 50H,A 25M'A 45-60M'A 50-57H'S 43-55M.S Percentage oflandbase Non-Ggregenerate only 43 10 28 25 51 17 Non-Gregenerate+CT^+other 41 80 50 21 25 17 Non-G regenerate + other(Non-CT) 10 13 47 14 12 Partial cut 6 2 Genetic regenerate + other 3 47 Conversion/rehabilitation 16 9 Reserved 4 3 6 Total' 100 100 100 100 100 100 Nonindustrial private management regime: H,A M.A M.A M.S M.A Age (years) 50 45-65 45-60 43-49 40-60 Percentageoflandbase Non-G regenerate only 62 95 98 61 83 Non-G regenerate + CT + other 3 5 1 Non-G regenerate + other(Non-CT) 1 3 7 Partial cut 33 Genetic regenerate + other 1 10 Conversion/rehabilitation 35 Reserved 3 Total 100 100 100 100 100 Note: WOR=westernOregon,WWA=westernWashington.PNWW=PacificNorthwestWest.Harvest(H)orminimumharvest(M); ageassumed(A)orestablishedfromsurvey(S). "Beuterandothers 1976. WDNR 1975. Sessions 1991. ^Adamsandothers 1992. 'Adamsandothers2002. ^Haynes2003. ^Non-G=Notgeneticallyimprovedplantingstock. hCT=Commercial thinning. 'Totalsmaynotaddupbecauseofrounding. been a decline in intentions to pursue highly complex systems such as the "inten- sive" regime (plant/PCT/fertilize/CT) in the Timber Trends study. These systems appear to have been replaced by simpler one- or two-treatment regimes (for exam- ple, plant/PCT orplant/PCT/fertilize). On nonindustrial private (NIPF) lands, management is largely limited to secur- ing regeneration with only limited use ofother treatments. Earlier western Oregon 4 EconomicFeasibilityofLongerManagementRegimesin theDouglas-FirRegion studies made it clearthatNIPF lands were in relatively poor management condition (low coniferrestocking afterharvest) (Sessions 1991). Recent studies (Adams and others 2002) show a large fraction ofthe landbase in some form ofpartial cutting or selection management. Forpublic lands, management regimes until the early 1990s followed a mix of the three general managementregimes described earlier. The adoption oftheNorth- west Forest Plan altered managementregimes on federal timberlands and, where timberproductions is allowed, changedmanagementto mimic natural successional processes by usingpartial harvests accomplished through multiple entries. The state timberlands are still generallymanagedon longerrotations than are private timberlands and are managed for avariety ofgoals including the production oftim- ber and environmental services such as habitat andwater quality. Rotation age, or age ofharvest, is one ofthe most closely scrutinized elements ofthese timbermanagementregimes. Harvest age affects the rate ofreturnreceived by the owner onregeneration, tending, and landholding investments; the maximum volume attainedby the stand; the characteristics ofwildlife habitat affordedby the stand over its life; andthe timepatterns ofatmospheric carbon uptake and release. Yet as important as rotation is inboth management decisions and in assessing the broader social and environmental impacts ofmanagement, relatively little has been objectively measured about actual harvest agepatterns onprivate lands inthe PNW. As shown in table 1, six ofthe seven timber supply studies discussed in connec- tionwith the structure ofmanagement regimes also provide information on antici- pated or intendedrotation ages. In most cases these are "minimum harvest ages," — the youngest age at which owners would considerharvesting a stand not average — anticipated harvest ages andthey commonly differby site, owner, and manage- mentregime. Forthe mostpart, available estimates were developed from expert opinion ratherthan surveys ofowners ortheir agents. Withthe exception ofthe Ses- sions (1991) study, the range ofharvest ages is broadly similar forboth industrial andnonindustrial ownerships, 40 to 65 years depending on site and regime. The WW limited information on the actual harvest age behavior ofowners inthe PN (Lettman and Campbell 1997) suggests that the area-weighted average harvest age is 58 years (see Haynes and others 2003 for computation method). The range is from 20 to more than 90 years, but more than two-thirds ofthe private cut comes from timberwhose age ranges from 40 to 70 years. Management Regimes, Assumptions, and Methods In the shifting discussions about land managementparadigms during the late 1980s and early 1990s, longermanagement regimes were proposedboth as away 5 RESEARCH NOTE PNW-RN-547 to produce higher quality timber (see Barbour and others 2003) and as a way to produce a broader array ofecosystem goods, services, and conditions, especially interms ofwildlife habitat. Much ofthis discussion is summarizedby Weigand and others (1994) who describedthe results from the High Quality Workshop held in spring 1993. The original proposal was describedby Gus Kuehne (President, Northwest Independent Forest Manufacturers) in fall 1990 as "High Quality For- estry (HQF): an alternative formanagement ofNational Forest lands."7The name HQF came from an emphasis on growing timber with old-growth-like characteris- tics (large diameter, relatively knot free, and a high number ofrings per inch) that wouldproduce a high proportion ofShop and Select grades oflumber. The HQF method called for extended harvest rotation of 150 to 200 years, severely reduc- ing clearcutting and emphasizing multiple management goals. The HQF method relied on precommercial thinning, pruning, and commercial thinnings to accom- modate the needs ofwildlife species and to maintain and improve timberyields. The intervals between intermediate cuts ranged from 15 to 30 years, depending on site and terrain. The record ofdecision (UDSA and USDI 1994) implementing the Forest Ecosystem Management Assessment Team (FEMAT 1993) strategy for habitat conservation shifted discussions about public land management paradigms to approaches otherthan longerrotations. Here I look at the differences between three management regimes. The first regime is currently employed on 47 percent offorest industry timberlands in WW PN (summarized from table 1). It consists ofplanting, PCT (at age 15 years), fertilization (usually on forest industry timberlands), and final harvest around 45 years. The second regime is one often favored by silviculturists who rely on rota- tion lengths set by culmination ofmean annual increment, which is around 80 years for Douglas-fir (Curtis and others 1998). This regime also includes planting. PCT, and a commercial thinning at about 50 years. This thinning was designed to reduce stand volumes by one-fifth and to capture for use the expected mortality between 50 and 80 years old. A third management regime was examined that represents Kuehne's long-rotation proposals. This regime includes planting, PCT, pruning, a series ofcommercial thinnings (at ages 50, 80, and 120, each designed to reduce stand volumes by one-fifth and to make use ofdead trees before the next entry), and final harvest at age 160. The thinning intervals are longer than originally envisioned by Kuehne because in early analysis ofhis proposals, light frequent thinnings as he envisioned were not economically feasible (see Weigand 1994). A rotation length of 160 years was chosen to continue the progression ofthe first two regimes. Hisproposal is includedasan appendix in Weigandand others (1994). 6 EconomicFeasibilityofLongerManagementRegimesin theDouglas-FirRegion The analysis method consisted ofcalculating the soil expectation value which is the present value ofall future netreturns from apiece ofland growing timber. Given a rate ofinterest, rotation length, and the value oftimber (or other) prod- ucts produced, the presentnetworth ofland can be computed. In this case, we are looking at the present net worth ofa stand grownunderthree rotation lengths each producing a differentproductmix (in terms ofthe grades oflumberthat can be produced from the available logs). This approach depends on several key assumptions. First, all prices and costs are in real terms (that is they are net ofinflation). There is little expectation forreal price appreciation in lumber and stumpage markets. Price projections from Haynes (2003) suggestprice appreciation of0.4 and 0.2 percentperyear overthe next 50 years forthe lumberand stumpage markets. Second, stand volume was computed byusing an empirical yield function shown in figure 1 forDouglas-fir derived from USDA Forest Inventory andAnalysis data forwestern Oregon. This represents o A3 0) o 3 c o o c 3 o Age (years) — Figure 1 EmpiricalyieldfunctionforDouglas-firwiththinningsat50, 80, and 120years. yields forrelatively full stocking observed in actual stands. Actual stand stocking is between 60 and 65 percent offull stocking onthe industrial and nonindustrial timberland in western Oregon suggesting that the realizedyields will be less than shown infigure 1. Third, there are price premiums associated with the intermediate Thepresentvalueofallfuturenetreturns,thesoilexpectationvalueorSEV, fromasimple plant-and-harvestmanagementregimecanbeexpressedas: SEV = Price x(V1o+lIunmteereasttrorattaet)iRoonta-tI;Polna-nt1in-gcost Plantingcost 7 RESEARCH NOTE PNW-RN-547 and intensive regimes reflecting largerproportions ofhigh-grade products resulting from larger trees with less taper and fewerbranches associated with thinned stands and longerrotations. Figure 2 illustrates the actual ratios betweenprices forhigh, medium, and low Douglas-fir lumber grades (see Haynes and Fight 1992, 2004 for a discussion). Sustained high-grade prices that are three to fourtimes low-grade lumberprices have influenced the advocates for HQF. Table 2 shows the actual pro- portions ofDouglas-fir lumberrecoveryby grade groups. This table was developed from data collected in empirical woodproducts recovery studies (see Stevens and Barbour 2000). High/medium High/low Medium/low o o ~\ i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i r Year — Figure2 Douglas-firpriceratios for gradegroups. — Table 2 Proportion of Douglas-fir recovery by grade groups Factory and Select No. 2 and Utility and Age class Selects structural better Economy Years Percent 40 21.1 70.4 8.6 80 2.3 22.3 66.8 8.6 160 7.5 24.7 59.2 8.6 A key assumption that shows differences among the three management alter- natives is the price projection for mixes ofstumpage coming from different stand ages. Table 3 shows the average lumberprice per thousand board feet (MBF). It was developed by combining lumber price projections by grade (from Haynes and Fight 2004) and the recovery data shown in table 2. The next step converts these 3