document Historic, archived Do assume not content reflects current scientific knowledge, policies, or practices. m ? i, F7C 26 Uf,; ?s t*i United States Extended Rotations and Department of Agriculture '^S^r Age Forest Service Culmination of Coast Pacific Northwest Research StationN Douglas-fir: Old Studies Research Paper PNW-RP-485 November 1995 Speak to Current Issues O Robert jCurtis 30 Author ROBERT O. CURTIS is a principal mensurationist, Forestry Sciences Laboratory, WA 3925 93d Ave., S.W., Olympia, 98512-9193 Curtis, Robert O. 1995. Extended rotations and culmination age of coast Douglas-fir: old studies speak to current issues. Res. Pap. PNW-RP-485. Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station. 49 p. Trends of mean annual increment and periodic annual increment were examined in 17 long-term thinning studies in coast Douglas-fir {Pseudotsuga menziesii var. menziesii (Mirb.) Franco) in western Washington, western Oregon, and British Columbia. Maximum ages included ranged from about 90 years on high sites to 117 years on a low site. None of the stands had clearly reached culmination of mean annual increment, although some appeared close; periodic annual incre- ments declined only slowly. Extended rotations combined with increased thinning harvests are promising components of any strategy to reduce conflicts between timber production and other forest values. These comparisons indicate that rotations can be considerably extended without reducing long-term timber production. A major problem in such a strategy is design of thinning regimes that can maintain some reasonable level of timber flow during any transition period. Keywords: Growth and yield, mean annual increment, rotation, Douglas-fir, Pseudotsuga menziesii, alternative silviculture, ecosystem management. Trends of mean annual increment and periodic annual increment were examined in 17 long-term thinning studies in Douglas-fir {Pseudotsuga menziesii var. menziesii (Mirb.) Franco) in western Washington, western Oregon, and British Columbia. Maxi- mum ages were about 90 years on high sites to 117 years on a low site. Problems in evaluating growth trends and culmination ages are discussed. None of the stands had clearly reached culmination of mean annual increment, although some seemed close. Periodic annual increments declined only slowly. Culmination has not yet been observed in systematically thinned stands that are relatively free from damage. The observed trends seem generally consistent with some other recent comparisons. Ex- tended rotations combined with increased use of thinning are promising components of any strategy to reduce conflicts between timber production and other forest values. These comparisons indicate that rotations can be considerably extended without reducing long-term timber production; value production probably would increase. A major problem in such a strategy is design of thinning regimes that can maintain a reasonable level of timber flow during the transition period while producing stand conditions compatible with other management objectives. Several examples of drastic thinnings seem promising. The continuing value of long-term permanent plot studies emphasized. is Contents 1 Introduction 2 Objectives 2 Methods 2 Definitions 2 Data Assembly 2 Data Summarization 4 Data Analysis 4 Results 6 Discussion 6 Problems in PAI Estimation 7 Possible Sampling Biases 7 Data Limitations 8 General Trends 10 Very Heavy Thinning 10 Possible Climatic Effects 10 Culmination Ages, Rotations, and the National Forest Management Act 12 The Timber Supply Problem 12 In Conclusion 13 Metric Equivalents 13 Literature Cited 16 Appendix 16 Fuller Hill 19 Hoskins LOGS 21 Delezenne 23 Port Blakely XT-1 25 Port Blakely XT-2 27 Port Blakely XT-3 29 Port Blakely XT-7 31 McCleary 33 Schenstrom 35 Snow Creek 37 Black Rock 39 Voight Creek 41 Wind River Spacing Test 43 Stampede Creek LOGS 45 Martha Creek 47 Lookout Mountain 49 Mount Walker — Introduction Extended rotations whether applied on the basis of the large even-aged units used — in the recent past, or to small even-aged patches of a few acres have been pro- posed as a component of strategies for mitigating current conflicts among forest uses and reconciling amenity, wildlife, and environmental values with timber production (Curtis 1994a, Curtis and Marshall 1993, Newton and Cole 1987, Weigand and others 1994). Two key questions involved in such proposals are those of (1) the effects of extended rotations on short-term timber supply and (2) the effects on long-term timber produc- tion. Extending rotations would reduce short-term timber flow during the transition period, although reductions could be offset to some unknown degree by increased production from thinnings and reduced pressures to remove land from the timber base. The probable effects of extending rotations on long-term timber production are not generally recognized. Coast Douglas-fir {Pseudotsuga menziesii var. menzieseii (Mirb.) Franco) is a very long-lived—species that maintains growth to advanced ages, and there is reason to believe that compared with the relatively short rotations — adopted by many owners moderate increases in rotation age would substantially increase long-term timber production. Several approaches to the question of probable long-term effects are possible. Curtis (1992a) showed that a bias exists in the Jong-used yield tables of McArdle and others (1961); incorporation of modern height curves changes the shape of the yield curves and indicates culmination at substantially greater ages than those given in the earlier publication. Curtis (1994b) compares estimates for selected management regimes produced by a number of widely used stand simulators. Although there are considerable differences among simulators, results indicate culmination ages considerably greater than the 40- to 50-year rotations now common in many ownerships, and suggest little change in production rates for an undetermined but considerable number of years beyond culmination. This report presents additional evidence from examination of actual growth trends observed in existing long-term thinning studies. In the period from about 1930 to 1960, a considerable number of thinning studies were established in the Pacific Northwest by foresters who were then thinking in terms of relatively long rotations and who saw commercial thinning as a major future activity. But, conversion of old stands was found to be a much more profitable activity than thinning; and with the move toward shortened rotations many owners lost interest in commercial thinning. Recent reductions in timber supply and improved markets for small material, coupled with pressures to reduce the use of clearcutting, have revived interest in commercial thinning. Data from many of these old studies are still available, and some of the installations still exist. Much of the data have never been adequately analyzed or reported. This report examines trends of mean annual increment and periodic annual incre- ment in long-term thinning studies and discusses their implications for management. 1 Objectives The study objective was to examine the historical record of stand development on a series of long-term thinning studies for indications of: . age of culmination shape of MAI and PAI curves . . effects of management regime and site differences on MAI and PAI curves . consistency of patterns among installations consistency with estimates from other sources . A secondary objective was to locate, assemble, and edit basic data from long-term thinning studies for use in the Stand Management Cooperative modeling project (not reported here). Methods Mean annual net increment (MAInet) is cumulative net volume production (live stand Definitions + thinnings), divided by years required to produce it; that is, average production rate from stand establishment to a given year. Periodic annual net increment (PAInet) is change in cumulative net production be- tween two successive stand measurements, divided by number of years in the period. It approximates current annual increment (CAI), the annual growth occurring in any given year. Gross periodic annual increment (PAIgross) also includes mortality be- tween the two successive measurements, and it thus is a measure of biological pro- duction as opposed to usable volume. The curves of MAI and PAI in volume or value over stand age have characteristic shapes (fig. 1). The maximum point ("culmination") of the MAI curve identifies the age at which average production per year is maximized. This is also the point at which the PAI curve intersects the MAI curve. Data Assembly Data were assembled from records of long-term experimental installations that included some type of systematic thinning or spacing treatment on relatively large plots and for which repeated tagged tree remeasurements of reasonable precision were available. Location of suitable installations was based on personal knowledge, past data inventories of the Stand Management Cooperative and the earlier DFSIM modeling effort (Curtis and others 1981), and inquiry to various landowners and research organizations. Data sources consisted in part of past and current experiments of the Pacific North- west Research Station. Weyerhaeuser Co., Port Blakely Tree Farms, British Columbia Ministry of Forests, and Oregon State University also contributed proprietary data and access to their installations. Several installations lacking recent measurements were remeasured in 1991 and 1992. Data Summarization Data quality differed widely; some data required rather arbitrary adjustments to resolve inconsistencies in successive measurements of diameters and heights. 2 Age (years) — Figure 1 Mean annual net increment (MAI) and periodic annual net increment (PAI) curves for well-stocked unmanaged stands. Values for site III are from table 2 in McArdle and others (1961), adjusted to a 6-inch merchantable top. Height-diameter curves were initially fit with a modification of Hyink and others' (1988) constrained height curve fitting procedure. This worked well for data with consistently high-quality height measurements but often gave erratic results for data containing poorly distributed height samples or major inconsistencies in consecutive height measurements, or that required interpolation across large gaps in the age distribution of the height sample. Most of the data were subsequently resummarized by using height-diameter curves smoothed over age (Curtis 1967). In most cases the form, H = 4.5 + exp (a + b/Age + c/D**n + e/D**n/Age) , where 0.5= <n = <1.5, was found satisfactory. Differences between PAI estimates for individual growth periods obtained by the two methods were sometimes large, but overall trends usually differed little and the differences do not alter inferences. All volumes were expressed as merchantable cubic volume inside bark to a 6-inch top (CV6) of trees 7.6 inches and larger in diameter; a value that more closely approx- imates usable volume than does the total stem volume often used in research studies. (Differences are small except for very young or very dense stands or very poor sites.) Calculations were based on the Bruce and DeMars (1974) equation for total stem volume, and conversion to CV6 through tarif system equations (Brackett 1973). 3 Data Analysis Many of the early studies were unreplicated, and definitions of planned treatment were often vague or unavailable. Although some installations were designed studies with replication, there was no consistency in design or treatment definitions among installations. Consequently no simple overall averaging or statistical analysis was possible. Ultimately, the data will be incorporated in the database used to develop response surfaces for the simulation model now under development by the Stand Management Cooperative. The data are treated here as a series of case histories, in which inferences are based on graphic comparisons of MAI and PAI curves and their apparent consistency among installations. Although much other information is present in the data, no attempt was made at general comparisons of other aspects of stand development. Individual graphs of MAInet, PAInet, PAIgross over age and RD over age were pre- pared for each plot (RD = basal area/sqrt(QMD), where QMD is quadratic mean dia- meter; Curtis 1982). RD is a measure of relative stand density, and its trend over time describes one aspect of a thinning regime; a value of RD70 approximates "normal" or full stocking, as shown in normal yield tables. The number of such graphs was far too large for publication, and in this report, plot data from replicated studies have been averaged within treatments. Data also have been averaged within treatment groups where treatments seemed similar in nature and results. Results The principal characteristics and results are summarized in the appendix for each of the 17 installations examined, with figures showing trends of RD, MAI, and PAI for each installation and treatment or treatment group. The overall result is that none of the stands examined have clearly reached culmina- We tion age, in the absence of major stand damage. have as yet no examples of culmination in systematically thinned stands, although some appear near culmination. Table 1 gives condensed information on the range in ages and sites represented, and on average diameters and dominant heights at the most recent measurement by treatment or treatment group. (There were often considerable differences in initial conditions and treatments among plots combined into groups, and one should not attempt to infer treatment effects from these values; they are intended only as a rough description of the range in conditions included.) 4