Eastern White Pine abundance in 19th century forests: a reexamination of evidence from land surveys and lumber statistics

Размещено на http://www.allbest.ru/

Eastern White Pine Abundance in 19th Century Forests: A Reexamination of Evidence from Land Surveys and Lumber Statistics

Forest managers in recent decades, especially on public lands, have become increasingly involved in managing forests to emulate natural disturbance processes more closely and to restore some characteristics of old-growth forests. Obtaining baseline information on old-growth forest dynamics is challenging in parts of North America where old-growth remnants are rare and commercially valuable species were systematically culled from stands at an early date. In these cases, historical records, such as early township surveys, can provide valuable information about presettlement forests and natural disturbance regimes. In the New England and Great Lakes regions, these records include data on the original abundance of eastern white pine (Pinus strobus), a species of intermediate shade tolerance that often regenerates after fires but which had been heavily culled from the forests by the mid- to late 19th century.

Historical records suggest a low proportion of pine in many presettlement northern hardwood-conifer forests (e.g., Siccama 1971, Manies et al. 2001, Whitney and Decant 2003). For Maine, Coolidge (1963) used early mill records and other sources to estimate a statewide standing pine timber volume of 40 bbf in the year 1600--a mean of 49.8 million bd ft/township, equivalent to about 2,160 bd ft or two large pines per acre. He recognized, however, a strong decreasing gradient in pine abundance from south to north and suggested that most townships in the Aroostook region of northern Maine probably only had a few million board feet per township.

These early estimates by Coolidge were largely supported by subsequent studies. Land surveys in northeastern and west central Maine from 1793 to 1835 indicated heavy dominance by spruce (Picea spp.) and northern hardwood species, and pine witness trees made up only 0.4--1.3% of the total (Lorimer 1977, Cogbill et al. 2002). Independent cruises of merchantable pine in 21 townships in northeastern Maine also in dicated a mean of only 6.7 million bd ft/ township, averaging 290 bd ft/ас, less than one-third of the volume of a single large tree. Tabulations of recently burned land, windfalls, and stands dominated by postfire species such as paper birch (Betula papyrifera) and aspen (Populus spp.) suggested relatively long rotation periods for stand-replacing fire and catastrophic windthrow (Lorimer 1977, Lorimer and White 2003).

In a recent Journal of Forestry article, Wilson (2005) estimated presettlement pine volumes based on lumber production statistics for the Penobscot River watershed in northern Maine from 1832 to 1872. Wilson argued that the amount of pine lumber produced suggests much higher volumes of standing merchantable pine, which he felt may warrant a reinterpretation of historic disturbance regimes. Wilson offered several explanations for the differences between the two data sets, including the possibility that the timber cruises were not a sufficient sample of the region, that surveyors may have been biased against selecting pine witness trees, or that logging had already reduced pine volumes by the time the cruises were made. He also suggested that surveyors' descriptions of forest vegetation along township lines were more consistent with his reconstructed pine density because pine was the third most commonly cited conifer.

Interpreting the composition and dynamics of presettlement forests is a difficult task best accomplished by using multiple independent lines of evidence. The lumber statistics are therefore potentially valuable as additional information on pine abundance. However, Wilson (2005) understated the difficulties in relating lumber statistics to the amount of standing timber, and the article unfortunately contains a number of problematic assumptions and misleading comparisons. It is not possible to address all these issues thoroughly in an article of limited length. But this article will examine (1) the key assumptions in Wilson's analysis; (2) what can reasonably be inferred about pre- settlement pine density from land surveys and lumber statistics, based on new evidence and analyses; and (3) what the best available estimates of pine density suggest about presettlement disturbance regimes.

This article uses township line descriptions and witness tree data from 196 townships surveyed between 1793 and 1827 in northeastern Maine (study area of Lorimer 1977), as well as independent timber cruise data contracted by the state land office between 1826 and 1849 for 107 townships (Maine State Archives 1793-1849). In the land surveys, dominant tree species were listed in order of decreasing abundance along 1-mi township line segments, with one witness tree recorded per mile (species identified but diameter not recorded).

Twenty of the townships with timber cruises lie within the boundaries of the 1793-1827 survey tract, while 87 others, most of which were not previously included in the analysis of Lorimer (1977), are located in adjacent regions of northern Maine (Figure 1). Timber cruising methods varied among different workers and were not described in detail, but all appeared to involve counts of pine trees multiplied by the estimated average merchantable volume per tree on different tracts. Pines were commonly counted along belt transects across the townships and on larger tracts of land viewed from climbing tall spruce trees on ridgetops. Pines considered merchantable in the 1830s and 1840s were mostly larger than 20 in. in diameter (Wood 1935; also confirmed in several cruise reports), although on some townships, separate estimates were provided for smaller trees as well. Volume estimates may have been somewhat conservative but did not reflect the often wasteful cutting practices of the day; the pine timber estimates were based on what "a prudent man would cut if he owned the land."

Wilson's Comparisons of Lumber and Land Survey Data

The crux of Wilson's analysis is that the 3.36 billion bd ft of Penobscot River pine lumber reported from 1832 to 1872--and the 6 billion bd ft of standing timber probably needed to generate that amount of lumber--is substantially higher than the 1.4 billion bd ft that he believed would have been estimated from land office timber cruises for the same watershed area. It is difficult to make close comparisons between two 19th century data sources gathered by different methods and in different locations; but even if one were to accept all of his assumptions, this is a small difference when examined on a stand-level basis. The lumber statistics imply a mean, of 661 bd ft of pine per acre. For large trees averaging about 30 in. dbh, this volume is equivalent to approximately 0.7 tree/ас. This compares with 290 bd ft or 0.3 tree/ас previously estimated from the timber cruises.

Wilson avoided these implications of low pine abundance by focusing on volumes at a very large scale, whereby small, local differences can accumulate to more impressive regional differences and by distributing all the pine volume uniformly over a small proportion of the watershed to achieve an estimated density of 5 large trees/ас. Even with the higher estimate of 6 billion bd fit of standing timber in the watershed to account for cull and waste in producing the lumber, this still only amounts to a regional average of 1.2 large pines/ас, still a rather small proportion of the forest composition, and of questionable significance for reinterpreting forest successional status. But there are reasons for believing that the differences between the data sets are not nearly as great as Wilson suggested.

First, there are major environmental differences between the Penobscot River watershed and the more northerly regions included in the 1793--1827 land surveys. Wilson justified the comparison because the two regions overlap. But 49% of the land survey tract lies outside the Penobscot watershed, mostly in the watersheds of the Aroostook and St. John Rivers in northeastern Maine, and 2.7 million ac of the Penobscot watershed (equivalent to 117 townships) were not included in the land survey tract (Figure 1). The Aroostook and St. John watersheds contain some of the best soils and farmland in Maine, whereas the Penobscot contains a great deal of shallow, stony soil, and rough terrain with rocky outcrops.

These geographical differences--and the implications for pine abundance--were obvious to the surveyors. As E. Hoard remarked, "in the Aroostook Country the soil in general is good & the [pine] timber mixed with hardwood and more scattering than on the East Branch [Penobscot] & but few large bodies of timber are to be found ... [the reason] must be, the soil is worth something, while on the East branch, the soil is worthless." After surveying a large tract in the Aroostook region in 1825, Joseph Norris summarized the timber resources by saying that "there is perhaps enough pine & spruce timber for the consumption by its inhabitants when it is settled, but none to spare." In contrast, he described the western Penobscot region as "presenting more valuable pine timber than any other tract of equal extent which we have before surveyed." The descriptions often emphasized the scattered distribution of pine on Aroostook townships, such as on Tl6 R10, where the pine timber was said to be "of decent quality, and like that in other townships afore described, it is considerably scattered over the township" or T9 R11, where the pine was "very much scattered among the hills, and will hardly pay for hauling" (Maine State Archives 1793-1849).

This regional variation also shows up in the witness tree data, which ranged from 0.5% pine in the Aroostook region to 2.7% pine in the western Penobscot region (Lorimer 1977). These regional differences have persisted to the present day. In the 1995 state inventory, white pine made up 3.9% of the number of trees more than 5.0 in. dbh in Penobscot County, compared with only 0.4% in Aroostook County (Griffith and Alerich 1996).

Interpretation of Land Survey Data

Merchantable Volume Estimates from the Timber Cruises. The additional cruise data from the 107 townships in Figure 1 indicate that the low estimates of pine volume from the timber cruises were not an artifact of a limited sample. The mean volume per township was 7.6 million or 320 bd ft/ас, only slightly higher than the estimate for northeastern Maine in the study by Lorimer (1977) and well below Wilson's "solid minimum benchmark" of 15 million bd ft/township. Not surprisingly, the western Penobscot watershed had some of the highest volumes (20--35 million bd ft on 6 townships), but these were counterbalanced by 20 Penobscot townships with less than 7 million bd ft. The overall average for the western Penobscot was 9.0 million bd ft/ township.

These merchantable volume estimates were made after estimated cull was already deducted and before waste and losses in utilization could occur, and so the volumes estimated by the timber cruisers should be intermediate between Wilson's minimum estimate based on processed lumber (661 bd ft/ас) and his higher estimate to allow for cull and waste (1,181 bd ft/ас). A critical point is that Wilson's estimates are not consistent with the fact that cruisers regarded 500 bd ft/ас in northern Maine as "considerable pine," substantially above the average amount (e.g., "there is considerable pine timber in the valleys, say 500 ft to the acre....").

These low pine volumes also are not explained readily by timber cutting. "Timber depredations" were reported on some townships, but these occurred mosdy in two areas along the Aroostook River and in the setded lower St. John River valley along the New Brunswick border. Townships that had logging were excluded from Table 4 in Lorimer (1977) except for a few townships (noted in the table) where logging was confined to culling of first-quality pine in a limited area along riverbanks in part of the township. Among all 107 townships, 96 had no reports of past logging, and these averaged 7.4 million bd ft, similar to the regional average. Thirty-seven of these had cruises conducted at an early date, between 1826 and 1835. Some of these townships had no suitable streams for driving logs, and on some others, the cruisers noted that blasting and other stream modifications would be needed before the timber could be removed. These 37 townships, most of which were in the western Penobscot watershed, averaged 9.6 million bd ft of pine, only slighdy higher than the western Penobscot average.

The Additional Volumes of Small Canopy Pines. Although pines 9--19 in. in diameter were not considered merchantable at the time, timber cruise records also contain information on their abundance for 28 townships. These include 5 townships for which cruisers reported direct estimates of small pine volume, 6 townships where they reported the proportion of pines falling below the merchantable threshold, and 17 townships where they recorded "no small pines of any note." These 28 townships, covering a total of 640,000 ac, are well distributed across northern Maine and had average' merchantable pine volumes of 8.1 million bd ft, suggesting that they are an adequate sample of the region and typical in their levels of pine abundance.

Records for these 28 townships suggest that smaller canopy pines 9--19 in. dbh made up about 40--58% of the total number of pines in the overall region but 20% or less of the total pine volume. Adding the small pine volume to that of the merchantable pines yields an average total volume of approximately 8.3--10.1 million bd ft/township for the entire region that includes northwestern Maine, ranging from 1.0 to 48.5 million bd ft among individual townships. Depending on the volume table used to make the conversions, total canopy pine densities for the whole tract would have averaged 0.6--1.0 pine trees/ас (range among townships of 0.03--7.8).

Another independent means of estimating densities of small pine is from the witness tree records. Although some witness trees were probably within the merchantable size range, survey records with witness tree diameter measurements in similar forest types elsewhere suggest that they were mosdy smaller trees ranging from 8 to 20 in. (Manies et al. 2001). Mature and old-growth spruce- hardwood forests in northern New England commonly had about 100 trees/ас within this size range (Chittenden 1905). Given the percentage of pine witness trees, the smaller pines would have averaged approximately 1.3 trees/ас on the 1793-1827 survey tract. However, for the Aroostook region, total pine density more than 8 in. dbh would have been only about 0.8 tree/ас. For the western Penobscot region, the witness tree data suggest an average of about 2.7 pines/ас below the merchantable threshold.

The Spatial Dispersion of Pine. How was pine actually dispersed across the pre- setdement landscape? Table 4 from the study by Lorimer (1977) indicated tremendous variation in pine volume and density among individual townships, and Table 5 showed that at a local scale, pine in postfire stands could somedmes occur in densides at least as high as 117 trees/ас. Thus, the calculation of average pine volumes and densides on a per acre basis in that study does not mean that pine was assumed to be uniformly distributed across each township, as Wilson (2005) implied. Mean volumes per acre were provided simply as a readers' aid in visualizing the level of abundance implied by the various township volumes and as a logical way of comparing average pine abundance in different locations and at a variety of spatial scales.

Wilson (2005) proposed two hypothetical scenarios of pine dispersion in which the total amount of timber from the lumber surveys was evenly distributed across 15-27% of the landscape, giving an average volume of about 5,000 bd ft/ас or five large emergent pines per acre on sites where it was present (Figure 5 in the study by Wilson 2005). However, because the land surveys contain spatially explicit information, it is unnecessary to speculate about pine dispersion. In the 1793-1827 land surveys, pines were recorded as being present in varying amounts (including all reports of at least "a few pines") on 35% of the 1-mi township segments. Areas of scattered pine were much more common than medium to heavy concentrations. Twenty-four percent of the landscape had only scattered pine or "some pine," which, based on cruises of areas with similar descriptions of pine abundance, averaged only 99-360 bd ft/ас. Only 3.4% of the descriptions referred to "considerable pine," which averaged 460 bd ft/ас. Areas of "much pine" (or pine listed first) made up an additional 7.6% of the descriptions and averaged a modest 966 bd ft/ас (Tables 1 and 2).

Evidence of pine volumes on a smaller scale is also available from four townships in the western Penobscot region where cruises were conducted on each of the 1-mi sections. The land office conducted these more. detailed inventories because the original cruises indicated volumes about three times higher than that of the average township. Even on these high-volume townships, only 1 of the 152 sections had merchantable pine volumes averaging 5,000 bd ft/aс or more, comparable with Wilson's Figure 5» and

Table 1. Specific tracts in the 19th-century cruise descriptions with both qualitative and quantitative pine volume estimates at the subtownship and township scale.

" For six cases with smaller tract sizes of 5,894--6,350 ас (У4 township), mean pine volume was 142 bd fc/ac. Also includes descriptions of "large quantity of pine," "vast number of pines," "well timbered," and "abundance of good pines."

Table 2. Frequency of qualitative pine abundance classes in the 1793-1827 land survey/ equivalent board-foot volumes (from Table 1 and Figure 2), and weighted pine volumes at the township level.

* Based on a township of 36 mi or 23,040 ac. Many townships in northern Maine are somewhat larger.

256 Journal of Forestry * July/August 2008

only 7% averaged 3,000-4,999 bd ft/ac. More than one-half of the sections on these high-volume townships had less than 500 bd ft/ac, which helps explain why timber cruisers generally regarded 500 bd ft/ac in northern Maine as "considerable pine" (Figure 2).

On small portions of the landscape, pine could attain higher volumes--these were the "compact bodies" referred to by surveyors and timber cruisers and usually occurred along rivers, streams, and the margins of lakes or swamps. In the township descriptions, 10 tracts across the region were singled out as having particularly valuable stands of pine. In the Penobscot watershed, the six choice tracts averaged 5,146 bd ft of pine per acre (range, 875--10,000) on an average tract size of 1,600 ac (range, 200-4,000). The best of the Penobscot stands had 4--10 merchantable pines per acre. The four tracts singled out for special mention in the Aroostook--St. John watershed had much lower volumes, averaging 605 bd ft/ac on an average tract size of 5,100 ac. No doubt, there were smaller areas that could attain more impressive volumes, especially if small pines were included, but these were likely to be groves of more limited extent. In two post- fire stands, Cary (1894, p. 44) reported volumes on a few small sample plots as high as 6,000-19,000 bd ft/aс but emphasized that these were "exceptional" plots "as well covered with pine as could be found."

Do the Township Line Descriptions Indicate Greater Abundance of Pine? Wilson (2005) suggested that the surveyors' descriptions of forest vegetation along township lines are more consistent with his reconstructed higher level of pine abundance, because pine was listed as the third most common conifer behind spruce and cedar {Thuja occidentalism Lorimer 1977). The line descriptions, however, are harder to interpret because they are qualitative, and pine occurrence was recorded much more consis- tendy than other species. Descriptions such as "mixed growth, with here and there a scattering pine" are common. No other species was recorded in such detail; one does not see references to "spruce and maple, with here and there a scattering beech." Furthermore, the large size of the pines, their economic value, and their relative scarcity seem to have influenced the reportage. Townships or parts of townships described in superlative terms as having "an abundance of good pines" or "a vast number of pines" had modest average volumes of less than 1,000 bd ft/ac (Table 1).

The more detailed reportage of pine abundance relative to the other species does, however, provide an opportunity for a systematic test of whether the implied abundance of pine in the line descriptions is higher than in the witness tree or cruise data. The pairing of verbal assessments of pine abundance with the timber cruises of the exact same areas (Table 1), along with data from the high-volume townships (Figure 2) and the "choice" pine tracts mentioned previously, can be used to estimate pine volumes from the qualitative descriptions along each 1-mi segment of the 1793--1827 land survey notes. Because the township line descriptions often included small and poor- quality pine ("scrubby pine," "rotten pine," "fourth-rate pine," and so on), this analysis should provide a more comprehensive estimate of pine abundance that includes many of the unmerchantable trees. These descriptions were also made before 1828 --well before the advent of significant logging--and independently by different personnel. These results show that when the volume categories are weighted by their frequency among the 1-mi township segments (Table 2), the estimated average pine volume for the 1793-1827 land survey tract was 6.4-7.2 million bd ft/ township, similar to the timber cruise estimates reported in Lorimer (1977).

Volume Estimates from Lumber Statistics

All lines of evidence from the 19th century land surveys are internally consistent in indicating modest amounts of merchantable pine in northern Maine, averaging 6-9 million bd ft/township. There is litde indication that merchantable volumes averaged 15 million bd ft or more per township. Spalding and Fernow (1899) indicated a typical cull deduction of 15-20% for old-growth stands, but even if we assume an average 35% cull deduction across all age classes, standing gross volumes for trees more than 20 in. dbh on all 283 townships in Figure 1 would not have averaged more than about 12 million bd ft/township, much less than the 27 million bd ft suggested by Wilson (2005). Therefore, a closer examination of the assumptions involved in translating lumber production records into standing timber is warranted.

A problem in using lumber statistics to indicate standing tree volumes is that the method requires that the pines be treated as a static resource. Wilson (2005) recognized growth of the residual stand over the 40-year period as a complicating factor but discounted it on the grounds that large pines would have been slow growing, and gross growth would be reduced by losses from mortality. He also recognized the possibility of ingrowth from faster growing young pines but reasoned that if such trees were common, they would not have been able to grow into large size classes after only 40 years.

However, growth and yield statistics for white pine show that ingrowth and residual growth of pine stands over a 40 year span is not inconsequential The history of logging in the region was one of repeated harvests on the same areas, with the first wave of cutting removing only the largest and finest pine tiees with a diameter greater than 20 m (roughly 140 years old), commonly scattered Ш a dense forest of other species Al though the oldest pines do have relatively slow growth, residual pines l40 years old would increase in volume 25-50% over a span of 40 years despite mortality {Pinchot and Graves 1896 Spaldmg and Fernow 1899)

To this net growth of residual stands we would also need to add the volume of trees reaching minimum merchantable size over the 40-year span A precise estimate of pme age distribution in the presettlement forest will probably never be Imown, but young pmes were clearly well represented A large area surrounding Chesuncook Lake and encompassing all or parts of nine con tiguous townships was described by survev-ors in the 1830s as "thickly covered by young thrifty timber' between 9 and 16 in in diameter A timber cruiser predicted that the pines on this tract "will be sought after and taken off within 20 years, but [are] worth nothing to manufacture at present' Pmes averaging a foot in diametei in 1832 would be approximately 70 years old on av erage sites and would double or triple in vol ume in the next 40 years (Pinchot and Graves 1896, Spalding and Fernow 1899) Smaller tracts of young pmes were also ex phcitly mentioned in 22 other townships scattered across the region

Furthermore, minimum merchantable diameter progressively decreased during the mid 19th century as the largest pmes were depleted (Spring 1904) Wood (1935) noted that the average log volume at Bangor decreased from 340 bd ft in 1833-1842 to 187 bdftm 1850-1857, adeclme of 45% in only 20 years By 1860-1870, minimum merchantable diameters in the western part ofthe land survey tiact had decreased further to 12-14 in (Hosmerl902) Trees 70-140 years old, which would ha\e been considered unmerchantable at the time ofthe land surveys and earlier lumber surveys, were be-mg included in the later lumber surveys An increased tolerance of cull and defect probably also allowed much additional timber to be included Even as early as 1853, Thoreau was informed by a logger in the western Penobscot region that "what was considered a tip top tree now was not looked at twenty years ago, when he first went in the business, but they succeeded very well now with what was considered quite inferior timber then" (Thoreau 1864, p l44)

Table 3 shows an example of how some of these factors might influence the total amount of pine timber felled from 1832 to 1872, compared with a single merchantable harvest m 1832 The 40-year harvest scenario results in 65% greater yield, even without taking into account the increased tolei-ance of defect About one half of this increase is attributable to a lowering of the minimum merchantable size Of the remainder, about one-half is because of net growth of the residual stand and ingrowth into the youngest merchantable class and one-half is attributable to closer utilization of a scarce resource in the final decade. Even if we assume a loss of 50% volume for all age classes and all decades because of waste and cull, the 40-year harvest still yields 54% more timber than in the . single decade of 1832.

Pine volumes in the Bangor lumber surveys were also inflated by the diversion of timber from the upper Allagash watershed to the Penobscot by means of the Telos Canal, cut in 1841. "Wilson noted this, but assumed that any timber transfers to the Penobscot were balanced by similar losses from the Penobscot to the Kennebec River via the sluice at Northeast Carry. This could not be the case because that sluiceway was not built until 1893 (Gary 1896, Coolidge 1963), long after the liquidation of old-growth pine. Coolidge (1963) noted that 286 mi2 of Allagash timber were funneled into the Penobscot via the Telos Canal. Based on the timber cruises, the aggregate pine timber volume amounted to at least 100 million bd ft, not even counting net growth over the 30-year period.

Cumulatively, these factors all have a sizable influence on merchantable pine volume inferred from lumber production statistics. Conservatively, average pine volume per township would be reduced from 15.2 to about 9.0 million bd ft, which is much closer to timber cruise estimates. There would, of course, still be sizable differences between lumber survey and cruise estimates because of losses and waste. It may be correct that about twice as much standing timber would be needed to generate the lumber volumes estimated at Bangor. But even gross standing volumes of 18 million bd ft/township would only average 781 bd ft/ас, still less than the volume of a single large pine.

It is probably unrealistic to expect much closer agreement between the two sources of data. There are likely several reasons for the remaining differences in pine volume estimates in the two historical sources. Geographical differences between the Penobscot watershed and the land survey tract are probably the single largest factor. Much pine logging in the 1830s and 1840s was in the southern Penobscot watershed (Wood 1935), where volumes were probably much higher (note the strong north- south gradient in pine abundance in the study by Cogbill et al. 2002). Other likely factors include the unknown estimation errors in both data sets and later utilization of pines formerly considered defective. However, as shown previously, even when allowances are made in the land survey data for cull volume and trees below the merchantable size threshold, none of the estimates in northeastern Maine come close to a gross standing volume of 27 million bd ft/township.

Implications for Presettlement Disturbance Regimes

Differences in pine abundance suggested by lumber statistics do not necessarily imply that disturbances were more frequent than has previously been proposed. Disturbance rotation periods are not determined by the volume of pine present on the landscape,-but only by the amount of area disturbed per unit time. The area occupied by pine in presettlement times is reasonably well known from the land surveys (Table 2). Whether the tracts of "much pine" averaged 2,000 bd ft/aс or less, as suggested by the timber cruise evidence, or 5,000 bd ft/ас as in Wilson's hypothetical scenario, is irrelevant to calculation of disturbance rotations. In either case, at least 7-6% of the landscape (Table 2) was disturbed during the approximate 270-year longevity of pine stands, which, by itself, would suggest incredibly long disturbance rotation periods. Much of the scattered pine may have originated after smaller disturbances, but even if all areas with at least scattered pine were considered to be of fire origin, the implied fire rotation period would still be 771 years. This is not much different from the 806 years estimated in the study by Lorimer (1977), even though the latter figure was based on areas with a birch-aspen component, a more reliable indicator of postfire conditions.

It is also a common misconception that long disturbance rotation periods imply that stand-replacement disturbance is rare. A fire rotation of 800 years implies nearly 35% turnover of the landscape during the lifespan of white pine. This would easily account for the observed proportion of the presettlement landscape with a pine component; but when the effect of windstorms and lesser disturbances are also factored in, rotation periods are shortened accordingly. For example, an 800-year fire rotation and 575-year rotation for moderate or heavy windthrow implies a disturbance rotation of only 335 years (Lorimer and White 2003). Although spatial variation in disturbance regimes is still poorly known, rotation periods representing the combined effects of different catastrophic events varied among specific habitats and biogeographic regions, perhaps ranging from 200 to 1,400 years (Lorimer and White 2003). Based on the historic dispersion of large tracts of birch-aspen and young pine, fire was likely much more common on rocky sites .and gravelly outwash habitat, especially along major rivers and travel routes, than for the region as a whole.

Modern management policies can not be rigidly constrained by historic conditions nearly 200 years ago, and the amount of pine in presetdement times is only one factor in guiding current regeneration practices. Investigation of natural disturbance regimes has also been an active topic in recent years, and conclusions may be refined as new evidence becomes available. But currently available evidence from other independent methods (including sedimentary charcoal, investigation of remnant old-growth stands, early lightning fire records, and data or descriptions from 19th century foresters) is consistent in suggesting historic intervals between catastrophic disturbance on many spruce-northern hardwood sites that were substantially longer than the life spans of the dominant species (Schauffler and Jacobson 2002, Lorimer and White 2003, Fraver and White 2005). The Penobscot lumber statistics appear consistent with this large body of evidence, especially when geographical differences between the Penobscot watershed and the broader region of northeastern Maine are taken into account.

Discussion

Toward Sustainable Forestry in California: Mobilizing the Radical Center Louis Blumberg

On a clear, brisk fall day in 2004, two SUVs bounced their way up a dirt road through a northern California forest. The passengers, mostly middle-aged men, included Sierra Club and other environmental activists, representatives of conservation nonprofits, professional foresters, and staff of government agencies. We were on the second day of a two-day trip to take a first-hand look at logging practices in this industrial forest owned by a private company. Our perspectives on forestry spanned a wide range, but all wanted to know whether it was possible to find common ground in the contentious, deadlocked debate over forest management in California. .

The vehicles topped a rise and descended through a redwood-and-fir forest that had been selectively logged two years previously, though it was hard to tell that cutting had taken place. The forest, replete with bird life, looked healthy. There was no sign of erosion, and the fire risk seemed low. The plentiful canopy was made up of trees of different ages, types, and sizes, while on the ground various grasses, shrubs, and saplings were flourishing.

We took a short walk to a small clearing and gathered in a circle in the filtered sunlight to talk about what we had seen. The discussion began tensely, with preconceived viewpoints much in evidence. But crucial questions emerged: Are there old-growth trees here that should be preserved? Is logging like this acceptable to us? Is it acceptable to the public, which in recent years has tended to look askance at any logging at all?

Then there was the practical question of whether we could find enough agreement to craft policies allowing this forest, and others like it, to remain both an economically viable enterprise and a healthy, living forest in the long run. Shaped by several of the activists standing in the clearing, a bill had already been introduced in the state legislature in Sacramento that would prohibit the cutting of all remaining old-growth trees on private land in California, including any in the forest we were visiting. The bill (SB 754 [Perata], 2003--04 session; failed to pass) defined old- growth redwood as at least 42 in. in diameter and 150 years old.

In the redwood forest where we were standing, many large trees appeared to meet the 42-in. limit. But the foresters in the group, two independent professionals and another employed by the landowner, pointed out that it would be time-consuming and costly to measure all of these big trees. What's more, the foresters agreed that some of the large trees were relatively young, growing quickly in fertile areas, but that others in less favorable areas were growing more slowly, yet were most likely more than 150 years old.

The professional foresters concluded that cutting some of these 42-inch trees, regardless of their age, would not damage the forest ecosystem. Instead, it would probably improve it by giving the remaining trees space and light to grow more quickly. To the foresters, this was an environmentally sound project.

The forester representing the landowner added that the selective logging here had worked out financially as well. In deference to concerns of neighbors, the timber company had lowered its revenue goal for this project, removed fewer trees then allowed by the state logging rules, and still showed an acceptable profit.

But today, this type of selective-harvest project isn't necessarily enough. All of us standing there in the clearing agreed that the proponents of the pending old-growth legislation, and probably the majority of the public, would not accept this type of project. We had reached what I still see as the core of the deadlock over sustainable forestry: A forest management policy may be ecologically sound and economically viable, but because some trees will be cut, it is not socially acceptable. (For a brief history of how we got here, see sidebar "From Clear Cuts to No Cuts".)

Later, as I reflected on that conversation, I recalled a book I had read (Halstead and Lind 2001) and concluded that conservationists need to educate and nurture a "radical center" based on a shared conception of sustainable forestry. The radical center I envision is not the "squishy middle." It is "centrist" because it stands between the extremes--the traditional, intransigent stakeholders on each side. It is "radical" (Latin "radix" means "root") because it aims at the roots--the public--and because to succeed, it will require a major shift in the prevailing public view. If forests are to flourish, logging must be rationally controlled, and that means that some trees will be cut. Persuading the broad public to accept this vision of sustainable forestry will be difficult. But we are doomed to perpetual conflict and gridlock if we do not try.

We begin by explaining that healthy, ecologically functioning forests "are not only compatible with civilization, but absolutely essential to its highest development" (Leopold 1918). We emphasize benefits like wildlife habitat, clean water, and recreation. We point out that, by storing carbon in the form of living trees, forests help counteract global warming. Although this conception of sustainable forestry must be linked to efforts to reduce overall consumption of wood and paper products [1], we acknowledge that some timber will need to be cut. We explain that, with proper planning and scientific oversight, carefully executed logging can be both ecologically sound and economically viable.

One policy reform essential to gaining public support, especially in California, is that clear-cutting should be allowed only where it is scientifically essential to restore damaged forests. In its place, timber companies would practice "variable retention" logging, with at least 20% of the wood volume on the site left uncut. Another important reform is the protection of remaining old-growth forests from logging, because they are both a heavily depleted resource and an irreplaceable legacy of yet-to-be-discov- ered secrets.

To ensure that forests survive and thrive over time and across the landscape, "industrial" or "working forests" should be managed for all the benefits they provide, not just timber. As Paul Hawken put it, sustainable forestry "... considers watershed services and biodiversity as well as cultural values; and provides equitable rewards to individuals, communities, and other owners" (Hawken 2007). Although employment will never reach numbers as in the old days, in a sustainably managed timber industry, workers will have stable, well-paying jobs.

In addition to implementing ecologically responsible forest management, the timber industry commits to another key principle. As the Pacific Lumber story shows (see sidebar), aggressive logging to maximize financial returns can produce short-term profits but ultimately leads to ecological, financial, and social bankruptcy. So in the "radical center," revenue goals should be aligned with environmentally sound outcomes. When they conflict, the long-term needs of the forest should take precedence.

New policies can be adopted to create incentives for sustainable forestry, build public trust, and create social license for appropriate timber harvest. One option in forest-rich California, where state regulations are complex and prescriptive, is extending the operating time for a permit to cut timber to those landowners who voluntarily adopt long-term practices more protective of the environment than required by the rules. This type of policy reform would encourage good practices, reduce management costs and regulatory uncertainty, and save time and money for landowners.

The market has a key role to play by rewarding forest owners for adopting long- term, sustainable practices through financial incentives. For example, forest certification, a system that has developed since 1993, could create-sufficient financial incentives to help develop new markets for sustainably produced wood products [2].-

Another strategy particularly applicable throughout the western U.S. would depend on the development of a-market or financial subsidies for using the smallest trees (those with little or no commercialvalue) and other- vegetation, or "biomass," as new commercial products like pellets for wood stoves or as a substitute for fossil fuel in power plants. By creating a new revenue stream, the "biomass" strategy would reduce pressure for the extensive cutting of larger trees, extend the useful life of landfills, help combat global warming by reducing fossil fuel use, and reduce the risk of wildfire.

Another incentive-based strategy that has attracted considerable attention recently and has multiple benefits is linked to efforts to combat global climate change by limiting carbon emissions. Pioneered by the federal EPA in 1993 as a means of reducing acid rain pollution in the eastern and midwestern U.S., a system known as "cap and trade" has become a key component of comprehensive climate change policy proposals. The best- known cap-and-trade system for carbon emissions is the Kyoto Protocol, followed by the related European Trading System. In the United States, 9 Northeast and Mid-Atlantic states are creating a cap-and-trade system for their energy sector, and California is developing such a system in its efforts to implement the state's Global Warming Solutions Act, passed in the fall of2006.

Forests naturally capture and store carbon, and properly managed forests can store vast amounts of it over time. If future agreements, legislation, and regulations are carefully crafted, forests could be the sources of significant amounts of "offset" credits (See sidebar "Forests and Carbon Credits."). This link between forests and climate change illustrates the great potential of financial incentives for stimulating change and spurring forest owners and timber companies to adopt long-term, sustainable forest practices.

Scientific advances in forest ecology, conservation biology, and other scientific disciplines are continually producing tools to help scientists and conservationists more accurately evaluate the effects of logging on land and water and to develop specific remedies and prescriptions to keep forests sound. Our understanding of forests will continue to grow, but the science needed to support the new conception of sustainable forestry is already at hand. Some encouraging signs are beginning to come from the forest industry, where a growing number of forest owners now recognize that the traditional approach of maximizing short-term revenue was not sustainable. Having witnessed the lack of social license for logging as expressed in more stringent regulations and bitter public protests, these few are adopting new financial perspectives and adapting their practices to create stable, viable enterprises for the long haul. Seeing the potential of garnering public support and access to alternative revenue streams generated through market incentives, others in the industry may eventually join in.

From Clear Cuts to No Cuts: How California Reached an Impasse

With little thought about where our wood comes from, the prevailing public presumption is that generally, trees should be protected in parks, not cut. A brief look at the past sheds light on how we reached this impasse.

When California was settled by descendants of Europeans in the last half of the nineteenth century, forests seemed limitless and stumps were signs of progress (Daly and Cobb 1994, p. 253). Given the technology and the relatively small population, the level of logging proved benign. Settlers cut what they needed without compromising: the forests' ability to produce sufficient wood for future human needs or their many other benefits. We still enjoy those other benefits today, including fish and wildlife habitat, clean water, recreation, wilderness, and, an increasingly important function in the current period of climate change, storing vast amounts of carbon.

But after the Second World War, demand for wood products driven by a rapidly growing population and improvements in mechanized equipment led to a great increase in the rate of logging. The amount of. timber cut annually in California quadrupled, and clear-cutting--removing every tree in a parcel-became the preferred method on both federal and private lands (Maine and Knight 1999, p. 5). Throughout the 1960s and 1970s, maximizing revenue continued to take precedence over ecological considerations.

By 1980 Congress had passed legislation attempting to require sustainable logging in federal forests, and California had adopted even stricter rules for privately owned forests. But the general public remained essentially unaware of and unconcerned about logging practices. With the most egregious practices like 100-acre clear cuts deliberately relegated to remote areas out of the public eye, the timber industry retained social license, demonstrated in part by strong Congressional support, to harvest aggressively. But that tacit support would soon prove fragile.

Public concern was awakened in the 1980s, when scientific research began to show that forest management practices were not ecologically sound. A parade of scientific reports documented how intensive timber cutting was damaging habitat, harming key wildlife species, and degrading forest ecosystems. New disciplines like forest ecology and conservation biology emerged to address the question of how forests could be managed to ensure a sustainable supply of both wood products and ecological benefits.

The result was a major shift in the concept of sustainable forestry held by professional foresters and some conservationists. The formerly prevailing view, that limiting the amount of timber cut to the amount grown would be a sustainable practice, was replaced by a new construct in which eco logical and social benefits would be given the same weight as economic returns. To conservationists this is the proper construct or vision, but even today it has not been put into practice in the woods.

Public opinion was changing as well. As the media reported, several decades of intense logging had left forests unable to fulfill their natural ecological roles in the landscape. Salmon populations declined precipitously as streams filled up with sediment, the product of decades of poor road construction and aggressive logging practices. By 1990 the decline of the northern spotted owl became the most discussed sign that logging practices were not ecologically sound [4]. Growing public concern, even outrage, led to lawsuits, court-imposed injunctions, and new laws and regulations mandating greatly reduced logging to protect the forest environment. On federal lands in California, for example, logging fell from two billion board feet to less than 500 million board feet annually between 1989 and 1994.

During this same period, the forest industry was going through a structural transformation. It was completing a technological transition toward greater mechanization, including computer-controlled processing, and would require fewer workers to produce the same amount of wood as in the past. In addition, the industry was consolidating ownerships and operations, and closing out-of-date mills (Brower and Chappie 1995, p. 68).

But the protection of Headwaters Forest did little to relieve the tension or limit the acrimonious debates among the company, state agencies, and local environmentalists over the amount and location of logging on Pacific Lumber's remaining forest holdings. The activists' furor was matched by Pacific out-of- date mills (Brower and Chappie 1995, p. 68).

These changes in the industry, along with the drastically reduced supply of timber due to stricter regulations, led to widespread unemployment and social disruption in logging towns. Reports on unrest in timber-dependent communities soon joined stories of environ mentally damaging forestry practices on the evening television news. Most of the urban public lost any remaining trust in the timber industry, which had joined the list of corporate villains. As a result, social license for almost any sort of logging vanished.

The most visible and polarizing conflict over logging in California involved Pacific Lumber Company and its large holdings in Humboldt County's redwood forests. In an effort to pay off massive debt created through a hostile takeover, the new management of Pacific Lumber doubled the company's previous logging levels, enraging local and eventually national environmentalists. The conflict centered on the company's Headwaters Forest, the largest remaining old-growth redwood forest in private hands in the United States, which environmentalists wanted to protect from any logging. In this highly polarized climate, an activist named Julia "Butterfly" Hill made international headlines by living in one of the forest's old-growth redwoods for two years. The environmentalists' campaign to protect this forest lasted ten years. Finally, in 1998 the federal and state governments purchased the 7,500-acre Headwaters Forest for $480 million, placing 3,000 acres of old growth and 4,500 acres of second growth in public ownership and off limits to logging.