The Natural History of the Present, Chapter 16

Chapter 16: Forests

Like fields, forests affect atmospheric chemistry, the chemistry and flow of local streams, local and global precipitation, climate. Root growth accounts for 50% of the net primary productivity in forests (the net yearly production), leaves for 25-35%, growth of wood the rest. About 20% of the mass of forest trees is roots, most in the top few inches of soil. The growth and death of roots delivers carbon to the soil, removing it from the atmosphere. Thus old-growth forests continue to store carbon, even after growth of their trunks has slowed. Storage varies with the type of forest. It falls precipitously in southern pine plantations after 25 years, when their growth begins to slow. The dry Ponderosa forests of the Rockies, with their frequent low level fires, may be carbon neutral, while the mature rain forests of the Pacific Northwest continue to store considerable carbon. In general, forests sequester between 0.5 and 1.0 ton of carbon per acre per year. The alluvial forests of the Mississippi Valley are capable of storing 2 tons of carbon per acre per year—and so, at $20-$50 a ton, or $40-$100 an acre, are worth in carbon storage what they are as farmland. Tropical peat swamp forests in Indonesia sequester 200 tons of carbon per acre in their soils and standing biomass, an argument for paying Indonesian landowners to not cut them down. Destruction of tropical peat swamp forests—for instance, for palm oil plantations to produce green diesel fuel for Europe— takes a century or so to create a carbon benefit and currently produces more carbon dioxide than the burning of fossil fuels in China. Because of the clearing of its tropical forests, per capita emisssions of carbon dioxide in Indonesia are close to those in the West.

Evergreen forests release large amounts of hydrocarbons in the summer, as part of their cooling mechanism; these chemicals become part of global chemistry and in the presence of nitrogen oxides (products of combustion that are always present in the modern atmosphere) and sunlight, produce photochemical smogs. (So President Reagan’s comment that forests produce smog.) At least 120 chemical compounds are present in Sierra Nevada mountain air. These are insect deterrents and thermo-regulating chemicals. The monoterpenes, which prevent and cure cancer, are among the most abundant. Monoterpenes enter the bloodstream through the lungs and the limbic system through the olfactory nerves. So the healthy effect of a breath of mountain air has some basis in fact.

A forested landscape has a lower albedo (absorbs more of the sun’s heat) and so is warmer in winter; the sheltering trees also make the temperature of the ground much less variable than that of bare ground or grassland. In summer, a forested landscape tends to be cooler because of transpiration by the trees. (On Vancouver Island, in a relatively cool climate, cutting the forests has produced a permanent temperature increase of 1º-2º C.) Over large areas, the albedo of the ground affects temperature, precipitation, and the speed of jet stream winds. Because of their absorption of sunlight, evergreen forests in snowy regions probably have a net warming effect on the climate, despite their absorption of carbon dioxide from the atmosphere. Snow melts some weeks later in forests (especially evergreen forests) and so spring run-off is later; and slower. A later runoff makes (for instance) more water available over a longer period in California rivers. Fog-catching trees, like redwoods, increase streamflow (2 to 3 times more water reaches the ground under redwoods than not); fog-catching forests on the Canary Islands recharge the islands’ aquifers, which dry up when the forests are cut (and cannot regrow because the ground is too warm for condensation to occur); fog-catching shrubs on the Galapagos Islands drip moisture onto soils where it waters herbaceous plants and collects into pools, providing water for giant tortoises. In general, forests lower runoff by intercepting and evaporating 35-50% of the precipitation that falls on them. They thus reduce peak flows (floods) and, by storing more water in their soils and recharging aquifers, increase the low flows of summer. The water that flows from them is cooler. In humid climates such regulation of streamflow is generally desireable; in drier climates, forests, especially introduced forests (eucalyptus in California, saltcedar or Tamarix in the American Southwest, Mediterranean pines in the fynbos shrublands of South Africa) may reduce surface water flow considerably (to zero in small streams), with serious results.

Forests control stream temperatures. Summer water temperatures are 10 degrees lower in shaded reaches of streams. Streams that flow through old-growth forest accumulate fallen logs at a rate of about 1 every 10 feet. The trees help stabilize the streams, creating pools. Their wood is turned by detritus feeders into nutrients for freshwater organisms, which become food for fish; the nutrients are retained in the swirling pools. Salmon reproduce and survive much better in log choked old growth streams, with their shade, slower currents, more complex habitat of pools, more complete nutrient capture by stream organisms, than in the more or less channelized streams of commercial plantations. Undeveloped lakes in Minnesota have about 500 fallen logs per kilometer, or about 1 every 6 feet; while developed lakes have about half that, and near cabin sites have only 1 log every 50 or 60 feet. (People remove them and cut the forest, the source of new logs, along the shore.) In lakes, logs provide habitat and shelter for fish and detritus for invertebrates. The logs become part of the lake’s food chain. Log-free parts of lakes have fewer fish. (Bass, for instance are 5-60 times more abundant near brush shelters.) In rivers, grounded logs catch other snags, form jams, sometimes accumulate sediment and become islands, force the river about them and form side channels. The slowly flowing side channels become habitat for ducks and juvenile fish. Fallen logs along rivers also protect the shoreline, create pools, slow water flow, enlarge (with partial dams) the floodplain and thus the riverine habitat, and provide detritus-based nutrients. Sailing out to sea, rafts of logs, hung with barnacles and other creatures, increase the size of the ocean edge (several hundred thousand per year may have sailed out from pre-contact North America); and perhaps introduce alien organisms to new environments (a job done much more efficiently now by the ballast water in ships). For unknown reasons, floating logs in the ocean attract schools of fish. Floating man-made constructions with homing beacons are used as lures in some West African fisheries. In West Africa many logs from illegal cutting are lost in river drives and float out to sea, ending up on beaches in numbers that interfere with nesting sea turtles.

In aboriginal forests much of the forest was old growth, or so-called primary forest. How much was old growth varied with the type of forest and the site. The mixed deciduous and conifer forests of Maine are thought to have been about 30% old growth. Records from early surveys in Maine indicate that 2% of stands were recently burned, 14% were birch and aspen (short-lived, early successional trees), 25% were young forest (75 to 150 years old), 32% were mature forest (150 to 300 years old), and 27% were old growth (greater than 300 years old). So about 60% were what we would call old forest. Natural forests are subject to many types of disturbance, which range from the deaths and collapse of individual trees to hurricanes or tornadoes blowing down whole tracts of forest. The type of destruction varies with the location of the forest. Natural fires in the northern hardwood/hemlock forests of Maine occur every 800 to 1400 years, an interval much longer than the maximum ages of the trees (250-500 years). Icestorms in interior New England occur perhaps twice a decade and serious hurricanes once a century. Windthrow, along with ice storms, heavy early snows, pathogens and fire (spreading from the more fire-prone coastal and northern forests) likely kept Maine’s forests young. In the Midwest the interval between disturbances is also much greater than the lifetimes of the trees (800 years for straight line winds from thunderstorms, 1000 years or more for tornadoes). The natural fire interval is shorter than in the east (summers are hotter and drier) but about 85% of the forests of the Upper Midwest were mature and old growth. Catastrophic crown fires occur every 150-500 years years in mesic stands of Douglas fir on the western slopes of the Cascades in Washington. Less destructive ground fires are more common. The trees are somewhat fire-resistant and also have adaptations to fire. Once the lower branches of a Douglas fir become shaded and begin using more resources than they produce, they lose their needles, die, and fall off; in time this removes the fuel ladder that lets ground fires reach up into the crowns of the trees. Huge floods occur perhaps once a century on creek bottoms inhabited by coastal redwoods; some trees are felled by the floods. New shoots sprout from their trunks and broken-off stumps. Trees still standing send out adventitious roots into the deepened muck, and benefit from the new soil. (Fifteen such floods have raised the floodplain along Bull Creek, a tributary of the Eel River of northern California, by 30 feet over 1000 years.) Such adaptive behavior lets the trees live for two millennia and outcompete the Douglas firs and California bays on rich alluvial flats. Stand and soil-removing fires follow insect infestation every 50 to 200 years in boreal spruce-fir forests, resulting in complete stand replacement and temporary soil impoverishment (many nutrients are vaporized).

North America now lacks large tree-breaking animals like the elephants that renew acacia woodlands in Africa, but insects like spruce budworm kill stands of balsam or spruce, which may then be replaced by birch or aspen (which are periodically defoliated by tent caterpillars). Budworm outbreaks, like that from 1950-1954 in Atlantic Canada, kill whole forests, when then burn and regrow. During a budworm outbreak, budworm larvae increase from 1000 per acre to 8 million per acre, and the populations of wood warblers that eat budworms increase by 10 times or more; but not enough to contain the insects. (A virus or the death of the forest does that.) By allowing the (less dominant) aspen and birch to replace the (more) dominant spruce and fir, budworms renew old or stressed conifer forests. Conifer productivity on a site reaches its maximun extent at relatively low levels of nitrogen. Hard to break down, conifer litterfall ties up (the unneeded) nitrogen, while the nitrogen in the leaves and branches of the more demanding hardwoods is quickly recycled. So patches of aspen and birch among the conifers raise the productivity of the forest and the spread of the trees after a budworm outbreak (and fire) help renew the soil. (Nitrogen availablility under an intact stand of hemlocks — another conifer — is low, but equals that of a hardwood forest when small groups of sugar maple share the canopy with the hemlocks.) Some writers claim hardwood trees and shrubs have a keystone function in conifer forests. They provide habitat for butterflies and moths, which attract birds, which also eat caterpillars that eat conifer foliage; they provide habitat for parasitoids that help control conifer-eating insects; and their sprouting ability stabilizes soils after disturbances like treefall of fire. The spruce-fir forests of Atlantic Canada, where budworms and hardwoods have been controlled by spraying, and the proportion of spruce is falling thanks to heavy cutting (the forests are being turned into monocultures of faster-growing balsam fir) are becoming less productive.

Beaver dams in northeastern conifer swamps also kill conifers, which are replaced by damp-tolerant aspen and alder, whose litterfall enriches the forest edges. The trees themselves enrich the habitat, as well as provide food for beaver. The turnover of the plants and animals in the ponds (which also store soil and nutrients moving downstream) increases the nutrients available for the trees. Browsing by mammals also shapes forests; an increase in browsing by deer, elk, moose, snowshoe rabbits, meadow voles, by decreasing the survival of the more palatable species (oak, white pine, eastern hemlock, Canada yew, northern white cedar), will change the future composition of a forest, and reduce a mixed herbaceous understory to ferns and grass. Heavy browsing by elk in Yellowstone tends to eliminate streamside willows and reduce beaver wetlands on small and medium-size streams; the streams straighten out, become more erosive, cut more deeply into their beds, thus lowering water tables. Browsing by elk also prevents reproduction by aspen (another beaver food). The reintroduction of the wolf to Yellowstone let both types of tree come back, partly by reducing the elk herd (wolves seem to keep the herd 20-30% below what weather and vegetation would allow), but mostly by changing elk behavior: elk are frightened of spending too much time in dense cover where they might be ambushed by wolves. With the return of willow and aspen, beaver came back, trout became more abundant in the deeper, slower streams and a greater variety of songbirds bred in the streamside vegetation.

People influence forests. Throughout the temperate zone, people began to shape forests long before they had reached equilibrium with their post-glacial environments (3000 to 6000 years ago in much of North America). Anthropogenic fire, such as the cool ground fires set in early spring or fall by Native Americans in the oak and pine woods of southern New England, produced an open park-like wood of large nut-bearing trees, mixed with areas of younger vegetation, amidst much larger areas (such as white cedar swamps and spruce-fir forests) that were never (or rarely) burnt. Anthropogenic fire probably created the open oak and hickory forests of the Middle West. Without fire, they are now returning to the more shade-tolerant beech and maple. Primary forests have large trees, with a standing biomass 3 to 6 times that of second growth (and several times the board footage, which explains the size of the early cuts). The most productive old-growth forests in the eastern United States carried 125 to 250 tons of wood per acre. In old photographs the tall trunks of Middle Western oaks and sycamores dwarf the nineteenth-century biologists standing beside them. The large old trees in these forests produced enormous amounts of nuts and seeds, the so-called mast.

Once they reach a certain age all the trees in a forest produce mast. Most trees produce some seeds after a few decades, but become large producers as they mature in size and have spreading branches and root systems capable of supporting larger numbers of seeds. (The story is the same as with fish, where large females produce many times the eggs of smaller animals.) Mast production is related to crown development, so burning, by thinning the forest, increases mast production; the more rapid turnover of nutrients in a burned forest may also help produce more seeds. In the cool Northeast, a red oak begins producing acorns at 25-50 years, and then continues production (which would reach a peak at perhaps 150 years) for 100-300 years. White oaks and sugar maples, longer-lived trees, produce large crops of mast for longer periods. The smaller seeds of ash, maple, elm, those of the evergreens, are food for birds and rodents. The several varieties of red crossbills, a finch of the boreal forests, specialize in the seeds of specific species of evergreens, to which the shapes of their beaks and tongues (the seed-extracting devices) are adapted. Seed production by conifers and birches (an alternate food) tends to be synchronous over a large area, but irregular, so the birds move long distances (hundreds to thousands of miles) over the boreal forest in their search for food, and adjust their breeding schedules to the supply of seeds. With sufficient food, they will breed in mid-winter, with temperatures -30° F. or below. Thus the populations of many northern seedeaters (crossbills, redpolls, pine siskins, pine and evening grosbeaks) are not regional but continental. The rodents that eat the tree seeds support populations of predatory birds and animals (weasels, minks, martens, owls), and recycle (like the birds) most of the nutrients in the seeds back to the trees through their droppings. The larger tree seeds (beechnuts, acorns, hickory nuts, pecans, formerly chestnuts) are eaten by both the rodents and the larger animals (deer, elk, turkeys, bear, probably buffalo). Such animals and the nuts themselves were once eaten by people; aboriginal people also ate the small rodents (mice, squirrels), which are extremely abundant in mast-producing forests. (About 222,000 mice and other small rodents live in a typical 10 square miles of eastern forest, which explains the presence of mouse skeletons in human coproliths.) Recently 3-5 billion passenger pigeons supported themselves on the mast of the Eastern forests (squirrels may have exceeded that number), nesting where the acorn crop from the last year remained on the ground in the spring. Their habit of migrating over wide fronts in parallel flocks has been interpreted as an adaptation to locating areas of sufficient food to nest. (As part of their scheme to keep populations of seed-eaters low, so some nuts will survive to sprout, most nut trees bear only every second or third year.) Much of the forestland that supported those birds has been converted to cornfields; corn, cattle and pigs have replaced the passenger pigeons and beechnuts. (Over time, in settled countries, the best forestland always becomes agricultural land. So wildlife abundance inevitably falls.)

Trees reach economic maturity long before they reach biological maturity and so, under modern economic management, are cut down soon after they become large producers of mast. This changes the nutritional relationship of the forest with its inhabitants, reducing the numbers of seed-eating animals and their fur-bearing predators, while increasing the numbers of animals that eat browse. (White-tailed deer, cottontail rabbits and grouse do well; turkeys, squirrels and bear are probably reduced. Many animals, including deer, turkeys and bear, eat both browse and mast). It eliminates the tall, old trees raptors use for nest sites, from whose dead lower limbs flycatchers hunt, and the old rotten trees in which woodpeckers and chickadees excavate nest-holes, and in whose large hollows bears and raccoons den. Economic maturity is determined by a change in the rate of increase in the mass of the tree. As trees approach their final heights, the yearly increase in new wood fiber slows. In long-lived woody plants, maintenance respiration increases with age; so there is a necessary reduction in net annual production of fiber, that is, of growth. Risk of loss of the tree from windthrow or disease also rises with age; so the trees are cut. Douglas firs, trees that reach ages of 500-700 years, that stand a century or two as dead snags, and whose fallen logs take up to 300 years to rot, and so whose individual influence on the forest spans a millennium, are cut down soon after they reach the majority of their height growth at 80 years. Redwoods that once grew for 2 millennia or more are also cut at 80 years. Old Douglas firs support a community based on arboreal lichens, whose mass can be 4 times that of the foliage. The tops of old-growth redwoods support another forest 150 feet up. Protruding redwood tops break off in storms and the upper branches bend up, turning the top into a small forest of two-foot thick vertical branches, dead stubs, soil (which collects in cavities left by broken-off stubs and in hollows in the branches), with banks of fern, blueberries, seedlings of Douglas fir and Sitka spruce, and nesting seabirds. (Part of the temperate rainforests that occupy west-facing coasts worldwide, most redwoods are found within 10 miles of the sea. They grow where their limbs and needles can comb water from coastal fogs—12 inches during the usual summer dry season—and out of the reach of salt spray. Marbled murrelets nest in their crown forests, ancient murrelets among their roots, the former now in steep decline as their nesting habitat has shrunk from 2 million to a few thousand acres.) Southern pines, grown mostly for paper pulp, are cut at much shorter rotations, 25-35 years (younger trees make better pulp). Like browsing, cutting changes the composition of the forest. The cutting rotation in the Northeast is now too short for hemlock, a tree that can live 600-900 years and is probably worth cutting at 150 years. (The crown forests of old hemlocks show some of the characteristics of old redwoods.) Frequent cutting in the Maine woods has favored balsam over red spruce (another slow-growing tree). In 1902 the volume of spruce was 7 times that of fir; today the two are virtually equal in volume. Frequent cutting in the forests of northern Pennsylvania favors trees that sprout (cherry, oak, maple) over softwoods like white pine and hemlock. Early logging probably favored American chestnut, which rose from 4-15% of forest volume in early surveys of New Jersey and Connecticut to 60% (becoming the most abundant tree species) in 1900. While oaks and hickories produced hundreds of pounds of nuts per acre, chestnuts produced thousands.

Cutting can also mimic natural process. In Sweden white spruce usually reproduces itself through small gaps, such as the death of a large tree. Such forests can be reproduced by selective cutting that leaves most of the forest intact and also leaves existing snags and fallen trees (until recently the United States Forest Service required their removal on logging sites) and some old trees. One would also leave most of the deciduous trees, which have important functions in conifer-dominated woodlands. Spruce-fir stands in boreal North America are often fire-adapted communities, reproduced by stand-replacing fires. Such communities can be maintained by a modified clearcutting, leaving wide bands of trees along watercourses, with large, scattered clumps of trees about old trees left in the cut as nesting sites, sources of seed and future snags. Tops and slash are left on the ground. Spruce and fir are usually used for pulp in the Northeast and so are cut as soon as they are large enough to make cutting economic (this size goes down as logging becomes more mechanized). Red spruce makes valuable lumber (it is used for pianos); it is a slow-growing tree and leaving clumps of red spruce among the balsam to mature increases the variety and value of the forest. Letting some aspen and birch mature (birch for sawlogs, aspen for pulp) increases the forest’s productivity and makes it more friendly to game animals. To minimize blowdown and drying of the forest, such clear-cut strips are usually limited to a width 1.5 times the height of the trees (or 100-150 feet). Such cuts attempt to mimic the effects of catastrophic disturbance in primary forests and leave the landscape conducive to the movement of seeds and animals. In the East, such cuts are not burned (similar, lodgepole pine cuts in the West probably benefit from being burned). Leaving merchantable trees in the woods increases the future health and productivity of the forest. It reduces the volume of wood harvested, and thus current profits to the landowner. In the future, losses in volume will be compensated by the more valuable logs from old trees.

As forests mature, they accumulate dying and fallen trees. Fallen trees cover up to 20% of the forest floor in mature Douglas fir woodlands in California. A recent checklist for old-growth forests in the East included 3-4 logs greater than 16 inches in diameter, per acre, on the forest floor: a much smaller number. Hollow trees become den or nesting trees for various animals and birds; large hollow trees are used by bats and chimney swifts as well as hibernating bears: Audubon counted 9000 chimney swifts leaving one hollow sycamore. (Bats now make use of attics and mines, chimney swifts use chimneys, both birds having become in the Northeast—like the nighthawks which nest on the flat roofs of commercial buildings, and whose notes drop from the skies of suburban evenings—animals of settled landscapes.) Downed trees create new habitat on the forest floor for invertebrates, rodents, birds and amphibians. Soil accumulating on the upslope side of a log provides habitat for burrowing insects and small mammals, while the downslope side provides shelter and nesting sites. An acre of Maine woodland has more biomass of salamanders than of moose: a fact only briefly surprising. Rotting logs have a greater mass of living tissue (in their plants and decomposers) than growing ones. Nitrogen-fixing bacteria living in the guts of Pacific dampwood termites fix nitrogen for the termites. The nitrogen ends up in the nutrient-poor ecosystem of the rotting log and fertilizes tree seedlings growing on it. The invertebrates of the logs and the forest floor (earthworms, centipedes, millipedes, spiders, beetles, mites, springtails, psocids, nematodes) are eaten by small mammals, birds and amphibians (shrews, toads, frogs, thrushes, salamanders; also by the less abundant snakes and turtles); the amphibians, reptiles, birds and mammals form one basis of the forest’s food chain. The fallen logs also act as dams to keep forest soil from moving during rains and thaws. The old trees of a primary forest create a moister, shadier environment. Their large limbs are habitat for lichens that synthesize their nutrients (including nitrogen-rich proteins) from the air. These nutrients enter the forest ecosystem through leaching and litterfall, and through the dung of the squirrels and rodents (often different species than the mast-eaters) that eat them. It is said that young Douglas firs first grow on the nitrogen banked in the soil by the nitrogen-fixing shrubs that follow forest fires, then (as adults) on the nitrogen synthesized by their lichens. A major food of the rodents of old growth woods are the fruiting bodies of the mycorrhizal fungi that are allied with the roots of the trees. The fungal spores pass through the rodents’ digestive systems unharmed and are deposited on the forest floor and in their tunnels and burrows, generating more fungi. These rodents are part of a new food chain of the primary forest that ends in predators of the deep woods, such as pine martens and spotted owls. The water that flows out of such woodlands, thanks to the shade and the lack of sediment, is cold and clear (sometimes tinged brown with tannins), nutrient poor, habitat for salmon and trout. Tree roots stabilize the streams; and the trees themselves, falling across the brooks and larger creeks, also stabilize them, creating pools. Salmon, running upstream, like the seabirds that nest among the trees, contribute nutrients from the sea.

That only a third of the mixed Maine woodland was in primary forest implies a fairly high level of natural disturbance. In much of the Appalachians and Midwest the disturbance regime was several times the lifetime of the trees (even that of the white pines, which live 500 years). The result was old forests with complex mixes of trees of different shade tolerances and ages. In natural forests non-catastrophic replacement is usually in patches. One or a group of trees that rise slightly above the canopy are blown down, or a tree collapses from old age and takes some others with it, and the shade tolerant saplings that have been growing in the understory shoot up. If light is sufficient, the seeds of the shade-intolerant species that have been waiting in the soil sprout and grow. (Yellow birch and white ash will grow in the gap left by a single tree, while red oak, black cherry, sweet birch and tulip tree need larger gaps.) Hurricanes, tornadoes and fires clear larger areas. What develops on a site after a catastrophic disturbance depends on what trees are left alive (the so-called seed trees), the presence of trees that sprout from roots (many deciduous trees and shrubs), the seeds in the soil-bank, and what is brought to the site by birds, mammals and the wind. So the forest one sees on a site depends on the site’s history, its surroundings (another history), and chance. In much of the conifer and mixed hardwood forest of the Boundary Waters Canoe Area, the original fire rotation time was thought to be 50-100 years. Such short rotations favor fire-adapted trees: those that sprout from roots, such as quaking aspen; or that have well-dispersed seeds, such as paper birch; those with serotinous cones carried high in the canopy, such as jack pine or black spruce; or trees whose thick bark lets them tolerate fires, such as red pine, which become fire-resistant after 50 years. Short fire rotations must be fairly ancient in the general area since the Kirkland’s warbler is adapted to breeding in fire-succeeded stands of jack pine; it will breed nowhere else and abandons older jack pine stands. If a new stand escapes fire for 40 years or more, jack pine and aspen slowly succeed to more fire resistant, very mixed stands of black spruce, balsam fir, paper birch and white cedar. Red pines, which are capable of masssive recruitment after a fire slowly die out in very old stands (very old: red pines are still present after 300 years). In this case pines will no longer reproduce if disturbance is limited to windthrow and spruce budworm; they need fire. The spread of hemlock into a stand lengthens the fire frequency indefinitely (they draw up groundwater to the surface and moisten the soil), and if a stand escapes fire for a very long time, a forest dominated by hemlock and the shade tolerant sugar maple may develop. By lengthening the fire frequency, the hemlock makes possible the success of the fire-susceptible maple. Both hemlock and sugar maple produce a forest floor unfavorable to the establishment of other species, so this forest is somewhat self-maintaining (but in fact no more of a “climax” than the more mixed forests it replaced: felled by a windstorm, it will be replaced by aspen, spruce, and birch).

Primary forest in the eastern United States is characterized by fairly large trees (16-20 inches or more in diameter according to a recent checklist; but certainly many first-growth trees were larger). Other characteristics include a relatively small percent of intolerant (pioneer) tree species in the canopy; a full compliment of spring ephemerals suitable to the site (flowers that bloom and set seed before the tree leaves come out; these plants are not too much bothered by occasional logging, but frequent logging, clear-cutting or grazing will reduce them; since they are pollinated and their seeds distributed by ants, they spread slowly and can take centuries to recolonize a site); by a full compliment of bryophytes (mosses, liverworts, lichens: plants of damp and shaded woods), including those characteristic of old growth; by large old logs on the forest floor (microhabitat for many birds, mammals, reptiles, invertebrates, decomposers, mosses, fungi, tree seedlings; forests in northern regions probably need 12 tons of coarse woody debris per acre to maintain site quality); and by large, standing snags. Many of the large trees of a primary forest are rotten. (A complaint of woodcutters in early Maine, who would chop a hole 3 feet into an old pine, 50 feet above the ground, to check its condition before felling it. The wood of such large old pines, so-called pumpkin pine, was valuable for being soft and easily worked.) Many species of birds and mammals in the eastern forests require standing dead trees for perching, foraging, nesting, roosting, and denning. The ivory-billed woodpecker is said to have specialized in debarking large, dead trees to eat the insects beneath the bark. The demise of such trees as southern bottomland forests were logged and converted to agriculture was a major factor in its extinction. (Audubon shows a family of ivorybills collecting beetles from under the bark of a small dead stub.) Many writers would claim that old growth or primary forest is not absolutely necessary in a functioning natural landscape, even along streams, but no one can be sure of this. Some of the forest’s inhabitants (those microbes, fungi, invertebrates) will go extinct without it and the landscape in a larger sense will be reduced. Other writers argue that tracts of old growth must be very large to function truly as primary forest. The woodland must contain the large carnivores (grizzly bears, wolves, mountain lions) whose presence indicates an intact food web. Such tracts require 250,000-300,000 acres of contiguous woodland, with connections to other such areas.

Logged forests are not natural forests. Economic considerations will never allow them to move far into the realm of old growth. Their soils may or may not achieve a net accumulation of carbon. Logging disturbs the soil and exposes it to sun and rain; increased microbial action in the warmed soil releases carbon and other nutrients. The decay of logging debris releases carbon. It takes 10 to 20 years for a logged second-growth forest to begin a net accumulation of carbon, 45 years for regrowth to compensate for the carbon released by the decay of debis left from cutting an old-growth forest. And the carbon in the logs removed returns to the atmosphere quite rapidly. A third to a half of the logs end up as waste (sawdust, planer mill shavings, cut ends of boards), which is burned or decays. The rest of the carbon in the logs returns to the atmosphere after a short detour through the man-made world (for much of the material, made into paper or pallets, less than 10 years; that made into furniture or buildings lasts somewhat longer — less than 50 years on average). But it should be possible to mimic somewhat the process of natural forest succession through logging: to provide a natural variety of habitats (if not in the same proportions as in the aboriginal forest); to leave some old trees and downed logs; to conserve forest soils; and to protect forest streams from too much sun and from the constant pulses of silt and nutrients that logging produces. All this of course will cost the landowners and loggers money, at least at first.

Cutting, like fire, always sets the forest on a new trajectory. In the humid forests of the Northeast and Middle West, selective cutting favors the shade-tolerant beech and sugar maple; clear-cutting the warmth-loving pines and oaks, and other pioneer species like aspen, pin cherry and paper birch. Clear-cutting in dry western forests can eliminate the mycorrhizal fungi on which the trees depend and make reforestation of west-facing slopes difficult. (Adapted to living with certain species of trees, the fungi must find a new host in two years or die.) Clear-cuts, and heavy selective cuts, create more or less even-aged stands, with more pioneer and early successional species, while light selective cutting creates forests of more shade-tolerant species. The early successional forests created by heavy cutting (taking all merchantable trees when entering the woods) are supposed to provide a constant, sustainable yield of wood, at the maximum potential of a site. But, an artifact of cutting, these forests are quite unstable. They seek to produce only fiber. Natural stands in the same areas would be a mosaic of old growth, maturing trees, tangles of young growth, species with different tolerances to light, with a full assortment of the birds, mammals, invertebrates and fish characteristic of the area. The forest in a given place, with its vertebrate and invertebrate inhabitants, would depend on the sorts of disturbance to which the stand had been subject; that is, on the site’s particular history.

Logging influences what trees grow in a forest, and in general speeds up (may double) their rate of growth. Logging also has many negative effects on a forest. Soils are compacted and erode. Soil oxygen levels are lowered. The stems and roots of trees are often wounded. Root compaction and oxygen deprivation slow growth for some time after a cut. Heavy cuts move water and nutrients into streams. In areas where the natural water temperature in summer is near the upper limit of their requirements, clear-cutting may raise stream temperatures above what salmon and trout tolerate. So fish disappear. Cuts along streams remove the trees whose roots hold the bank together. With the additional water and soil moving into them, the streams widen, scour, straighten out and take some time (usually more than a decade) to return to a more stable condition; then the logging starts again. Heavy cutting on steep slopes in the Pacific Northwest makes the slopes liable to landslides following insignificant rains; slides begin three years after logging, when the roots that held the slope together have rotted. Cutting along streams removes the logs that would have become coarse woody debris in the water and on the forest floor. Such logs retard water flow, and help stabilize streams. Logging on the banks fills streams with tops and slash. The cumulative effect on streams can be catastrophic for fish, as is the case with many salmon and trout streams in coastal California, Oregon and Washington. Slope tremendously influences the erosive capacity of water, which increases by the fifth power of its velocity. As the speed of flowing water (down a skid trail, through a culvert) doubles, it is able to carry 32 times the sediment and move particles that are 64 times heavier. Much of the sediment in west coast streams comes from logging roads and skid trails. When these are treated properly (which includes closure to vehicles) after logging is over, the sediment loads decrease, the streams clear, and salmon and trout return to them. Hunters and drivers of recreational vehicles object to such closures, which makes the back country only accessible by horse or foot. (Included with the sediment are some of the hydraulic fluids from the logging eqipment, 70-80% of which escapes in leaks, spills and line and filter failure, and some of the chain oil and engine oil from the saws: arguments for using biodegradable oils.)

A forestry that tried to recreate the forest as an ecosystem would place old growth in the commercial forest along streams. Very lightly logged primary forest (a tree per acre every 5-10 years) would cover the banks of deeply incised streams, and cover the streambank out for 50 to 100 feet otherwise. (A tree per acre every 5-10 years is 4-8 trees per acre for entries at 40 year intervals; depending on the market, logs from such trees might be worth $1000 or more at the mill, or $300 in stumpage to the landowner.) Probably less than 20% of eastern forests have soils good enough to make the risk of growing trees greater than 100-150 years old worthwhile (sites in coves, valleys, the toe slopes of hills, in soils derived from nutrient-rich bedrock: most of the best forest soils in the East are now in agriculture); so one puts old growth where, whether a good commercial risk or not, it will do the most biological good. Then fingers of old growth penetrate the logged woods: primary forests are multi-story forests in the Northeast and upper Middle West, where most trees top out at 80-100 feet, and large white pines penetrate the canopy to reach 150 feet or more. These fingers of old growth will not function well if they back entirely upon clearcuts (winds will blow the trees over, too much sun dry the forest floor). So the forest that abuts the old growth is a mix. If we follow the Maine example, one third would be in early old growth (cut at 125-175 years; perhaps let mature further in better sites); poorer sites would be cut at 80-100 years. The value of the water running off the land and the land’s value as a carbon store will affect cutting choices. Two-thirds of the forest would be maturing and young forest. If the water that runs off the forest is worth $30 an acre, when close to the presettlement original amount and condition, forest management that keeps the water in that condition should be worth that. Carbon storage in trees and soils depends on the site, the forest and its age; if it amounts to half a ton per acre per year and stored carbon is worth $50 a ton (probably a high figure), then carbon storage is worth $25 an acre. Over a 150 year rotation, $55 an acre a year ($8250 dollars) is likely comparable to the stumpage value of the timber.

Logging is always somewhat destructive to a forest. By opening the canopy, logging increases light, temperature and wind speed; it lowers relative humidity; disturbs and compacts soils (especially the top few inches where most of the tree roots grow; this is why the worst time to log is spring, the best time in winter on frozen ground). Logging damages nearby trees and causes wounds in roots and stems that leave trees susceptible to decay or disease. After logging, erosion increases in the forest as a whole and especially on the 5-10% of it that is in roads, skid trails and landings. Something like 65% of the lubricants and coolants used in logging machinery end up on the forest floor. But logging also speeds up natural processes, raising the 2% return in growth on unlogged stands to perhaps 4%. Young stands of northern hardwoods will naturally thin themselves from 1000 trees per acre at four inches diameter at breast height to 40-60 trees per acre at 20 inches diameter. (Since a site can support only so much basal area of tree per acre, as the individual trees get thicker, there must be fewer of them.) This process takes 200 years in nature, but half that if the stand is thinned periodically. Another way to put it is that for the trees in a fully stocked stand to gain an inch in diameter, 20% of them must die. So this is the rationale for periodic logging in ecologically managed forests. Thinning of long-rotation hardwoods in the Northeast usually starts at 30-40 years when they have reached (coincidentally) 30-40% of their mature height. Letting them remain crowded up to then forces the trees to grow taller. (Deciduous trees grow upward in response to light and will adopt a rounder, shorter habit if not forced upward by competition; conifers grow from an apical bud that responds to gravity and so continue upward no matter what the competitive situation.) Thinning lets the crowns develop (the crown should occupy 35% of the height of the tree) and the boles increase in diameter faster. Trees 8 inches in diameter are usually thinned to 200 to the acre. Thinning then continues at 10-40 year intervals, depending on how the site is being managed. In general the longer intervals, which involve fewer entries of heavy machinery into the woods, are better for the forest. Most commercial forestland in the East is young, and logging to re-create a mixed, old-growth forest would have to focus on the trees to be left rather than on the trees removed (the usual focus of loggers). The larger trees in a stand may be the same age as the smaller ones but more vigorous. So one leaves those, whatever their place in the succession. Early and mid-successional species that have been overtopped and suppressed are removed. In general, one leaves vigorous, late successional species with good form and no wounds, that are capable of reaching the overstory. The trees that grow earlier in a succession, such as white birch and red maple and the shrubs, put their growth into stems, flowers and fruits rather than roots; their purpose is to reproduce before they are shaded out. A heavy fruiting may kill them. More long-lived species, such as sugar maples and oaks, put their energy into roots, stems, branches and leaves, and fruit later and at longer intervals. Logging aims at an uneven-aged forest. Several old trees are left per acre, no matter what their status or form: they will become snags, hollow trees, and finally coarse woody debris on the forest floor. Creating such forests takes time, in most cases a century before they become capable of producing a steady flow of timber. In an ideal world, returns from the value of the water flowing off the land and from carbon storage would provide additional income over this time. Logging of old growth in a given woodland would not happen all at one time, since one aims to provide roughly the same proportions of each habitat over time: this lets the various inhabitants of the forest (as well as the loggers and lumber mills) survive.

In an ideal eastern forest, one would cut only those trees unlikely to survive until the next cutting cycle (a variation on John Muir’s sawmill for dead trees). This will result in a forest of shade-tolerant trees: a “climax forest.” Cutting mature trees is likely to damage the surrounding ones however. An alternative is cutting the forest in little clear-cuts, 0.25 acre to 2 acres in size (2 acres is large), somewhat linearly, along the slope. This allows for reproduction of more warmth-loving and intolerant species (oaks, birches, pines). It tends to reproduce the existing forest. About 15% of the forest is left standing in the cut: some green trees for shade and seed, some large trees to become overstory trees, some to become snags and logs on the forest floor. Tops less than 6 inches in diameter and branches are left in the woods. The microhabitat the standing trees and downed logs provides reduces erosion and loss of soil nutrients, the partly shaded habitat is also more friendly for soil microbes and invertebrates, mycorrhizal fungi, small vertebrates and amphibians. The slash helps prevent browsing of saplings by deer. A different mix of trees will grow into the architecture of the remaining forest than into a clearcut. (In former Forest Service guidelines for western forests, stumps were removed and the slash piled and burned, in preparation for replanting ranks of fir or spruce. Except where it is very dry, most forests will reseed naturally, but not to the monocultures of plantations.). As the young trees grow, they are thinned. Timely thinning speeds tree growth but also removes carbonaceous material and nutrients that would add to forest soils (leaving tops and branches from the thinned material helps). The point of thinning in the developing forest should be to create a mixed stand of trees (mixed in age and species), not to harvest the largest and most valuable timber (generally the goal in commercial thinning). The first thinning is done when the trees reach 8 to 12 inches in diameter. This thinning produces poles, sawlogs and firewood. Trees too small to use are left in the woods. One selects for trees suitable for the site and opens up the forest around the better formed, better adapted (to the site), more vigorous trees. Usually commercial selective cuts involve the removal of 50% or more of basal area and all large trees, but cuts in the ecological forest would remove considerably less than that, and leave a cross section of tree ages. Since selective cutting will tend to favor the more shade tolerant trees (such as hemlock and sugar maple in the East), one would also in good seed years cut some larger gaps to allow regeneration of less shade-tolerant species (yellow birch, oaks, pines). Cutting over the long term inevitably reduces the nutrient capacity of a site, but the term is long (probably longer than the climate that supports the forest) and restoration forestry can reduce this loss to near zero. Ecological forests are managed for their processes, but after a time also produce a steady flow of forest products (poles, sawtimber, veneer logs, fuelwood, chips, pulp). A point to be made is that in the modern world the healthiness of a forest may constitute its greatest value (its value as real estate).

There are many different types of forest in the United States. In the latter half of the nineteenth century, logging and fire converted much of the pinery of the upper Middle West into an aspen forest (aspens sprout vigorously from their roots); or in places where the topsoil was burned away by the fires, to shrublands of bird-planted pin cherry, blueberry, grasses, raspberry and shadbush. From 1875 to 1900, logging fires averaged 500,000 acres a year in Michigan and Wisconsin. Much of this area had little value for agriculture and ended up as national forest. (The pinery of the northern forest was originally several hundred billion board feet.) The aspen forest has been maintained as a monoculture by cutting on a short-term rotation (30 to 50 years) for paper pulp and for chips for composition board. Clearcutting the sites removes the conifers and other hardwoods in the stand, and regenerates the aspen from root sprouts. To recreate a more mixed forest, probably a more stable forest (and one with more commercial potential), loggers would remove the mature aspen but retain the conifers and the other hardwoods. Released from competition, these trees would shoot up. The aspen would regenerate, but less strongly, as they must compete with the residual trees. The result would be structural and habitat diversity. Future cuts would treat the forest as a developing old growth woodland, clear-cutting the aspen stands as they matured (their greatest value for grouse and deer tends to be when young), and the evergreens and hardwoods as they reached the first stages of old growth. Along watercourses trees would be let mature further. Some of the canopy would be left standing in the future clearcuts as a legacy of the stand architecture: seed trees, snags, hollow trees, some large old trees, some young and middle-aged trees (living trees might be left in clumps). The clear-cuts would be small and patchy. The alternating dominance of aspen and conifers would mimic the lifecycles imposed on the forest by tent caterpillar and spruce budworm, the insect pests of each species. It would also exploit the opposite effects of aspen and conifers on the nitrogen cycle. Much of the nitrogen that accumulates in conifer needle litter does not break down and is vaporized in stand-clearing fires — especially in boreal woodlands (no stand of black spruce hs been found that did not originate in a fire) — while nitrogen in aspen litterfall is exploited in the rich, easily decomposed leaf litter of the aspens.

Some eastern forests burn naturally. Many more were regularly burned by the Native Americans. Burning altered the forest cover along much of the East Coast, its effects extending up the major river valleys. In the Southeast, burning created the savannahs of long-leaf pine characteristic of the pre-contact coastal plain. (Perhaps 3% of this once enormous habitat remains.) In the Middle West late summer burning produced open forests of oaks and hickories and pushed the edge of the continental grasslands hundreds of miles east. The sunny dry forests of the mountain West naturally burned. Some were burned by Indians. At European contact, probably 20 million acres of the trans-Mississippi West burned every year, much of it grassland and scrub, but 6 million acres of it forest. Virtually all western forests arose out of fire: the cool coastal rainforests of Vancouver Island and British Columbia were perhaps an exception. The types of fires varied. Under natural conditions about 40% of western forests experience low-intensity fires every 1-30 years. These are so-called understory fires. Low intensity or understory fires create forests of large, well-spaced, sun-loving, fire-resistant trees, such as Ponderosa and Jeffrey pine and western larch in the Rockies; giant sequoia, redwood, and some types of oak forests further west. Ponderosa pine forms a relatively pure climax at middle elevations in the western United States. Ponderosa is moisture dependent and grows more thickly at higher elevations (more mixed with Douglas fir), more spaced out at lower and drier ones. The trees in historic old-growth middle elevation forests were 200-400 years old, with 30-40 trees to the acre (one tree per 200 square meters). Grasses and fire resistant herbs and shrubs formed an open understory, which was browsed by elk and deer. Frequent ground fires in such forests scorch the lower branches of the trees (which eventually die and fall off), kill saplings and large shrubs, and thin (often drastically) the pine seedlings and those of shade-tolerant trees, such as true fir and Douglas fir. They burn off the litter layer of undecomposed needles, fallen branches and the dead grasses on the forest floor. Thus they keep fuel loads low and, by pruning trees and killing seedlings, remove the fuel ladder that lets fires reach the crowns of the trees. They maintain the open forest. When such forests are logged, most of the old trees are removed. Fires are suppressed. Shade tolerant Douglas fir, true fir and lodgepole pine seed in and grow up into thick stands of young trees, among the remaining older trees. The tree density is 40 times that of historic old growth stands. Competion for moisture and nutrients in the thick forests stresses the older trees. Without fire to keep competion down, the old larch and Pondersosa pine decline in vigor. Diameter growth slows drastically after 30 years and foliage grows sparse after 80 years. The trees loose the ability to manufacture the resins necessary to combat bark beetle infection (the beetles are drowned in their tunnels by the sticky resins secreted by the tree) and are killed by the beetles or fire. In general, crowded stands are more susceptible to bark beetles, spruce budworm, dwarf mistletoe, root rot, and catastophic crown fires. Because of such poor management in western forests, stand replacing fires are twice as common as historically. In overgrown forests with a history of frequent fire, where fire has been suppressed, insects and fire take control. The resulting catastrophic fires will kill both old growth and young trees and endanger human settlements in the forest.

Ponderosa pines are adapted to light fires. They have deep roots, thick bark, open crowns, large fleshy buds, and long needles spread out to avoid rising heat (not the short densely spaced needles of black spruce that make the tree flare up like a torch). Severe fire in Ponderosa forests makes them hard to regenerate. Ponderosas have heavy seeds that fall within 150 feet of the tree, so unless planted by birds or squirrels, the trees take time to spread; the soil after a hot fire forms a water-repellant surface inimical to vegetation; and the rest of the forest vegetation, also not adapted to severe fires, recovers slowly. One way to regenerate a fire-resistant stand of old-growth Ponderosa pines is to log from below. One removes most of the small and medium sized trees and all the shade tolerant trees (such as Douglas fir, some of which may be large: removing these large trees helps pay for the treatment). All the large pines that are vigorous enough to survive are left. After logging, when conditions allow, the forest is burned, to reduce the fuel load and further thin the Ponderosa and fir seedlings. One wants a mix of different age classes of trees amidst an open understory that will allow periodic prescribed burning. This forest over several decades will become an open forest dominated by shade-intolerant Ponderosa pines, with some young growth and some trees of other species. Such forests are logged every 25-30 years to remove excess small, medium and large trees (those that wouldn’t survive to the next cutting cycle). Some dying old trees are left. The small openings created by each entry allow a new age class of pines to develop. The forests are burned every 10-35 years to keep fuel loads low, recycle nutrients, control Douglas fir and stimulate herb and shrub regeneration. In time, such forests will provide a constant flow of commercial timber: large sawlogs, smaller logs and poles for latilla and viga makers (the traditional southwestern ceiling beams and lattices); chip wood for boiler fuel. The constant flow of wood provides predictable amounts of material for sawmills, pulp mills, wafer board and plywood plants, post and pole plants. Restoring overgrown forests in 1% of Montana a year (not just Ponderosa forests, but all types of forests) would increase timber productivity by 50%. (In 2000, thanks to a century of poorly managed timber cutting, wood production in Montana was 15% of that in 1990.) Such restoration forestry costs about $200 an acre initially. With the second or third cut, the forest breaks even, and after 100 years should turn a profit, which continues indefinitely, and rises with good management. (Fighting crown fires is also expensive, up to $1700 an acre, with no profit in sight.) The United States spends over $2 billion a year fighting fires in the West. This is money that in many cases pays for a history of forest mis-management, a tax on short-term capitalist management and scientific ignorance. Without putting money into good management (an amount equal to that used to fight forest fires would restore 10 million acres a year), such costs will only rise. One tries to recreate the historical forest not only to recreate the past but because this forest is more sustainable, with a lower risk to its biota and a lower risk of catastrophic fire. It makes economic sense; as several writers have pointed out, a forest’s maintenance of ecological process constitutes its greatest value (human or otherwise), its ability to maintain a predictable flow of wood and support a rural population constitutes its greatest social value, while control of its fires raises the value of the land about it as real estate.

The cooler, moister, higher elevation forests of the West are subject to more intense fires (so-called intermediate intensity fires), every 30-100 years. In such forests good fire conditions exist for a short time: a few days or weeks each summer. The fires are intense enough to kill most fire-susceptible trees but the mountainous areas burn unevenly, so much of a site (up to half) remains unburned. Such forests become very mixed in species and in tree ages. Here, logging followed by the suppression of fire leads to dense forests of shade tolerant trees which are susceptible to stand replacing fires. A more sustainable forestry in such overgrown stands, which, as they get older, are favored by spotted owls, martin and fisher, would remove the shade-tolerant trees in a patchy pattern, retain the sun-loving trees, and encourage (with a controlled burn) regeneration of fire-dependent herbs, trees, and shrubs. Such treatments would take place every 40-100 years. In the case of spotted owl habitat, a central (fire-unstable) old growth forest (say of firs) would be surrounded by more open forests of large trees (larch and pine), which would serve as firebreaks. Such firebreaks are usually a quarter of a mile wide. When fire destroys the central old growth, or after it is logged, parts of the firebreak are let develop into the thick, shade tolerant, old growth forest favored by owls. Surrounding undisturbed old growth forests with a matrix of semi-natural wooded habitat makes it easier for the animals of those forests move to other old growth areas.

Moist cold forests on the upper slopes of mountains (the last 20% of western forests) burn occasionally, at 100-400 year intervals, in stand replacing fires. In many ways, they resemble the boreal forests of Canada, altitude here compensating for latitude. Stand replacing fires occur when conditions are unusually dry and fuels are available, often during major droughts. Such forests include lodgepole pine; and the white-bark pine, whose plump seeds help support mountain populations of grizzly bears and which is replanted in areas cleared by wind and fire by Clark’s nutcrackers storing pine seeds for the winter. Such forests become mosaics of burned and unburned patches, of hundreds to thousands of acres. The trees are adapted to catastrophic fires. Lodgepole pine cones open in the heat of fire. Green cones in fire killed western larch and coastal Douglas fir drop viable seed. After a burn, elderberries, currants, gooseberries, bitterberries, ceanothus, wild geraniums and hollyhocks are planted in clearings by birds; aspen and fireweed seeds blow in; alders replenish the nitrogen in the soil. The forest is full of standing dead trees. Depending on the intensity of the fire, willow, aspen and mountain maple may sprout from rootstocks; and other plants grow from seeds in the soil seedbank. Clear-cutting such forests is not a replacement for fire. Clear-cutting leaves no standing dead trees to become breeding places for insects that attract birds; that release nutrients; and whose shade helps in regeneration of trees, herbs and shrubs. Fires affect soils differently from bulldozers and skidders. Fires leave irregular patches and strips of surviving trees and an ashy seedbed for the bird and wind planted seeds from outside the burn. Clearcut stands lack the diversity of fire cleared stands. A better way to log such stands is to remove two-thirds of the trees, leaving most of the overstory trees (such as Douglas fir and western larch) and the rest of the stand in irregular clusters. Burning the slash after logging kills most of the remaining trees and leaves snags and a suitable seedbed for regeneration of the conifers, herbs, shrubs and deciduous trees characteristic of recent burns.

The whitebark pine of the high slopes is suffering tremendous mortality from blister rust, an Asian fungus that took almost a century to reach whitebark stands from its initial foothold on the east coast (where it remains a problem in white pine), and may require intensive human intervention to reproduce the rust-resistant trees. The whitebark’s large, oil-rich seeds let grizzly bears gain enough weight to go into winter in good condition and keep the bears in autumn at high elevations out of the way of humans. The bears rob caches of cones hidden by squirrels, then crush the cones between their paws to get out the seeds. (Another late fall bear food is cutworms, whose moths are blown into the mountains from Kansas wheat fields, and whose larvae are found under mountain rocks.) A poor competitor with other trees, whitebark pines grow on the thin, stony soils and in the difficult climates of alpine elevations. They shade the snowpack and stabilize the rocky soils, protecting the quality of the snowmelt and rain that flows in brooks and streams down to the lowlands. Clark’s nutcrackers collect whitebark seeds and cache them several miles away in open areas for the winter. Those that are not retrieved (this number may fall as the pines grow more scarce, and feeding by nutcrackers become a problem for the pines) grow into new stands of trees.

Sustainability is a human concept that may or may not work in nature. Attempting permanance, it implies a constant flow of animals, crop plants or timber from a place. Sustainability in agriculture means maintaining agricultural soils and reducing erosion to the levels at which the soil is rebuilt (essentially zero); and making agriculture one part of the larger ecosystem. Sustainable forestry in the long run probably means maintaining all components of the ecosystem. Such components include disturbances, such as fires, large herbivores and their predators (moose and wolf), the small mammals, amphibians, arthropods and fungi of the forest floor, the soil microflora, the arboreal lichens and mosses of old forests. The purpose in retaining large canopy trees during initial logging operations, and lengthening the rotations, is to produce a forest with more large, old trees, a multi-storied canopy, a greater variation in tree size (these all lead to a more complex forest architecture), with large woody debris on the forest floor, tighter nutrient cycles, shaded refugia for mycorrhizal fungi and nitrogen-fixing bacteria. (One could argue that the fungi, with which many trees have obligate relations, by connecting trees of different ages and species, are a major factor in controlling what happens in the forest.) Such a forest provides for beneficial predator-prey relationships among forest vertebrates and habitat for plants and animals that require structural complexity and late seral conditions, as well as for the needs of those that inhabit young forest. Some birds in Pacific Northwest forests jump in density as the forest enters a state of old-growth, indicating a (non-linear) change in the quality of the habitat. (But bird variety in mature forests is often low.) In woodlands with a history of fire, maintaining open forests of old sun-loving trees, mixed with some shade-tolerant trees and with younger stands, minimizes the destructive potential of fires. (One can’t eliminate fire, only plan around it; the success of Smokey the Bear has caused our present dilemma.) Such forests also maintain the place of the forest globally, which a young forest, maintained on a short rotation, doesn’t: the forest as a store of carbon, as an inoculae for mycorrhizal fungi, lichens and forest bacteria, as a positive influence on the water that flows through them, and on the coastal estuaries that lie below.

Forests managed for their process as well as for timber will produce less total wood fiber. Parts of them should be left undisturbed and separate forestlands should be connected by corridors and buffer zones, that allow plants and animals to migrate. Leaving large trees standing and lengthening rotations to let the forest become old both reduce the amount of wood harvested per acre. The net return per acre however remains the same or rises, since larger logs from old trees have a much higher value. (About half the wood cut in the U.S. goes to paper, though sawtimber is worth 2-5 times as much, good sawtimber more than that.) So the contribution of the forest to the economy remains the same, or rises; but only after a period of forest recovery that would last for about a century. Such forests will require a new infrastructure to use their logs of different species and sizes (saplings, poles, sawlogs). Some thinning operations would be better done by hand, or with horses or light tractors, than by heavy machinery. Modern industrial forestry, faced with forests of declining value, uses harvesting machinery of immense power and mobility in order to make a profit harvesting wood of low quality and low quantity (per acre). The post-harvest machinery tends to homogenize the harvested logs into composite products (like oriented strand board, which has replaced plywood, which requires better logs, for many uses: in this case, resin replaces fiber). Such plants are now being constructed around the perimeter of the Cumberland Plateau in the southern Applalachians, in order to use the second growth hardwood forests of the plateau. The forests will be clearcut to feed the plants, at terrible cost to the biota and watercourses of the region. An alternative, regeneratively harvested forest would produce a mix of graded materials for plywood, oriented strand board, flooring, cabinetry, moldings, wall panelling, furniture, construction lumber, tool handles, fencing and so on. Such a forest would require a more far-sighted and curious forestry, some public funding, and patience.

One faces essentially the same question as in agriculture: whether to focus on the yield per acre per year, or on the forest as habitat and process, a landscape that also produces wood. As with food, one can ask how much wood do we need? Wooden pallets use 1.5 billion board feet of lumber a year in the United States (about 40% of the hardwood cut) and most of them are thrown away after a single use. Re-manufactured pallets could provide the lumber for 300,000 to 600,000 houses a year; or a considerable amount of hardwood flooring, material for composition board, and wood shavings for packing material. A tax on logs (making pallets more expensive) would encourage reuse. A rise in landfill fees would make reuse of lumber in demolished buildings profitable. (Construction debris constitutes 15-40% of landfill contents in developed countries; an increase in Danish landfill fees pushed reuse of construction materials from 12% to 82% over 10 years.) A better recycling effort would reduce the wood needed for paper and cardboard by up to 90%. (Paper can be recycled 9-10 times; 90% recovery probably isn’t possible but 75% probably is. Reducing wood use by recycling would also reduce our carbon footprint.) Alternative fiber crops such as kenaf (an African plant), hemp, hybrid poplar, native trees that sprout from stumps, grown and harvested as saplings by machinery, produce much more fiber per acre than natural forests (2-5 times more in mechanically coppiced forests), and are easier on the land than cotton or corn. Hemp, which can be made into paper or cloth, produces more fiber per acre at a much lower environmental cost than cotton. So does bamboo. In short, we don’t need much of the wood that modern forests produce; but under a capitalist regime land must produce value. Wood fiber (not whole ecosystems) is what forests produce, and it is presently economically advantageous to use the fiber and throw it away. Returning some of the value of regeneratively managed forests to the landowner (their value in slowing soil erosion, protecting streams, controlling water and nutrient runoff, filtering the air, storing carbon, their role in providing habitat for useful wildlife, in regulating local microclimate and rainfall) would help make managing forests for their processes profitable. Such value, in dollars per acre per year, may be worth more than the wood, but that value only becomes apparent when the forests are degraded or gone. (The monies could be raised through a tax on logs.) The appearance of regenerative forests also increases property values. Costa Rica pays its farmers to maintain and replant their forests for all these reasons. In a capitalist world my notions are likely forest dreams; enforceable on public land if the government and the people desire.

The usual fate of forestland in the modern world is to become building lots. Inland from the sea along the Gulf of Mexico in the southern United States were so-called wet pine savannahs. These were grassy meadows with scattered pines growing on waterlogged clay soils, amidst tupelo-cypress swamps. The wet savannahs formed a narrow band between the drier piney uplands and the brackish marshes of the Gulf. The savannahs and the uplands were partly maintained by fire. The savannah soils were naturally acid and nutrient-poor and filtered the water flowing into the coastal marshes. They were the preferred habitat of the Mississippi sandhill crane, now much reduced in numbers. These wetlands, like many southern coastal wetlands, were ditched and drained by timber companies after the Second World War to grow slash pine on short rotations for paper pulp. Fire on the plantations was supressed. Drainage removed the filtering effect of the savannahs, increasing the nutrient flow to coastal waters. After a few rotations, the timberland was sold off for building lots, as air conditioning made life in the Deep South more comfortable, and a rising standard of living made life there more available.

There are other stories. The Menominee Indian Reservation in northeastern Wisconsin includes 220,000 acres of forestland. The forest has been commercially harvested for 140 years, with approximately 2 billion feet of lumber removed from the forest over that time. The volume of standing timber now is greater than when the reservation was established in 1854. The Menominee management program predates current concepts of ecosystem management. The Menominee must log (to support the tribe) and try to manage the forest so as to maintain a more or less even flow of sawtimber and pulpwood. They try to maximize the quality as well as the quantity of sawtimber, with a sustained yield of material, and a diverse mix of native tree species. Harvests are based on excess stocking of overstocked stands. When the stocking rate becomes too great for the site, trees are cut. Trees are let grow as long as they remain healthy and vigorous. Thus the forest contains many large old trees of great value (habitat for the insects and lichens of old-growth forest, great producers of mast). The Menominee cut trees no faster than the forest can regrow, rather than in response to market conditions (say, a rise in the price of red oak logs). Loggers must attend a class in how to cut and skid timber and contracts are terminated if the cutting methods are not satisfactory. About 65% of the forest is managed as a mixed-age forest, with a 15 year cutting cycle. The remainder is managed as even-aged forest. To a casual observer the forest looks pristine. Partly because of the old trees, it contains a much more diverse mix of plants and animals than the surrounding commercial timberland. The Menominee forest has healthy streams. Whether such forests are net accumulators of carbon isn’t certain: the above-ground part of the forest has grown in mass and the soils have probably continued to increase their stored carbon, but 2 billion board feet of wood (about 170 million cubic feet or, counting waste, branches and tops, about double that in woody material) have been removed, and much of its carbon returned to the atmosphere.

In the Acadian Forest Region of the Canadian Maritimes (a division of the Northern Hardwood Region), some writers claim a well stocked 100 acre woodlot will provide full time employment for 4 people, if the logs are sawn into lumber and dried and sold from the lot. The sawdust and slabs, if not burned for heat, are used on the wood roads. (Using these materials on roads slows their conversion to carbon dioxide. Converting them to biochar in a small furnace at 180º C. with a citric acid catalyst makes a soil amendment—biochar charcoal—that greatly increases the productivity of agricultural soils and keeps the carbon out of the atmosphere for several thousand years.) Four employees may be an overestimate but is comparable to those employed in Swiss woodlands. Much depends on the productivity of the site and on the price of lumber. I would think one person per 100-200 acres is a more reasonable estimate for much of the Northern Hardwood Region. This however is much less than the acreage required to keep a modern woodcutter busy (800 acres per job in Canadian industrial forestry). Such woodlands, like the Menominee forest, are managed for their long-term productivity. Sawing the lumber on site lets work in the woods be done when conditions (weather, the state of the ground) are appropriate. Logging might only occupy three months a year. The rest of the year would be spent sawing lumber, working up firewood, thinning, working on roads, perhaps making maple syrup, or renting simple summer cabins. The investment per person in such enterprises is much lower than in modern wood production. Jobs per log or board foot are higher. (Not usually a good thing in the modern world, but it is costs per board foot that matter.) One theory of civilisations’ collapse proposes that as the complexity of a society rises, its costs per unit of investment also rise, and the marginal returns on investment decrease. That is, investment produces declining (or negative) returns. Finally, this decrease in returns (from private investment in manufacturing and agriculture, from public investment in education, a military establishment, public safety and public works) bankrupts the society, which can no longer meet its obligations. Improving the yield of woodland, or of the agricultural landscape, so as to provide a constant flow of high value material provides one way out this impasse.