Summary and Synthesis: Past and Future Changes in the Alaskan Boreal Forest

Historically the boreal forest has experienced major changes, and it remains a highly dynamic biome today. During cold phases of Quaternary climate cycles, forests were virtually absent from Alaska, and since the postglacial re-establishment of forests ca 13,000 years ago, there have been periods of both relative stability and rapid change (Chapter 5). Today, the Alaskan boreal forest appears to be on the brink of further significant change in composition and function triggered by recent changes that include climatic warming (Chapter 4). In this chapter, we summarize the major conclusions from earlier chapters as a basis for anticipating future trends. Alaska warmed rapidly at the end of the last glacial period, ca 15,000–13,000 years ago. Broadly speaking, climate was warmest and driest in the late glacial and early Holocene; subsequently, moisture increased, and the climate gradually cooled. These changes were associated with shifts in vegetation dominance from deciduous woodland and shrubland to white spruce and then to black spruce. The establishment of stands of fire-prone black spruce over large areas of the boreal forest 5000–6000 years ago is linked to an apparent increase in fire frequency, despite the climatic trend to cooler and moister conditions. This suggests that long-term features of the Holocene fire regime are more strongly driven by vegetation characteristics than directly by climate (Chapter 5). White spruce forests show decreased growth in response to recent warming, because warming-induced drought stress is more limiting to growth than is temperature per se (Chapters 5, 11). If these environmental controls persist, projections suggest that continued climate warming will lead to zero net annual growth and perhaps the movement of white spruce to cooler upland forest sites before the end of the twenty-first century. At the southern limit of the Alaskan boreal forest, spruce bark beetle outbreaks have decimated extensive areas of spruce forest, because warmer temperatures have reduced tree resistance to bark beetles and shortened the life cycle of the beetle from two years to one, shifting the tree-beetle interaction in favor of the insect (Chapter 9).

Timber Harvest in Interior Alaska

The most active period of timber harvesting in the history of Alaska’s interior occurred nearly a century ago (Roessler 1997). The beginning of this era was the year 1869, when steam-powered, stern-wheeled riverboats first operated on the Yukon River (Robe 1943). Gold was discovered in Alaska in the 40-Mile River area in 1886, a find that was overshadowed 10 years later by the discovery of gold in the Klondike, Yukon Territory. By 1898, Dawson City, Yukon Territory, was reported to have 12 sawmills producing a total of 12 million board feet of lumber annually (Naske and Slotnick 1987). Over the next 50 years, more than 250 different sternwheeled riverboats operated in the Yukon drainage, covering a large part of Alaska and Canada’s Yukon Territory (Cohen 1982). This transportation system required large amounts of fuel. Woodcutters contracted with riverboat owners to provide stacked cordwood at the river’s edge, at a cost of $7.14 in 1901 (Fig. 18.1; Cohen 1982). Between 100 and 150 cords of wood were required to make the 1400-km round trip from the upper Yukon to Dawson City (Trimmer 1898). Over time, woodcutters moved inland from the rivers’ edges, significantly impacting the forest along many rivers of the Yukon drainage (Roessler 1997). The growth of the town of Fairbanks required wood for buildings and flumes as well as for fuel. In Fairbanks’s early days, all electrical generation was by wood fuel at the N.C. Company’s power plant. From the founding of the town in 1903 through the 1970s, white spruce harvested in the Fairbanks area was used exclusively by local sawmills, which produced small amounts of green and air-dried lumber. In 1984, however, the Alaska Primary Manufacturing Law was struck down by the U.S. Supreme Court, removing the legal barrier to round-log export of timber harvested from State lands. During the late 1980s and 1990s, many high-quality logs from State and private land timber sales were exported, primarily to Pacific Rim countries. Declining markets ended this trend in the late 1990s, and there have been no significant exports since the market collapse.

Mammalian Herbivory, Ecosystem Engineering, and Ecological Cascades in Alaskan Boreal Forests

The mammalian herbivores of the taiga forests include members of the largest (moose) and smallest (microtines) vertebrates that inhabit North American terrestrial biomes. Their abundance in a particular area fluctuates dramatically due to seasonal use of particular habitats (moose) and external factors that influence demographic processes (microtines). The low visibility of herbivores to the casual observer might suggest that these animals have minimal influence on the structure and the function of boreal forests. On the contrary, seedling herbivory by voles, leaf stripping by moose, or wholesale logging of mature trees by beaver can profoundly change forest structure and functioning. These plant-herbivore interactions have cascading effects on the physical, chemical, and biological components of the boreal ecosystem that shape the magnitude and direction of many physicochemical and biological processes. These processes, in turn, control the vertical and horizontal interactions of the biological community at large. Herbivores act as ecosystem engineers (Jones et al. 1994) in that they reshape the physical characteristics of the habitat, modify the resource array and population ecology of sympatric species, and influence the flux of energy and nutrients through soils and vegetation. Additionally, many herbivores are central to a variety of human activities. Both consumptive and nonconsumptive use of wildlife represents a pervasive aspect of life in the North. In this chapter, we examine the interactions of mammalian herbivores with their environment, with an emphasis on moose, and attempt to delineate the biotic and abiotic conditions under which herbivores influence the phenotypic expression of vegetation. We also examine the role of herbivores, and of wildlife in general, in the context of human perceptions and interactions with their environment. Human-environment interactions are both direct and indirect and pertain to a variety of social expressions. The relationship between humans and wildlife has economic, cultural, and psychological dimensions, which underscore the importance of these animals in a broader social, as well as ecological, context. Northern ecosystems such as the boreal forest are characterized by extreme seasonality and pronounced change in resource availability between summer and winter. Not surprisingly, these conditions are reflected in the population dynamics of the animals that inhabit these environments, particularly in smaller-bodied herbivores.

Dynamics of Phytophagous Insects and Their Pathogens in Alaskan Boreal Forests

Boreal forests support an array of insects, including phytophagous (plant-eating) insects, saprophagous (detritus-eating) insects, and their associated parasites, predators, and symbionts. The phytophagous species include folivorous leaf chewers and miners, phloeophagous cambial and sapwood borers, stem gallers, and root feeders. Biological diversity and distribution of insect species exhibit predictable patterns among vegetation types (Werner 1994a). In this chapter, we discuss how phytophagous species of insects differ among plant communities and how various populations of insects react to disturbances that alter forest stand composition and density. The distribution of insects differs among plant communities depending on the ecosystem type and plant height (Table 9.1; Werner 1983, 1994a). Grasses, mosses, small tree seedlings, and other herbaceous plants located on the forest floor have the highest arthropod densities. Shrubs have the lowest densities, and trees are intermediate. The herbaceous layer is inhabited primarily by scavengers, predators, and saprophages but has few defoliators (Werner 1983). Taller shrubs contain more species of phytophagous insects than do herbs, but trees have the most species of phytophagous insects, parasites, and predators (Werner 1981, 1983). Few saprophages and scavengers (carabid beetles), however, occur on shrubs and trees (Werner 1986a). Associations of plants and phytophagous insects in boreal ecosystems are similar to temperate assemblages in that insect species differ in the range of food plants that they utilize (Bernays and Minkenberg 1997, Futuyma et al. 1993, Thorsteinson 1960). Because of low plant diversity, however, many boreal phytophagous insects feed on several species of plants (Werner 1981). For example, the spear-marked black moth, Rheumaptera hastata (L.), feeds primarily on paper birch, but during periods of high populations it also feeds on alder, willow, and rose species but not on poplar (Werner 1977, 1979). When population outbreaks of phytophagous insects deplete their preferred host plants, less desirable species are sometimes consumed or starvation occurs (Werner 1981, 1986a). The biomass of phytophagous insects is greater on broad-leafed than on conifer trees (Werner 1983). Species of Coleoptera, Hemiptera, Homoptera, Hymenoptera, and Lepidoptera are common on broad-leafed trees, whereas only a few taxa of Homoptera, Hymenopera, and Lepidoptera are associated with conifers such as spruce or larch (Table 9.2; Werner 1983, 1994a).

Fire Trends in the Alaskan Boreal Forest

Fire is ubiquitous throughout the global boreal forest (Wein 1983, Payette 1992, Goldammer and Furyaev 1996, Kasischke and Stocks 2000). The inter- and intra-annual patterns of fire in this biome depend on several interrelated factors, including the quantity and quality of fuel, fuel moisture, and sources of ignition. Fire cycles in different boreal forest types vary between 25 and >200 years (Heinselman 1981, Yarie 1981, Payette 1992, Conard and Ivanova 1998). Although the increased presence of humans in some regions of boreal forest has undoubtedly changed the fire regime (DeWilde 2003), natural fire is still a dominant factor in ecosystem processes throughout this biome. Boreal forest fires are similar to those of other forests in that they vary between surface and crown fires, depending on forest type and climatic factors. Surface fires kill and consume most of the understory vegetation, as well as portions of the litter or duff lying on the forest floor, resulting in varying degrees of mortality of canopy and subcanopy trees. Crown fires consume large amounts of the smaller plant parts (or fuels) present as leaves, needles, twigs, and small branches and kill all trees. These fires are important in initiating secondary succession (Lutz 1956, Heinselman 1981, Van Cleve and Viereck 1981, Van Cleve et al. 1986, Viereck 1983, Viereck et al. 1986). Unlike fires in other forest types, smoldering ground fires in the boreal forest can combust a significant fraction of the deep organic (fibric and humic) soils in forests overlying permafrost (Dyrness and Norum 1983, Landhauesser and Wein 1993, Kasischke et al. 2000a, Miyanishi and Johnson 2003). During periods of drought, when water tables are low, or prior to spring thaw, organic soils in peatlands can become dry enough to burn, as well (Zoltai et al. 1998, Turetsky and Wieder 2001, Turetsky et al. 2002).

Watershed Hydrology and Chemistry in the Alaskan Boreal Forest: The Central Role of Permafrost

Hydrological processes exert strong control over biological and climatic processes in every ecosystem. They are particularly important in the boreal zone, where the average annual temperatures of the air and soil are relatively near the phase-change temperature of water (Chapter 4). Boreal hydrology is strongly controlled by processes related to freezing and thawing, particularly the presence or absence of permafrost. Flow in watersheds underlain by extensive permafrost is limited to the near-surface active layer and to small springs that connect the surface with the subpermafrost groundwater. Ice-rich permafrost, near the soil surface, impedes infiltration, resulting in soils that vary in moisture content from wet to saturated. Interior Alaska has a continental climate with relatively low precipitation (Chapter 4). Soils are typically aeolian or alluvial (Chapter 3). Consequently, in the absence of permafrost, infiltration is relatively high, yielding dry surface soils. In this way, discontinuous permafrost distribution magnifies the differences in soil moisture that might normally occur along topographic gradients. Hydrological processes in the boreal forest are unique due to highly organic soils with a porous organic mat on the surface, short thaw season, and warm summer and cold winter temperatures. The surface organic layer tends to be much thicker on north-facing slopes and in valley bottoms than on south-facing slopes and ridges, reflecting primarily the distribution of permafrost. Soils are cooler and wetter above permafrost, which retards decomposition, resulting in organic matter accumulation (Chapter 15). The markedly different material properties of the soil layers also influence hydrology. The highly porous near-surface soils allow rapid infiltration and, on hillsides, downslope drainage. The organic layer also has a relatively low thermal conductivity, resulting in slow thaw below thick organic layers. The thick organic layer limits the depth of thaw each summer to about 50–100 cm above permafrost (i.e., the active layer). As the active layer thaws, the hydraulic properties change. For example, the moisture-holding capacity increases, and additional subsurface layers become available for lateral flow. The mosaic of Alaskan vegetation depends not only on disturbance history (Chapter 7) but also on hydrology (Chapter 6).

The Role of Fine Roots in the Functioning of Alaskan Boreal Forests

The patterns of production described in Chapter 11 tell only half of the story about boreal forest production because a large proportion of the carbon (C) acquired by plants is allocated belowground in ways that have traditionally been extremely difficult to quantify. Work in the Bonanza Creek LTER provides considerable insight into the patterns, causes, and consequences of this belowground C allocation. Belowground allocation has a number of important ecosystem consequences beyond the simple fact that C allocated belowground comes at the expense of aboveground growth. Belowground and aboveground tissues differ substantially in the rates of C and nitrogen (N) incorporation into new tissue, the ratio of growth to respiration, and the rate of tissue decay. For example, despite the small biomass of fine roots relative to aboveground tissues in forest ecosystems, disproportionate amounts of C and N cycle annually through fine roots, which grow, die, and decompose very rapidly and have high N concentrations (Hendrick and Pregitzer 1992, Ruess et al. 1996, 2003). The objectives of this chapter are to (1) summarize our understanding of the structure and function of fine-root systems in forest types within the Bonanza Creek Experimental Forest, (2) compare our findings with the results of studies of other boreal and temperate ecosystems in order to develop a broader understanding of fine-root function, and (3) identify critical research gaps in our understanding of the role of fine-root systems in boreal ecosystem function. Fine roots grow more rapidly than the rest of the root system in a forest and are responsible for the bulk of nutrient and water acquisition. Until recently, fine roots were defined rather arbitrarily as roots less than 1–2 mm in diameter, while roots larger than this were considered coarse roots. Only one data set for fine and coarse root biomass has been published for interior Alaskan forests (Ruess et al. 1996), which shows (1) live fine-root biomass ranging from 221 g m-2 in floodplain white spruce stands to 832 g m-2 in upland birch-aspen stands, (2) a positive correlation between fine-root and coarse-root biomass, with coarse-root biomass averaging 50% greater than fine roots, and (3) no relationship between aboveground biomass and fine or coarse root biomass.

Mammalian Herbivore Population Dynamics in the Alaskan Boreal Forest

The population dynamics of boreal mammals differ strikingly from those of mammals in temperate and tropical ecosystems in their extraordinary fluctuations in abundance (Elton 1924). These fluctuations lead to strong top-down direct effects in which herbivores reduce the biomass of their preferred foods, such as birch and willow, and predators reduce the biomass of herbivores (Chapter 13; Sinclair et al. 2000). These effects are clearly demonstrated in experiments that exclude herbivores or their predators. Some authors have argued that bottom-up influences of food supply on herbivores are negligible because food augmentation to herbivores in the presence of predators had no detectable effect in reducing herbivore decline (Sinclair et al. 2001). Several members of the mammalian herbivore guild are also important as a human subsistence resource. Dynamics of moose (Alces alces) and snowshoe hare (Lepus americanus) can be altered by human harvest. Overexploitation by humans may reduce moose populations to densities where they can be predator-limited—the so-called predator pit (Messier 1994). In this chapter, we present information on dynamics of some mammalian herbivores in the Alaskan boreal forest and potential drivers that are responsible for these dynamics. We omit discussions of the dynamics of porcupines (Keith and Cary 1991), red squirrels (Boonstra et al. 2001a), and beavers (Donkor and Fryxell 1999), as studies of these species have not been conducted in Alaska’s boreal forests. Moose are thought to have arrived in Alaska during the Illinoian glaciation, about 400,000 yr B.P. (Pewe and Hopkins 1967). They may have retreated to refugia in central Alaska during subsequent glacial advances (Peterson 1955) and expanded at times when climate was warmer. Moose populations in North America have more than doubled over the past 30–40 years, to approximately 890,000 animals (Kelsall 1987). The Koyukuk River drainage in the northern interior, for example, is presently known for its large moose populations. However, the oral tradition of moose hunting in the Koyukuk is relatively recent. Native elders recall that, in their youth, moose were extremely rare and that moose did not figure prominently in the local subsistence economy until the 1930s.

Controls over Forest Production in Interior Alaska

State factors provide a powerful conceptual basis for understanding current patterns and potential changes in forest productivity (Chapter 1). The BNZ-LTER program has focused on investigations of ecosystem structure and function related to state factors (time, topography, and climate) that account for dramatic spatial variation in productivity and provide a basis for predicting future temporal variation (e.g., climate change). Other state factors are either relatively uniform across the region (e.g., potential biota) or co-vary with topography (i.e., parent material) and are difficult to study as clearly independent factors. State factors have many direct and indirect effects on productivity, and these controls may co-vary in a complex fashion across the landscape. The nitrogen productivity concept provides a mechanistic framework for understanding the effects of environmental variation on forest productivity (Ågren 1985). The nitrogen productivity of a tree or forest stand is defined as the amount of production per unit of foliar nitrogen (gram biomass production per gram foliar nitrogen) in the canopy of the tree or stand. At steady-state nutrition, the growth rate is proportional to the amount of foliar nitrogen and the N-productivity. Biological and chemical processes that occur in soils are an excellent example of the way in which multiple interacting factors influence productivity through their effects on N supply. In interior Alaska, several state factors have a hierarchical influence on forest production. These factors are time (Chapter 7), parent material (Chapter 3), topography (Chapter 2), and macroclimate (Chapter 4). These factors have both direct and indirect effects, many of which vary over time and space. In this chapter we emphasize the influence of the four relatively direct state factors: parent material, topography, time, and climate and a critical “resource” (Chapter 1), soil, which represents the indirect interaction of multiple state factors. The parent material in lowland locations is primarily alluvium or loess over alluvium; thick silt, glacial deposits, or eolian sands are present in some areas (Chapter 3). Organic soils also occur on level surfaces that rarely or never flood. Both alluvial and organic soils usually contain a fine-grained mineral substratum. In addition, limited lowland areas in interior Alaska contain very thick loess deposits.

Successional Processes in the Alaskan Boreal Forest

Superimposed on the topographic and climatic gradients in vegetation described in Chapter 6 are mosaics of stands of different ages reflecting the interplay between disturbance and succession, that is, the ecosystem changes that follow disturbance. The nature of disturbance governs vegetation succession, and vegetation properties, in turn, influence disturbance regime. Both disturbance and succession are controlled by state factors and by stochastic variation in local conditions such as weather and the abundance of herbivores. Even in this relatively simple biome, the interactions among site, chance, and disturbance history result in a vast array of possible successional trajectories following a disturbance event, generating at least 30 forest types in interior Alaska (Viereck et al. 1992). Despite this broad range of possible dynamics, certain patterns recur more frequently than others (Drury 1956, Viereck 1970). In this chapter, we discuss selected successional pathways that commonly occur on river floodplains and on permafrost-free or permafrost-dominated upland sites in interior Alaska. River floodplains occupy only 17% of interior Alaska, but they account for 80% of the region’s commercial forests and therefore have attracted considerable attention from forest managers (Adams 1999). These forests provide an excellent example of primary succession, that is, the succession that occurs on surfaces that have not been previously vegetated. Although many successional pathways are possible on interior Alaska’s floodplains (Fig. 7.1; Drury 1956), the trajectory that actually occurs in a particular place is usually determined by the patterns of colonization during the first decades (Egler 1954). This, in turn, depends primarily on physical environment, flood events, and seed availability. For example, fine-textured sediments, which are common along the gradual grade of the Tanana River near Fairbanks (Chapter 3), retain more moisture than gravelly substrates and favor establishment of thinleaf alder (Alnus incana subsp. tenuifolia) following the initial colonization by willow (Salix). Alder is therefore a more important component of this successional sequence than along some other rivers. In this chapter, we focus on the alder-mediated pattern of floodplain succession, which has been the major focus of the LTER research program. The common and scientific names of species mentioned in this chapter are given in Table 6.1.