Phellinus Weirii and Other Native Root Pathogens As Determinants of Forest Structure and Process in Western North America1

P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

Annu. Rev. Phytopathol. 2000. 38:515–39

PHELLINUS WEIRII AND OTHER NATIVE ROOT PATHOGENS AS DETERMINANTS OF FOREST STRUCTURE AND PROCESS IN WESTERN NORTH AMERICA1

E.M. Hansen Department of Botany and Plant Pathology, Oregon State University, Corvallis, Oregon 97331; e-mail: [email protected] Ellen Michaels Goheen USDA Forest Service, SW Oregon Forest Insect and Disease Service Center, Central Point, Oregon 97502; e-mail: [email protected]

Key Words Phellinus weirii, laminated root rot, Douglas-fir, forest ecology, forest succession ■ Abstract The population structure and ecological roles of the indigenous pathogen Phellinus weirii, cause of laminated root rot in conifer forests of western North America, are examined. This pathogen kills trees in slowly expanding mortality centers, creating gaps in the forest canopy. It is widespread, locally abundant, and very long-lived. It is among the most important disturbance agents in the long intervals between stand-replacing events such as wildfire or harvest in these ecosystems and shapes the structure and composition of both wild and managed forests. Trees are infected and killed regardless of individual vigor. Management of public lands is changing dramatically, with renewed emphasis on natural forest structures and processes but pathogens, especially root rot fungi, remain a significant challenge to “ecosystem management.”

Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org CONTENTS

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. INTRODUCTION ...... 516 THE PRIMEVAL FOREST ...... 517 PHELLINUS WEIRII ...... 519 Pathology ...... 519 Population Structure ...... 521 Impacts on Forest Structure ...... 522 Impacts on Succession and Diversity ...... 524

1The US Government has the right to retain a nonexclusive, royalty-free license in and to any copyright covering this paper.

515 P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

516 HANSEN GOHEEN

Impacts on Nutrient Cycling ...... 526 Effects of Tree Vigor ...... 527 PATHOGENS AND OLD-GROWTH FORESTS ...... 529 THE FOREST TODAY AND TOMORROW ...... 530 Recent Changes in the Forest...... 530 Trends in Forest Management ...... 531 Ecosystem Management ...... 532 Modeling Root Disease ...... 534

INTRODUCTION

In agricultural ecosystems pathogens are considered “pests” interfering with the production of a healthy, valuable crop. Pathogens first evolved, however, free of human expectations in much more complex natural ecosystems, and the destruction and loss we ascribe to them today are just one interpretation of their successful evolutionary strategy. Tree pathogens are integral components of forest ecosystems around the world, altering forests in many ways, both subtle and profound. In forests managed for economic value, pathogens force changes in management practices, reduce profitability, or even threaten economic viability. Here, however, we take an ecological perspective on the interactions between plant pathogens and the forest communities they inhabit. We focus on the population structure and ecological roles of one indigenous root rot pathogen, Phellinus weirii (Murr.) Gilbertson, in Douglas-fir and mountain hemlock forests of the Pacific Northwest. This and similar root decay fungi play similar roles in the other forest types of western North America, but we draw our main examples from the forests we know best. Our views of forest pathogens are inevitably dominated by a few exotic pathogens in vulnerable forests and a strong professional legacy—the idea that a goal of forest management is regulated, disease- and decay-free forests on the nineteenth- century European model (34). Even today there is little appreciation for the significant effects that indigenous pathogens have on natural forests. Recent reviews do a poor job of distinguishing native from exotic pathogens, and wild from disturbed Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org ecosystems (4).

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. Pathogens affect forests most dramatically by killing trees. Plants, of course, are not defenseless against pathogens, and in ecosystems where plants and pathogens have evolved together, the evolutionary success of both is assured. Because evolution acts on populations, not individuals, however, some indigenous pathogens can and do kill single trees or even groups of trees in natural ecosystems without threatening the forest as a whole. They kill big trees and in the process change the diversity of the forest community, and they kill small and young and weak trees, maintaining the ﬁtness of the ecosystem. Many forest pathogens do not kill trees directly, but still affect the forest community by altering competitiveness and reproductive success of trees, and nutrient cycling and primary productivity of the P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

PHELLINUS WEIRII AND FOREST STRUCTURE 517

forest ecosystem. They change local population structure of individual species and landscape-scale patterns of forest succession. Some pathogens play critical roles in determining range limits and habitat occupancy. In some ecosystems, including the vast coniferous forests of western North America, pathogens are among the most important disturbance agents in the long intervals between stand-replacing events such as wildfire or harvest; the wild forests we know today look and function as they do because of pathogens. Pathogens have been invisible to or misunderstood by most forest ecologists, unless entire forests are destroyed by exotic diseases such as chestnut blight. There is a widespread presumption that indigenous pathogens kill trees only if they are physiologically stressed or weakened by other agents or in response to mismanagement of forest lands. Some pathogens fill these scavenger roles in forest ecosystems, but others kill vigorous dominant trees, altering the very structure and composition of the forest. In the coniferous forests of western North America, Phellinus weirii has this effect. Phellinus weirii kills trees in slowly expanding mortality centers, creating gaps in the forest canopy. Gap or patch dynamics is an important component in current studies of forest ecosystem dynamics (37), with gaps in the canopy created by “disturbance agents.” Pathogens differ from other disturbance agents such as lightning or hurricanes in fundamental ways, however, and pathogen-induced gaps have different consequences to forest communities than other types of gaps. Pathogens usually affect species differentially, that is they exhibit host specificity, and pathogens act slowly. A lightning strike or a tornado kills all the tree species in a discrete patch instantaneously, but Phellinus weirii, for example, kills Douglas-fir but not western hemlock in a patch that slowly increases in size throughout the life of the stand.

THE PRIMEVAL FOREST

Conifers dominate the temperate forest ecosystems of western North America. Forests extend nearly continuously about 1000 km across mountainous terrain from the eastern slopes of the Rocky Mountains to the Paciﬁc Ocean, broken in places by the Columbia Plateau, the Great Basin, and large and small river valleys. Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org Coniferous forests cover 82,000,000 ha in Oregon, Washington, Idaho, and British

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. Columbia, about 52% of this region. Two main north-south mountain ranges, the Coast Ranges and the Cascade Mountains, dissect the area and through their influence on maritime temperature and precipitation, geology, and soils, delimit several very distinctive ecosystems (11). Forests west of the Cascade Mountains in Oregon and Washington and the Coast Mountains in British Columbia (westside forests) are strongly influenced by the Pacific Ocean. Annual precipitation in the Douglas-fir forests ranges from 150 to more than 500 cm a year. The maritime climate, with mild, wet winters and dry summers, favors evergreens over deciduous tree species. Conifer forests of the Northwest are more productive and accumulate greater standing biomass P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

518 HANSEN GOHEEN

than other forests. Great size and age of dominant tree species are typical of western coastal forests and make them unique among the forests of the world (51). Western hemlock [Tsuga heterophylla (Raf.) Sarg.], a shade-tolerant tree, is the principal late-successional species in most westside ecosystems, but Douglas- fir [Pseudotsuga menziesii (Mirb.) Franco] dominates most forests. In the drier forests of southwestern Oregon and northern California, ponderosa pine (Pinus ponderosa Dougl. ex. Loud.), sugar pine (P. lambertiana Dougl.), and hardwood species such as Pacific madrone (Arbutus menziesii Pursh), California black oak (Quercus kelloggii Newb.), Oregon white oak (Q.garryana Dougl. ex Hook.), and tanoak [Lithocarpus densiflora (Hook. & Arn.) Rehd.] join Douglas-fir in the early successional forest, and Abies species (true firs) substitute for hemlock as shade- tolerant species. Along the crest of the Cascade Mountains, covering the broad divide between the volcanic peaks, is a mixed subalpine forest. Large areas are occupied by mountain hemlock [Tsuga mertensiana (Bong.) Carr.] at elevations around 1700 m. Sites are severe; summers are short and dry, snowpack is deep and persistent from November through June, and the moderately deep soils of volcanic tephra origin are young (most recent volcanic ash deposits are about 6600 years old) and nutrient-poor. Annual precipitation, mostly as snow, varies with local topography and ranges from about 160 to 280 cm. Infrequent stand-replacing wildfires lead to establishment of a seral lodgepole pine (Pinus contorta Dougl. ex. Loud.) forest that is gradually replaced by the long-lived mountain hemlock. Other conifers including western white pine (Pinus monticola Dougl. ex. D. Don), whitebark pine (P. albicaulis Engelm.), and Abies species are present, especially in the more diverse forest established in root rot mortality centers. East of the Cascade Mountains (eastside forests), more continental weather systems bring a shorter growing season, with cold winters and hotter summers. The forests are stratified by elevation grading from shrub steppe upward through ponderosa pine, Douglas-fir, and mixed conifer forests to subalpine communities and finally tundra above timberline in the Rocky Mountains. The range of forest types includes the warm, moist, productive cedar-western hemlock-white pine forests of the Interior Northwest (annual precipitation 56–170 cm), hot, dry ponderosa pine forests at lower elevations (annual precipitation 35–76 cm), cold spruce-fir- Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org whitebark pine forests at high elevations, and a diverse group of mixed conifer

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. forests that occur where the climate is moderate. Forests are naturally dynamic communities of organisms, with trees growing old, competing, dying and being killed, and being replaced by other trees. Under- story vegetation changes in response to changes in the overstory. In the Douglas-fir forests of the western slopes of the Cascade Mountains, stand-replacing wildfire or its modern surrogate, clearcut harvesting, controls the overall pattern of succession (40). The natural fire return interval averages between 250 and 400 years in these forests. Immediately following a stand-replacing fire, local plant species diversity is greatly diminished. Diversity increases rapidly in the early years as herbaceous plants and shrubs colonize the open ground. Plant diversity is usually P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

PHELLINUS WEIRII AND FOREST STRUCTURE 519

at its highest point in these early seral communities (29). Douglas-fir is the first conifer to colonize after wildfire and grows up through the early seral herbaceous and shrub vegetation, often forming dense, nearly even-aged stands. As the crowns of the trees expand and meet, light to the forest floor is gradually reduced, and the lower vegetation dies or retreats to gaps in the canopy. Diversity decreases and stays low for 100 years or more, depending on intermediate disturbances. The light-demanding Douglas-fir cannot normally reproduce in its own shade, but its potential life span of greater than 600 years is much longer than the usual fire return interval. Relatively shade-tolerant trees such as western hemlock can establish and survive under Douglas-fir if there is a seed source, but hemlock seedlings are confined to the understory until some disturbance agent (usually a pathogen) kills the Douglas-fir and allows enough light to stimulate height growth of the hemlock. As mortality of the Douglas-fir increases across the centuries, plant diversity slowly increases in the light gaps and western hemlock trees can grow into the lower canopy. However, even old-growth forests are usually dominated by the long-lived Douglas-fir except where some disturbance accelerates the succession. A forest is a community of organisms that is dominated by trees. Pathogens are integral members of forest communities. For example, the total biological diversity of the Douglas-fir forests of the western United States includes the 439 species of fungi listed on Douglas-fir in the Host-Pathogen Index (8). Most characteristic, and with the greatest impact on forest structure, are the Basidiomycete wood decay fungi, and most influential of these is the “Douglas-fir form” of Phellinus weirii.

PHELLINUS WEIRII

Phellinus weirii, the causal agent of laminated root rot in Douglas-fir and other conifers, is considered the most economically damaging pathogen in the valuable Douglas-fir forests west of the Cascade Mountains. Losses to root rot greatly exceed losses to fire and insects in this region. Perhaps it is not surprising that P.weirii also plays a significant role in the ecological functioning of these forests. Laminated root rot is found from the Klamath Mountains of northern California to near the northern limit of Douglas-fir in British Columbia, and east to Idaho (Figure 1) Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org (35, 43). It is a part of many forest communities, well adapted to the climatic and

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. disturbance regimes of western forests, and often occupies a signiﬁcant portion of the landscape.

Pathology Nearly all western conifers are recorded as at least occasional hosts for P. weirii, but the tree species differ greatly in susceptibility (9, 43). Douglas-ﬁr, mountain hemlock, and grand/white ﬁr (Abies grandis/A. concolor) are the most susceptible. Western hemlock, in contrast to mountain hemlock, is often infected but usually not killed whereas western redcedar (Thuja plicata Donn) is little affected by the P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

520 HANSEN GOHEEN

Figure 1 Range of the Douglas-ﬁr form of Phellinus weirii in the Paciﬁc Northwest (43) (shaded area).

Douglas-ﬁr form of the fungus. Pines are resistant and the fungus is not found in pure pine stands, although pine trees may be infected when growing mixed with Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org more susceptible conifers. Hardwoods are immune.

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. P. weirii spreads from tree to tree as ectotrophic mycelia. It penetrates the root through intact bark, killing phloem and cambium as it enters and initiates decay in the xylem (49). Both lignin and cellulose are utilized. Infected trees may live in a slowly declining state for many years, gaining necessary water and nutrients from adventitious roots formed near the root collar (1, 41). Trees eventually topple from weakened supporting roots, although they usually still have green crowns when they fall. The fungus continues to spread to adjacent trees across root contacts. The mortality front advances about 30 cm a year (33, 36), but the advance is usually uneven, reﬂecting the irregular distribution of trees in the forest and the sporadic nature of winds and bark beetles that topple or kill the root-rotted trees. Mortality P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

PHELLINUS WEIRII AND FOREST STRUCTURE 521

increases steadily in Douglas-ﬁr stands 30 to 150 years old (6) but spread is slower in older stands because trees are farther apart and it takes many decades for the large old trees to be killed by the fungus.

Population Structure Phellinus weirii is a complex of three distinct biological species. There are two reproductively isolated populations in North America: the Douglas-fir form that causes laminated root rot in Douglas-fir and other conifers; and the cedar form, cause of butt rot in western redcedar. It is widely accepted that these different diseases are caused by different fungi. There are also reports of P.weirii from Japan and Siberia. Molecular evidence and compatibility tests suggest, however, that Asian populations are reproductively isolated from North American populations although closely related to the Douglas-fir form (22). Ten Latin binomials have been applied to Phellinus weirii since its first description in 1907. Most recently, Larsen et al (27) concluded that Douglas-fir form and Asian forms are conspecific, and best called Inonotus sulphurascens. Phellinus or Inonotus heinrichii Pilat has priority over I. sulphurascens (22), however, and because generic concepts are also changing, a pragmatic decision seems to have been reached to continue to call the species complex Phellinus weirii until all the changes can be made at once. Phellinus weirii (all three populations) has a bipolar pattern of sexuality (22, 27) but spores are not an important part of the life cycle as we understand it. The population structure is clonal, with large, old, vegetatively spreading individuals or genets slowly moving through the forest (5), and surviving saprophytically between tree generations. Survival has been recorded up to 100 years in the remains of old- growth trees (49), although the mass of viable inoculum is greatly reduced after 50 years (20). Active growth resumes when roots of susceptible trees grow in contact with old infected roots. Laminated root rot appears in mature forests as scattered mortality centers, each occupied by a distinct genet of the fungus (6). Larger centers may be composites of two or more genets that have grown together, but the mycelial individuals maintain their separate identities through a somatic incompatibility system (19). Somatic Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org incompatibility, different from the sexual compatibility system, is regulated at one

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. or two multiallelic loci (23). Genets increase slowly in size throughout the life of the stand. Mortality centers expand at the margins and are progressively colonized by a variety of plants taking advantage of the light gap. In westside forests, Douglas-fir regeneration does not survive the competition in the centers, and most of the species that do survive are immune to or tolerant of infection by P. weirii. The fungus eventually dies from most of the old infected roots in the center while it continues to spread on the edges. After a stand-replacing disturbance such as wildfire, there is usually a delay of a few years to a decade or two before Douglas-fir reestablishes itself on the site, allowing still more time for the pathogen to die. As a result, large old genets are P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

522 HANSEN GOHEEN

fragmented and reappear as small, separated mortality centers in the new stand and the total area occupied by the pathogen remains roughly constant across the tree generations. This was demonstrated in a retrospective study on four, one-ha plots in a 60-year-old Douglas-fir stand naturally regenerated after harvest and wildfire (44). Mortality centers in the old stand were larger (up to 0.5 ha) than in the young stand, and generally without living Douglas-fir trees, as evidenced by complete lack of stumps. In the young stand, mortality centers were still small, mostly confined to fewer than ten trees, and centered around individual old stumps or snags from the previous stand. The same genet of P. weirii was isolated from both old stumps from the previous stand and recently killed trees of the current generation (unpublished data). Healthy Douglas-fir trees occupied the areas that corresponded to the middle of the mortality centers in the previous stand. The dynamics of advance and subsequent fragmentation of genets of P. weirii were carefully documented in old mountain hemlock forests in the high Cascades by Dickman & Cook (7). In this area, large circular disease centers are distinct on aerial photographs (Figure 2) and on the ground. Lodgepole pine dominates in the decades after stand-replacing wildfire, with susceptible mountain hemlock not dominating the stands again for about 200 years. Large old genets from the previous forest eventually reappear as scattered small mortality centers, ramets of the original genet, that resume slow spread through the aging mountain hemlock stand. Assuming the current spread rate has been constant in the past, individual genets were estimated to be up to 1340 years old (7). Lattin (28), accounting for the fire return interval and the long periods of fungus inactivity when sites were occupied by lodgepole pine, estimated that individual genets were probably twice as old, having persisted across ten or more tree generations.

Impacts on Forest Structure Forest structure, its stature, evenness, layering, and composition, is altered by Phellinus weirii. Young Douglas-fir forests that regenerated after wildfire are often very uniform and dense, except for the gaps created by laminated root rot. In many stands, such gaps are frequent. The fungus has killed trees on an estimated eight percent of the Douglas-fir forest area throughout the region (43), and in large local Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org areas the measured incidence is much greater. In ten counties of northwest Oregon,

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. incidence of P. weirii on ten-acre plots was 12% (12). In British Columbia, more than 80% of the second-growth Douglas-fir stands are infested (1). Seven large surveys in western Oregon, covering a total of 169,000 ha, found an average of 8.6% (range 4.7% to 16%) of the Douglas-fir forest area in root rot centers. In these laminated root rot mortality centers, an average of 50% of the Douglas-fir trees had been killed by P. weirii (13). In westside Douglas-fir stands, individual laminated root rot centers are often small (about 1/20 ha), although mortality centers up to one hectare in size are not uncommon. In detailed surveys of four 2.5-ha plots, Childs (6) mapped an average of 3.6 mortality centers per ha. In young stands there are usually more small centers (sometimes only scattered individual trees are P1: FHA August 2, 2000 17:33 Annual Reviews AR107-21

PHELLINUS WEIRII AND FOREST STRUCTURE 523 87 m on original print). = in mountain hemlock forest at 1700 m elevation 200m. (1 cm = Phellinus weirii Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org Mortality centers caused by Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. central Oregon Cascades. Scale bar Figure 2 in the P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

524 HANSEN GOHEEN

killed), but these merge into larger centers as the fungus advances through time (42). In mountain hemlock forests, the average size of laminated root rot centers is much larger. In a 28,000-ha area in the central Oregon Cascades, 2713 mortality centers were mapped, covering about 10.9% of the area. The average center was 93 m in radius (28). The impact of laminated root rot on stand structure is enhanced by interactions with other mortality agents, including wind and bark beetles. Wind topples root-rotted trees as well as sound trees exposed to gusts on the edge of root rot mortality centers. In a 500-acre stand of 70-year-old Douglas-fir in the Oregon Coast Ranges, windthrow from severe winter storms was almost exclusively associated with root rot centers. The Douglas-fir beetle (Dendroctonus pseudotsugae Hopkins) is closely associated with laminated root rot centers in westside forests (16). This beetle breeds in the boles of recently dead or dying Douglas-fir trees. Beetle populations are maintained year to year in root-rotted trees. Following ex- ceptional winter winds with large numbers of blown-down trees, beetle populations may increase to levels sufficient to overcome healthy trees. In surveys of mature Douglas-fir stands in western Oregon, more than 90% of down, root-rotted trees larger than 20 cm diameter had been infested by the Douglas-fir beetle. About 77% of the beetle-killed standing trees were on the edges of laminated root rot mortality centers (13). As a result of these interactions with wind and bark beetles, laminated root rot mortality tends to increase in pulses, although the spread of the fungus below ground is presumably much more constant. The structure of forests with root rot becomes increasingly uneven as stands age. The gaps in the stand do not remain empty, and competition for newly released resources, especially light, in root rot centers determines the effects on composition and succession. Gaps fill quickly with herbaceous and shrub vegetation, and are more gradually colonized by trees, younger and smaller than the surrounding stand. The result is increased structural diversity and creation of two or three vegetation layers. In young stands especially, root rot centers stand out from a distance as green islands of light amid the dark conifer forest.

Impacts on Succession and Diversity

Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org Phellinus weirii changes plant community diversity and successional trajectories,

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. but the direction of change depends on the relative susceptibility of the tree species present. In the high Cascade Mountains, mid-late–successional mountain hemlock is killed by P. weirii and the expanding gaps are ﬁlled with early seral tree species (mostly root rot–resistant pines) that can reproduce in the light of the mortality centers. Mountain hemlock also reproduces, and is eventually killed by the fungus. Species diversity, both richness and evenness, signiﬁcantly increased in the gaps and at the stand-level by the action of laminated root rot (33). Succes- sion is reversed, but because mountain hemlock continually regenerates in the large mortality centers, and is eventually killed, a unique community association results. P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

PHELLINUS WEIRII AND FOREST STRUCTURE 525

In Douglas-fir forests, the impacts of laminated root rot on local species diversity are not so predictable (24, 25). In old-growth Douglas-fir forests, where seed sources or advance regeneration of late-successional tree species such as western hemlock or western redcedar are often available, succession may be advanced as these shade-tolerant trees assume dominance behind the mortality front. The result is often a decrease in local species diversity because hemlock and cedar form very dense, dark canopies, allowing fewer understory species to grow than are present beneath an old Douglas-fir overstory. Although western hemlock is the so-called “climax” species in much of the Douglas-fir region, extensive areas with hemlock as the canopy dominant are seldom encountered, except in areas where Phellinus weirii has killed the Douglas-fir. In forests where hemlock and cedar are not available to colonize the gaps, diversity may be increased or unchanged, depending on the mix of shrub and herbaceous species that can take advantage of the light. Other species fill the role of western hemlock in other locations. In the Siskiyou Mountains in Southwest Oregon, mortality centers may fill with laminated root rot–resistant Port-Orford-cedar. Incense-cedar and sugar pine are observed regenerating in high numbers in laminated root rot mortality centers in some southern Oregon Cascade plant communities. In eastside forests, birch and aspen join surviving and regenerating conifers to create a unique mixed conifer-hardwood association. Although the specific consequences differ depending on the species composition and seral stage of the forest, laminated root rot is a major factor in shaping forest structure, composition, and process wherever it is found. Some of the changes in plant community composition caused by laminated root rot are persistent across tree generations. Many shrubs that persist in the root rot gaps are killed to the ground by stand-replacing fire, but sprout back from the roots. Ingersoll and colleagues (26) documented differences in plant species five years after stand-replacing disturbance (harvest) between areas that were intact forest in the previous stand and areas that were root rot gaps. Although total plant cover and species diversity were similar in former gaps and former forest, the species composition of the two areas differed significantly. Other observations suggest that hardwood species such as bigleaf maple that persist in the Douglas-fir forest in root rot gaps may sprout again after stand-replacing fire. Differences in early seral plant composition can influence later forest succession patterns (18), thus Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org amplifying future changes as the root rot reexerts itself in the new forest.

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. Animals respond to the changes in forest community. Where mortality centers fill with herbaceous vegetation, insects, birds, and mammals dependent on that vegetation, or on the snags and down trees, congregate. Mountain beaver (Aplodontia rufa), a burrowing mammal unrelated to the aquatic beavers, is common at early successional stages in some Douglas-fir forests. Once Douglas-fir crowns close and understory vegetation decreases, however, populations of this herbivore de- cline, surviving principally in canopy gaps where herbaceous food plants are still available. After fire or harvest returns the early successional flora, mountain beaver numbers increase again from these root rot refugia (PE Hennon & EM Hansen, unpublished). P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

526 HANSEN GOHEEN

Impacts on Nutrient Cycling Temperature and moisture relations are changed in the canopy gaps created by laminated root rot, sometimes greatly increasing rates of nutrient release and cycling from killed trees and litter. The effects have been documented in the mountain hemlock forests near Waldo Lake in the Oregon Cascades. Soil temperature, moisture, and nitrogen status were monitored along transects extending from intact 275-year-old mountain hemlock forest, across the advancing mortality fronts, and into the regenerating forest in the middle of the laminated root rot centers (31). Summer soil temperatures and water content were greater just behind the advancing front than in either the adjacent old forest or the older parts of the regenerating forest. This corresponded with a doubling or more of nitrogen mineralization rates along the advancing mortality front. In a follow-up study along the same transects, Waring and colleagues (52) measured decomposition rates, root chemistry, and tree vigor (wood production per unit leaf area). They found that decomposition rates, soil nitrogen availability, and root nitrogen levels increased behind the front, and declined again beneath closed canopy portions of the regenerating forest. Tree vigor was greater in young regenerating trees than in either the old uninfected forest or the older trees further behind the root rot mortality front. Regeneration is delayed in the zone immediately behind the mortality front, perhaps because of the harsh environment of the exposed soils. Seedlings become abundant only 10 m, about 30 years, behind the advancing front. Biomass increases with time in the regenerating stand, but does not approach pre-disease levels in 100 years, and net ecosystem production of carbon is negative or zero for much of that time, as decomposition of killed trees proceeds, regeneration grows very slowly, and a portion of the young trees are killed as the pathogen reasserts itself in the older regrowth (2). These effects on nutrient cycling are undoubtedly important on a landscape scale as well. Mortality centers cover a signiﬁcant portion of the landscape and are continuously moving. Phellinus weirii remains active in the more diverse regenerated forest behind the front, killing scattered susceptible trees as they grow in contact with old infected roots, triggering repeated bursts of decomposition and nitrogen availability. Similar processes occur in response to laminated root rot in

Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org the Douglas-fir forests, but the changes may not be as pronounced or as significant to the ecosystem as in the harsh environment of the mountain hemlock forest. Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. Douglas-fir trees are taller and individual mortality centers are smaller, so less direct sunlight reaches the forest floor in the gaps. Mortality fronts in the Douglas-fir forest are not marked by bare ground, since herbaceous and shrubby vegetation fills the gaps quickly on these more productive sites. Root rot centers are characterized by accumulations of coarse woody debris, the remains of trees killed by the pathogen. Large pieces of dead wood are important to a number of forest processes. They are slowly decomposed by a succession of fungi aided by wood-boring and -shredding insects. A large tree bole may be recognizable on the forest floor for 300 years or more as it is slowly reduced to P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

PHELLINUS WEIRII AND FOREST STRUCTURE 527

lignin and humus. It acts as a reservoir of water through the dry summer and thus is an important substrate for western hemlock seedlings and their mycorrhizal fungi. It provides shelter for numerous insects and several amphibians. Nitrogen content of decomposing wood increases with time through the scavenging action of decomposing fungi and nitrogen fixation by free-living bacteria that are concen- trated in the decaying logs. Over time, the net nitrogen contribution is significant to the n-cycle of the Douglas-fir forest (30).

EffectsofTreeVigor It has been proposed that the stimulation of nitrogen availability behind the mortality front is part of a feedback loop of alternating susceptibility and resistance to laminated root rot in mountain hemlock and Douglas-ﬁr forests (31, 52). The argument is based on an assumption of a cause and effect relationship between two observations in mountain hemlock mortality centers: (a) Root rot mortality in the mountain hemlock trees regenerating after the killing front has passed is not evident for many years, until a dense young stand of mixed species has reestab- lished; and (b) young trees growing on the open ground near the front are more vigorous, with a greater growth efﬁciency than the older trees that die further behind the front under the increasingly stressful, nitrogen-limiting conditions of the new closed forest. The investigators concluded that enhanced nutrient status and vigor of mountain hemlock reproduction behind the advancing edges of disease centers endowed the trees with increased resistance to infection. Matson & Waring (32) tested the hypothesis in a growth room inoculation experiment. Moun- tain hemlock seedlings were wounded and exposed to P. weirii inoculum, then grown under various combinations of nutrient and light deprivation. Inoculated, stressed seedlings developed more chlorotic foliage than control seedlings. The experiment led Waring and colleagues (52) to suggest that improving the environmental conditions for tree growth through vigor-increasing treatments such as fertilization or thinning would increase tree resistance to P. weirii. Their reasoning grew from the widespread but false belief that pathogens in natural ecosystems only kill weakened plants, articulated in a contemporaneous paper, “Characteristics of Trees Predisposed to Die” (50). Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org Several forest pathologists objected to the conclusions, in part because no direct

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. evidence of successful infection was observed and the fungus could not be reiso- lated from the seedlings (38). In general, pathologists have not observed laminated root rot mortality conﬁned to less vigorous trees or lower crown classes (21, 35). Local distribution of laminated root rot appears to be correlated with environmental conditions and historical disturbances that favor continuous site occupancy by susceptible conifers. In a direct test of the hypothesis that tree stress increases susceptibility to laminated root rot, wounded and unwounded roots on high- and low-vigor trees on four study sites in Oregon were inoculated with P. weirii (15). Vigor treatments in Douglas-ﬁr forests included thinned versus non-thinned trees, and codominant P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

528 HANSEN GOHEEN

TABLE 1 Phellinus weirii inoculation success on high and low vigor trees in four stands in Oregon. Necrotic ﬂecking Inoculation (percent of Extent of success Ectotrophic successful incipient Study Site (percent) growth (cm) inoculations) decay (cm)

Woods Creek (Douglas-fir) Low vigor 86 32 71 25 High vigor 90 33 72 28 HJ Andrews (Douglas-fir) Low vigor 83 22 77 18 High vigor 78 25 87 25 Waldo Lake (Mountain hemlock) Low vigor 96 56 84 48 High vigor 89 52 85 45 Sisters (Douglas-fir) Low vigor 75 23 61 10 High vigor 68 19 57 14

or dominant trees versus suppressed trees. The study site in the mountain hemlock forest was near the transects used by Waring and colleagues (52), and high- and low-vigor trees outside and inside the mortality centers were inoculated. Tree vigor was measured as the ratio of annual basal area increment per unit of sapwood area (53). Inoculation success averaged 83% after two years on the 710 roots (Table 1). There were no consistant or signiﬁcant differences in inoculation success or subsequent fungus growth between trees in high- or low-vigor treatments at any of Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org the study sites. Colonization was highly variable between individual trees within

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. vigor categories. Necrotic ﬂecking in the phloem was present beneath surface- colonized bark on 74% of the successfully inoculated roots. P. weirii was recov- ered from necrotic phloem in 32 of 200 roots (16%). Incipient decay caused by P. weirii was observed in the xylem of inoculated roots that had been wounded. Stain in the xylem ranged up to 52 cm, and P. weirii recovery from stained wood was 49%. We reject the hypothesis that vigorous Douglas-ﬁr and mountain hemlock trees are more resistant to infection by P. weirii than stressed trees. The apparent lower infection rates of mountain hemlock reproduction behind advancing root disease mortality fronts in high-elevation Cascade stands is not due to vigor differences, P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

PHELLINUS WEIRII AND FOREST STRUCTURE 529

but can be explained by a combination of other factors. Careful investigation shows that young, vigorous mountain hemlock reproduction is dying on these sites, al- beit in a very scattered fashion. The ﬁrst regenerating trees to be killed are small, and do not support tree-to-tree spread of the fungus. Mortality becomes obvious only when larger, older trees with extensive root systems are killed and there is opportunity for spread to adjacent trees. Several tree species, including some that are highly resistant to laminated root rot, occur with mountain hemlock in these areas (33) and interrupt tree-to-tree spread. In the coarse volcanic soils on these sites, many infected mature trees are uprooted. This may result in smaller pieces of colonized roots left to decay in the soil, thus reducing the inoculum potential of the site.

PATHOGENS AND OLD-GROWTH FORESTS

Many of the unique features of very old forests, including multi-storied stands, increased species diversity, large accumulations of snags and coarse woody debris, as well as the associated animals, are the consequence of pathogen activity. Root rots, stem decay fungi, and dwarf mistletoes all increase with stand age and often interact to shape the old-growth forest. As described above, root rot mortality centers continue to expand through the life of the stand, and fill with a different tree community than found outside the mortality center. Mortality centers, often many hectares in size, may lose definition in some communities, as fewer and fewer susceptible old individuals remain to infect on the perimeters, and scattered mortality continues in regenerating susceptible and tolerant trees within the area of infection. The altered community is usually more open, with fewer and more scattered large trees. In many cases, disease-tolerant trees in the “root rot climax” plant associations (48) are inherently smaller and shorter-lived than the susceptible trees they replaced. Western hemlock following Douglas-fir and lodgepole pine after mountain hemlock are typical. Stem decay fungi are more abundant in old forests. In most cases, young trees are as susceptible as old trees, but the chances of wound decay from falling tree or fire scars, or infection from true heart-rotting fungi such as Phellinus pini or Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org Echinodontium tinctorum accumulate with time and the volume of decayed wood

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. increases. The incidence of decay is often very high in old forests. In Douglas-ﬁr stands greater than 250 years old, Boyce & Wagg (3) found 17% of the stem volume decayed, primarily by P. pini, and net annual growth was negative after about age 250 years, with more volume lost to increasing decay than added through radial growth. In old stands of Abies spp. in mixed conifer forests, nearly every stem may be decayed, often with E. tinctorum. Pathogenic stem decay fungi also kill trees in western forests. Phellinus hartigii on hemlock and P. cancreformans on grand ﬁr are canker rots that kill the cambium and decay sapwood as well as heartwood. Phellinus pini slowly encroaches on the functional sapwood leading to mortality after 100 years or so (3). Stem P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

530 HANSEN GOHEEN

decay fungi, sometimes working together with carpenter ants and termites, are important to woodpeckers and the succession of birds that follow them in trees, and to mammals from squirrels to bears and fishers that use cavities created by decay and fire for dens. Dwarf mistletoes, caused by several host-specific species of Arceuthobium, are very widespread in the west, especially in older, multi-storied stands. These parasitic seed plants spread most effectively from tall trees onto shorter trees, and cause increasing growth loss and mortality over time, especially when infections are high in the crown. Witches brooms are characteristic symptoms of a number of species, and provide essential nesting and roosting structure for several birds, including the northern spotted owl (Strix occidentalis caurina). Each of these agents exerts an increasing negative effect on net primary production of forests as they attain great age. When they act together in the same stand, the result may be dramatic. Holah et al (25) briefly described an old-growth Douglas- fir stand where laminated root rot had advanced succession to the shade-tolerant western hemlock. Many of the oldest hemlocks, however, had been killed by P. hartigii canker rot, and the remainder were severely stunted by Arceuthobium tsugense, hemlock dwarf mistletoe. The resulting plant community was dominated by the shrub Rhododendron macrophylum. On harsh sites, such as the high- elevation mountain hemlock forest where P. weirii is so dramatic, stand biomass stays well below the surrounding old forest perhaps indefinitely, because lodgepole pine is naturally a rather small and short-lived tree, and because of continu- ing root rot mortality in the regenerating mountain hemlock (2). In the primeval forest, such dramatically deteriorated stands were relatively uncommon, because stand-replacing wildfire usually returned after several hundred years to reset the successional clock and reduce the pathogen population.

THE FOREST TODAY AND TOMORROW

The forest still grows much as it always has in large areas of western North America. The total forest area is nearly unchanged. Most areas, including nearly all of the privately owned forest land, have seen harvest activity, however, and the entire Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org region has been impacted to a greater or lesser extent by ﬁre control. A third

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. change is the introduction of several exotic pathogens and insects to the forests of the west. The impacts of these modern human inﬂuences on ecosystem structure and function and on interactions with indigenous pathogens vary by forest type.

Recent Changes in the Forest In westside forests, clearcut harvesting followed by slash burning and planting Douglas-ﬁr seedlings has been standard harvest/regeneration practice for about 60 years. The result is a landscape with a mosaic of small (16–80-ha) stands of varying ages, but mostly less than 100 years old. Management mimics the natural P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

PHELLINUS WEIRII AND FOREST STRUCTURE 531

stand-replacing wildfire, natural Douglas-fir regeneration sequence, although on a shorter time scale. Laminated root rot behaves much as it always has. Its signifi- cance in bringing structural and species diversity to the young forests is enhanced, however, with plantation management. Harvest and fire exclusion have had their greatest ecological impact in mixed conifer and pine forests east of the Cascade Mountains and in southern Oregon and northern California, where Armillaria ostoyae and Heterobasidion annosum join P. weirii in shaping the forest landscape. The species composition of these forests has been dramatically shifted from early seral (more valuable) pines and larch to mid- and late-seral Douglas-fir and Abies species. Since the latter species are generally more susceptible to all three root pathogens, the result has been a general increase in root rot incidence and impact. Armillaria is specifically favored by thinning and selective harvest activities (39); the pathogen rapidly colonizes the root systems of cut stumps spreading from previously localized root infections, and, with this increase in available energy, is able to infect and kill surrounding trees. Heterobasidion annosum (54) also benefits directly from selective harvests. The fresh stumps provide new infection courts for this spore-disseminated pathogen. Stumps and roots are colonized and the fungus spreads to adjacent trees across root contacts. The introduction of white pine blister rust (Cronartium ribicola) to western forests has had a dramatic direct impact, with an estimated 95% of western white pine trees killed since 1920. It has also changed ecosystem response to root rot. In Idaho and western Montana, stands dominated by root rot–resistant white pine have been replaced by root rot–susceptible species, and throughout its range white pine has been reduced to a temporary colonizer of root rot mortality centers. This removes one of the few tall-stature, long-lived trees from the mix of species that can follow root rot mortality fronts, increasing root rot activity in the more susceptible remainder, and accentuating the general decrease in site productivity that is associated with large, old root rot mortality centers in many western forests. Phytophthora lateralis threatens to have a similar effect in some forests of southwest Oregon and northern California where it threatens Port-Orford-cedar.

Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org Trends in Forest Management

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. Changes in the economics of the timber industry, past harvest practices, and politi- cal and legal pressure from environmental lobby groups are forcing a reassessment of forestry objectives and practices throughout North America. The consequences have been felt most immediately in the management of Federal forests, but all landowners are affected. An overdue reemphasis on forest environmental values is being awkwardly and uncertainly translated into changed forest management practices. Too often the targets are short-term aesthetic concerns (ban clearcutting, stop wildﬁres) without thoughtful consideration of long-term ecological consequences. There is increasing pressure for no active management, especially of public forest lands, but in this time of pervasive human impacts, no management is in itself P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

532 HANSEN GOHEEN

a management decision, with significant long-term ecological consequences, especially if we continue to suppress wildfires. Clearcutting and even-aged management is widely perceived as destructive to forest ecosystems and the trend is toward all-age management and selective harvest, even in the westside Douglas-fir forests that grew naturally as relatively even-aged cohorts after stand-replacing wildfire. In eastside pine and mixed conifer forest types, these trends may be appropriate and, if skillfully managed, can restore a more natural forest composition to large areas, if fire and thinning can be used to regulate stocking and species mixes. In westside Douglas-fir, they will favor shade-tolerant species, western hemlock and Abies grandis, and lead to a different forest composition than previously known, unless stand-replacing wildfires are allowed to burn. The impacts of these changes on root rot fungi and their roles in the ecosystem will vary from place to place, but in general have not been considered. On Federal lands, forest silviculturalists and pathologists have at least begun to struggle with the implications of these changes.

Ecosystem Management Ecosystem management is now policy for Federal forest land managers in the United States. Ecosystem management has been described as “the careful, skill- ful, and integrated use of ecological principles at various scales to produce, restore, or sustain ecosystem integrity and provide desired conditions, uses, products, values, and services over the long term” (45). Emphasis is placed on understanding system structure and function and on emulating ecological processes. Diver- sity and sustainability are guiding principles. Planning is done over long time frames and at watershed or wider scales. The positive role of fire is recognized. Insect and disease impacts are, in general, considered to be beneficial. Reserve areas are expanded. Rotation lengths are often increased. Thinning treatments are often scheduled more to accelerate development of old-growth characteristics than to increase wood production. Generally, uneven-age systems, such as single tree and group selection, and modified even-age systems that retain some trees, such as extended seed tree or shelterwood cuts, are preferred. Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org In the Pacific Northwest, controversy over the northern spotted owl and old-

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. growth Federal forests has led to the adoption of the Northwest Forest Plan (46). This plan provides standards and guidelines for management of seven different land allocations, including Riparian Reserves to protect the health of the aquatic system and its dependent species, Late Successional Reserves to maintain functioning late-successional and old-growth habitat, and Matrix where most timber harvest and other silvicultural activities will occur. The Northwest Forest Plan rec- ognizes, in principle, the critical roles that pathogens play in the development of multi-storied stands, increased diversity, large accumulations of snags, and coarse woody debris. However, the Plan does not distinguish among pathogens and their various inﬂuences in different ecosystems and does not adequately address the P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

PHELLINUS WEIRII AND FOREST STRUCTURE 533

complexities. Native pathogens are considered beneficial or inocuous in all cases except where impacts are considered to be large scale (1.6 ha or more with less than 40 percent crown closure as a result of disease impacts). This view presumes that disease dynamics are the same now as they were in historic times. Implications of past management practices, such as widescale establishment of disease-susceptible species like white and grand fir on sites with laminated root rot, which may take decades to become obvious, are often not addressed. The impact of fire exclusion in fire-regulated ecosystems on the distribution and severity of forest tree diseases also is often not considered. Predicting disease impacts decades into the future can be challenging. Required assessments such as watershed-level analyses and Late Successional Reserve As- sessments are supposed to address any specific disease concerns for an area. It is unfortunate that pathology input has not always been sought; these assessments provide the important opportunity to evaluate the roles of pathogens at varying scales and address the kinds of situations where they are or may become significant management concerns. The South Cascades Late Successional Reserve Assessment provides an example of the evaluation process (47). This Reserve covers approximately 300,000 ha in southern Oregon. Large portions of the area are highly fragmented; many young plantations occur. Accelerating the development of late-successional character and increasing connectivity are important management goals for the Reserve. Root disease is a factor in determining species to favor in stocking control measures done to accelerate development of late-successional character. Many plantations with root disease–susceptible species planted in areas of high P. weirii inoculum are now part of the Late Successional Reserve network. This will result in large areas where the desired development of trees of large size may be slowed or altogether precluded by root rot. Root disease–related management recommendations for the Reserve include: (a) In stands younger than 25 years old, where greater than ten percent of the area is determined to contain root-diseased trees, thinning should be avoided unless root disease–susceptible species can be discriminated against in favor of root disease–resistant species; (b) planting root disease–resistant species should be considered in openings; (c) the presence of root disease in older stands should be a factor in determining species to favor in intermediate entries done Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org to accelerate development of late-successional character. Root diseases in these

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. older stands may be beginning to provide canopy openings important to late- successional character, but additional entries in the stand may exacerbate root disease impacts beyond desirable levels. Further recommendations include avoid- ing intermediate entries in stands older than 25 years where root disease is present on more than 25% of the area. Managers should consider planting root disease– resistant species in openings if susceptible hosts make up a high proportion of the stand and a high degree of canopy closure is desired. Finally, in established stands where root disease is present at lower levels and intermediate stand entries are planned, managers should discriminate against susceptible species during the entry. P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

534 HANSEN GOHEEN

Modeling Root Disease Resource managers have a general understanding that forest pathogens cause growth loss and mortality of individual trees and that pathogens influence stand- level diversity, structure, and composition. Consistently describing and depicting these influences over time and space across a range of management alternatives is often very challenging, however. Predicting pathogen effects becomes even more challenging as stand-level prescriptions become more complex and analysis areas increase in size from a few adjacent stands to entire landscapes or watersheds. Forest pathologists have responded to these challenges by developing predictive tools including the Western Root Disease Model (WRDM) extension to the Forest Vegetation Simulator (FVS) (10, 55). The WRDM simulates growth impacts, mortality, spread within root disease centers, and spread into the healthy portion of a stand for Armillaria root disease, laminated root rot, or both the S and P types of annosus root disease. It takes into account known differences in host susceptibility. Stand density, tree size, stump size, and species composition interact to influence root disease spread and survival of inoculum in a stand. The model was designed to be flexible; the user can adjust model parameters to portray local conditions that may differ from westwide defaults. The WRDM also models root disease interactions with bark beetles and can depict density-dependent or windthrow-related bark beetle activity that is not associated with root disease. The user can simulate root disease–management activities such as pulling stumps or applying borate compounds as stump protectants. Once current root disease conditions are described to FVS, the model accumulates root disease impacts as long as inoculum remains viable and susceptible species remain on the site. WRDM-specific outputs include information on area affected, numbers of infected trees, and numbers, sizes, and species of trees killed. Volume losses specific to root disease are also tabulated. Comparisons are often made between projec- tions where root disease impacts are included and those where root disease impacts are ignored. The WRDM is often used to evaluate management alternatives at the stand-level. It has also been used at planning area and landscape scales to simulate the influence of root diseases on vegetative structure, forest succession, wildlife

Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org habitat connectivity, visual quality objectives, watershed health, and commodity outputs. The WRDM was recently used to calibrate forest successional models Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. as a part of the Interior Columbia River Basin Assessment. It is currently being used to understand laminated root rot impacts in a network of Late Successional Reserves in the Oregon Coast range. One of the immense challenges that land managers face if they wish to simulate root disease influences on stands is lack of information describing current root disease conditions. The often subtle nature of root disease symptoms makes these pathogens difficult for field crews to identify; as a result, accurate root disease information is often missing from forest inventories. Funding for detailed, P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

PHELLINUS WEIRII AND FOREST STRUCTURE 535

plot-based stand examinations that could be used to acquire this information is decreasing. Many assessments rely on “walk-through” types of reconnaissance for evaluating vegetative conditions. There are many options in the WRDM for initializing current root disease conditions in a stand depending on whether there were data collected at the tree-level, plot-level, stand-level, or in some combination. The WRDM must be supplied with at least a minimal description of the area affected, the proportion of trees affected, and how the root disease is distributed. As a result, pathologists have placed much effort on methods to obtain these de- scriptions including evaluating aerial photography and devising stand-level rating scales that can translate into direct input to the WRDM (14, 17). Another challenge results from the complex nature of forest ecosystems and their resident insects and pathogens. Root diseases often occur together in com- plexes in many areas of the West. Currently, the WRDM allows the simulation of one root disease at a time in any given stand. The WRDM takes into account the relationships between bark beetles and root disease pathogens, but no model exists that looks at the interactions of other common insects and pathogens. The need for a multi-pest modeling structure is often discussed. For example, a large-scale insect and disease analysis currently under way in the southern Oregon Cascades points out the need to carefully consider multiple pathogen interactions (14). In this area, as many as four root diseases occur in a given stand and there is a high incidence of Douglas-fir dwarf mistletoe. Resource objectives include commodity outputs, development and maintenance of late seral conditions, and connectivity. It is not easy for silviculturists to meet their management objectives while balancing host susceptibility to the pathogens and site productivity. As a demonstration of the Western Root Disease Model, projected volumes and trees per acre under current root disease conditions and when root disease is ignored were compared. A 10.8-ha stand affected by laminated root rot in the southern Oregon Cascades was inventoried. The stand is fire-regenerated and is approximately 90-years-old with scattered 150- to 250-year-old residuals. Mortality centers were mapped to estimate the total stand area in root disease. Tree-level information was collected to indicate whether trees were infected by P. weirii, within 30 feet of infected trees (“suspect” trees), or healthy. The stand was dominated by white fir (75% of the trees per acre and 68% of the volume), followed by Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org Douglas-fir (11% of the trees per acre and 14% of the volume) and Shasta red fir

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. (10% of the tree per acre and less than 1% of the volume). Minor components of sugar pine (less than 1% of the trees per acre and 11% of the volume) and incense- cedar (3% of the trees per acre and 4% of the volume) were scattered throughout the area. The ratio of infected and suspect trees to healthy trees was approximately 2 : 1 for the entire area. The mapped area of root disease centers was approximately 7.2 ha. The information was then entered into the West Cascades variant of the Forest Vegetation Simulator with the Western Root Disease Model extension and a 100- year projection of stand growth (1996–2096), with and without root disease, was P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

536 HANSEN GOHEEN

Figure 3 Modeling laminated root rot impacts in a 90-year-old white ﬁr/Douglas-ﬁr stand in southwest Oregon. Trees per acre with and without root disease (RD) over a 100-year Western Root Disease Model (WRDM) projection, 1996–2096.

made. Model results indicate an estimated 75% reduction in both stocking and volume for the stand over the next 100 years relative to a healthy stand when no ingrowth was considered (Figures 3, 4). Information like this provides graphic input to silviculturalists and administra- tors challenged to implement ecosystem management in the real world. Pathogens are a natural part of the ecosystem, but they may interfere with the achieve- ment of ecological objectives just as they have limited timber productivity in the past. Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only.

Figure 4 Modeling laminated root rot impacts in a 90-year-old white ﬁr/Douglas-ﬁr stand in southwest Oregon. Cubic foot volume with and without root disease (RD) over a 100-year Western Root Disease Model (WRDM) projection, 1996–2096. P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

PHELLINUS WEIRII AND FOREST STRUCTURE 537

Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED 1. Bloomberg WJ, Reynolds G. 1985. Growth northwest Oregon, 1976. USDA For. Serv. loss and mortality in laminated root rot in- Res. Note PNW-381. 7 pp. fection centers in second-growth Douglas- 13. Goheen DJ, Hansen EM. 1993. Ef- fir on Vancouver Island. For. Sci. 31:497– fects of pathogens and bark beetles on 508 forests. In Beetle-Pathogen Interactions in 2. Boone RD, Sollins P, Cromack K Jr. 1988. Conifer Forests, ed. TD Schowalter, GM Stand and soil changes along a moun- Filip, pp. 175–98. New York: Academic. tain hemlock death and regrowth sequence. 252 pp. Ecology 69:714–22 14. Goheen EM. 1998. Root diseases in the 3. Boyce JS, Wagg JWB. 1953. Conk rot of southern Oregon Cascade Mountains: a old-growth Douglas-fir in western Oregon. method for large scale impact assess- OR. For. Prod. Lab. Bull. 4. 96 pp. ment, modeling, and treatment prioriti- 4. Castello JD, Leopold DJ, Smallidge PJ. zation. Proc. Int. Conf. Root Butt Rots, 1995. Pathogens, patterns, and processes Bordeaux, 9th, pp. 191–98. Paris: INRA in forest ecosystems. BioScience 45:16– Editions (France) Les Colloques No. 89. 24 15. Goheen EM, Hansen EM. 1993. Tree vigor 5. Childs TW. 1963. Poria weirii root rot. and susceptibility to infection by Phellinus Phytopathology 53:1124–27 weirii: results of field inoculations. In Proc. 6. Childs TW. 1970. Laminated root rot Int. Conf. Root Butt Rots, 8th, pp. 45–51. of Douglas-fir in western Oregon and Uppsala: Swed. Univ.Agric. Sci. Washington. USDA For. Serv. Res. Pap. 16. Hadfield JS, Goheen DJ, Filip GM, PNW-102. 27 pp. Schmitt CL, Harvey RD. 1986. Root Dis- 7. Dickman A, Cook S. 1989. Fire and fungus eases in Oregon and Washington Conifers. in a mountain hemlock forest. Can. J. Bot. USDA For. Serv., PNW Reg., For. Pest 67:2005–16 Manag., Portland, OR. 27 pp. 8. Farr DF, Bills GF, Chamuris GP, 17. Hagle SK. 1985. Monitoring Root Disease Rossman AY. 1989. Fungi on Plants Mortality: Establishment Report. USDA and Plant Products in the United States. For. Serv., North. Reg. Missoula, MT. Rep. APS Press: St Paul, MN. 1252 pp. 85–27. 13 pp. 9. Filip GM, Schmitt CL. 1979. Susceptibil- 18. Halpern CB, Franklin JF. 1990. Physiog- ity of native conifers to laminated root rot nomic development of Pseudotsuga forests

Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org east of the Cascade Range in Oregon and in relation to initial structure and distur- Washington. For. Sci. 25:261–65 bance intensity. J. Veg. Sci. 1:475–82

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. 10. Frankel SJ, technical coordinator. 1998. 19. Hansen EM. 1979. Sexual and vegeta- User’s guide to the Western Root Disease tive incompatibility reactions in Phellinus Model, version 3.0. USDA For. Serv., Gen. weirii. Can. J. Bot. 57:1579–82 Tech. Rep. PSW-GTR-165. 164 pp. 20. Hansen EM. 1979. Survival of Phellinus 11. Franklin JF, Dyrness CT. 1973. Natu- weirii in Douglas-ﬁr stumps after logging. ral Vegetation of Oregon and Washington. Can. J. For. Res. 9:484–88 USDA For. Serv., Gen. Tech. Rep. PNW-8. 21. Hansen EM. 1986. Inoculation of Douglas- 417 pp. ﬁr roots with isolates of Phellinus weirii on 12. Gedney DR. 1981. The occurrence of lami- sites differing in root rot severity. Can. J. nated root rot on non-federal timberland in For. Res. 16:619–23 P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

538 HANSEN GOHEEN

22. Hansen EM, Angwin PA, Dreisbach TA, in the Oregon Cascades. Ecology 65:1511– Gernandt D, McWilliams MG. 1998. 16 Species limits for Phellinus weirii. In Proc. 32. Matson PA, Waring RH. 1984. Effects of Int. Conf. Root Butt Rots, 9th, Bordeaux, nutrient and light limitation on mountain pp. 119–27. Paris: INRA Editions (France) hemlock: susceptibility to laminated root Les Colloques No. 89 rot. Ecology 65:1517–24 23. Hansen EM, Stenlid J, Johansson M. 1994. 33. McCauley KJ, Cook SA. 1980. Phellinus Somatic incompatibility in Heterobasid- weirii infestation of two mountain hemlock ion annosum and Phellinus weirii. In Proc. forests in the Oregon Cascades. For. Sci. Int. Conf. Root Butt Rots, 8th, pp. 323–33. 26:23–29 Uppsala: Swed. Univ. Agric. Sci. 34. Meinicke EP. 1916. Forest Pathology in 24. Holah JC, Wilson MV, Hansen EM. Forest Regulation. USDA Bull. No. 275. 1993. Effects of a native forest pathogen, 62 pp. Phellinus weirii, on Douglas-fir forest 35. Mounce I, Bier JE, Nobles MK. 1940. A composition in western Oregon. Can. J. root rot of Douglas-fir caused by Poria For. Res. 23:2473–80 weirii. Can. J. Res. C 19:522–33 25. Holah JC, Wilson MV, Hansen EM. 36. Nelson EE, Hartman T. 1975. Estimating 1997. Impacts of a native root-rotting spread of Poria weirii in a high eleva- pathogen on successional development of tion mixed conifer stand. J. For. 73:141– old-growth Douglas fir forests. Oecologia 42 111:429–33 37. Pickett STA, White P, eds. 1985. The Ecol- 26. Ingersoll CA, Wilson MV, Thies WG. ogy of Natural Disturbance and Patch Dy- 1996. Effects of Phellinus weirii gaps namics. Orlando: Academic. 472 pp. on early successional vegetation follow- 38. Shaw CG III, Goheen DJ, Hessburg ing timber harvest. Can. J. For. Res. 26: PF. 1988. Letter to the editor. Forestry 322–26 61:369–73 27. Larsen MF, Lombard FF, Clark JW. 1994. 39. Shaw CG III, Kile GA. 1991. Armillaria Phellinus sulphurascens and the closely re- Root Disease. USDA For. Serv. Agric. lated P. weirii in North America. Mycolo- Handb. No. 691. 233 pp. gia 86:121–30 40. Spies T. 1997. Forest stand structure, com- 28. Lattin PD. 1997. Spatial distribution of position, and function. In Creating a Phellinus weirii infection centers in the Forestry for the 21st Century, ed. KA Central Oregon Cascades: inferring eco- Kohm, JF Franklin, pp. 11–30. Washing- logical processes from patterns. MS thesis, ton DC: Island Press. 475 pp. Oregon State Univ. 157 pp. 41. Thies WG. 1983. Determination of growth 29. Long JN. 1977. Trends in plant species reduction in Douglas-fir infected by Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org diversity associated with development Phellinus weirii. For. Sci. 29:305–15

Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only. in a series of Pseudotsuga menziesii/ 42. Thies WG, Nelson EE. 1997. Laminated Gaultheria shallon stands. Northwest Sci. root rot: new considerations for surveys. 51:119–30 West. J. Appl. For. 12:49–51 30. Maser C, Tarrent RF, Trappe JM, Franklin 43. Thies WG, Sturrock RN. 1995. Laminated JF. 1988. From the Forest to the Sea: A root rot in Western North America. USDA Story of Fallen Trees. USDA For. Serv. For. Serv. PNW-GTR-349. 32 pp. PNW-GTR-229. 153 pp. 44. Tkacz BM, Hansen EM. 1982. Damage by 31. Matson PA, Boone RD. 1984. Natural dis- laminated root rot in two succeeding stands turbance and nitrogen mineralization: of Douglas-ﬁr. J. For. 80:788–91 wave-form dieback of mountain hemlock 45. USDA For. Serv. 1993. A Draft Glossary P1: FHA August 1, 2000 13:14 Annual Reviews AR107-21

PHELLINUS WEIRII AND FOREST STRUCTURE 539

for Ecosystem Management. Pacific North- 51. Waring RH, Franklin JF. 1977. Evergreen west Reg. Portland, OR. 12 pp. coniferous forests of the Pacific Northwest. 46. USDA For. Serv., USDI Bur. Land Manag. Science 204:1380–86 1994. Record of Decision for Amendments 52. Waring RH, Cromack K Jr, Matson PA, to Forest Service and Bureau of Land Boone RD, Stafford SG. 1987. Responses Management Planning Documents within to pathogen-induced disturbance: decom- the Range of the Northern Spotted Owl. position, nutrient availability, and tree Portland, OR. 75 pp. vigour. Forestry: 60:219–27 47. USDA For. Serv., USDI Bur. Land Manag. 53. Waring RH, Thies WG, Muscato D. 1980. USDI Fish Wildlife Serv. 1998. South Stem growth per unit of leaf area: a mea- Cascades Late Successional Reserve sure of tree vigor. For. Sci. 26:112–17 Assessment. 255 pp. 54. Woodward S, Stenlid J, Karjalainen R, 48. van der Kamp BJ. 1991. Pathogens as Huttermann A, eds. 1998. Heterobasid- agents of diversity in forested landscapes. ion annosum: Biology, Ecology, Impact For. Chron. 67:353–54 and Control. Wallingford, UK: CAB Int. 49. Wallis GW, Reynolds G. 1965. The initia- 589 pp. tion and spread of Poria weirii root rot of 55. Wyckoff WR, Crookston NL, Stage AR. Douglas fir. Can. J. Bot. 43:1–9 1982. User’s Guide to the Stand Prognosis 50. Waring RH. 1987. Characteristics of trees Model. USDA For. Serv. Gen. Tech. Rep. predisposed to die. BioScience 37: 569–74 INT-GTR-133. 112 pp. Annu. Rev. Phytopathol. 2000.38:515-539. Downloaded from www.annualreviews.org Access provided by U.S. Department of Agriculture (USDA) on 09/02/16. For personal use only.