BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER

and Erosion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Effective precipitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Stream/spring flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Hillslope runoff and erosion . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Hydrologic impacts of western juniper control treatments . . . . . . 37 Wildlife . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Large Herbivores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Birds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Winter habitat and food source . . . . . . . . . . . . . . . . . . . . . . . . 39 Breeding habitat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Old-growth woodlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Small Mammals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Management Considerations . . . . . . . . . . . . . . . . . . . . . . . . . 41 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 3 Restoration and Management . . . . . . . . . . . . . . . . . . 42 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Assessment of Western Juniper Control Justifications . . . . . . 42 Does western juniper removal restore plant communities? . . . . . 42 Does western juniper removal increase forage production and quality? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Does western juniper removal reduce erosion and increase water capture on site?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Does western juniper removal increase discharge and subsurface flow? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Does western juniper removal improve wildlife habitat? . . . . . . . 43 Does western juniper removal increase biological diversity?. . . . . 43 Mechanical Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Chainsaw cutting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Heavy machinery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Fire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .48 Plant composition and seed pools . . . . . . . . . . . . . . . . . . . . . . 48 Western juniper cutting and prescribed fire combinations . . . . . . . 49 Burning in aspen for juniper control . . . . . . . . . . . . . . . . . . . . . 49 Post-treatment management and disturbance . . . . . . . . . . . . . . 49 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Chemical. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Livestock Grazing Following Western Juniper Treatment . . . 50 Economics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Weeds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Ecology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Weed response following treatment. . . . . . . . . . . . . . . . . . . . . 52 Species of concern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Guidelines for Management . . . . . . . . . . . . . . . . . . . . 54 Asking the Right Questions . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Management Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Fire. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Mechanical: chainsaws . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Mechanical: heavy machinery. . . . . . . . . . . . . . . . . . . . . . . . . 57 Chemical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Seeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Weeds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Uses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Ethnobotany. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Current Uses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Knowledge Gaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Biology and Ecology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Literature Cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Appendixes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Appendix 1. Western Juniper Old-growth Cover Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Appendix 2. Summary Results from Research Investigating Response of Western Juniper to Various Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Appendix 3. Benefits and Comments Extracted from BLM Reports on Western Juniper Treatments Conducted in Oregon . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Appendix 4. Pre- and Post-treatment Results for Two Different Ecological Sites . . . . . . . . . . . . . . . . . . . . . . . . . 76 4 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER T he rapid expansion of western juniper into neighboring plant communities during the past 130 years has caused considerable concern because of increased soil erosion, reduced stream flows; reduced forage production; altered wildlife habitat; changes in plant community composition, structure, and biodiversity; and the replacement of mesic and semi-arid plant communities with woodlands. However, the impacts of post- settlement woodland expansion are not always clear or consistent across sites and have led to debate and legal challenges over control projects and management plans for western juniper. This publication represents a synthesis of what is known about the history, biology, ecology, and management of western juniper. Western juniper occupies 9 million acres in central and eastern Oregon, northeastern California, southwestern Idaho, and northwestern Nevada, and occurs in a few outlying stands in southern Washington. Presettlement changes in woodland abundance and distribution are largely attributed to long-term changes in temperature, amounts and distribution of precipitation, and the extent and return intervals of fire. Evidence supporting rapid post-settlement expansion is derived from old surveys, photographs, the distribution of relict presettlement woodlands, and tree-ring chronologies. Western juniper represents the northwestern portion of the piñon and juniper region in the Intermountain West. The tree is submonoecious and develops male cones in early spring, which attain full size the first summer and mature during the second summer. Female cones persist on trees for nearly 2 years. Seeds are dormant and germination potential is greatly enhanced by prolonged cool- moist stratification, which is cumulative from year to year. Seed dispersal of western juniper occurs through gravity, overland flow, and animals. At least 12 species of birds feed on the fruits and as a group are the most important disseminator’s of western juniper seed. Western juniper grows on a wide variety of parent materials and soils including materials derived from aeolian (e.g., pumice sands), sedimentary, and igneous sources (e.g., rhyolite, andesite, basalt). Soil textures range from clay to sandy and soil temperature regimes from mesic to frigid. Western juniper communities may be separated into presettlement (old-growth) or post-settlement (expansion) communities. We suggest 1870 as a cut-off to separate the two age classes. Western juniper is a long-lived species (more than 1,000 years). However, old-growth represents only a small proportion of the population throughout most of its range with the exception of the Mazama Ecological Province. Old-growth trees and stands can easily be separated from post-settlement stands based on morphological and stand structure characteristics. The majority of post-settlement communities are still in a state of transition. The stage of woodland succession (defined in this publication as Phases I, II, and III) directly affects plant community structure, composition, seed pools, wildlife habitat, and ecological processes including hydrologic and nutrient cycles. The phase of woodland development also affects the selection of management treatment, response following treatment, follow-up management, and treatment cost. As the tree layer increases in dominance, the shrub and herb layer decline. The degree that the herb layer is depleted is dependent upon soil depth to a restrictive layer. The minimum time for the tree overstory to begin suppressing the understory is 45– 50 years and to approach stand closure 70–90 years on cool wet sites and 120–170 on dry warm sites. Western juniper expansion into sagebrush grassland Summary BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 5 can affect the spatial distribution of soil organic matter, carbon, and nutrients. The loss of nutrients will also increase if woodland development results in accelerated erosion. Changes in hydrologic processes and water balance as tree abundance and dominance increase are not well understood. Evidence suggests that juniper can impact infiltration rates, sediment loss, and soil water storage and depletion rates. Accelerated soil water depletion rates in western juniper-dominated stands can decrease the length of the growing season by as much as 4–6 weeks. However, the impacts of western juniper on the water balance at the watershed or basin level have not been determined, nor have the effects of woodlands on subsurface flow into streams and springs. A large variety of wildlife species use early transitional states of woodlands that still contain an understory of shrubs and herbs. However, as structural diversity declines with increasing tree dominance, wildlife abundance and diversity also decline. Western juniper has significantly increased in density and distribution since the late 1800’s and if left unchecked can have significant impact on soil resources, plant community structure and composition, water and nutrient cycles, and wildlife habitat. As a result, control of western juniper has been a major concern of land management since the early 1960’s. In the 1960’s through the early 1970’s chaining and dozing were the most common forms of western juniper control. In the 1970’s, chainsaws became a widespread tool used for juniper control. In the 1990’s, the use of prescribed fire for juniper control also increased. This document evaluates different western juniper control practices including fire, mechanical, chemical, seeding, and post-treatment grazing. Specific concerns regarding the justifications used to support western juniper removal are also discussed. A large concern in woodland control treatments is weed infestation. Weed response following woodland conversion projects is site-specific and depends heavily on the initial floristics of each plant community. The ecological site (especially where it fits along the gradient of warm- dry to cool-moist), initial floristics, and the stage of woodland development are very important factors that will influence the response of a site following thinning or total removal of trees. A framework of questions are defined that will help land managers and private landowners select the most appropriate management action. There are some commercial uses of juniper but profit margins are often marginal. To date, products include firewood, chips for particle-flake board and animal bedding, decking, interior paneling, doors, cabinetry, rustic furniture, picture frame molding, small gifts, Christmas decorations, and the female cones are used as flavoring for gin. A great deal has been learned about the ecology, biology, history, and management of western juniper over the past several decades. However, not all of the questions have been answered in some areas, reducing but not totally limiting our ability to manage western juniper on an ecosystem basis. Photo by Lynn Ketchum 6 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER Introduction Distribution and History of Woodland Expansion Western juniper (Juniperus occidentalis var. occidentalis Hook.)1 has occupied its current range for several thousand years. Its rapid expansion into neighboring communities the past 130 years has caused considerable concern because of increased soil erosion; potential reduced stream flows; reduced forage production; altered wildlife habitat; changes in plant community composition, structure, and biodiversity; and the replacement of mesic and semi-arid plant communities with woodlands2. However, western juniper has been reported to be a valuable source of wildlife habitat throughout the literature and has aesthetic appeal. The wood has been used for various products including firewood, fence posts, and commercial energy production. The impacts of post-settlement woodland expansion are not always clear or consistent and have led to debate and legal challenges over control projects and management plans for western juniper. This document represents a synthesis of what is known about the history, biology, ecology, and management of western juniper. We hope to dispel some of the myths, identify knowledge gaps, sort out some of the issues related to woodland expansion, and increase the overall understanding of western junipers place and function in the northern Great Basin. This synthesis will provide guidance for defining long-term goals, setting management priorities, and developing management plans and strategies related to western juniper This publication is separated into six major sections: 1) distribution and history of woodland expansion, 2) life history and biology, 3) ecology; 4) hydrology, 5) restoration and management, and 6) management guidelines. Subsections within each category allow readers to easily refer to specific subject areas related to western juniper. We cite some literature associated with other juniper species to help put western juniper communities into a larger context of juniper and piñon woodlands in the American West and to support ecological concepts and management action. However, the focus of this paper is on western juniper. Distribution Juniper and piñon woodlands currently occupy over 74 million acres in the western United States (West 1999). The northwestern portion of the piñon and juniper region is represented by western juniper. Western juniper occupies 9 million acres in central and eastern Oregon, northeastern California, southwestern Idaho, and northwestern Nevada, and occurs in a few outlying stands in southern Washington (Table 1, Fig. 1) (USDA Forest Service 1981, Gedney et al. 1999, Miller and Tausch 2001, Azuma et al. 2004). Western juniper is usually the only conifer species occupying a site except where western juniper woodlands adjoin ponderosa pine (Pinus ponderosa) forests. Precipitation across most of the western juniper zone varies between 10 and 15 inches (Gedney et al. 1999), most of which falls during the winter and spring (October through June). However, western juniper can grow in areas receiving as little as 7 inches or exceeding 20 inches of precipitation annually. It grows over a wide array of environments and occupies elevations ranging from 600 to 8,000 ft (Sowder and Mowat 1958, Miller and Rose 1995, Gedney et al. 1999, Miller et al. 2000). Nevertheless, most of western juniper woodlands and savannas are found between 2,000 and 6,000 ft (Gedney et al. 1999). Western juniper is usually not found above 7000 ft because its foliage is damaged by extreme winter temperatures (Miller and Rose 1995). A second variety of western juniper, Sierra juniper (J. occidentalis var. australis), extends along the eastern slopes of the Sierra Nevada mountain range south of Susanville, California, and east and south from the Anchorite Hills (south of Walker Lake in Nevada near the California border) to the Panamint and San Bernardino mountains (USDA Forest Service 1981)(Fig. 1). This species is usually found growing as widely scattered trees mixed with other conifers at elevations between 4,100 and 9,100 ft. Recent work has documented small pockets of Sierra juniper growing in the mountains of central and eastern Nevada (Charlet 1996). Although stands typically occur well above Utah juniper (J. osteosperma) in this region, mixed stands including hybrids of Sierra and Utah juniper are occasionally found along drainages at lower elevations (Charlet 1996, Terry et al. 2000). 1All scientific names used throughout the text are from Cronquist, A.A., et al. 1972–1996. Intermountain-flora: vascular plants of the Intermountain West, USA. 2 Western juniper woodlands are defined as having more than 10 percent tree canopy compared to savannas that have less than 10 percent tree canopy (Gedney et al. 1999). BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 7 Figure 1. Distribution map of western juniper (Juniperus occidentalis var. occidentalis) and Sierra juniper (J. occidentalis var. australis) (derived from Griffin and Critchfield 1972, Charlet 1996, Gedney et al. 1999, and USGS 1:250,000 maps; developed by Steve Petersen, Department of Rangeland Ecology and Management, Oregon State University, Corvallis, Oregon). Table 1. Estimated area occupied by western juniper woodlands (tree canopy more than 10 percent cover) and shrub steppe savannas (tree canopy less than 10 percent). Community State structure Acres Reference California woodland 1,284,000 Bolsinger 1989 savanna 797,000 Bolsinger 1989 Idaho woodland 250,000 Chojnacky 1995 Nevada mixed 100,000 estimate1, no reference Oregon woodland 2,239,000 Gedney et al. 1999 savanna 2,818,000 Gedney et al. 1999 Washington savanna trace Total 7,488,000 (9,000,000)2 1Estimated from USGS 1:250,000 maps by R.F. Miller, EOARC, Oregon State University. 2An in-progress juniper inventory has increased the acreage in Oregon an additional 1.5 million acres bringing the total to nearly 9 million acres (Azuma et al. 2004) Presettlement Expansion The distribution and density of western juniper changed significantly across the Intermountain West around the late Pleistocene and into the Holocene. Changes in woodland abundance and distribution are largely attributed to long-term changes in temperature, amount and distribution of precipitation, and the extent and return interval of fire (Davis 1982, Thompson and Hattori 1983, Mehringer 1987, Van Devender et al. 1987, Wigand et al. 1995). During much of the Pleistocene, 45,000– 12,500 years BP (before present), western juniper had a much more southerly distribution, with the northern boundary near Kings Canyon National Park, California (Cole 1983). Towards the end of the Pleistocene, 12,000–15,500 years BP, its northern- most distribution was the Winnemucca Lake Basin in Nevada (Thompson 1984). It was also located on the eastern shore of Pluvial Lake Lahontan (Thompson et al.1986). Only prostrate juniper (J. horizontalis) and common juniper (J. communis) occupied southeastern Oregon at the end of the Pleistocene (Wells 1983). As temperatures warmed during the early Holocene, western juniper began migrating north into its present range. Macrofossils (leaves, twigs, and seeds) from pack rat middens found in caves at the Lava Beds National Monument in northern California date its arrival around 5,300 years BP (Mehringer and Wigand 1984). In Oregon, the earliest evidence of western juniper (pollen from pond and lake sediment cores) was dated 6,600 years BP in the Fort Rock Basin in south-central Oregon (Bedwell 1973) and 4,800 years BP at Diamond Craters in eastern Oregon (Wigand 1987). Since the arrival of western juniper in central and eastern Oregon, northeastern California, and southeastern Idaho, its abundance and distribution have fluctuated (Mehringer 1985, Mehringer and Wigand 1990, Miller and Wigand 1994). Following a very dry period during the mid-Holocene, 7,500– 5,000 years BP, western juniper rapidly expanded into its new range. Precipitation increased while temperatures remained warm between 5,000 and 4,000 years BP (Davis 1982, Mehringer 1986, Wigand 1987). Between 4,000 and 3,000 years BP climatic conditions were relatively wet and cool. Western juniper continued to increase, but retreated from higher elevations and expanded to lower elevations during this period. Western juniper reached most of its current geographic range approximately 3,000 years BP (Wigand et al. 1995). Severe drought and major fires during the late Holocene, 2,500–1,500 years BP, resulted in regional declines in western juniper (Mehringer and Wigand 1987, Wigand et al. 1995). Around 1,200 years BP summer precipitation increased, resulting in increases in abundance of both grasses and western juniper. A drying period between 900 and 700 years BP again reduced woodland abundance (Wigand et al. 1995). 8 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER Post-settlement Expansion During the past 130 years, western juniper has been expanding within its geographic range at unprecedented rates compared to any other time period during the Holocene (Miller and Wigand 1994, Miller and Tausch 2001). Historical expansions of western juniper and other piñon and juniper species throughout the West are well documented in the literature (Cottam and Stewart 1940; Burkhardt and Tisdale 1976; Tausch et al. 1981; Tausch and West 1988, 1995; Miller and Rose 1995, 1999; Gedney et al. 1999; O’Brien and Woudenberg 1999; Soulé and Knapp 1999; Tausch and Nowak 1999; Coppedge et al. 2001; Soulé et al. 2004). For western juniper, evidence supporting rapid post-settlement expansion is derived from old surveys, photographs (i.e., Fig. 2), the distribution of relict presettlement woodlands, and tree ring chronologies. Limited evidence suggests western juniper began increasing its range following the end of the Little Ice Age in 1850 (Mehringer personal communication, Johnson 2005). However, its rapid increase in abundance and expansion since the late 1800’s (Table 2) has largely been attributed to anthropogenic factors (Miller and Wigand 1994, Knapp et al. 2001b, Miller and Tausch 2001). Western juniper is a long-lived species3 and presettlement woodlands have been in place for hundreds and thousands of years (EOARC4, unpublished data). However, presettlement western juniper stands outside of the Mazama Ecological Province5 are estimated to account for only 10 percent or less of present day woodlands (Miller et al. 1999a, Johnson 2005). Most woodlands have developed during the past 130 years. Western juniper woodlands in eastern Oregon with more than 10 percent canopy cover increased from 456,000 acres in 1936 (Cowlin et al. 1942) to 2.2 million acres in 1988 (Gedney et al. 1999). Other evidence supporting the post-settlement expansion of western juniper is the sharp rise in pollen in the mid-1900’s, which Mehringer (1987) detected in lake sediment cores. The presence of old stumps and logs, which can persist on a site for hundreds of years in this semi- The Little Ice Age, 700–150 years BP, was the wettest and coolest period during the last half of the Holocene. Increased grass cover during this period (Wigand et al. 1995) probably supported higher fire frequencies (Gruell 1999, Miller and Rose 1999), which limited woodland distribution and abundance (Wigand 1987, Miller and Wigand 1994). The abundance of juniper pollen has gradually increased since 1500 A.D., fluctuating in the early 1800’s and sharply increasing in the mid-1900’s (Mehringer 1987). Since the end of the Little Ice Age around 1850, annual temperatures have been slowly but steadily rising (Ghil and Vautgard 1991). Relict juniper woodlands, tree age chronology data, down and dead trees and stumps, and historic documents (i.e., surveys) generally indicate that presettlement western juniper trees were typically confined to rocky ridges, low sagebrush (Artemisia arbuscula) flats, and pumice soils where fine fuels were too low in abundance to carry fire (Burkhardt and Tisdale 1976, Vasek and Thorne 1977, Holmes et al. 1986, Miller and Rose 1995, Waichler et al. 2001). The physiognomy of most stands was sparse and savanna-like (less than 10 percent tree canopy cover) on the rocky shallow soils and open-canopy woodlands (10–25 percent tree canopy cover) in the pumice region. 3The oldest western juniper aged to date is 1600 years old, located on Horse Ridge, Oregon. 4 EOARC Eastern Oregon Agricultural Research Station, Burns, Oregon, jointly operated by Oregon State University and USDA Agricultural Research Service. 5An ecological province is a subdivision of a region having a distinctive combination of geological features and ecological sites (Anderson et al. 1998). 6Alliance is a physiognomically uniform group of plant associations sharing one or more dominant or diagnostic species, which, as a rule, are found in the uppermost stratum of the vegetation (Grossman et al. 1998). Figure 2. Keystone Ranch east of Prineville, Oregon, in Crook County on Ochoco Creek. Majority of trees are juniper with a few ponderosa pine. The smaller trees in the foreground of Figure 2a appear to be about 10 to 25 years old, and larger trees 60 to 70 years. Photo by Stu Garrett. Figure 2a. Keystone Ranch, about 1890. Figure 2b. Keystone Ranch, 1989. BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 9 arid climate, are good indicators as to whether woodlands were present on a site prior to the 1860’s. Old stumps and logs in post-settlement western juniper woodlands associated with aspen (Populus tremuloides), riparian, and the majority of the mountain big (Artemisia tridentata var. vaseyana) sagebrush alliance6 across the High Desert and Klamath Ecological Provinces are absent or rare (Miller and Rose 1995, 1999; Miller et al. 2000; Miller and Tausch 2001; Wall et al. 2001). The strongest evidence for the post-settlement expansion of western juniper is from tree-ring chronologies (Table 2, Fig. 3). These chronologies, which describe the age composition and establishment of woodlands over time, show a rapid increase in establishment since the 1870’s (Miller and Tausch 2001, Soulé et al. 2004). In southeastern Oregon and southwestern Idaho, peak establishment in some closed woodland stands7 occurred between 1900 and 1920 (Fig. 3b) (Miller and Rose 1999, Table 2. Decadal initiation of western juniper expansion, location of study, and period of peak establishment based on tree-ring data. Sample size = number of trees sampled (from Miller and Tausch 2001). Cover type Initiation Peak Location Sample size Reference Sagebrush 1860’s 1880–1920 e OR <1,000 Gedney et al. 1999 Mountain big sagebrush 1890’s 1902–1936 Silver Lake, OR 228 Adams 1975 1870’s 1910–1940 Owyhee Mt, ID Burkhardt & Tisdale 1976 1880’s 1900–1910 Prineville, OR >1,000 Eddleman 1987 1850’s Juniper Mt & South Mt, ID >500 Johnson 2005 1850’s 1900–1920 Juniper Mt, ID >1,000 Johnson 2005 1880’s Steens Mt, OR >500 Johnson 2005 1880’s Steens Mt, OR >1,000 Miller & Rose 1995 1870’s Kiger Gorge, OR 240 Unpublished data 1870’s Hart Mt, OR Gruel 1999 1870’s 1905–1925 Paisley, OR >1,000 Miller & Rose 1999 1870’s Lava Beds NM, CA 715 Miller et al. 2003 Wyoming & low sagebrush 1880’s 1890–1910 nw CA <100 Young & Evans 1981 Low sagebrush 1870’s Paisley, OR 500 Miller & Rose 1995 Aspen 1890’s 1910–1940 se OR, ne CA, nw NV >1,000 Wall et al. 2001 EOARC unpublished data). This was a period of above-average precipitation. A similar peak was reported for woodlands of Utah juniper in Nevada (Tausch et al. 1981). A similar pattern of western juniper encroach- ment has occurred in aspen communities throughout the range of western juniper (Fig. 3d) (Miller and Rose 1995, Wall et al. 2001). In southeastern Oregon, northeastern California, and northwestern Nevada, 12 percent of the aspen stands (n = 100) measured were completely replaced by western juniper (Wall et al. 2001). These stands were identified as previously being dominated by aspen based on the high density of dead aspen logs in the understory. In addition, post-settlement western juniper was the dominant tree species in 23 percent of the stands and common to codominant in 42 percent of the aspen stands measured. Western juniper began invading aspen stands in the 1890’s, with peak establishment occurring between 1900 and 1940 (Table 2). No western juniper in these aspen stands exceeded 130 years in age. In much of its range, western juniper has increased the area it occupies by an estimated 10-fold in the past 130 years (Miller et al. 1999a) and has the potential to occupy far more area than it now does (West and Van Pelt 1986, Betancourt 7Closed stands are sites where western juniper is the dominant vegetation layer and the primary species controlling ecological processes on the site. The physiognomy of closed stands are characterized by a dominant overstory layer of mature trees and often suppressed subcanopy trees, and a shrub canopy of less than 5 percent cover. 10 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 1987, Miller et al. 2000). Most of the 9 million acres occupied by western juniper is still in transition from shrub-steppe to western juniper woodland (Miller et al. 2000) and the species continues to expand its range and increase in density (Miller and Rose 1995, 1999; Knapp and Soulé 1998; Wall et al. 2001), even in the absence of livestock grazing (Soulé et al. 2004). Factors affecting post-settlement expansion Factors most frequently attributed to the increase in both density and area of piñon and juniper are climate, the introduction of livestock, industrial increases in atmospheric CO2, and the reduced role of fire (Fig. 4). Climatic influences From 1850 to 1916, winters became milder and precipitation was greater than the current long-term average in much of the Great Basin (Antevs 1938, Wahl and Lawson 1970, LaMarche 1974, Graumlich 1987). There is some indication that woodland expansion was initiated between 1850 and 1870 in some areas prior to European settlement (Fig. 3b) (P.E. Mehringer, Department Figure 3a. Lava Beds National Monument in northern California. Figure 3c. Chewaucan River basin in the Paisley Ranger District, Fremont National Forest in south-central Oregon. Figure 3d. When juniper encroachment began (based on the three oldest western juniper in the stand) in 96 aspen stands in southeastern Oregon, northeastern California, and northwestern Nevada. Chronologies are based on tree ring data. � Figure 3b . Combined chronologies from 42 miles of transects in Steens Mountain south- eastern Oregon, and South Mountain and Juniper Mountain in southwestern Idaho. of Anthropology and Geology, Washington State University, personal communication; Johnson 2005). However, expansion across the majority of areas sampled occurred in the late 1800’s (Table 2; Fig. 3a, c, d). Annual tree ring growth in western juniper is strongly related to local climatic conditions (Pohl et al. 2002). Soulé et al. (2004) reported that western juniper annual ring growth across five sites in eastern Oregon were above-average from the late 1800’s through the early 1900’s. This wet period coincides with post-settlement establishment and the peak period of woodland establishment for closed stands (Table 2). Wet, mild conditions promote vigorous growth in western juniper (Fritts and Wu 1986, Holmes et al. 1986). Livestock grazing Introduction of livestock in the 1860’s and the large increase of animals from the 1870’s through the early 1900’s (Oliphant 1968, Miller et al. 1994) coincide with the initial expansion of western juniper woodlands. Season-long grazing by the large numbers of domestic livestock during this period is believed to have reduced fine fuel loads, thus contributing to a significantly reduced role of fire in the northern Great Basin (Burkhardt and Tisdale BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 11 1976, Miller and Rose 1999, Miller and Tausch 2001). Fire occurrence and fire size declined dramatically in the late 1800’s. Miller and Rose (1999) reported a large decrease in fire occurrence in southeastern Oregon shortly after large numbers of livestock were introduced in the late 1860’s (Fig. 3c). The lack of fire and decreased competition from herbaceous species probably contributed to an increase in shrub density and cover, thus providing a greater number of safe sites for western juniper establishment (Miller and Rose 1995, 1999). The role of livestock as a mechanism for western juniper seed dispersal appears to be minimal (Burkhardt and Tisdale 1976). Atmospheric CO2 Rising levels of atmospheric CO2 seem to have enhanced the increase in woody species throughout the West (Johnson et al. 1993, Knapp and Soulé 1999b). Increases in atmospheric CO2 levels do not coincide with the initial increase or peak periods of western juniper establishment (Table 2). However, elevated atmospheric CO2 during the last half of the 20th century may be an important contributing factor accelerating tree canopy expansion and establishment in some areas (Knapp and Soulé 1996, 1998, 1999b; Soulé et al. 2004). Annual sapwood growth in western juniper has been significantly greater since the 1950’s compared to prior years (Knapp et al. 2001a, b), suggesting accelerated growth. The authors suggest elevated CO2 levels may have a drought-ameliorating effect by increasing water use efficiency. Fire Fire is considered to have been the most important factor in limiting conifer encroachment into shrub-grassland communities in the Intermountain West prior to European settlement (West 1999, Miller and Tausch 2001). However, only a few studies have documented fire regimes across shrub-steppe communities and woodlands throughout this region. Unlike ponderosa pine, junipers seldom repeatedly scar in response to fire; thus it is difficult to determine or describe presettlement fire regimes across many shrub- steppe and woodland communities. Fire scars on western juniper are occasionally found, but most presettlement trees do not grow on sites representative of more productive deeper-soil sites, which now support expanding post- settlement woodlands. Old-growth western juniper is commonly found on relatively fire-safe sites (i.e., rocky surfaces, shallow soils, limited effective moisture) characterized by low production with limited fine fuels (Burkhardt and Tisdale 1976; Vasek and Thorne 1977; Young and Evans 1981; Holmes et al. 1986; Miller and Rose 1995, 1999). Evidence that woodland expansion was limited by fire events prior to settlement includes: (1) sites supporting old-growth trees are usually fuel-limited, (2) most young stands occupy the more productive communities where fine fuel loads could carry a fire, and (3) the time sequence of woodland expansion is synchronous with the decline in fire occurrence. In productive mountain big sagebrush plant associations in the Northwest, such as those characterized by Idaho fescue (Festuca idahoensis), MFRIs (mean fire return intervals8) typically ranged between 10 to 25 years (Table 3) and large fires every 38 years. Potential natural vegetation resulting from these short fire return intervals would probably have been dominated by Idaho fescue with an open, scattered canopy of mountain big sagebrush. MFRIs were determined from fire scars collected on ponderosa pine or Douglas-fir (Pseudotsuga menziesii) growing in or adjacent to mountain big sagebrush communities (Fig. 5). In two studies, where presettlement MFRIs were 12–15 years, fire-free intervals varied between 3 and 29 years (Gruell 1999, Miller and Rose 1999). However, fire occurrences were less frequent in the more arid plant associations in the mountain big sagebrush alliance. Based on tree growth, age structure, and the scarcity of presettlement trees or the presence of large dead wood, the maximum MFRI in the mountain big sagebrush/Thurber needlegrass (Stipa thurberiana) plant association was probably 50–70 years. Fire return intervals up to 50 years were probably adequate to limit western juniper encroachment into the mountain big sagebrush alliance (Burkhardt and Tisdale 1976, Miller and Rose 1999). The probability that western juniper will establish and successfully mature greatly increases Figure 4. Conceptual model illustrating factors influencing the expansion of western juniper since the late 1800’s and throughout the 1900’s (Miller and Tausch 2001). 8MFRI = arithmetic average of the number of years between fire events determined for a designated area during a designated time period. 12 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER Figure 5. Ponderosa pine with over a dozen pre-1900 fire scars in a densely wooded community at Lava Beds National Monument, northeastern California. The pre-1900 plant community was an open ponderosa pine stand with an understory dominated by Idaho fescue. The mean fire return interval was between 8 t� y of mountain mahogany soon to be overtaken by western juniper (greater than 100 trees/acre). Table 3. Presettlement mean fire-return intervals (MFRI=average number of years between fire events) in sagebrush and aspen cover types associated with western juniper. Change indicates the decade when the MFRI increased (Miller and Tausch 2001). Plant MFRI Decade of Association (yrs) change Location Reference Mountain big sagebrush/ Idaho fescue 20 Late 1800’s Lava Beds National Monument, CA Martin & Johnson 1979 11 1910 Owhyee Mt, ID Burkhardt & Tisdale 1976 Idaho fescue 12–15 1870’s Chewaucan–Paisley, OR Miller & Rose 1999 Idaho fescue 13 Late 1800’s Hart Mt, OR Gruell 1999 Idaho fescue 13–15 1870’s Pine Mt, OR Miller et al. 2001 Idaho fescue 16 1860’s Summer & Silver Lake, OR1 Miller et al. 2001 Idaho fescue 12 1880’s Fort Rock, OR1 Miller et al. 2001 Idaho fescue 17 1870’s Devils Garden, CA Miller et al. 2001 Idaho fescue 6 1870’s Silver Lake, OR1 Miller et al. 2001 Idaho fescue 16.5 Silver Lake northwest, OR1 Miller et al. 2001 Idaho fescue, with some ponderosa pine 8–10 1870 Lava Beds National Monument, CA Miller et al. 2003 Western Juniper/ western needlegrass 150+ Lava Beds National Monument, CA Miller et al. 2003 Low sagebrush/ Not reported 1860 northwestern CA Young & Evans 1981 Sandberg bluegrass 138 1870 Chewaucan–Paisley, OR Miller & Rose 1999 Aspen 602 1870’s eastern OR, northeast CA, & Wall et al. 2001 northwest NV 1General location of stand studied. 2Stand replacement interval based on aspen age structure, disturbance may not be fire. BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 13 as MFRIs become more than 70 years. A fire free period of more than 70 years will also increase the potential for leaving large-diameter charred wood consisting of heartwood (that can persist on the site for more than 100 years), resulting from the development of mature trees on the site. Small trees consisting of mostly sapwood usually decompose in several decades. In northern California, a plant community identified as a western juniper/ mountain big sagebrush/western needlegrass (Stipa occidentalis) plant association burned in 1856 (Miller et al. 2003). Intact charred wood and fire-killed trees are still present on the site. On Juniper Mountain in eastern Oregon, trees killed by fire in 1717 still persist in the stand. A number of studies in mountain big sagebrush communities in the Intermountain West have reported significant declines in fire events since the late 1800’s (Table 3) (Miller and Tausch 2001). Several studies have shown a close relationship between the early expansion of western juniper in the late 1800’s and the sudden decline in fire occurrences in the mountain big sagebrush alliance (Figs. 3c, 6) (Miller and Rose 1999; Miller et al. 2001, 2003). MFRIs reported for the low sagebrush/Sandberg bluegrass (Poa sandbergii) association (Table 3) were considerably longer than for neighboring mountain big sagebrush communities (Young and Evans 1981, Miller and Rose 1999). Fire-free periods of 90 (Young and Evans 1981) and 138 years (Miller and Rose 1999) were reported for this plant association in northern California and south-central Oregon and it is not unlikely that fire-free periods exceeded 150 years for some sites. This plant association can be characterized by a low density of widely scattered old-growth western juniper, which suggests infrequent fires. Tree growth rates are relatively slow with the average age of a 3-m-tall tree ranging from 75 to 90 years. Fire return intervals of 100 to 150 years would probably be adequate to maintain a low-density stand of widely scattered trees in this plant association. In the absence of fire, western juniper will slowly increase in density in this plant association. Fire also played an important role in the maintenance of healthy mixed-age aspen stands in the semi-arid West (Bartos and Campbell 1998). In the northwestern Great Basin, Wall et al. (2001) reported that encroachment of western juniper into these communities began around 1900. Fire was probably the primary disturbance factor limiting western juniper invasion into these aspen communities. Based on the composition and distribution of age of aspen in two large stands in southeastern Oregon, presettlement mean disturbance intervals were determined to be 16 years within portions of these stands. Wall et al. (2001) estimated that total stand replacement in these two aspen communities occurred around 60–100 year intervals. Climate and fire In eastern Oregon, large presettlement fires in sagebrush-steppe communities were usually preceded by at least one year of above-average growing conditions (Miller and Rose 1999). In these semi-arid ecosystems, fuels are often limited in abundance and continuity. A series of wet years allows fuels to accumulate and become more contiguous. Wetter than average conditions in the late 1800’s would have resulted in the accumulation of fine fuels. However, high livestock stocking rates and season-long or heavy grazing during this period reduced fine fuel accumulations and thus significantly decreased the potential for fire (Burkhardt and Tisdale 1969, Miller and Rose 1999). The combination of reduced fire occurrences (Miller and Tausch 2001) and optimal climatic conditions for conifer establishment (Fritts 1974, Fritts and Wu 1986) at the turn of the century were probably the two dominant factors that initiated post-settlement western juniper expansion. Figure 6. Tree densities and age chronologies of western juniper and date of last fire for three cinder buttes sampled on the Lava Beds National Monument, northern California. The lack of large dead juniper wood on these sites suggests mature juniper did not occupy these sites prior to 1900. Potential natural vegetation is Idaho fescue grassland with widely scattered ponderosa pine and mountain big sagebrush and mean fire return interval prior to 1900 was 8–10 years (Miller et al. 2003). 14 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER Western Juniper Varieties Morphology of western juniper W estern juniper is submonoecious9. Trees are pyramidal to round in shape and typically reach 13–32 ft in height at maturity, but will occasionally reach 65 ft in height. Trunks are usually composed of a single erect stem 13.7–27.6 inches in diameter (maximum of 74.8 inches) (Vasek 1966, Cronquist et al. 1972). The largest reported western juniper, located in the Lost Forest in northern Lake County, Oregon, is 78 ft tall with a trunk circumference of 19 ft. Bark is typically gray but can turn reddish in some old trees (more than 300 years). Mature western juniper leaves are 0.039–0.118 inches in length, compressed to the stem and overlapping the next leaf (Fig. 7). Leaves occur as opposite pairs or in whorls of three. Each scale has a conspicuous resin gland on the dorsal side of the leaf (Fig. 7). In contrast, juvenile leaves are not compressed to the stem and are spiny tipped. Seed bearing can begin as early as 10–20 years of age, but significant fruit production usually starts at 50–70 years of age (Miller and Rose 1995). The yellowish-brown male cones are 0.12–0.16 inches long and occur at the end of a branchlet (Fig. 7). Male cones develop during the late summer and early fall and shed their pollen early the following spring (Vasek 1966). Female cones are bluish to bluish-black at maturity, covered with a resinous pulp, and contain two to three seeds (occasionally one seed). These cones begin development in early spring, attain full size the first summer and mature during the second summer. Female cones persist on trees for nearly 2 years. Morphological characteristics of western juniper and Utah juniper are usually distinct. Utah juniper lacks the resin gland on the back of the leaf scale and the female cones are brownish with a mealy to fibrous covering (Cronquist et al. 1972). However, in northwestern Nevada, where the distribution of the two species overlap, differences become less apparent due to hybridization (Vasek 1966, Terry et al. 2000). Morphology of Sierra juniper Sierra juniper, a variety of western juniper, is located primarily south and southeast of the range of western juniper. Sierra juniper is distinguished from western juniper in that it is mostly dioecious10, has reddish-brown bark rather than gray bark (Cronquist et al. 1972), can attain a larger size at maturity, and grows in different plant associations, higher elevations, and different climatic conditions. However, the bark on older western juniper trees also often attains a reddish color. Charlet (1996) reported that Sierra juniper material collected in Nevada was distinct from western juniper and suggested a taxonomic reevaluation of the variety. Further, the largest recorded Sierra juniper is 83 ft tall and 40 ft in trunk circumference, located in the Stanislaus National Forest, east-central California. Seed Production, Dissemination, Germination, and Establishment Although seed production occurs in most years (Sowder and Mowat 1958), western juniper seed- crop production is highly variable across sites and years. The environmental variables that trigger the initiation of male and female cones have not been identified. Research on factors influencing seed production and seedling establishment will be required to predict annual seed-crop production and better understand woodland dynamics (Chambers et al. 1999b). Life History and Biology 9Submonoecious—male and female cones are borne on the same individual; however, some trees will produce pre- dominantly male or female cones. 10Dioecious—male and female cones are borne on different individuals. Figure 7. Western juniper male cones and foliage showing white dried resin exuded from the resin gland located on the dorsal side of the leaf scale. BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 15 Seed dispersal of western juniper occurs through gravity, overland flow, and by animal transport. At least 12 species of birds feed on the fruits and as a group are the most important disseminator of western juniper seed (Fig. 8) (Gabrielson and Jewett 1940, Maser and Gashwiler 1978). American robins (Turdus migratorius) and Townsend’s solitarie (Myadestes townsendi) often winter in woodlands and consume the female cones (Lederer 1977, Podder and Lederer 1982, Reinkensmeyer 2000). Townsend’s solitarie can consume over 80 female cones/day. Mountain bluebirds (Sialia currucoides), cedar waxwings (Bombycilla cedrorum), and Steller’s (Cyanocitta stelleri) and western scrub- jays (Aphelocoma californica) have been observed consuming female cones. Most birds have limited gut-retention times and fly short distances to perch and process the fruit, thus limiting the distance of most seed dispersal (Schupp 1993, Chambers et al. 1999b). After feeding on Ashe juniper (Juniperus ashei) fruits, American robins flew an average distance of 145 ft to a post-foraging perch, which could be another tree, shrub, or on the ground beneath a woody canopy (Chavez-Rameriz and Slack 1994). In Spain, Santos and Telleria (1994) reported birds feeding on juniper berries were more likely to visit large stands of trees and less likely to feed in small isolated juniper stands. Coyotes (Canis latrans), cottontail rabbits (Sylvilagus sp.), and several rodent species also consume and scatter western juniper seeds (Chambers et al. 1999a). Mule deer (Odocoileus hemionus) have also been observed to eat western juniper fruits during winter months when preferred foods are unavailable (Leckenby 1968, Trout and Thiessen 1968). However, successful establishment of seed dispersed by mammals is probably limited, because seeds are deposited at high densities in microsites where establishment is poor (Schuppe 1993; Schuppe et al. 1997; Chambers et al. 1999a, b). Western juniper seeds are initially dormant immediately following seed drop (Johnson and Alexander 1974). Germination potential is greatly enhanced by prolonged cool-moist stratification, which is cumulative from year to year (Young et al. 1988). This suggests germination of a particular seed crop may span several years. Seeds of several other juniper species are also long lived with an extended dormancy, resulting in highly persistent seed banks (Chambers et al. 1999a, b). Little information is available on percent seedling survival or climatic conditions that influence seedling establishment. However, two studies indicate survival rates for western juniper seedlings are high (Burkhardt and Tisdale 1976, Soulé et al. 2004). In addition, Soulé et al. (2004) reported that wet cool summers may lower western juniper seedling mortality. Much of successful western juniper seedling establishment occurs beneath shrubs (Burkhardt and Tisdale 1976, Eddleman 1987, Miller and Rose 1995, Soulé and Knapp 2000, Soulé et al. 2004). This may be attributed to a disproportionate amount of seed dropped by perching birds and/or more favorable growing conditions beneath the shrubs compared to the interspace. Growth rates of young trees beneath mountain big sagebrush canopies were greater (1.34 inches/year) than in the interspace (0.95 inches/year) (Miller and Rose 1995). Compared to bare soils in the interspace, soils beneath a sagebrush canopy can have nearly twice the moisture content and nitrification (Roberts and Jones 2000). Cooler temperatures and higher relative humidity beneath the sagebrush canopy also provide more favorable growing conditions for juvenile foliage, which has poorer stomatal control and lower water use efficiency than adult foliage (Miller et al. 1992). Safe microsites that modify the environment may be responsible for greater seedling survival rates under relatively dry conditions. Many seeds also germinate beneath the tree canopy; however, survival and growth rates are low because of high intraspecific competition from the overstory tree. No evidence suggests competition from associated shrubs or herbs limits the success of western juniper seedling establishment (Burkhardt and Tisdale 1976, Miller et al. 2000). However, an increase in bare ground and mature western juniper cover was negatively correlated with successful tree establishment across the mountain big sagebrush alliance in Oregon and California (Miller et al. 2000). This may be the result of intraspecific competition from overstory trees and limited safe sites for seedling establishment as woodlands approach late successional stages. Figure 8. Mountain bluebirds consuming juniper berries early in the spring. Photo by Rick Vetter. 16 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER Roots During the first 10 years of growth, western juniper directs most of its effort into developing a taproot with only limited lateral root development (Kramer 1990). After 10 years, lateral root develop- ment increases, accounting for about 65 percent of the root biomass in trees 30–35 years old. Root: shoot ratios for young trees vary from 0.55 to 0.76 (Miller et al. 1990). Taproot development declines as trees begin to lose their juvenile foliage (Young et al. 1984) on shallow soils. However, taproots have been observed on some sapling and mature trees growing in deep soils (Fig. 9). Trees develop a massive fine root mat system with age. Young et al. (1984) reported most of tree roots were located in the upper 30 inches of the soil profile in a soil that is 40 inches deep. Large lateral roots commonly extend a distance that equals the height of the tree, but in some cases can extend as much as three times tree height. Growth Following germination, aboveground growth is relatively slow, averaging 1.18–1.58 inches/year in height for the first 10 years and increasing to 3.54–6.57 inches/year for older trees up to 100 years old (EOARC, unpublished data). Root development appears to far exceed aboveground growth during early development. Leaf canopy development remains relatively slow during the first 35–45 years. At the age of 45–50 years, the rate of tree canopy development increases (Fig. 10). Current year’s sapwood development begins during the spring and usually ends in early to late August, depending on the site and annual precipitation (Peter 1977). In wet years, ring growth can continue through August. Branchlet and leaf growth are greatest during June and July (Miller et al. 1992). Western juniper typically approaches its maximum height at 80–100 years of age across its geographic range (EOARC, unpublished data). Depending on site potential and competition from other trees, mean height of western juniper at 80 years of age will vary from 19.7 to 49.2 ft (Gedney et al. 1999). Site index curves that describe tree age and height relationships for western juniper varies widely across sites. Height for trees 80 years old at breast height (approximately 4.25 ft above the ground) ranged from 15 ft on scablands to 35 ft on sites associated with ponderosa pine (Sauerwein 1982). In central Oregon, mean height growth rate varied from 3.5 to 6.6 inches/year for dominant trees (Eddleman 1987). Several authors have developed regressions estimating western juniper leaf area, leaf biomass, and total standing crop using tree basal and sapwood areas (Gholz 1980, Miller et al. 1987). In a fully developed woodland in eastern Oregon, Figure 9. Juniper roots in deep soils can have both large lateral and fine roots in the upper 24 inches, and a deep taproot. Taproots are often missing in shallow soils. Figure 10. Canopy area and age for individual western juniper trees showing an increase in canopy expansion for trees over 45 years of age (from Miller and Tausch 2001). Photo by Hugh Barrett. BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 17 Figure 11. Leaf scales on the top (abaxial) [A and B at 150x] and lower (adaxial) [C at 250x] surfaces of western juniper. Most of the stomata are located on the adaxial leaf surface and a few are at the base of the abaxial surface, which is covered by the lower overlapping scale (Miller and Schultz 1987). Gohlz (1980) estimated foliage biomass of 4,550 lb/ acre, total standing crop biomass of 23,300 lb/acre, and a leaf area index (LAI) of 2 (2 units of leaf area to 1 unit of ground area) for stands with a mean density of 608 trees/acre). Primary production was 1,200 lb/acre, about half that of adjacent ponderosa pine communities and 10 percent of Douglas-fir communities in the Cascades. Leaf Morphology Young western juniper trees (usually less than 25 years) have needle-like leaves, which are different than leaves on older trees (De Laubenfels 1953). Leaves on mature trees are triangular with minutely serrated margins and have a low surface- to-volume ratio (Fig. 11a) (Miller and Schultz 1987). Leaf margins are slightly cupped, which seals one leaf against the other and forms a chain-like cylinder. The leaf epidermis is heavily cuticularized (waxy covering on the leaf surface), which greatly reduces water loss through the leaf surface. Most of stomates are located on the protected side of the leaf surface facing the stem (Fig. 11c). Stomates on the outer surface are located at the base of the leaf and are covered by the adjacent subtending leaf (Fig. 11b). The leaf morphology of western juniper allows for maximum drought avoidance through low leaf area, low surface-volume ratios, thick cuticle layer, and protected stomata. Mean maximum leaf conductance (transpiration, measured as inches of water/second passing through the leaf surface to the atmosphere) per unit leaf area was lower (0.03–0.05 inches/second) than values reported for several other conifer species (0.05–0.16 inches/second) (Miller and Schultz 1987). Water Use and CO2 Assimilation Ecophysiological (Moore et al. 1999) and morphological (Miller and Shultz 1987) adaptations allow western juniper to tolerate relatively large environmental changes. In addition, allocation of resources in young trees partially explains the species ability to compete successfully with other native species (Miller et al. 1990). By reducing allocation of resources to branches and trunks, juvenile and small adult western juniper allocate larger portions of dry mass to foliage and roots to optimize photosynthetic capacity and uptake of water and nutrients than mature trees. 18 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER During winter, cold soil temperatures limit water use by western juniper (Miller and Schultz 1987). As soil temperatures drop below 40˚F, water uptake at the root surface significantly decreases. As soil temperatures increase in March, trees begin to actively transpire and grow. In warmer climates, such as in the John Day or Mazama ecological provinces, more moderate soil temperatures may allow western juniper to transpire water during any month in the winter. In central Oregon, Jeppesen (1977) reported greater winter soil water loss at 20- inch soil depth in woodlands compared to thinned stands. Leaf conductance is strongly influenced by soil temperature and vapor pressure deficit during the spring. During the summer, soil water availability and vapor pressure gradient11 are the primary factors influencing water use and CO2 assimilation in western juniper (Miller et al. 1992, Angell and Miller 1994, Moore et al. 1999). Stomata closed when stem water potentials decreased to –2.0 MPa (mega pascals) (Miller and Schultz 1987). In a dry year, the greatest amounts of water were transpired during April and May, compared to June and July in a wetter-than-average year (Fig. 12) (Angell and Miller 1994). In a moderately stocked stand of 30 trees/acre and 1.6 LAI, the water-use model predicted western juniper would extract 2 inches of soil water in a dry year and 5.6 inches in a wet year. These predictions suggest soil water depletion rates will significantly shorten the growing season on the site, a point confirmed by Bates et al. (2000). They reported the growing season of the understory was shortened by as much as 6 weeks in uncut western juniper stands, compared to adjacent cut stands. Juveniles with the awl-shaped leaves have higher leaf conductance, transpiration, and greater total CO2 assimilation per unit of leaf weight during the growing season than sapling and mature trees (Miller et al. 1992). The change from juvenile to mature foliage reduces the amount of carbon assimilated per unit leaf area but also reduces the amount of water lost to transpiration by 40 percent (Miller et al. 1993). Insects Artichoke-like galls located on the branchlets of western juniper (frequently misidentified as reproductive structures) are formed by midge larvae Walshomyia spp. (Purrington and Purrington 1995)(Fig. 13). Moth larvae Heinrichiesa sanpetella were found to inhabitat 40 percent of these galls, over-wintering and pupating there in early spring. Other moth caterpillars that feed on western juniper are the sequoia sphinx (Semiothisa spp., Sphinx sequoiae), cedar streak (Lithophane logior), and Mitoura grynes barryi (Miller 1995). Other insects known to feed on western juniper include long- horned beetle (Styloxus bicolor), juniper bark beetle (Phloeosinus serratus), round-head borers (Callidium califonrnicum and C. juniperi), wood-boring beetle (Melonophila miranda), and grasshoppers (Melanoplus sp.). The western juniper bark beetle is typically attracted to wounded or felled trees (personal communication, Jane L. Hayes, USDA US Forest Service Research Station, La Grande, OR). Insect attacks usually do not result in the killing of live trees, however in the 1920’s and 1930’s in addition to drought, areas of western juniper were killed by insects in central Oregon (Furniss and Carolin 1977). Current work has identified 25 species of bark and woodboring beetles feeding on western juniper (Hayes, unpublished work in progress). During the grasshopper outbreaks near the John Day Fossil Beds in eastern Oregon in the late 1970’s, the tops of some western juniper trees were nearly totally defoliated. On dead or dying juniper, round- Figure 12. Modeled daily transpiration for western juniper during a drought year (1990) and wet year (1984) (from Angell and Miller 1994). 11 The water vapor concentration gradient from inside the stomata to the open atmosphere. BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 19 head borers or long-head beetles (Creambydidae spp.) deposit eggs in the bark (Swan 1996). Upon hatching, the larvae bore into the wood, deriving nourishment from the soluble carbohydrates in wood particles and/or fungal tissue. Associated Nonvascular Plants Mistletoe Mistletoes that commonly infect western juniper are juniper mistletoe (Phoradendron juniperinum), and dense mistletoe (P. densum) (Geils et al. 2002). Juniper mistletoe is the primary species found on western juniper and is the most widespread mistletoe infecting juniper species throughout the West. Juniper mistletoe has leafless stems and pinkish-white colored berries about 0.16 inches in diameter. Dense mistletoe occurs in the southwestern range of western juniper. This species has white- to straw-colored berries 0.16 inches in diameter and is easily differentiated from juniper mistletoe in that it has leaves. Birds feed on the fleshy mistletoe berries and are the primary dispersers of the sticky seeds. Birds that commonly feed on the berries include American robins, Townsend’s solitaires, cedar waxwings, flycatchers, mountain bluebirds, and thrushes (Sutton 1951). The mistletoe foliage is high in nutritional value (Urness 1969). Juniper mistletoe usually occurs in a patchy distribution with only a few heavily infected trees. Although it can stress the tree by absorbing relatively large amounts of water and nitrogen, the tree is rarely killed. Mosses, fungi, and lichens Limited information is available on the ecology and life histories of nonvascular plants associated with western juniper. We also know very little about the effects of western juniper expansion or removal on biological crusts. Tortula ruralis is commonly associated with mature western juniper trees where it grows beneath the tree canopies. Four species of wood-rotting fungi, Antrodia juniperina, Pyrofomes demidoffii, Diplomitoporous rimosus, and Phellinus texanus, may cause heart rot in western juniper (Knapp and Soulé 1999a). These fungi Figure 13. Artichoke-like gall located on the branchlet of a western juniper (frequently misidentified as a reproductive structure) is formed by midge larvae Walshomyia species. typically enter openings in the heartwood or in dead sapwood. Knapp and Soulé (1999a) reported a widespread occurrence of heart rot (suspected to be Antrodia juniperina) between 1730 and 1749 in western juniper across eastern Oregon and northeastern California. Heartwood rot is most commonly found in trees more than 150 years old (EOARC, unpublished data). Western juniper roots can be infected with symbiotic fungi mycorrhizae (Trappe 1981). Roberts and Jones (2000) reported higher levels of vesicular-arbuscular mycorrhiza fungi under western juniper canopies than under sagebrush or grass canopies. Two species of foliose lichens commonly associated with western juniper are Letharia columbiana and L. vulpina. These lichens are brilliant fluorescent yellow-green or chartreuse in color, and highly branched. Both species are nearly identical in form except that L. vulpina lacks the small disk-like fruiting bodies (soredia). Both species can occur on a single tree and are often most abundandant on dead, barkless branches or snags. 20 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER Soils W estern juniper grows on a wide variety of parent materials and soils (Driscoll 1964a). Parent materials are derived from aeolian (e.g., pumice sands), sedimentary, and igneous (e.g. rhyolite, andesite, basalt) sources. Soil textures can range from heavy clays to sandy soils. Soil depths vary from bare rock to more than 3 ft, and soil temperature regimes are mesic and frigid12 (limited cryic13). Western juniper roots are able to penetrate fractured basalt bedrock, allowing it to occupy rock outcrops and soils less than 12 inches deep. The wide range of soils has a large impact on potential erosion, woodland development, overstory-understory interactions, and response to disturbance across the range of western juniper. Western Juniper Communities Numerous classifications have been proposed for western juniper plant associations14 and communities (Driscoll 1964a, b; Hall 1978; Hopkins 1979; Johnson and Clausnitzer 1992). In addition, the Natural Resource Conservation Service (NRCS) is developing ecological site classification with western juniper in the plant association name. However, it is not always clear in these classifications if western juniper was a part of these communities prior to European settlement or has encroached since settlement. Western juniper communities may be separated into presettlement (old-growth) or post-settlement (expansion) communities. We suggest 1870 as a cut-off to separate the two age classes. The date separating pre- and post-settlement is based on the approximate time when fire regimes changed (1870’s) and the arrival of livestock in eastern Oregon, southwestern Idaho, and northeastern California (late 1860’s) (Oliphant 1968, Miller et al. 1994, 1999a). Common associated diagnostic species Western juniper is associated with a wide range of plant communities including forest, riparian, aspen, and shrub-steppe. Within these community groups, it has actively expanded into numerous plant alliances and associations defined by ponderosa pine, aspen, willow (Salix spp.), mountain big sagebrush (Artemisia tridentata ssp. vaseyana), Wyoming big sagebrush (A.t. ssp. wyomingensis), basin big sagebrush (A.t. ssp. tridentata), low sagebrush, stiff sagebrush (A. rigida), bitterbrush (Purshia tridentata), and mountain mahogany (Cercocarpus ledifolius). Scattered stands of western 12Mesic: a soil temperature regime that has mean annual soil temperature of 46-59° F, and more than 43° F difference between mean summer and mean winter soil temperatures at 20 inches below the surface, or at a densic, lithic, or paralithic contact, whichever is shallower. Frigid: a soil temperature regime with warmer summer temperatures than cryic, with mean annual soil temperatures less than 46° F, and more than 43° F difference between mean summer and mean winter soil temperatures at 20 inches below the surface, or at a densic, lithic, or paralithic contact, whichever is shallower. 13Cryic: a soil temperature regime that has mean annual soil temperature of 0° F but lower than 8° F difference between mean summer and mean winter soil temperatures at 20 inches below the surface at 20 inches below the surface, or at a densic, lithic, or paralithic contact, whichever is shallower. 14Plant association is defined by the dominant/diagnostic overstory and understory species (e.g., mountain big sagebrush/ Idaho fescue plant association). juniper in Siskiyou, Trinity, Shasta, and west Lassen counties in California are associated with Oregon white oak (Quercus garyana), buckbrush ceanothus (Ceanothus cuneatus), and several other conifer species (Vasek and Thorne 1977). Common understory diagnostic species are Columbia needlegrass (Stipa columbiana), needle-and-thread needlegrass (S. comata), western needlegrass, Thurber needlegrass, Idaho fescue, bluebunch wheatgrass (Agropyron spicatum), Ross sedge (Carex rosii), and Sandberg bluegrass. Old-growth (pre-settlement) communities It is estimated that less than 10 percent of existing western juniper individuals established prior to the 1870’s (USDI–BLM 1990, Miller et al. 1999a, Johnson 2005). However, the proportion of old-growth varies across ecological provinces and few presettlement stands have been inventoried or separated out from post-settlement stands. Old-growth western juniper is associated with a variety of soils, landforms, and plant associations throughout its range. Old- growth communities typically occupy rock outcrops and soils that are shallow, rocky, and often high in clay or sand, and fine-textured sedimentary soils. Examples include the shallow claypan soils occupied by low sagebrush and Sandberg bluegrass common in the High Desert, Klamath, and Humboldt ecological provinces. However, it is also associated with the ashy-sandy-pumice soils associated with mountain big sagebrush, western needlegrass, and needle-and-thread grass in the Mazama Province and sedimentary soils in the John Day province. The common factor linking this wide array of soils and landforms that support old-growth stands Ecology BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 21 Figure 14. An 800-year-old western juniper tree with spreading rounded top and large lower limbs, Connelly Hills, south-central Oregon. Figure 15. Old-growth western juniper/low sagebrush/Sandberg bluegrass plant association, occupying a shallow heavy clay soil on the Modoc Plateau in northern California. Figure 16. Bark characteristics of three different aged trees: At 75 years, bark is thin and flaky; at 152 years, bark layer is thickening and beginning to develop vertical furrows; and at 270 years, bark is thick,fibrous with well-developed furrows. Figure 16a. A 75-year-old tree. Figure 16b. A 152-year-old tree. Figure 16c. A 270-year-old tree. is their low production potential, which limits the accumulation of fuels. Thus, fire events were typically limited to one or several trees, or stand replacement, and mixed-severity fire events were infrequent (more than 150 years). For a definition of old-growth western juniper woodland, see Appendix 1. Single tree perspective Old-growth is a relative term, and has been based on morphological characteristics, actual age, or general period of establishment (pre- and post-settlement). As trees age they change morphologically. Compared to younger trees, old trees have approached their maximum size, height growth has ceased, and the tree crowns may be in various stages of decline. As trees mature, their inverted-cone-shaped canopy becomes increasingly unsymmetrical in appearance with rounded tops and spreading canopies that may become sparse and contain dead limbs or spike tops (Figs. 14, 15). In addition, the bark on the trunk becomes deeply furrowed, fibrous (Fig. 16), and reddish in color. Bark on trees less than 150 years is scaly and furrows are shallow or lacking. Branches near the base may be very large (more common in open stands), and branches are covered with bright green arboreal fruticose lichens. The cambium layer (live wood tissue) may also die around portions of the tree trunk, leaving only a narrow strip connected to a single live branch. An additional characteristic that helps distinguish older trees is limited terminal leader growth on branches in the upper 25 percent of the tree canopy. Younger trees, between 80 to 22 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 130 years, typically have terminal branch leader growth ranging between 2 to 6 inches in the upper portion of the canopy. Many of these traits that separate old and young trees usually begin to develop at 150 (+ 30) years of age but can vary across different sites. For example, on Green Mountain in northern Lake County, Oregon, trees older than 200 years still retain symmetrical inverted-cone-shaped canopies. However, terminal and lateral leader growth in the upper canopy was less than 0.5 inches and vertical furrows in the bark were present. Community perspective Old-growth is usually defined at the community level based on structural components that are easily identifiable. In the absence of major disturbance, structural characteristics that increase over time include morphological characteristics of old trees, standing and down dead, canopy decadence (dead branches), abundance of lichen in the tree canopies, hollows, and cavities. For example, a stand that is 150–400 years old generally has little standing or down dead wood. However, as stands Figure 18. Old-growth western juniper woodland in the Mazama Ecological Province. Figure 17. An old growth juniper growing on a rocky ridgetop with a young post- settlement stand in the background on Steens Mountain, Oregon. mature to over 500 years, standing and down dead wood in the community accumulates due to slow decomposition rates. Structure of the tree layer in old-growth western juniper communities (e.g., percent cover, tree density, size, etc.) will vary with site conditions and the history of past disturbances. Most old-growth stands can be separated into three general categories based on stand structure: (1) isolated stands of one to several trees located on rocky outcrops and ridges (Fig. 17); (2) low sagebrush grasslands with widely scattered trees (i.e., savannas, Fig. 15), and (3) woodlands with tree canopy cover typically less than 20 percent (Fig. 18), but occasionally exceeding 35 percent. Old-growth savannas probably account for the largest land area of old-growth but tree densities are usually very low. Trees are widely dispersed and primarily shrubs and herbs influence the tree interspace, with little interference from western juniper roots. A typical example is a low sagebrush-Sandberg bluegrass tableland (Fig. 15). Old-growth woodlands are defined as stands where tree root competition is dominant over shrubs and dominant or co- dominant over grasses in extracting resources in the tree canopy interspace. The most extensive area of old-growth woodland occurs in the aeolian sands in the Mazama Province and the northwestern edge of the High Desert Province. Composition and structure of these old-growth communities varies widely across the range of western juniper. Mean density of overstory trees varies from 80 trees/acre in the pumice region (Waichler et al. 2001), 96 trees/ acre in southwestern Idaho (Burkhardt and Tisdale 1969), and 146 trees/acre on Juniper Mountain in southeast Oregon (EOARC, unpublished data). Cover of tree canopies ranges from 10 to 60 percent on these sites. Old-growth types Three primary regions of old-growth can be differentiated by soils derived from three different parent materials: igneous, sedimentary, and aeolian soils. The three types usually differ in community structure and composition. Igneous soils. Soils derived from igneous parent materials dominate much of the landscape in the High Desert, Klamath, southwestern portion of the Snake River, and the Owyhee Plateau region in the Humboldt ecological provinces. In these provinces, old-growth western juniper typically grows in widely spaced stands on shallow, rocky, heavy clay soils, or rock outcrop, which support limited fuels to carry fire (Fig. 15, Fig. 17) (Vasek and Thorne 1977, Miller and Rose 1995, West 1999). Old-growth western juniper is estimated to make less than 10 percent of the western juniper population across this region (Miller et al. 1999a, Johnson 2005). The low sagebrush claypan communities probably account for the greatest land area occupied by old- growth western juniper across these provinces. The dominant grass in these low sagebrush BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 23 tablelands is Sandberg bluegrass, with Idaho fescue frequently growing beneath the tree canopy. These communities often occupy extensive flats with slopes typically less than 5 percent, but sometimes approaching 30 percent. The rocky, shallow heavy clay soils originate from ancient volcanic flows of basalt, andesite, and rhyolite. Although soils are shallow (less than 20 inches) on these tablelands or nearly non-existent on rock outcrops, western juniper roots are capable of penetrating fractured bedrock, usually basalt (EOARC, unpublished data). Tree canopy cover on the low sagebrush clay- pan sites is highly variable and may approach 20 percent, but usually is less than 5 percent (EOARC, unpublished data). These communities are usually rich in herbaceous species with a high diversity of forbs. Trees are usually uneven aged. On the Devils Garden in northern California, 63 percent of the presettlement trees aged varied between 200 and 500 years old. More than 30 percent were older than 500 years. The oldest trees aged to date on igneous parent material ranged between 1,000 and 1,400 years old, and were located north of Fredrick’s Butte in southeastern Deschutes County, Oregon. Low presettlement tree densities in these communities may be attributed to limited tree establishment, slow growth rates due to poor site conditions, and occasional fires. Mean fire intervals of 80–150 years were probably adequate to create a stand of widely scattered western juniper trees (Young and Evans 1981, Miller and Rose 1999), but single-tree lightening fires were more common occurrences across these western juniper-low sagebrush claypan communities. Tree densities in these communities have increased since the late 1800’s (Young and Evans 1981; Miller and Rose 1995, 1999). On deeper (more than 20 inches) igneous soils, fire limited the development of old-growth western juniper woodlands (Miller and Tausch 2001). These soils typically support mountain big sagebrush grassland communities. Mean fire intervals of 10–25 years occurred in the more productive plant associations of this alliance (Houston 1973, Burkhardt and Tisdale 1976, Martin and Johnson 1979, Miller and Rose 1999). However, Juniper Mountain, located east of Alkali Lake in Harney and Lake counties, Oregon is an exception. This site may serve as a model as to what the more productive mountain big sagebrush plant associations would look like if fire had played a minor role in the sagebrush ecosystem (Fig. 19). On the north and northeast aspects tree canopy cover ranged between 35 and 60 percent. On south and southwest aspects tree cover ranged between 25 and 40 percent. Preliminary work indicates the age of overstory trees ranged between 350 and 600 years (EOARC, unpublished data). Understory trees 3–5 ft tall were between 100 and 200 years old. Shrub cover accounted for less than one percent of the understory cover. Dominant herbaceous species were Idaho fescue on the north aspect, Thurber needlegrass on the south aspect and bluebunch wheatgrass on the west aspect. In August of 2001, a stand-replacement fire occurred on the northeast aspect of Juniper Mountain. Sedimentary soils. Little work has been conducted on old-growth western juniper on these soils. Most of these soils occupied by old-growth western juniper occur in the John Day Ecological Province with limited amounts occurring in other provinces. These soils usually support a low density of trees and a sparse understory incapable of carrying fire. The accumulation of both down and standing dead and decadent trees on many of these sites indicates the presence of very old stands. Dead trees may remain standing for hundreds of years. Old-growth also occurs on shallow rocky soils and rock outcrops in this province. Old-growth stands probably account for less than 5 percent of the western juniper woodland component in these provinces. Aeolian soils. The aeolian soil region, primarily composed of pumice sands, is located in the Mazama and northwestern portion of the High Desert ecological provinces. This region supports extensive old-growth western juniper woodlands (Fig. 18). Although not inventoried, these woodlands are estimated to account for more than 10 percent of the area occupied by western juniper woodlands in the Mazama Province. These stands are characterized by very sandy pumice soils derived from the eruption of Mount Mazama, 7,600 years ago. In the northwestern corner of the High Desert Province, soils are mixtures of wind-blown sands from Pleistocene lakebeds and pumice from Mount Figure 19. The largest and most dense old-growth western juniper woodland in the High Desert Province, located on Juniper Mountain about 4 miles southeast of Alkali Lake, Oregon. The dense north-face side of the mountain burned by a lightning-initiated stand-replacement wildfire in 2001. 24 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER Mazama and Newberry Crater. Stand structure varies across these provinces, but is generally open, with tree canopy cover typically ranging from 10 to 15 percent in central Oregon (Waichler et al. 2001). Density of live trees ranged from 15 to 25/acre, standing dead were usually less than or equal to 6/acre, and down dead ranged from 1 to 7/acres. At the Lava Beds National Monument in northern California, tree densities in old-growth stands were 50/acre. The oldest western juniper tree aged to date is 1,600 years old and is located in this pumice region on Horse Ridge in Deschutes County, Oregon (Fig. 20). Past fires in this zone were typically small, burning single to several trees within a stand. However, old fire scars on these landscapes indicate occasional, extensive stand-replacement fires did occur. In the Lava Beds National Monument in northeastern California, a large stand-replacement and mixed-severity fire occurred in old-growth stands. Western needlegrass and needle-and-thread are usually the dominant grasses characterizing stands with very old trees. However, bluebunch wheatgrass and Idaho fescue are occasionally the diagnostic understory species. Plant associations included in the old-growth western juniper type are usually very low in both forb diversity and abundance. In the Bend-Redmond area, which lies below 3,500 ft in elevation, rabbitbrush (Chrysothamnus spp.) and cheatgrass (Bromus tectorum) will dominate the understory on sites that have been overgrazed or mechanically disturbed. Woodland Succession Most plant communities occupied by post- settlement western juniper are in a transitional state, ranging from open stands of trees with a dominant understory of shrubs and herbs to mid- or late succession, where trees are beginning to dominate the site and tree canopies are approaching full coverage (Miller et al. 2000). It is important to identify the woodland transitional state in resource evaluations or inventories and when developing management strategies. The state of woodland development directly affects plant community structure, composition, seed pools, wildlife habitat, and ecological processes including hydrologic and nutrient cycles. The stage of woodland succession will also directly affect the selection of management treatment, response following treatment, follow- up management, and treatment cost. In addition, continued changes in structure and composition in developing woodlands over time should be considered when developing resource plans and setting management priorities. Identification of the woodland stage of succession We have separated woodland succession into three transitional phases (Fig. 21): • Phase I, trees are present but shrubs and herbs are the dominant vegetation that influence ecological processes (hydrologic, nutrient, and energy cycles) on the site (Fig. 22a); • Phase II, trees are codominant with shrubs and herbs and all three vegetation layers influence ecological processes on the site (Fig. 22b); • Phase III, trees are the dominant vegetation and the primary plant layer influencing ecological processes on the site (Fig. 22c, d). There are several characteristics that can be used to define the phase of woodland development, regardless of the plant association or site potential (Table 4) (Miller et al. 2000). These traits relate to the degree of western juniper dominance on the site. Early signs of western juniper domination on a site are canopy mortality of the shrubs in the interspace and the reduction of leader growth (Fig. 23) on sapling size (less than 10 ft tall) trees. Figure 20. This 1,600-year-old western juniper tree is the oldest aged to date. Located on Horse Ridge in central Oregon, east of Bend. BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 25 Rates of woodland development The rate of woodland succession from initial encroachment to fully developed woodlands is a function of the rates of tree establishment and growth. There is a high degree of variability in woodland succession rates across and within plant associations. In eastern Oregon, all three transitional phases of western juniper stand development can be observed where encroachment began in the late 1800’s (Miller and Rose 1999, Miller et al. 2000). The minimum time for the tree overstory to begin suppressing the understory is 45–50 years and to approach stand closure 70–90 years on cool wet sites (i.e. mountain big sagebrush/Idaho fescue and or Columbia needlegrass) (Fig. 24) (Johnson 2005). On warm dry sites (mountain big sagebrush/Thurber needlegrass and/or western needlegrass), stand closure occurs in 120–170 years. Dense fully developed post-settlement woodlands that initiated establishment in the late 1800’s had reached Phase III by the 1950’s and early 1960’s, based on tree growth rates (Fig. 25). In closed stands in southwestern Idaho, a significant decrease in growth of annual tree rings occurred during the 1950’s, suggesting the onset of intra-specific competition. In adjacent open stands of trees in Phases I and II, tree-ring growth did not decline. The primary factor controlling the number of years between initial encroachment and stand closure is establishment rate of tree seedlings. This is largely determined by seed input and the abundance of safe sites for seedling establishment. There may be a lag period of tree establishment immediately following fire, because of the reduction in shrubs (Erdman 1970, Burkhardt and Tisdale 1976). Shrubs provide desirable microsites for tree establishment (Burkhardt and Tisdale 1976, Miller and Rose 1995) and perching sites for avian seed dispersers. Stand structure in closed stands Canopy cover and density of overstory trees at stand closure varies among and within plant associations (Table 5). The density of large dominant trees in fully developed woodlands can vary from as low as 32 trees/acre on dry sites to more than 500 trees/acre on cool moist sites. Height and basal diameters are usually smaller in the denser stands of western juniper. Tree densities can exceed over 500 trees/acre if subcanopy trees are included. In closed woodlands shrub cover is typically less than 1 percent on the drier sites. On sites with higher effective precipitation that support both wax current (Ribes cereum) and snowberry (Symphoricarpos oreophilis), mean shrub cover is reduced to less than or equal to 5 percent in closed woodlands. Understory dynamics Shrubs As western juniper begins to dominate a site, shrubs begin to decrease (Figs. 21, 26) (Burkhardt and Tisdale 1969, Adams 1975, Bunting et al. 1999, Miller et al. 2000, Roberts and Jones 2000, Schaefer et al. 2003). This has a significant impact on ladder fuels, ground- and shrub-nesting birds, seed pools, and structural complexity of the plant community. At a site near Silver Lake, Oregon, 71 percent of the trees established during 1900–1936 (Adams 1975). The rapid decline in bitterbrush and sagebrush on these sites began in 1948. In the John Day Province near Prineville, Oregon, shrub cover in untreated western juniper plots was 0.4 percent compared to 9.4 percent cover in adjacent plots cut 18 years earlier (Eddleman 2002d). The decline in mountain big sagebrush is not proportional to the increase in western juniper. As western juniper approaches 50 percent of maximum potential, cover of mountain big sagebrush declines to about 20–25 percent of maximum potential (Miller et al. 2000). Tausch and West (1995) also reported a disproportionate decline; shrubs declined to one-fourth of maximum when single-leaf piñon (Pinus monophylla) and Utah juniper cover reached 50 percent of maximum in Nevada. Figure 21. A conceptual model illustrating the relationship between shrub canopy cover, tree canopy cover, relative growth rates (i.e., ratio of annual ring width:mean ring width), and management strategies during the three phases of woodland development. 26 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER Figure 22a. Subordinate—Phase I. Figure 22. Three phases of woodland succession in mountain big sagebrush communities. Figure 22b. Co-dominant—Phase II. Figure 22c. Dominant—Phase III on a south aspect with a soil restrictive layer at 16–18”. Figure 22d. Dominant—Phase III on a north aspect and deep well-drained soil. BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER 27 Grasses and forbs Although it is often stated that the herbaceous layer declines as western juniper increases in dominance, only a few studies have evaluated this relationship for western juniper. Two types of experiments support the hypothesis that western juniper overstory significantly affects production, diversity, and cover of the herbaceous layer: (1) spatial, comparing different transitional states within plant associations (Bunting et al. 1999, Miller et al. 2000), and (2) temporal, comparing herbaceous response over time between cut and uncut western juniper plots (Bates et al. 2000, Eddleman 2002d). Miller et al. (2000) reported that the relationship between herbaceous cover and western juniper canopy cover differed among plant associations. Herbaceous vegetation in plant associations characterized by Thurber needlegrass, which often had a restricted subsoil layer or strong argillic horizon, was the most sensitive to increasing tree dominance (Fig. 27a). Mean herbaceous cover, in early states of woodland development, was 16 percent, compared to 5 percent in late stages of development. However, herbaceous cover was not significantly different between different stages of woodland development in plant associations characterized by Idaho fescue (Fig. 27b). In central Oregon, the presence of western juniper was associated with an increase in bare ground and smaller, more widely spaced grass clumps on relatively shallow soils (Roberts and Jones 2000) and a significant decrease in ground cover (Knapp and Soulé 1998). This was consistent with results from southwestern Idaho, where herbaceous cover also decreased in the mountain big sagebrush alliance as western juniper dominance increased (Bunting et al. 1999). However, changes in species richness across the transitional phases of woodland development were not consistent. In southwestern Idaho and southeastern Oregon, species richness did not change as western juniper increased in dominance (Bunting et al. 1999, Miller et al. 2000). In contrast, species richness declined in Thurber needlegrass communities in Oregon and in Idaho fescue communities in northeast California (Miller et al. 2000). Herbaceous species diversity and richness also significantly increased following western juniper removal on a mountain big sagebrush/Thurber needlegrass plant association (Bates et al. 2000). Characteristics (post-settlement stands) Phase I (early) Phase II (mid) Phase III (late) Tree canopy (% of max. potential) Open, actively expanding <10% Actively expanding 10 to 30% Expansion nearly stabilized >30% Leader growth (dominant trees) (cm/yr) terminal >10 lateral >10 terminal >10 lateral 5 to >10 terminal >10 lateral <5 Crown lift1 (dominant trees) Absent Absent Lower limbs dying or dead where tree canopy >40% Potential berry production Low Moderate to high Low to near absent Tree recruitment Active Active Limited Leader growth (understory trees) (cm/yr) terminal >10 lateral >8 terminal 5 to >10 lateral 2 to >8 terminal <5 lateral <2 Shrub layer Intact Nearly intact to significant thinning >75% dead 1 Crown lift is the mortality of lower tree limbs, usually due to shading by neighboring trees. Table 4. Stand characteristics differe� needlegrass (maximum juniper cover= 25–41%, Idaho fescue (maximum juniper cover 34–58%), and Columbia needlegrass (maximum cover= 60–75%) (derived from Miller et al. 2000). 28 BIOLOGY, ECOLOGY, AND MANAGEMENT OF WESTERN JUNIPER Figure 23. Terminal leader growth is obvious throughout the outer c