Sparta Vegetation Management Project Wildlife Specialist Report

Transcription

1 Sparta Vegetation Management Project Wildlife Specialist Report Whitman Ranger District, Wallowa-Whitman National Forest Baker County, Oregon November 8, 2016 Author: _/s/ Laura Navarrete Laura Navarrete District Wildlife Biologist Wallowa-Whitman National Forest 1

2 Table of Contents MANAGEMENT INDICATOR SPECIES (MIS) 5 Old-Growth Associated Species 5 Connectivity of LOS Habitat 9 American Marten 13 Northern Goshawk 17 Pileated Woodpecker 24 Primary Cavity Excavators (PCE) 30 Effects on PCE species 43 ROCKY MOUNTAIN ELK 48 LANDBIRDS INCLUDING NEOTROPICAL MIGRATORY BIRDS (NTMB) 54 UNIQUE HABITATS 61 SALE AREA IMPROVEMENT PROJECTS 62 Snag Creation 62 WILDLIFE REFERENCES 63 2

3 Introduction This analysis describes the terrestrial wildlife species found in the project area and the potential effects of the Sparta alternatives on these species. Rather than addressing all wildlife species, discussions focus on Forest Plan management indicator species (MIS), threatened, endangered and sensitive (TES) species, Forest Plan featured species, and landbirds (see individual species lists below). The existing condition is described for each species, group of species, or habitat. Direct, indirect and cumulative effects of alternatives are identified and discussed. Supporting wildlife documentation is located the Project Record, and includes detailed data, methodologies, analyses, conclusions, maps, references and technical documentation used to reach conclusions in this environmental analysis. Regulatory Framework The three principle laws relevant to wildlife management are the National Forest Management Act of 1976 (NFMA), the Endangered Species Act of 1973 (ESA), and the Migratory Bird Treaty Act (MBTA) of 1918 (as amended). Direction relative to wildlife is as follows: NFMA requires the Forest Service to manage fish and wildlife habitat to maintain viable populations of all native and desirable non-native vertebrate wildlife species and conserve all listed threatened or endangered species populations (36 CFR ). ESA requires the Forest Service to manage for the recovery of threatened and endangered species and the ecosystems upon which they depend. Forests are required to consult with the US Fish and Wildlife Service if a proposed activity may affect the population or habitat of a listed species. MBTA established an international framework for the protection and conservation of migratory birds. This Act makes it illegal, unless permitted by regulations, to pursue, hunt, take, capture, purchase, deliver for shipment, ship, cause to be carried by any means whatever, receive for shipment, transportation or carriage, or export, at any time, or in any manner, any migratory bird. Forest Service Manual (FSM) direction provides additional guidance: identify and prescribe measures to prevent adverse modifications or destruction of critical habitat and other habitats essential for the conservation of endangered, threatened and proposed species (FSM (6)). The Forest Service Manual also directs the Regional Forester to identify sensitive species for each National Forest where species viability may be a concern. Under FSM , the manual gives direction to analyze, if impacts cannot be avoided, the significance of potential adverse effects on the population or its habitat within the area of concern and on the species as a whole. The principle policy document relevant to wildlife management on the Forest is the Wallowa-Whitman Land and Resource Management Plan (USDA Forest Service 1990), referred to as the Forest Plan for the remainder of this analysis. The Forest Plan provides standards and guidelines for management of wildlife species and habitats. Standards and guidelines are presented at the Forest level (LRMP, pp to 4-56) or Management Area level (LRMP pp to 4-98). The 1995 Regional Forester s Eastside Forest Plan Amendment #2 (Eastside Screens) amended Forest Plans for the National Forests in Eastern Oregon and Eastern Washington, including the Wallowa-Whitman National Forest. Amendment # 2 established interim wildlife standards for old growth, old growth connectivity, snags, large down logs, and northern goshawks. The Regional Forester has periodically distributed letters clarifying direction in Amendment #2 (Regional Forester, October 2, 1997; October 23, 1997; and June 11, 2003). Additional management direction is provided for the conservation of migratory landbirds. This direction is consolidated in the Forest Service Landbird Strategic Plan and further developed through the Partners in 3

4 Flight Program. The Oregon-Washington Partners in Flight Conservation Strategy for Landbirds in the Rocky Mountains of Eastern Oregon and Washington (Altman 2000) identifies priority habitats, and focal species and habitats for the Blue Mountains of Oregon. Analysis Methods Two different scales of analysis are used in this document to analyze the effects of the treatment activities on wildlife, and include the following: Sparta Project Area perimeter at 17,951 acres on National Forest System lands. The cumulative effects area encompassing the Sparta Project varies by species and is described within sections dedicated to individual species analyses. The project area boundary occurs within the Eagle Creek watershed. The existing condition is described for each species, group of species, or habitat. Direct, indirect and cumulative effects of alternatives are identified and discussed. Incomplete or unavailable information, scientific uncertainty, and risk are disclosed where applicable. Management Indicator Species (MIS) The geographic ranges of the MIS are larger than the project area, thus the analysis of adequacy of habitats for viable populations of MIS needs to be done at a scale larger than the individual project. Habitat must be provided for the number and distribution of reproductive individuals to ensure the continued existence of a species generally throughout its current geographic range (FSM ). Provisions for contributing to viable populations are determined at the level of the Forest Plan through management requirements, goals and objectives, standards, guidelines, prescriptions, and mitigation measures to ensure that habitat needs of MIS will be sufficiently met during plan implementation at the project level (FSM ). Analysis for each MIS includes an assessment of consistency with the provisions identified in the Forest Plan. Cumulative effects of proposed management activities on habitat capability for MIS are evaluated (FSM ). Best available science is considered in this analysis in assessing project impacts to MIS. Landbirds including Neotropical Migratory Birds (NTMB) Landbirds, including neotropical migratory birds, are discussed since many species are experiencing downward population trends. Discussion can be found in the section titled Landbirds including Neotropical Migratory Birds (NTMB). Analysis Tools and Surveys Species presence/absence determinations were based on habitat presence, past wildlife surveys, recorded wildlife sightings, the Oregon Natural Heritage Information Center wildlife sightings database (2008), scientific literature, and status/trend and source habitat trend documented for the Interior Columbia Basin (Wisdom et al. 2000). Vegetation analysis and estimates of stand conditions were completed using silviculture analysis tables, results described within the Sparta Forest Vegetation Management Report, aerial photo interpretation, vegetation database, and/or ground reconnaissance. Analysis Methodology Alternative 1, the No Action Alternative is used as a benchmark to compare and describe the differences and effects between taking no action and implementing action alternatives. The No Action Alternative is 4

5 designed to represent the existing condition; resource conditions are then projected forward in time to estimate resource changes expected in the absence of the proposed management activities. Effects on species will be determined by assessing how the No Action Alternative and action alternatives affect the structure and function of vegetation relative to current and historical distributions. Some wildlife habitats require a detailed analysis and discussion to determine potential effects on a particular species. Other habitats may either not be impacted or are impacted at a level which does not influence the species or their occurrence. The level of analysis depends on the existing habitat conditions, the magnitude and intensity of the proposed actions, and the risk to the resources. Present and reasonably foreseeable future activities used in cumulative effects analysis are listed in the EA, Appendix D. Where the species cumulative effects analysis area is larger than the two subwatersheds encompassing projects listed in Appendix D, other sources are used to quantify these activities. MANAGEMENT INDICATOR SPECIES (MIS) The LRMP identifies 5 wildlife species, or groups of species, as MIS (Table 1) (US Forest Service 1990). These species serve as indicators of the effects of management activities by representing habitat for a broad range of other wildlife species. The habitat requirements of MIS are presumed to represent those of a larger suite of species using the same type of habitat. All MIS are present in the project area. Table 1. MIS and Their Primary Habitats. Species American marten Northern goshawk Pileated woodpecker Primary cavity excavators 1 Rocky Mountain elk Habitat Old-growth and mature forest Old-growth and mature forest Old-growth and mature forest Snags and logs Cover and forage 1 Northern flicker; black-backed, downy, hairy, Lewis, three-toed, and white-headed woodpeckers; red-naped and Williamson s sapsuckers; black-capped, chestnut-backed, and mountain chickadees; and pygmy, red-breasted, and white-breasted nuthatches. Old Growth Habitat: American Marten, Northern Goshawk, and Pileated Woodpecker Introduction The American marten, northern goshawk, and pileated woodpecker are MIS of old growth habitat (U.S. Forest Service 1990). Old-growth habitat is categorized and analyzed in 2 categories according to the LRMP: 1) late old-growth structure; and 2) MA15 Old-Growth Preservation. MA15 is a land allocation under the LRMP (U.S. Forest Service 1990) intended to provide quality habitat for wildlife species associated with old growth characteristics. Old growth is a structural classification used to implement direction in the Forest Plan Amendment #2 (Screens; U.S. Forest Service 1995) and refers to multi-strata stands with large trees (Old Forest Multi-Stratum- OFMS) and single-stratum stands with large trees (Old Forest Single Strata- OFSS). Although the two terms have different administrative implications, both are intended to provide habitat for old growth associated wildlife species. 5

6 Old growth habitat and old growth management indicator species will be discussed separately below to provide an overview of old growth habitat in general within the project area and at the landscape scale along with the effects of the East Face project on each of the species dependent on this habitat. Correct determination of the scale of analysis is the cornerstone of habitat analysis (Morrison et al. 2006). The choice of spatial scale must be based on the species relationship with the landscape and should consider the scale at which to apply our results for management purposes (Morrison et al., 2006). Wildlife habitat is commonly analyzed at the watershed scale because it provides a systematic way to understand and organize ecosystem information and thus enhances the ability to estimate direct, indirect, and cumulative effects of management activities (Regional Interagency Executive Committee 1995). However, the watershed scale may be too fine to analyze viability for wide-ranging species unless it can be placed within the broader context of how the watershed contributes to overall species viability (Regional Interagency Executive Committee 1995). Impacts to old growth and old growth dependent MIS species within the Sparta project area were determined by analyzing effects to their habitat at several spatial scales starting with the watershed then framing that within the context of the Wallowa-Whitman National Forest and the Blue Mountains Ecological Province. These scales take into account the species relationship with the landscape as well as being practical for management purposes. MIS population viability assessments have been conducted for American marten, pileated woodpecker, and northern goshawk at the Blue Mountains and WWNF. These assessments are incorporated by reference within the existing condition and effects analysis for each species. For more indepth information on the methodology behind these assessments, please refer to the full-length assessments in the project record and the associated peer-reviewed literature scales (Penninger and Keown 2011a, Penninger and Keown 2011b, Penninger and Keown 2011c). The R-6 Regional Forester s Forest Plan Amendment #2 contains standards and guidelines (S&Gs) that address the historical range of variability (HRV). Because the distribution, quality, and quantity of habitat largely determine the potential for a wildlife species to exist at viable levels, HRV becomes an important habitat indicator for resident species. By managing habitat within the historical range of variability it is assumed that adequate habitat will exist for associated wildlife species since they existed at viable levels under those conditions previously. The larger the gap between current and historical conditions, the less likely that adequate habitat is being provided to sustain those associated species. Old Growth Habitat Declines in single stratum large trees structure (late-seral ponderosa pine) has been well documented (Wisdom et al. 2000, Squires et al. 2006), while mid-seral shade-tolerant forests seem to be at nearly twice their historical levels. These changes benefit some species but negatively affect others. The winter wren, Swainson s thrushes, pileated woodpeckers and American marten favor dense, multi-storied forests. These species are rarely associated with open ponderosa pine and open mixed-conifer types, which historically were widespread in many dry landscapes. Other wildlife species, however, such as the white-headed woodpecker and flammulated owl are associated with open, old-growth ponderosa pine (Sallabanks et al. 2001) and their populations have possibly declined as result of the loss of this forest type (Csuti et al. 1997, Wisdom et al. 2000). Thinning reduces competition-induced- mortality in a stand, and can likely enhance habitat for species associated with late seral conditions, particularly if critical structural components, such as dead wood, are provided and if stands are managed to provide vertical and horizontal heterogeneity. Effects of thinning on a given species of wildlife may vary across a range of temporal and spatial scales. For example, large tree crowns may ultimately improve habitat for some small mammals and some species of birds to nest and forage, but increased spacing between crowns may temporarily decrease habitat suitability and inhibit 6

7 dispersal. Hayes et al. (1997) states that knowledge of many species is inadequate to predict responses at multiple time frames, but it is important to consider short- and long-term as well as stand- and landscapelevel perspectives when evaluating the implications of thinning. Regional Forester Amendment #2 of June 12, 1995 established interim riparian, ecosystem, and wildlife standards for timber sales (these standards are referred to as the Eastside Screens ). The Eastside Screens require that a range of variation approach be used when comparing historical reference and current conditions, incorporating the best available science. The range of variation approach assumes that native species have evolved with the historical disturbance regimes of an area and so a forest will continue to sustain populations of those species if current conditions fall within the historic range of variation (Powell 2010). The following range of variation analysis uses methods described in Range of Variation Recommendations for Dry, Moist and Cold Forests (Powell 2010), which is now considered the best available science. Five forest structural stages are identified within these three potential vegetation groups; Stand Initiation (SI), Stem Exclusion (SE), Understory Retention (UR) and Old Forest Single Stratum (OFSS) and Old Forest Multi Strata (OFMS). LRMP standards and guidelines The Regional Forester s Eastside Forest Plan Amendment #2 (SCREENS) contains standards and guidelines for old growth (U.S. Forest Service 1995). Standards and guidelines include maintaining all existing remnant late and old seral and/or structural live trees >21 dbh. According to the LRMP, areas allocated to MA15 have no scheduled timber harvest although salvage may occur following catastrophic destruction if more suitable replacement stands exist. The SCREENS also provides direction for connectivity. Old growth stands are directed to be connected in a least two different directions by the shortest length, minimum 400 ft. wide corridor which maintains canopy cover in the upper one-third of the site potential. If this standard cannot be met, proposed treatments are dropped. Existing Conditions MA-15 Old Growth Preservation- There are 487 acres of MA15 allocated land in the project area boundary. Suitable old growth habitat generally contains large diameter live trees, large snags and down wood; old forest multi story (OFMS) provides old growth habitat along with understory re-initiation (UR), though UR typically lacks the density of large structure. Late Old-Growth Structure Both OFMS (multi-storied LOS) and OFSS (single-storied LOS) are present in the Sparta project area. Existing OFMS and OFSS within the project area total 2,601 and 658 acres, respectively. However, for several potential vegetation groups (PVGs), the HRV analysis identifies departure from HRV. All PVGs show that the amount of multi-storied LOS is above the range estimated to have existed historically. In contrast, single-storied LOS stands are currently deficient within dry, cold and moist PVGs (Table 2). 7

8 Table 2 - Comparison of HRV to existing by potential vegetation group (PVG) in the Eagle Creek-Paddy Creek and Little Eagle subwatersheds PVG Existing Acres % of PVG Historical Range % Old Forest Multi Stratum (OFMS) moist upland 1,152 26% 15-20% dry upland 2,912 16% 5-15% cold upland % 10-25% Old Forest Single Stratum (OFSS) moist upland 146 3% 10-20% dry upland 875 5% 40-60% cold upland 31 4% 5-20% Direct/Indirect Effects on Old Growth Alternative 1 Under this alternative, the risk of uncharacteristic wildfire or disease/insect outbreaks would continue to increase naturally over time because there would be no changes to stand stocking levels or fuel loads from active management. Existing MA15 and old growth would be at risk if uncharacteristic wildfire and/or disease and insect outbreaks occurred. Old forest single story structure would continue to be deficient across all potential vegetation groups. Alternative 2 and 3 Alternative 2 and 3 propose to increase the availability of OFSS structure stages by treating primarily dry OFMS through a combination of commercial and non-commercial treatments to remove small diameter understories and reduce canopy cover where appropriate, emphasizing ponderosa pine restoration. Alternative 2 would increase available OFSS and generate the largest benefit for species dependent upon open forest with large trees by increasing available OFSS by about 55% (526 acres) and set up future stands to transition into OFSS condition by increasing single-story non-los acres (Table 3). Alternative 3 would not directly increase OFSS but would help remove the risk of stand replacing fires through non-commercial treatments. Those OFMS stands not selected for treatment under Alternative 3 would retain Douglas-fir and grand fir in stands for which ponderosa pine restoration is the emphasis, thereby perpetuating the condition of encroachment by mixed conifer species (Table 4). Neither Alternative would move total acres of moist or dry OFMS stands out of the Historic Range of Variation. Further discussion of effects of old growth treatment can be found under the Neotropical Migrant discussion. 8

9 Table 3. Silvicultural Effects on Single-Storied Large Tree Structure within Moist and Dry Forest Types within the Sparta Project Area. Post- Existing Post- Single-Storied Large Existing Moist Treatment Change in OFSS Dry PVG Treatment Dry Tree Structure PVG OFSS Moist PVG LOS acres OFSS PVG OFSS OFSS Alternative Alternative Table 4. Silvicultural Effects on Mesic Mixed Conifer Multi-Storied Large Tree Stands within the Sparta Project Area Multi-Storied Existing Existing Post- Post- Change in Dry Change in Large Tree Dry OFMS Moist Treatment Treatment OFMS acres Moist OFMS Structure (acres) OFMS Dry OFMS Moist OFMS (%) acres (%) Alternative 2 2, , (21%) -23 (11%) Alternative 3 2, , (0%) -0 (0%) Cumulative Effects Cumulative effects for old growth are specific for the structure stage and are not tied to old growth species habitat needs. Effects are analyzed at the same scale as the HRV analysis, within the Eagle-Creek Paddy Creek and Little Eagle subwatersheds. There are no projects on Forest Service land within these subwatersheds proposed within the foreseeable future. Private land exists within these subwatersheds and there is potential for landowners to alter structure stage in the foreseeable future. Private land commercial harvest activities are expected to continue to maximize commercial output and mitigate wildfire danger. These treatments are not expected to maintain old growth conditions and old growth habitat is expected to decrease on private land. Cumulative effects of implementing Alternative 2 or 3 would be positive in relation to increasing OFSS structure stage within the subwatersheds. CONNECTIVITY OF LOS HABITAT Existing Conditions Connectivity between MA15 allocated old growth and LOS stands was assessed utilizing field reconnaissance, aerial photographs and GIS mapping. The level of connectivity between MA15 and LOS stands varies across the project area. Areas of non-forested vegetation in combination with past timber harvest have created gaps of varying size in the central and southeastern portions of the project area. Several LOS stands are currently isolated by their adjacency to stands lacking a substantial forest cover component. There are connections to large areas of roadless and wilderness to the north of the project area. The LOS found in the southern portion of the project area is connected to the north, but not to the south where a natural gradation from forested to non-forested vegetation types occurs. Refugia The concept of refugia has been incorporated into project planning. Refugia is described by Camp et al. (1997) as forested patches that have missed one or more disturbances that shaped the surrounding vegetation on the landscape. In the Sparta project area these typically occur on north and northeastern aspects which have higher moisture retention than other aspects. Higher moister levels generally lead to denser vegetation and higher fuel moisture. Higher fuel moisture influences fire behavior by causing a fire to burn in more of a mosaic pattern or to spare forested patches altogether. 9

10 The connective corridor network for Sparta was delineated to not only connect LOS and MA15 areas, but to also link to many of the refugia stands. Camp et al. (1997) further indicated that late-successional refugia rarely formed extensively connected corridors within their study area (Wenatchee Mountains). In addition, highly connected late-successional habitats may allow endemic insect and disease levels to reach epidemic proportions as well as increase the potential for high-intensity wildfire within refugia. In order to balance the function of connective corridors with the risk of epidemic insect and disease events, silvicultural treatments were specifically designed within connective corridor stands to reduce tree stocking while retaining additional canopy closure and structural complexity. Differences between silvicultural prescriptions designed to retain the function of connective corridors are discussed in more detail in the direct and indirect effects section below. The Alternative description section of the EA contains the proposed commercial timber harvest units that occur within connective corridors for comparison between alternatives. The connectivity network was established based generally on stand boundaries and connects, to the extent possible, all LOS and MA15 stands within and outside the project area according to direction in the Forest Plan Amendment #2. The Pine Eagle Consensus Group proposed a management strategy to the Wallowa-Whitman NF in 1989 intended to incorporate biological diversity into management of the forest landscape. This group proposed the designation of management areas in the Land and Resource Management Plan that would identify old growth core areas, transition habitat, and connection habitat. This would result in a network of old growth core reserves, transition areas managed on long timber rotations, and corridors that connect this series of core and transition areas together. The Pine Eagle Consensus Group also provided a map of what this network would look like for the area that is now the Sparta analysis area. The Consensus Group s map was compared to the LOS, MA15 and connective corridor network designed for the Sparta analysis area to assess similarities and differences between the two approaches. A series of old growth reserves (Management Area 15) was established in the 1990 Land and Resource Management Plan. Essentially in 1993 the Regional Forester s Forest Plan Amendment #2 did what the Consensus Group had proposed four years earlier. This Amendment (referred to as the Eastside Screens) directed all Oregon and Washington National Forests east of the Cascade Range to use a historical range of variability (HRV) approach for managing forests. This effectively protected old growth habitat in amounts reflective of pre-european settlement times, and provided for connectivity corridors between management areas 15 and isolated old growth patches identified through the HRV process. Many of the silvicultural treatments pursued today are designed to accelerate the development of mature and old growth forest characteristics while improving stand resilience and resistance to insects and diseases. This approach is consistent with the Consensus Group s idea of managing transition areas on longer timber rotations. The result is a network of connected mature and old growth habitat that is remarkably similar to the Consensus Group s recommendations. Direct and Indirect Effects Introduction Although the Sparta project involves several types of activities that could alter the quality and function of connective corridors, this effects analysis focuses on the commercial silvicultural treatments. Commercial timber harvest is the tool used to accomplish at least part of the silvicultural objectives. Timber harvest is the primary activity that would reduce canopy closure and decrease structural complexity within treated stands. Other activities such as prescribed fire, non-commercial thinning, and mechanical fuels reduction can affect the quality and function of connective corridors, but to a much lesser degree than timber harvests. Also, the structural components effected by these less impacting activities can be replaced (grow back, recover) 10

11 quickly relative to timber harvest. For example, non-commercial thinning, mechanical fuels reduction, and prescribed fire generally target the reduction of smaller diameter materials from forest duff to woody materials under 10 in diameter. An exception is prescribed fire which can consume all sizes of woody material, live and dead. Fire is an inexact tool, so there is the possibility that some larger woody structures will be consumed, and new ones created as trees are killed. However, prescriptions for fire are designed to retain the larger diameter woody materials, and consume smaller diameter materials. Treatment prescriptions in any units within LOS connective corridors would retain snags, large down wood, and multiple canopy layers (if appropriate for the site). Basal area would also be maintained within the upper half of the management zone, which would approximate canopy closures in the upper 1/3 site potential. Stocking levels would be managed at the upper management zone for basal area except where tree quality and crown conditions are such that this level of stocking is unattainable, in these areas, 20% of the stand would be retained in untreated clumps. Trees with as little as 20% live crown would be retained if needed to maintain basal area levels. All snags greater than or equal to 12 inches dbh would be retained. Down logs would be retained at 200 lineal feet per acre, minimum lengths of logs 20 feet or largest available and minimum of 12 inch small end diameter logs or largest available. Silvicultural prescriptions in connective corridor units would reduce competition between residual trees, increase tree growth rates, and increase trees ability to defend against insects and diseases, while retaining levels of canopy closure and structural complexity to facilitate movement of wildlife between old-growth habitat patches. Alternative 1 This alternative would have no direct effect on connectivity between LOS habitat patches. The current level of connectedness would persist, and would improve in quality in the absence of large scale disturbances. In the absence of silvicultural treatments that reduce tree stocking, the connective corridors will continue to increase in canopy closure and structural complexity. This condition in cold and moist upland forests can enhance connectivity for species like American marten. Conversely, dry upland forests, which make up the majority of the project area are inherently less structurally complex than cold and moist upland forests. In the absence of silvicultural treatments to reduce tree stocking, these stands would continue to allow the establishment of shade tolerant grand fir, increased canopy closure, and increased stress to competition for resources. In the long-term (30+ years) these drier stands would be subjected to increased risks from wildfire, insects and diseases that will kill trees in numbers and distribution that could negatively affect connectivity between patches of single strata LOS habitat. To forego prescribed burning, non-commercial thinning, and mechanical fuels reduction would perpetuate higher tree densities, higher fuels loading, ladder fuels, and tree species compositions that are not sustainable for the biophysical setting. These indirect effects may contribute to uncharacteristic insect, disease, and wildfire events. The effects of wide scale tree mortality from these disturbances would have a much greater negative effect to connectivity than the prescribed treatments under Alternatives 2 and 3. These negative effects could render the LOS and connective corridors unsuitable for the wildlife species that depend on them as habitat. Alternative 2 Alternative 2 would reduce the quality of connectivity corridors on 1,844 acres by reducing the canopy closure and structural complexity. Table 3 compares the acres that are proposed for silvicultural treatments by alternative. Silvicultural prescriptions in connective corridor units would reduce competition between residual trees, increase tree growth rates, and increase trees ability to defend against insects and diseases, while retaining levels of canopy closure and structural complexity to facilitate movement of wildlife between old-growth habitat patches. 11

12 Alternative 2 would apply prescribed fire to approximately 286 acres within connective corridors, but outside of commercial timber harvest units. The large majority of these acres are within the dry upland forest types where periodic fire historically performed a maintenance function that shaped the structure, tree species composition, and wildlife value of these stands. Some snags and logs may be consumed by prescribed fire, while new snags and logs are recruited from fire killed trees. The burning, non-commercial thinning, and mechanical fuels reduction in connective corridors will not have a measurable negative effect on the quality or function of the corridors. Alternative 3 This alternative would reduce the quality of connectivity corridors on 1,694 acres. Table 3 provides a comparison of acres treated within connective corridors by alternative. Silvicultural prescriptions in connective corridor units would reduce competition between residual trees, increase tree growth rates, and increase trees ability to defend against insects and diseases, while retaining levels of canopy closure and structural complexity to facilitate movement of wildlife between old-growth habitat patches. Alternatives 3 would apply prescribed fire to approximately 286 acres within connective corridors, but outside of commercial timber harvest units. The large majority of these acres are within the dry upland forest types where periodic fire historically performed a maintenance function that shaped the structure, tree species composition, and wildlife value of these stands. Some snags and logs may be consumed by prescribed fire, while new snags and logs are recruited from fire killed trees. The burning, non-commercial thinning, and mechanical fuels reduction in connective corridors will not have a measurable negative effect on the quality or function of the corridors. Cumulative Effects Alternative 1 The no action alternative will not contribute to the cumulative effects of past, present and foreseeable future activities. Any effects of forgoing silvicultural treatments and prescribed burning would occur later in time, and are addressed as indirect effects above. Alternative 2 The reduction in connective habitat quality that results from silvicultural treatments would be greater than for Alternative 3 in both acres effected and degree of effects on specific acres. It is unknown whether the level of treatments in Alternative 2 would compromise connectivity to a level that leads to isolation or fragmentation of wildlife habitat. The riparian habitat conservation area network, M15 areas, wild and scenic river corridor (Eagle Creek), and the remaining forest matrix would combine to facilitate varying degrees of connectivity between distant LOS habitat patches. The incremental effects of prescribed burning, non-commercial thinning, and mechanical fuels reduction, would be immeasurable relative to the quality and function of connective corridors. Alternative 3 The reduction in connective habitat quality that results from silvicultural treatments would be relatively short lived as tree canopies respond to the reduced competition, and seedlings establish in response to increased sunlight reaching the forest floor. The quality of connective habitat in treatment units would likely recover to pre-treatment conditions within fifteen years. In the interim, the network of connectivity corridors that is not being treated (riparian habitat conservation areas, MA15 areas, and the matrix of forested habitats) would facilitate movement of LOS associated wildlife species between source habitat patches. Alternative 2 would reduce the quality of connective corridors on 150 more acres than Alternative 3. 12

13 This approach of addressing connectivity habitat is consistent with direction in the Regional Forester s Forest Plan Amendment #2 to retain canopy closure in the upper 1/3 of site potential, and other criteria that define connective corridors. The incremental effects of prescribed burning, non-commercial thinning, and mechanical fuels reduction, would not compromise the quality or function of connective corridors. Table 3. Connectivity Corridor Summary. Connectivity Elements Acres of Silvicultural treatments in connective corridors Alternatives Alternative 1 Alternative 2 Alternative 3 0 1,844 1,694 Connectivity Prescription N/A Retain basal area at UMZ Retain basal area at UMZ AMERICAN MARTEN Life history, risk factors, conservation status and population trend, as well as habitat condition and species viability are described in detail in the American Marten (Martes americana) Management Indicator Species Assessment, Wallowa-Whitman National Forest (Penninger and Keown 2011a). Portions of that assessment are summarized below. Distribution American marten ranges through Canada and Alaska, south through the Rocky Mountains, Sierra Nevada, northern Great Lake Region, and northern New England. In Oregon, the species occurs in the Coast Range, Cascades, and Blue Mountains (Penninger and Keown 2011a). Habitat The marten is a MIS associated with old-growth habitat. It is an indicator of the abundance and distribution of mature and old-growth forest, and the contigtuity of cover between old-growth stands. Marten are closely associated with forested habitats that have complex physical structure near the ground (Buskirk and Powell 1994, Slauson et al. 2007) including coarse wood (Stevens 1997). Open areas, such as regeneration logging units, recent severely burned areas, and natural openings are avoided, especially during the winter. Forested riparian habitats are used disproportionately higher than they are available, which indicates their importance as travel corridors (Bull et al. 2005). Bull and Heater (2000) found that natal dens in the Blue Mountains were either cavities in hollow trees or underground burrows, and post-natal den sites were also in hollow logs and slash piles. After bearing young, the female would move the kits to post-natal dens. Cavities in trees and down logs were primarily in grand fir, with western larch, Engelmann spruce, and subalpine fir also used. A key consideration for marten foraging is the structural complexity that supports a diverse prey base, provides for opportunities for marten to be successful in capturing prey, and allows for marten to hunt while minimizing their exposure to predation. Large down woody material, multiple canopy layers, high canopy closure, and over all higher structural complexity contribute to effective foraging habitat for marten (Penninger and Keown 2011a). Broad-Scale Habitat Assessments Wales (2011) used Bayesian Belief Network (BBN) Models to conduct viability assessments for various wildlife species of interest at the Blue Mountains and WWNF scales, including American marten. Using a threshold of 60% canopy closure and minimum stand diameters of 20 inches in Cold Moist and Cold Dry Potential Vegetation Groups, Wales compared current habitat conditions to those estimated to have occurred historically. The threshold of >40% of the historical amount of source habitat in a watershed was used to 13

14 identify watersheds with a relatively high amount of source habitat that would contribute to species viability. Watersheds that contain > 40% of the estimated historical median amount of source habitat (1,136 acres) are believed to provide for habitat distribution and connectivity, and better contribute to species viability across the Forest. Historically, it is most probably (59% probability) that marten habitat was broadly distributed and of high abundance, and that marten were well distributed within the mixed conifer forests of the Blue Mountains anf the WWNF. The abundance and distribution of habitat likely provided for a high degree of connectivity within the elevations and forest cover types that provided source habitat for martens (viability outcome A). Currently, marten habitat is more abundant in some parts of the Blue Mountains and less abundant in others, but less contiguous than historically. It is most probably (54% probability) that marten habitat is broadly distributed and of high abundance, but there are gaps where suitable environments are absent or only present in low abundance (viability outcome B). At the regional scale (Blue Mountains), Wales (2011) found that source habitat amounts that occurred historically in the Blue Mountains totaled 277,715 acres. Source habitats within the Blue Mountains currently total 257,942 acres; or 93% of historical levels. On the WWNF, 144,347 acres of source habitat are estimated to have occurred historically. Currently the WWNF contains 129,943 acres of source habitat (90% of historical) (Penninger and Keown 2011a). Like most coarse scale vegetation data sets, the one used in the viability assessment is imperfect. However, it indicates landscape patterns that reasonable estimate habitat conditions for marten at larger scales. Conditions within Eagle Creek Watershed The Sparta project area is located within the Eagle Creek watershed, included within cluster W3 (Wales 2011). The analysis determined that cluster W3 historically contained 13 watersheds (76%) with greater than or equal to 40% source habitat. Under the current condition, W3 contains the same number of watersheds (13) with 40% or greater source habitat, indicating that landscape capability within this cluster is similar to that provided historically. Analysis of conditions within the Eagle Creek watershed as modeled by Wales shows that source habitat totals 30% of potential habitat acres. The historical median for source habitat as a percentage of potential is 16 percent, indicating a high level of habitat quality and quantity and placing the Eagle Creek watershed among the 6 watersheds with the greatest amounts of source habitats in the Blue Mountains. Analysis for Sparta and Eagle Creek Watershed Habitat parameters described by Penninger and Keown (2011a) and Wales (2011) were applied to existing vegetation data to identify marten source habitats within the Sparta project area and Eagle Creek watershed. Furthermore, field review of selected stands modeled as source habitat was conducted in September 2011, resulting in changes in habitat classification of habitats modeled within the project area. Based on the field review, it appears that stands capable of supporting marten exist primarily above 4,700 feet elevation and those below 5,000 feet in the Sparta project area are restricted to moist stands on northerly aspects due to the dry nature of stands facing other aspects at these elevations. Because the Sparta project area is predominately south-facing, presence and development of larger core habitat areas for marten is more likely to occur above 5,000 feet where stands of more neutral aspects are more moist. Review of existing conditions shows that suitable marten habitat on WWNF lands in the Sparta project area is relatively sparse, with source habitats totaling only 151 acres. Using parameters identified by Penninger and Keown (2011a), approximately 12,011 acres of source habitat and 4,735 acres of secondary habitat modeled and adjusted after field review in the Sparta project area, is estimated to occur on WWNF lands within the Eagle Creek watershed (5 th -level HUC). The majority of source and secondary habitat (98%) within the Eagle Creek watershed occurs north of the Sparta project area. The Sparta project area represents the very southern periphery of marten habitat in the Eagle Creek 14

15 watershed. Marten have been observed along the northern edge of the project area, specifically in the area of Martin Bridge (ODFW, personal communication). It is assumed that marten can use portions of the Sparta project area for movement but due to the limited amount of source habitat, denning is unlikely. Direct and Indirect Effects Alternative 1 There will be no direct or indirect adverse effects to American marten from the Alternative 1 because no timber harvest, fuels treatments, or transportation activities will occur. Existing marten source and secondary habitat would remain unchanged. Alternatives 2 and 3 In general, silvicultural treatments have the potential to affect marten habitat suitability by reducing stand canopy closures and understory tree densities. Silviculture treatments proposed under alternative 2 would treat about 19 acres (12% of existing) of source habitat found within the project area (Table 4). Stands within the Sparta project area show uncharacteristically high levels of dwarf mistletoe and silviculture treatments propose to remove infected trees to some degree. Trees 21 dbh and larger will be retained on the landscape, but will be treated in a way to encourage snag creation and reduce mistletoe infection of the stand. Dwarf mistletoe, more specifically the brooms produced by dwarf mistletoe infection, has been identified as a structural component of marten habitat. In northeastern Oregon, Bull and Heater (2000) reported that 19% of tree platforms used by marten for resting were formed by mistletoe. In addition, mistletoe brooms provide resting structure for potential prey species including red squirrels and flying squirrels. Removal of mistletoeinfected trees would degrade habitat suitability by reducing available resting structure for marten and potential prey species that use these structures. However, snag creation can provide resting and denning habitat for marten as well as increase future down logs. The structural complexity of treated stands will be more simplified, thus providing lower quality habitat. Alternative 3 has no proposed silviculture treatments within marten source habitat. Application of fuel treatments outside of stands proposed for timber harvest has the potential to reduce understory and down wood densities, but is unlikely to substantially reduce stand canopy closures. Katie Moriarty (2014) compared marten movement within open, simple stands treated with fuels treatments and untreated complex stands. She found that martens selected home ranges with a disproportionate amount of complex stands and avoided openings. This implies that on a landscape level, only the percentage of openings affected the placement of marten home ranges; however, simple stands were marginally avoided compared to complex stands. Marten movement within simple stands vs. complex stands suggests that marten use simple stands for travel and for intermittent foraging but not for denning. Therefore, prescribed fire-only treatments are expected to degrade, but not remove, marten habitat. Prescribed fire is proposed in 19 acres of source habitat under all alternatives. Table 4. Proposed Silvicultural Treatments in Modeled Marten Habitats Treatment Type by Alternative, Acres (Percent of Corresponding Habitat Type), Sparta Project Area Habitat Type (Existing Alternative 2 Alternative 3 acres) on WWNF lands Silv. Rx Fuel Only Silv. Rx Fuel Only Source Habitat (151 acres) 19 acres (12%) 0 (0%) 0 (0%) 19 acres (12%) Marten Habitat at the Watershed Level Marten habitat contained within the Sparta project area occurs at the very southern periphery of marten habitat within the watershed and it is not expected that marten would use the area for anything other than 15

16 foraging. Treatments under Alternative 2 and 3 would degrade about 0.1% of source habitat available in the watershed. Post-treatment availability of source habitats would continue to exceed the threshold of 40% of the historical amount in the Eagle Creek watershed under all action alternatives, thereby continuing to contribute to species viability at the watershed scale. In addition, post-treatment amounts of source habitat as a percentage of potential habitat would remain at 32%, well above the historic median of 16% described by Penninger and Keown (2011a). Marten Habitat at the WWNF Scale Estimated habitat impacts at the project area and watershed scales (described above) are based on source habitat parameters modeled according to Penninger and Keown (i.e. 50% canopy closure and 15 inch DBH criteria). Existing marten source habitat on the WWNF as modeled by Wales (2011a) totals 129,943 acres. As a result of proposed activities under the Sparta project, source habitats would be degraded on 19 acres under Alternatives 2 and 3. Because source and secondary habitats at the Forest level were modeled according to more conservative thresholds described by Wales (i.e. 60% canopy closure and 20 inch DBH criteria), it is reasonable to assume that the source habitat impacts would actually be less than the estimate based on the 50% canopy closure and 15 DBH criteria. Therefore, source habitat impacts at the Forest level would equate to less than 0.003% under Alternatives 2 and 3. Cluster analysis used to describe existing distribution of source habitats across the WWNF indicates that these habitats are well distributed across the Forest (Penninger and Keown 2011a). Post-treatment levels of source habitat under all Sparta action alternatives are expected to result in no change in the number of watersheds in Cluster W3 containing >40% source habitat that contribute to marten habitat distribution. Landscape Permeability Treatments proposed under each action alternative may decrease existing habitat permeability due to reduced canopy closure, decreased structural complexity, and increased disturbance on specified and temporary roads. Impacts from temporary roads are expected to be short-term (up to 10 years), with impacts scattered in time and space as treatments are implemented. Permeability reductions would be localized in the southern portion of the Eagle Creek watershed, within the northern portion of the Sparta project area. Areas of higher permeability, located north of the project area, would remain unaltered by project activities. Cumulative Effects Past, present and reasonably foreseeable future actions were analyzed for cumulative impacts to the species. Effects of past activities including road construction, fire suppression, prescribed fire, and timber management on WWNF lands have been incorporated into the existing conditions for amounts and locations of marten habitats in the analysis areas and into the viability analysis Appendix D of the EA was reviewed for actions that might affect marten habitat within the Eagle Creek watershed. Cumulative impacts of ongoing and foreseeable actions are projected out to 20 years from the present. Ongoing and future livestock grazing is expected to have no effect on marten habitat because cattle tend to avoid areas with high amounts of down wood. On Forest Service lands within and outside the project area, firewood cutting will continue to reduce available snags and logs, but the effect is limited to areas adjacent to open roads. Timber harvest on private inholdings is expected to continue at some level, with anticipated reductions of trees larger than 10 inches DBH, but generally marten habitat does not occur on private inholdings in the Sparta project area. Wales et al. (2011) estimated that approximately 144,300 acres of source habitat existed on the WWNF historically. At the time of the analysis, approximately 129,900 acres (90% of estimated historical conditions) of source habitat occured on the WWNF. Since the viability assessment was run 14 16

17 Vegetation/Fuels Restoration projects have been analyzed across the Wallowa-Whitman. Some have been implemented and some are still undergoing the NEPA process, but are anticipated being implemented in the foreseeable future. These combined projects, including the Sparta Vegetation Management project, anticipate commercially impacting 2,427 acres of marten source habitat and non-commercially impacting 4,472 acres of marten source habitat. Taking these 6,899 acres of impacted source habitat into account, this results in approximately 123,001 acres (85% of estimated historical conditions) of source habitat existing on the WWNF. Cumulatively, vegetation management activities on the Wallowa-Whitman are not expected to change the viability outcome found by Wales et al. and marten source habitat will remain well distributed and highly abundant with some gaps where suitable environments are absent or only present in low abundance (viability outcome B). Conclusion Because this project impacts less than 0.003% of suitable habitat across the Forest, the overall direct, indirect and cumulative effects will result in a small negative effect to marten habitat. The decrease in habitat quality due to the Sparta Vegetation Management Project will be insignificant at the scale of the WWNF. The Sparta project may reduce habitat permeability at a localized scale, but impacts at the WWNF scale would be immeasurable. Post-treatment availability of source habitats would continue to exceed the threshold of 40% of the historical amount in the Eagle Creek watershed under all action alternatives, thereby continuing to contribute to habitat distribution and species viability on the WWNF. NORTHERN GOSHAWK The goshawk is a MIS associated with old-growth habitat. It is an indicator of the abundance and distribution of mature and old-growth forests. It is associated with dense-canopied mixed conifer and lodgepole pine forest types. Potential goshawk nesting habitat is defined as multi-story large tree (OFMS) mixed conifer stands with high canopy closure (60-90%), a north or northeast aspect, slopes <30%, a sparse understory, and near water (Marshall 1992). Moore and Henny (1983) reported that a portion of nesting goshawks (15%) utilized mistletoe brooms as nesting substrate in northeastern Oregon. McGrath el al. (2003) described that goshawks in study areas within the eastern Cascades and Blue Mountains, Oregon and Washington, also utilized stem-exclusion and understory-reinitiation forest structural stages with high basal areas, northerly aspects, and at lower slope positions. Goshawks defend territories of several hundred acres that usually contain 2-4 alternate nests, which are often used for more than 1 year and sometimes intermittently for decades (Reynolds et al. 1992, Marshall 1992). Typical nest distribution in eastern Oregon is about 3.2 mi between active nests or about 1 nest per 6,800 acres (Reynolds and Wight 1978, Csuti et al. 2001). Goshawks forage for birds and mammals in open understories below the forest canopy and along small forest openings (Marshall 1992, Bull and Hohman 1994). Goshawk prey is equally divided between mammals and birds, and quail, owls, small hawks, ducks, thrushes, squirrels, and shrews are among their prey (Marshall 1992, Csuti et al. 2001). Life history, risk factors, conservation status and population trend, as well as habitat condition and species viability are described in detail in the Northern Goshawk (Accipiter gentilis) Management Indicator Species Assessment, Wallowa-Whitman National Forest (Penninger and Keown 2011c). Portions of that assessment are summarized below. Nest Site Habitat Preferred nest stands typically have high canopy closure. Reynolds et al. (1982) described goshawk nest sites in an eastern Oregon study as having a canopy cover ranging from percent with a mean of 60 percent. Reynolds et al. (1992) reported that preferred nest stands have a minimum of 40% canopy closure; and the nest sites within these stands have >60% canopy closure and open understories. Bull and Hohmann (1994) reported average canopy cover as 81% at nest sites on the WWNF. Ten of 12 nest trees were in old growth stands that either had not been harvested or had been high-graded in the past. The other two nest 17

18 sites were in stands that had been partially logged and had remnants of old growth (Bull and Hohmann 1994). Daw and Destefano (2001) reported that in northeastern Oregon goshawks nested more frequently in stands with dense canopy and late forest structure (trees >21 inches [53 cm], canopy closure >50%) and rarely in stands with mid-aged forest structure. Late forest structure was more abundant within circles of 30 acres (12 ha) and 60 acres (24 ha) around nests than at random points (Daw and Destefano 2001). Post-Fledging Area (PFA) The PFA surrounds the nest area and is defined as the area used by the family group from the time the young fledge until they are no longer dependent on the adults for food (up to two months) (Reynolds et al. 1992, Kennedy et al. 1994). PFAs typically include a variety of forest types and conditions. Reynolds et al. (1992) called for maintaining the PFA in forest conditions intermediate between the high foliage volume and canopy cover of the nest stands and the more open foraging habitats. PFAs have patches of dense trees, developed herbaceous and/or shrubby understories and habitat attributes (snags, down logs, small openings) that are critical for goshawk prey (Reynolds et al. 1992). Although PFAs generally include a variety of forest conditions, the vegetation structure resembles that found within nest stands (Reynolds et al. 1992). Foraging Goshawks are classified as prey generalists (Squires and Reynolds 1997) and forage for small birds and mammals in open understories below the forest canopy and along small forest openings (Marshall 1992, Bull and Hohman 1994). Main foods include small mammals, ground and tree squirrels, rabbits and hares, large passerines, woodpeckers, game birds, and corvids (Squires and Reynolds 1997). DeStefano et al. (2006) found that birds comprised a high proportion of goshawk diet on the WWNF with northern flickers and American robins utilized the most. Important habitat components that support goshawk prey species include downed logs, woody debris, large trees, openings, herbaceous and shrubby understories, and a mixture of various forest vegetative structural stages (Reynolds et al. 1992). Home Range The foraging area encompasses the breeding pair entire home range. In North America, home ranges during nesting vary between 1,408 13,098 acres (570 5,300 ha), depending on sex, habitat characteristics, and choice of home range estimator (Squires and Reynolds 1997, Boal et al. 2003). The male s home range is usually larger than the female s (Hargis et al. 1994, Kennedy et al. 1994, Boal et al. 2003). Home ranges, excluding nest areas, appear not to be defended and may overlap adjacent pairs (Squires and Kennedy 2006). Shapes of home ranges vary from circular to almost linear and may be disjunct depending on habitat configuration (Hargis et al. 1994). Threats and Risk Factors A number of factors are cited by researchers and managers as potentially detrimental to current and future goshawk viability. These include, but may not be limited to, habitat alteration, direct human disturbance, pesticides and other contaminants, and harvest for falconry. However, the primary concern throughout the range of the goshawk is habitat alteration due to timber and fire management practices (Squires and Kennedy 2006). Timber harvest is the principal threat to breeding goshawk populations (Squires and Reynolds 1997). Potential threats to habitat caused by various silvicultural treatments include forest fragmentation, creation of even-aged and monotypic stands, potential increases in area of younger age classes, and loss of tree species diversity (Desimone and DeStefano 2005, Squires and Kennedy 2006). Harvest of mature forests decreases goshawk nesting habitat (Reynolds 1983). Desimone and DeStefano (2005) also report that reduced proportion of mature closed-canopy forest, which is subsequently replaced by early successional forest, reduces the probability of an area as a potential nesting habitat. More severe alterations such as clearcuts and moderately high alteration such as partial removal of stands resulting in <50% canopy closure reduce the 18

19 likelihood of goshawks reoccupying the area due to deterioration in the quality of potential nest areas (Desimone and DeStefano 2005). Fire suppression may lead to increased susceptibility of stand replacing fire and insect and disease outbreaks, which can result in the deterioration or loss of nesting habitat (Graham et al. 1999, cited in NatureServe 2010, Wisdom et al. 2000). Loss of foraging habitat due to dense conifer understory as a result of fire suppression. Dense understories may obstruct flight corridors used by goshawks to hunt prey (Wisdom et al. 2000). Predation is another threat to the goshawk. Following canopy reduction by logging, goshawks are often replaced by other raptors including red-tailed hawk, great horned owl, and long-eared owl (Crocker-Bedford 1990). Great horned owls are reported as especially significant as they prey on both adult and nestling goshawks (Boal and Mannan 1994, Erdman et al. 1998, Rohner and Doyle 1992). Habitat Analysis Wisdom et al. (2000) defined habitat requirements (source habitats) and assessed broad-scale trends of 91 species of terrestrial vertebrates within the Interior Columbia Basin, including that of the goshawk. They evaluated change in source habitat from pre-european settlement (historical, circa ) to current (circa ) conditions for each species. Source habitats are defined as those characteristics of macrovegetation (vegetation that can be measured accurately using 100 hectare [247 acres] pixel) that contribute to stationary or positive population growth for a species in a specified area and time and to long-term population persistence. At the Interior Columbia Basin and Blue Mountains ERU scales the amount of source habitat of the goshawk has declined from historical conditions. The greatest declines have occurred in the Interior ponderosa pine forest types. Historically, the ponderosa pine forest type provided the greatest proportion of goshawk habitat at both the Interior Columbia Basin and Blue Mountains ERU scales. Timber harvest that began early in the 20th century is the primary cause of this decline. Additional declines were also noted in western larch and mixed conifer woodlands. In contrast, increased habitat occurred within Douglas-fir and Grand fir-white fir types as a result of fire exclusion. Analysis of Source Habitat on the WWNF Wales et al. (2011) analyzed source habitat of numerous wildlife species of interest in the Blue Mountains and WWNF in support of the Blue Mountains Forest Plan Revision. Source habitats are defined by Wales et al. as those stands that provide for a stable or increasing population and for all the life history needs of the goshawk including nesting, roosting, foraging, resting, travel, and dispersal. Vegetation parameters used to identify source habitat include Dry Ponderosa Pine, Dry Douglasfir, Dry Grand Fir, Cool Moist, and Cool Dry PVGs where average tree size is equal to or greater than 15 inches DBH. In addition, source habitats within Dry Ponderosa Pine, Dry Douglas-fir, Dry Grand Fir PVGs contain canopy closures equal to or greater than 40%, whereas canopy closure in Cool Moist and Cold Dry PVGs is equal to or exceeds 60 percent. Potential habitat is defined as stands within dry Douglas-fir, dry grand fir, cool moist and cold dry potential vegetation groups that have the capability to provide source habitat but that currently do not provide the tree size, canopy cover, or structural conditions. Given time and lack of human intervention or disturbance these areas may provide source habitat. Wales et al. (2011) estimated that approximately 466,679 acres of source habitat existing on the WWNF historically. Currently, approximately 440,696 acres (94% of estimated historical conditions) of source habitat occurs on the WWNF. Goshawk Viability on the WWNF 19

20 A species viability assessment was conducted for the goshawk in the Blue Mountain region of northeast Oregon and Washington, as well as for the WWNF following Regional guidance (Wales et al. 2011). The viability outcome model was utilized to assess species viability. The model provides a large-scale index of the capability of the environment to support population abundance and distribution. The viability outcome model used the BBN models previously described to calculate viability outcome scores. The viability outcome scores were derived from the WI scores, weighted watershed index (WWI) scores and the habitat distribution index. It is assumed that that species with high viability outcome scores would have a high probability of having populations that are self-sustaining and well distributed throughout their historic ranges. The current viability outcome index for the WWNF is primarily an A for which there is a 57% probability that goshawk habitats are broadly distributed and of high abundance. Results showed a lesser probability (27%) that habitats are broadly distributed and of high abundance, but with gaps (Outcome B ). The outcome scores indicate that suitable environments of the goshawk are broadly distributed and of high abundance and that the goshawk is likely well distributed throughout the WWNF. In conclusion, the viability assessment indicates that source habitat of the goshawk is still available in adequate amounts, distribution, and quality to maintain goshawk viability in the Blue Mountains and on the WWNF. (Penninger and Keown 2011c). Source Habitat at the Watershed Scale The watershed index (WI) scores incorporate the habitat departure, percentage of source habitat that is latesuccessional, and habitat effectiveness calculations previously discussed. The WI provides a measure of change in the amount of source habitat from historical conditions (departure) and the influences of habitat quality (amount of late-successional forest) and habitat effectiveness (influence of roads and trails). The calculation provided an index for each watershed with the values ranging from 0-3 (low: >0-1, moderate: >1- <2, high: >=2). Historically, it was estimated that the watershed index was approximately 2.94 for all the watersheds in the Blue Mountains and on the WWNF that were capable of providing goshawk habitat. The results are that nearly all (86%) of the watersheds on the WWNF had WI scores that were high. High WI values are indicative of low departure from the historical median amount of source habitat and relatively high amounts of source habitat that are also late successional. While there are generally high WI values, habitat effectiveness is still lower than historical habitat effectiveness due to the presence of roads and trails. WI for the Eagle Creek watershed is ranked High. The WWI provides a relative measure of the potential capability of each watershed to contribute to the viability of the goshawk. It was calculated from the WI by weighting it by the amount of source habitat that was currently available in each watershed. It is a measure of the current capability of an area to provide for the sustainability of a species compared to what the capability was historically. Watersheds with a WWI greater than 60 provide the greatest contribution to viability. WWI ranked High for the Eagle Creek watershed. The habitat distribution index is an assessment of how the watersheds with a relatively high amount of source habitat are connected and distributed across the planning area. It is calculated by the interaction of two variables; 1) the percentage of the total number of watersheds that met a threshold amount of source habitat, and 2) the number of clusters, or groups of watersheds, with at least one watershed that met the threshold amount of source habitat. The threshold of >40% of the historical amount or source habitat in a watershed was used to indicate watersheds with a relatively high amount of source habitat that would contribute to species viability. Watersheds that contain > 40% of the estimated historical median amount of source habitat are believed to provide for habitat distribution and connectivity, and better contribute to species viability across the forest. To assess how watersheds that exceeded 40% threshold were distributed across the forest the WWNF was divided into four areas. Each area contained a cluster of watersheds. Sub-basins (4th HUC) and forest ownership patterns were used to identify the watershed clusters. 20

21 Considering the 40% threshold, it s estimated that historically thirty-five of the forty-nine watersheds (71%) analyzed on the WWNF exceeded the 40% threshold. In these watersheds more than 40% of the potential habitat met the vegetative conditions of source habitat. Not all watersheds have the potential to provide source habitat for goshawk. Currently, thirty-two (91% of historical conditions; 65% of the watersheds on the WWNF) of the thirty-five watersheds that historically met this threshold, currently meet the threshold. Likely due to fire suppression, two watersheds on the WWNF currently provide more than 40% source habitat that did not historically provide habitat, bringing the total number of watersheds on the WWNF providing source habitat to 34 (69% of the watersheds on the WWNF). The analysis by cluster indicates that habitat is still well distributed across the WWNF and provides opportunity for habitat connectivity. Each cluster has a relatively high number of watersheds that meet or exceed 40% of the historical median amount of source habitat. The Eagle Creek watershed currently contains >40% of historical median source habitat and was included in cluster W3, within which 88% of watersheds currently have 40% of historical median of source habitat. Surveys within the Sparta Project Area Surveys to determine presence of goshawks were conducted within portions of the Sparta project area during the summer of 2011 and the summer of Responses were detected at two historical sites. No nests were located so future surveys will be concentrated in those areas before treatment. Per the Eastside Screens, if a nest is found, a no treatment 30 acre nest core buffer and a 400 acres PFA buffer will be implemented. Habitat Condition within the Sparta Project Area Habitat characteristics and methodology selected for this analysis are described in the Northern Goshawk (Accipiter gentilis) Management Indicator Species Assessment, Wallowa-Whitman National Forest (Penninger and Keown 2011c). Goshawk habitats modeled by Penninger and Keown (2011c) and Wales (2011c) include source and potential habitats. Source habitats are defined as those stands that provide for a stable or increasing population and for all the life history needs of the goshawk including nesting, roosting, foraging, resting, travel, and dispersal. Vegetation parameters used to identify source habitat include Dry Ponderosa Pine, Dry Douglas-fir, Dry Grand Fir, Cool Moist, and Cool Dry PVGs where average tree size is equal to or greater than 15 inches DBH. In addition, source habitats within Dry Ponderosa Pine, Dry Douglas-fir, Dry Grand Fir PVGs contain canopy closures equal to or greater than 40%, whereas canopy closure in Cool Moist and Cold Dry PVGs is equal to or exceeds 60 percent. Potential habitat is defined as stands within dry Dry Ponderosa Pine, Douglas-fir, Dry Grand Fir, Cool Moist and Cold Dry PVGs that have the capability to provide source habitat but that currently do not provide the tree size, canopy cover, or structural conditions. Given time and lack of human intervention or disturbance these areas may provide source habitat. Table 5. Existing Goshawk Source and Potential Habitats. Habitat Type Sparta Project Area Eagle Creek 5 th HUC Source 1,891 acres 27,058 acres Potential 5,789 acres 40,323 acres Direct and Indirect Effects Alternative 1 - No Action There will be no direct adverse effects to old-growth associated MIS from the No Action Alternative because no timber harvest, fuels treatments, or transportation activities will occur. Existing source habitat would remain unchanged. However, the no-action alternative maintains possible unsustainable conditions in lateseral stage montane forests where there have been large transitions from shade-intolerant to shade-tolerant tree species, described as a management issue for Group 6 habitats by Wisdom et al. (2000). 21

22 Action Alternatives 2 and 3 Sparta Project Area Both timber harvest and prescribed fire treatments within and outside timber harvest units would occur in northern goshawk source habitat under all action alternatives. Intermediate harvest treatments are expected to increase average stand diameter due to removal of trees primarily in smaller size classes, but across all size classes for Alternatives 2 and 3. Due to the possibility of snag removal during harvest and potential consumption of down logs during post-treatment underburning and in prescribed fire-only units, treatments that retain sufficient canopy closures are still expected to degrade, but not remove, source habitat. Although some habitat elements may be reduced where habitat is degraded, sustainability of habitats is expected to increase as stand density reductions lower the risk of disturbance such as stand-replacement fire, especially in Dry Forest types. Table 6 shows acres and percent of source habitat proposed for treatment under each alternative. Treatments proposed under Alternatives 2 and 3 would impact the greatest amount of goshawk source habitat. Harvest activities would occur within 395 acres of source habitat in alternative 2 and 105 acres in alternative 3. These harvest activities may alter 6-21% of goshawk source habitat within the Sparta project area for approximately 20 years until canopy closure recovers and snags and logs begin to be recruited. Although the treated acres may no longer meet the definition of source habitat, they would still be available for goshawk foraging, roosting, and travel between other habitat patches. Fuel management activities (precommercial thinning, hand piling and prescribed fire would occur within 125 acres of source habitat in alternative 2, and 127 acres in alternative 3. Fuel management will reduce structural complexity in the understory in 7% of goshawk source habitat in the project area, but it will still meet the requirements for source habitat. Table 6. Summary of Proposed Treatments in Goshawk Source Habitat Habitat Alternatives Elements Acres of Treatment in Habitat % of source habitat in project area Alternative 1 (Existing) Alternative 2 Alternative 3 Harvest Fuels Rx only Harvest Fuels Rx only 1, % 21% 7% 6% 7% Broom structures created by dwarf mistletoe have been identified as potential nesting substrate for goshawks. Moore and Henney (1983) reported that approximately 15% of goshawk nests studied in northeastern Oregon were located on mistletoe-created platforms. All action alternatives propose some level of mistletoe reduction in Douglas-fir and ponderosa pine (see American marten discussion, above) by killing heavily infected trees. Where treatments are proposed within currently suitable goshawk nesting habitat, the proposed action would degrade habitat suitability by reducing available nesting platform structure, though the trees would still be available on the landscape in the form of snags. This effect would be most pronounced for Alternatives 2. Retention of mistletoe-infected trees 21 inches DBH or greater under both action alternatives would allow some treatment units to retain mistletoe platforms in the largest trees most likely to support goshawk nesting. However, most goshawk nests usually occur within close proximity to water so the majority of potential nesting structures will be retained under all alternatives because RHCAs will retain all existing mistletoe brooms. Aspen restoration on scattered individual 1-acre sites would remove conifers on a total of 14 acres under Alternative 2 and 13 acres under Alternative 3, a portion of which may occur in suitable goshawk habitat. Activities that increase overall human presence and project-related noise levels, including system road 22

23 reconstruction as well as timber harvest, may temporarily displace goshawks locally in the short-term (i.e. during implementation), but are not expected to impact distribution within the project area in the long-term. In addition to impacts to available habitats, each action alternative poses potential for direct impact to nesting individuals. Both timber harvest and prescribed fire could cause individual harm or mortality if operations destroy a nest tree occupied by young of the year. If goshawk nesting is discovered prior to, or during implementation, a no-activity nest area of at least 30 acres will be designated for active nests. Because goshawks were detected at several locations during 2011 and 2016 surveys, and because the existing nest site was not confirmed with 100% certainty, additional goshawk surveys in these locations would occur prior to implementation of proposed silvicultural and prescribed fire treatments (see PDFs, Appendix A). Goshawk Habitat at the Watershed Level Watershed indices reported by Wales (2011c) and further assessed by Penninger and Keown (2011c) for the existing condition showed that the Eagle Creek watershed currently contains a high amount of source habitat. Treatments proposed under Alternative 2 would reduce the amount of source habitat available in the watershed by approximately 1.5 percent (Table 7). Source habitat would be reduced by 0.4% under Alternative 3. Post-treatment availability of source habitats would continue to exceed the threshold of 40% of the historical amount in the Eagle Creek watershed under all action alternatives, thereby continuing to contribute to species viability at the watershed scale. Table 7. Goshawk Source Habitats by Alternative, Eagle Creek Watershed. Existing and Post-Treatment Habitat Acreage Amounts by Alternative (% of Existing) Habitat Type Existing Condition Source 27,058 Goshawk Habitat at the WWNF Scale Alternative 2 Alternative 3 26,663 (98.5%) 26,953 (99.6%) Existing goshawk source habitat on the WWNF as modeled by Wales et al totals 440,696 acres. As a result of projected habitat loss under the Sparta project, source habitats at the Forest-level would decline by less than 1 percent under all action alternatives. Cluster analysis used to describe existing distribution of source habitats across the WWNF indicates that these habitats are well distributed across the Forest. Post-treatment levels of source habitat under all Sparta action alternatives result in no change in the number of watersheds in Cluster W3 containing >40% source habitat that contribute to goshawk habitat distribution. Cumulative Effects Cumulative effects for goshawk are analyzed for the Eagle Creek watershed. Past, present and reasonably foreseeable future actions were analyzed for cumulative impacts to the species. Effects of past activities including road construction, fire suppression, prescribed fire, and timber management on WWNF lands have been incorporated into the existing conditions for amounts and locations of marten habitats in the analysis areas. Although some commercial treatments may occur within goshawk suitable habitat, the scale of potential impacts is not substantial in comparison to source habitats currently estimated to exceed 27,000 acres. Appendix D of the EA was reviewed for actions that might affect goshawk habitat in the Eagle Creek watershed. Ongoing and future livestock grazing is expected to have a minimal effect on suitable habitats. Additional grazing may occur in treated stands within the project area, but is not expected to alter suitable characteristics. On Forest Service lands within and outside the project area, firewood cutting will continue to reduce available snags and logs, but the effect is limited to areas adjacent to open roads. Access within the 23

24 watershed and across the WWNF may change pending the outcome of the Forest Travel Management Plan. Timber harvest on private inholdings is expected to continue at some level, with anticipated reductions of trees larger than 10 inches DBH. Lands to the south of the project area will continue to consist of open grassland habitats in private ownership. Wales et al. (2011) estimated that approximately 466,679 acres of source habitat existed on the WWNF historically. At the time of the analysis, approximately 440,696 acres (94% of estimated historical conditions) of source habitat occurred on the WWNF. Since the viability assessment was run 14 Vegetation/Fuels Restoration projects have been analyzed across the Wallowa-Whitman. Some have been implemented and some are still undergoing the NEPA process but are anticipated being implemented in the foreseeable future. These combined projects, including the Sparta Vegetation Management project, anticipate commercially impacting 6,523 acres of goshawk source habitat and non-commercially impacting 18,877 acres of goshawk source habitat. Taking these 25,400 acres of impacted source habitat into account there is approximately 415,296 acres (89% of estimated historical conditions) of source habitat existing on the WWNF. Cumulatively, vegetation management activities on the Wallowa-Whitman are not expected to change the viability outcome found by Wales et al. and goshawk source habitat will remain well distributed and highly abundant (viability outcome A). Conclusion Because this project impacts less than 1% of source habitat across the Forest, the overall direct, indirect and cumulative effects will result in a small negative effect to goshawk habitat. The loss of habitat will be insignificant at the scale of the WWNF. Post-treatment availability of source habitats would continue to exceed the threshold of 40% of the historical amount in the Eagle Creek watershed under all action alternatives, thereby continuing to contribute to habitat distribution and species viability on the WWNF. PILEATED WOODPECKER The pileated woodpecker is a MIS associated with old-growth habitat, and represents species dependent on large diameter snags and logs in older-aged forests. It prefers dense, multi-story old growth habitats with high snag densities. It will use both coniferous and deciduous trees, but tends to be most common in oldgrowth ponderosa pine-mixed conifer forests in eastern Oregon (Csuti et al. 2001). Large snags are used for nesting, roosting, and foraging; logs are used for foraging. Nesting occurs in live and dead trees that are at least 2-3 ft in diameter. Carpenter ants, which are found in decaying wood, are its main food item, but it also eats the larvae of wood-boring beetles, termites, berries, and acorns. More vegetable matter is consumed during the winter. Life history, risk factors, conservation status and population trend, as well as habitat condition and species viability are described in detail in the Pileated Woodpecker (Dryocupus pileated) Management Indicator Species Assessment, Wallowa-Whitman National Forest (Penninger and Keown 2011c). Portions of that assessment are summarized below to describe existing condition for the species. Nesting The pileated woodpecker nests in large, dead trees. Typical nest trees have little bark remaining, few limbs, are more than 30 feet tall, and have broken tops (Bull 1987). In northeastern Oregon, Bull and Meslow (1977) reported all nest trees from their study area were greater than 23 inches DBH. Bull (1987) reported a strong selection of trees larger than 21 inches DBH for nesting. Large diameter trees are required to accommodate the large nest cavity (8 inches diameter and 22 inches deep). In a study conducted on the Starkey Experimental Forest, all but one of 123 different nests were located in dead trees. Ponderosa pine (73%) and western larch (25%) are most commonly used on the east side of the Cascade Range (Bull 1987). Pileated woodpeckers select nest snags from clusters of snags in dense types or stands (Bull and Meslow 24

25 1977). The species excavates new nest cavities each year, with trees containing relatively sound wood in the early stages of decay from heart-rot fungi preferred (Aubry and Raley 2002). Roosting Pileated woodpeckers roost inside tree cavities at night. Roost trees are presumed to be used to reduce predation and to conserve energy by minimizing heat loss in the winter. Trees used for nesting and roosting differ. In northeastern Oregon (Union, Baker, and Umatilla Counties) the majority of roost trees were in grand fir (62%), both live and dead, that were extensively decayed by Indian paint fungus (Echinodontium tinctorium) and had a hollow interior. Conks of Indian paint fungus were seen on 92% of the roost trees in grand fir (Bull et al. 1992). Roost trees are used all year with pileated woodpeckers using multiple roost trees. Bull et al. (1992) reported an average of 7 roost trees being used by an individual bird over a 3-10 month period (range 4-11 roost trees). Some birds used the same roost every night while others changed every few weeks. Old-growth stands of grand fir, with >60% canopy closure and little or no logging activity are preferred sites for roosting (Bull et al. 1992). Foraging Pileated woodpeckers feed primarily on insects in dead wood in snags, logs, and naturally created stumps (Bull and Meslow 1977, Bull et al. 1986). Numerous authors have reported carpenter ants (Camponotus spp.) as the primary food of pileated woodpeckers (Beal 1911, Hoyt 1950, Hoyt 1957, Bull and Jackson 1995). Down logs provide an important colonizing substrate for ants (Stevens 1997). Carpenter ants comprised 95% of the diet of pileated woodpeckers on the Starkey Experimental Forest (Beckwith and Bull 1985). In addition they are known to consume fruits, nuts, woodboring beetle larvae, and other insects. In northeastern Oregon, snags and logs >15 inches diameter were preferred, as were Douglas-fir and western larch (Beckwith and Bull 1985, Bull and Holthausen 1993, Bull et al. 1992), but foraging was documented on dead woody material as small as 2 inches in diameter. Stands with higher densities of snags and logs were preferred for foraging (Bull and Meslow 1977). Home Range Bull (1975) estimated nest territory size based on observations, vocalizations, and feeding locations. Estimates of area utilized by various pairs during reproduction ranged from 250 ha to 500 ha ( acres). Bull and Meslow (1977) reported that the minimum density for one pair was 1,620 acres of forest. During nesting, territories ranged from 320 to 600 acres (Bull and Meslow 1977). Bull (1987) reported an average nesting territory size of approximately 544 acres (220 hectares). Bull and Holthausen (1993) found that pileated woodpeckers in northeastern Oregon have an average home range size of 900 acres (range 800 1,575 acres). Bull et al. (2007) utilized data from many of these previous studies, comparing density of nesting pairs and traditional home ranges on 30 years of data on two study areas and 15 years of data on five study areas on the WWNF and Umatilla National Forests. The average territory size (including habitat and non-habitat), considering all seven study areas was 1,561 ha (3,857 acres) and average amount of suitable habitat within the territory was 765 acres. Broadscale Habitat Analysis Habitat trends of the pileated woodpecker were assessed at the Interior Columbia Basin, Blue Mountains ecological reporting unit (ERU), and WWNF scales using information provided by Source Habitats for Terrestrial Vertebrates of Focus in the Interior Columbia Basin (Wisdom et al. 2000) and the species viability assessment conducted by Wales et al. (2011) in support of the Blue Mountains Forest Plan revision. Wisdom et al (2000) define source habitat as those characteristics of macro-vegetation (vegetation that can be measured accurately using 100 hectare [247 acres] pixels) that contribute to stationary or positive population growth for a species in a specified area and time and to long-term population persistence. Source habitats contribute to source environments, which represent the composite of all environmental conditions 25

26 that result in stationary or positive population growth for a species in a specified area and time. Source habitats are distinguished from habitats simply associated with species occurrence; species occurrence by itself does little to indicate the capability of the environment to support long-term persistence of populations. In comparing existing amounts of source habitat to historical amounts, Wisdom et al. (2000) reported that, with the exception of the western larch cover type, all cover types in the Blue Mountains ERU that constitute pileated woodpecker source habitat have strongly increased. This is in contrast to the Basin (private and public lands in the interior Columbia Basin) which has experienced a decrease in the percentage of area providing source habitat. Western larch represents approximately 10 percent of the total live tree volume in the Blue Mountains, with a gradient of more to less abundance from north to south. Although Wisdom et al. (2000) documents a strong decrease in source habitat in the western larch cover type, this apparent decrease is not likely significant in regard to the abundance and distribution of source habitat for pileated woodpeckers in the Blue Mountains. Stands dominated by western larch are not rare, but they represent a minor proportion of the forested landscapes in the Blue Mountains (Penninger and Keown 2011c). Wales et al. (2011) analyzed source habitat of numerous wildlife species of interest in the Blue Mountains and WWNF in support of the Blue Mountains Forest Plan Revision. The analysis was conducted at the stand-scale, a much finer scale than the source habitat analysis conducted by Wisdom et al. (2000). Forested stands modeled as pileated woodpecker source habitat consist of Dry Douglas-fir, Dry grand fir, Cool Moist, and Cold Dry PVGs where average tree size is equal to or greater than 20 inches DBH. In addition, source habitats within dry Douglas-fir and dry grand fir PVGs contain canopy closures equal to or greater than 40%, whereas canopy closure in cool moist and cold dry PVGs is equal to or exceeds 60%. Source habitats provide for a stable or increasing population and for all the life history needs of the pileated woodpecker including nesting, roosting, foraging, resting, and travel. Areas modeled as potential habitat consist of stands within dry Douglas-fir, dry grand fir, cool moist and cold dry potential vegetation groups that have the capability to provide source habitat but that currently does not provide the tree size, canopy cover, or structural conditions. Given time and lack of human intervention or disturbance these areas may provide source habitat. Historically (considering the HRV), it is estimated that at any one point in time there were approximately 359,608 acres on the WWNF that provided source habitat. Currently, there are approximately 206,374 acres (57% of historical conditions) that meet the source habitat definition. Forty of the forty-nine watersheds (82%) analyzed for the pileated woodpecker on the WWNF have less source habitat than the estimated historical condition. The source habitat analysis determined that the abundance of closed-canopied forest with >20 DBH trees in the dry Douglas-fir, dry grand fir, cool moist, and cold dry forest types on the WWNF has declined from historical conditions. This is similar to the Basin-wide findings from the source habitat analysis conducted by Wisdom et al. (2000) but in contrast to the findings in the Blue Mountains ERU where the abundance of source habitat was strongly increasing. Snag size and density analysis conducted as part of the focal species assessment indicate that overall, densities of snags > 19.6 DBH were in the low and moderate categories on the WWNF, meaning that the majority of source habitat has fewer than 2.4 snags per acre. Historically, it is believed that the majority of source habitat for the pileated woodpecker was in the high density class and had between 2.4 and 7.3 snags per acre (Penninger and Keown 2011b). Road densities were variable but primarily in the low or moderate category. Open roads are believed to be a risk to snags due to firewood cutting and hazard tree felling activities. Snag densities are frequently lower adjacent to open roads than areas away from roads. Historically, the road density class was zero. Viability at the Forest Scale A species viability assessment was conducted for the pileated woodpecker in the Blue Mountain region of northeast Oregon and Washington, as well as for the WWNF following Regional guidance (2010). A viability outcome model was utilized to assess species viability. The model provides a large-scale index of 26

27 the capability of the environment to support population abundance and distribution. The viability outcome model used the BBN models previously described to calculate viability outcome scores. The viability outcome scores were derived from the watershed index (WI) scores, weighted watershed index (WWI) scores and the habitat distribution index. It is assumed that that species with high viability outcome scores would have a high probability of having populations that are self sustaining and well distributed throughout their historic ranges. Results show a 66% probability that suitable environments are distributed frequently as patches and/or exist at low abundance (outcome C ). For comparison, historically, it was 70% probable that habitat was broadly distributed and of high abundance. However, the viability assessment indicates the WWNF still provides for viability of the pileated woodpecker. In addition, the species is distributed across the WWNF and there are adequate amounts, quality, and distribution of habitat to provide for pileated woodpecker population viability (Penninger and Keown 2011c). Conditions at the Watershed Scale The watershed index (WI) scores incorporate the habitat departure, snag density, and road density calculations previously discussed. The WI provides a measure of change in the amount of source habitat from historical conditions (departure) and the influences of habitat quality (snag abundance) and risk factors (road density). The variables were weighted based on peer review of what was believed to have the strongest relationship and influence on the pileated woodpecker. Habitat departure was most heavily weighted, followed by snag density, and then road density. The calculation provided an index for each watershed with the values ranging from 0-3 (low: >0-1, moderate: >1-<2, high: >=2). For the WWNF, 39% of watersheds ranked Low, 29% ranked Moderate, and 33% ranked High. The Eagle Creek watershed, which encompasses the Sparta project area, ranked high in both watershed index and weighted watershed index scores. As part of a cluster analysis, historical and current source habitat amounts were assessed for watersheds on the WWNF. A threshold of>40% of the historical amount or source habitat in a watershed was used to indicate watersheds with a relatively high amount of source habitat that would contribute to species viability. Watersheds that contain > 40% of the estimated historical median amount of source habitat are believed to provide for habitat distribution and connectivity, and better contribute to species viability across the forest. The Eagle Creek watershed currently exceeds the 40% threshold and is part of a cluster where 87% of watersheds currently exceed the same threshold. Results indicate favorable pileated woodpecker source habitat conditions at the watershed and cluster scales as well as distribution of habitats. Current Condition within the Sparta Project Area Surveys to determine presence of pileated woodpeckers were conducted within portions of the Sparta project area during the summer of The species was detected in 4 locations. Analysis of habitats modeled by Penninger and Keown (2011c) shows approximately 1,441 acres of source habitat within the Sparta project area and about 18,569 acres of source habitat within the Eagle Creek watershed. Distribution of source habitats is patchy throughout the project area, with notable concentrations located north of Sparta in the Eagle Creek watershed Direct and Indirect Effects Alternative 1 There will be no direct adverse effects to pileated woodpeckers from alternative 1 because no timber harvest, fuels treatments, or transportation activities will occur. Existing source habitat for pileated woodpeckers would remain unchanged. The no-action alternative maintains potentially unsustainable conditions in warm, dry LOS forests where there have been large transitions from shade-intolerant to shade-tolerant species. In the near-term, these denser forests with greater structural complexity may be highly attractive to pileateds. However, large uncharacteristic wildfires could eventually render pileated habitat unsuitable. 27

28 Alternatives 2 and 3 Both timber harvest and prescribed fire treatments within and outside timber harvest units would occur in pileated woodpecker source habitat under all action alternatives. Thinning harvest treatments are expected to increase average stand diameter due to removal of trees primarily in smaller size classes, but across all size classes for Alternatives 2 and 3. Treatments that retain canopy closures that meet the definition of source habitat would remain as source habitat. However, due to the possibility of minor snag reductions for logging safety, and potential consumption of down logs and snags during post-treatment burning and in prescribed fire units, treatments that retain sufficient canopy closures are still expected to degrade, but still function as source habitat. Although some habitat elements may be reduced where habitat is degraded, sustainability of habitats is expected to increase as stand density reductions lower the risk of disturbance such as standreplacement fire, especially in warm, dry forest types. Table 8 shows acres and percent of source habitat proposed for treatment under each alternative. Table 8. Summary of Proposed Treatments in Pileated Source Habitat. Source Habitat Alt 1 (Existing) Acres Affected Alternative 2 Alternative 3 Harvest Rx Fire Harvest Rx Fire Acres 1, % of source habitat in project area Treatments proposed under Alternative 2 would impact the largest amount of pileated source habitat. Harvest activities would occur within 295 acres of source habitat in alternative 2 and 83 acres in alternative 3. These harvest activities may alter 6-20% of pileated source habitat within the Sparta project area for approximately 20 years until canopy closure recovers and snags and logs begin to be recruited. Prescribed fire and fuels activities would occur within 76 acres of source habitat in both alternative 2 and 3. Prescribed fire will reduce structural complexity in the understory in 5% of pileated source habitat in the project area but it will still meet the requirements for source habitat. Retention of all snags except for safety concerns minimizes the potential for direct impacts to nesting pileated woodpeckers. Under alternatives 2 and 3 a portion of trees over 21 dbh in stands heavily infested with mistletoes will be girdled. These trees will remain on the landscape and provide additional roosting, foraging and nesting opportunities for pileateds. In the long-term, accelerated tree growth due to lower stocking densities is expected to develop large trees, and consequently large snags, at a faster rate than untreated areas. While long-term availability of total snag numbers may decrease, available snags will on average be larger in treatment units compared to untreated areas. Aspen restoration on scattered sites would remove conifers on a total of 14 acres under Alternative 2 and 13 acres under Alternative 3, a portion of which may occur in suitable pileated woodpecker habitat. Activities that increase overall human presence and project-related noise levels, including system road reconstruction as well as timber harvest, may temporarily displace pileated woodpeckers locally in the short-term (i.e. during implementation), but are not expected to impact distribution or productivity within the project area in the long-term. Pileated Woodpecker Habitat at the Watershed Level Watershed indices reported by Wales (2011b) and further assessed by Penninger and Keown (2011b) for the existing condition showed that the Eagle Creek watershed contains a high amount of source habitat and a lesser amount of secondary habitat. Treatments proposed under Alternative 2 and 3 would reduce the amount of source habitat available in the watershed by 1.6 and 0.5 percent, respectively. (Table 9). 28

29 Post-treatment availability of source habitats would continue to exceed threshold of 40% of the historical amount in the Eagle Creek watershed under all action alternatives, thereby continuing to contribute to species viability at the watershed scale. Table 9. Existing and Post-treatment Source Habitat Amounts, Eagle Creek Watershed. Habitat Type Source 18,569 Pileated Woodpecker Habitat at the WWNF Scale Existing and Post-Treatment Habitat Acreage Amounts by Alternative (% of Existing), Eagle Creek Watershed Existing Condition Alternative 2 Alternative 3 18,274 (98.4%) 18,486 (99.5%) Existing pileated woodpecker source habitat on the WWNF as modeled by Wales (2011b) totals 129,943 acres. As a result of projected habitat loss under the Sparta project, source habitats would decline by an estimated 295 acres under Alternative 2 and about 83 acres under Alternative 3. This results in a reduction in source habitat of 0.3% at the Forest level under Alternative 2 and a reduction of 0.1% under alternative 3 within the WWNF. Cluster analysis used to describe existing distribution of source habitats across the WWNF indicates that these habitats are well distributed across the Forest (Penninger and Keown 2011b). Post-treatment levels of source habitat under all Sparta action alternatives result in no change in the number of watersheds in Cluster W3 containing >40% source habitat that contribute to pileated woodpecker habitat distribution. Cumulative Effects Past, present and reasonably foreseeable future actions were analyzed for cumulative impacts to the species. Effects of past activities including road construction, fire suppression, prescribed fire, and timber management on WWNF lands have been incorporated into the existing conditions for amounts and locations of source habitats in the analysis area. Appendix D of the EA was reviewed for actions that might affect pileated habitat within the Eagle Creek watershed. Cumulative impacts of ongoing and foreseeable actions within the next 5 years from the present which overlap in time and space with the Sparta project and create a potentially measureable effect were considered. Ongoing and future livestock grazing is expected to have no effect on suitable habitats. Additional grazing may occur in treated stands within the project area, but is not expected to alter source habitats. On Forest Service lands within and outside the project area, firewood cutting will continue to reduce available snags and logs, but the effect is primarily limited to areas adjacent to open roads. Access within the watershed and across the WWNF will change when the Forest Travel Management Plan is implemented. Limiting public motor vehicle use to designated roads, trails and areas has the potential to reduce the miles of open roads where firewood gathering can reduce snags and logs. Timber harvest on private inholdings is expected to continue at some level, with anticipated reductions of trees larger than 10 inches DBH. Lands to the south of the project area will continue to consist of open grassland habitats in private ownership. Wales et al. (2011) estimated that approximately 359,608 acres of source habitat existed on the WWNF historically. At the time of the analysis, approximately 206,374 acres (57% of estimated historical conditions) of source habitat occurred on the WWNF. Since the viability assessment was run 14 Vegetation/Fuels Restoration projects have been analyzed across the Wallowa-Whitman. Some have been implemented and some are still undergoing the NEPA process, but are anticipated being implemented in the foreseeable future. These combined projects, including the Sparta Vegetation Management project, anticipate commercially impacting 3,020 acres of pileated source habitat and non-commercially impacting 9,857 acres of pileated source habitat. Taking these 12,877 acres of impacted source habitat into account, this results in 29

30 approximately 193,497 acres (54% of estimated historical conditions) of source habitat existing on the WWNF. Cumulatively, vegetation management activities on the Wallowa-Whitman are not expected to change the viability outcome found by Wales et al. and pileated source habitat will remain distributed frequently as patches and in low abundance (Viability outcome C). Conclusion Because this project impacts less than 0.5% of suitable habitat across the Forest, the overall direct, indirect and cumulative effects will result in a small negative effect to pileated habitat. The reduction of habitat would be immeasurable at the WWNF scale. Post-treatment availability of source habitats would continue to exceed the threshold of 40% of the historical amount in the Eagle Creek watershed under all action alternatives, thereby continuing to contribute to habitat distribution and species viability on the WWNF. PRIMARY CAVITY EXCAVATORS (PCE) The Forest Plan identifies 15 primary cavity excavators as management indicator species (MIS) for the availability and quality of dead and defective wood habitat: northern flicker; black-backed, downy, hairy, Lewis, three-toed, and white-headed woodpeckers; red-naped and Williamson s sapsuckers; black-capped, chestnut-backed, and mountain chickadees; and pygmy, red-breasted, and white-breasted nuthatches. Pileated woodpecker is also a primary cavity excavator, but due to its reliance on larger snags and trees, it is addressed separately as an MIS. The abundance of cavity-using species is directly related to the presence or absence of suitable cavity trees. Habitat suitability for cavity-users is influenced by the size (diameter and height), abundance, density, distribution, species, and decay characteristics of the snags. In addition, the structural condition of surrounding vegetation determines foraging opportunities (Rose et al. 2001). Not every stage of the snag s demise is utilized by the same species, but rather a whole array of species use the snag at various stages or conditions. Uses of snags include nesting, roosting, foraging, perching, courtship, drumming, and hibernating. Existing Conditions Black-backed Woodpecker Distribution: Across Canada, north to central Alaska; Rocky Mountains south to northern Wyoming; Cascades Range south to the Sierras. In Oregon, the species is distributed in the high elevations of west Cascades; east slope of Cascades through the Blue, Ochoco and Warner Mountains (Marshall et al. 2003). Habitat: Source habitat is old-forest stages of subalpine, montane, lodgepole, lower montane and riparian forests as well as young forest stages of lodgepole pine. Snags are an important habitat component. The species is known to nest in ponderosa pine, lodgepole pine, and western larch in the Blue Mountains. Blackbacked woodpeckers require conditions that produce bark and wood-boring beetle sources including fire-, wind-, or insect-killed mature or older pines and other trees with flaky bark, although it is positively associated with higher densities of small (9-15 DBH) trees and snags. Use of snags for foraging declines after 2-3 years. Nesting habitat includes mature and old trees infested with disease or heart-rot, or dead trees in the early stages of decay (Wisdom et al. 2000). Black-backed woodpeckers use managed and unmanaged forests with high levels of insect-infested trees. In burned stands, the species reaches highest densities in un-salvaged, recent (1-5 years) post-fire habitat with high densities of snags. Older burns do not support the high levels of wood-boring beetles that attract them to the recent burns. Threats and Risk Factors: Removal of fire-killed or insect-infested trees, altered frequency of standreplacing fire, decline in availability of medium to large snags infected with heart rot (Wisdom et al. 2000, NatureServe 2010). 30

31 Northern Three-toed Woodpecker Distribution: Holarctic distribution: Scandinavia, northern Europe, and Asia. In North America from Alaska to Newfoundland; Rocky Mountains south to northern Arizona and New Mexico; Cascades to southern Oregon; absent from the Sierras. In Oregon, it occurs in the high elevations of the Cascades and Blue Mountains (Marshall et al. 2003). Habitat: Source habitat is old forests in subalpine and montane forest (lodgepole pine, grand fir-white fir, Engelmann spruce-subalpine fir, white-bark pine and mountain hemlock). Specific habitats used consist of mature and overmature conifer stands with bark beetles, disease, and heartrot as well as recent standreplacement burns with abundant wood-boring insects. Snags used for nesting generally range from 9 to 20 inches DBH. Because snags are used for foraging as well as nesting, large burns and beetle-infested stands are strongly favored over unburned or noninfested stands (Wisdom et al. 2000). Threats and Risk Factors: Decline in late-seral subalpine and montane forests, especially Engelmann sprucesubalpine fir. Removal of fire-killed or insect-infested trees, altered frequency of stand-replacing fire, decline in availability of medium to large snags infected with heart rot (Wisdom et al. 2000, NatureServe 2010) Downy Woodpecker Distribution: Widespread permanent resident from Alaska across Canada, south to southern California across to the Gulf coast to south Florida. In Oregon, distributed across the state in appropriate habitats at low to moderate elevations (Marshall et al. 2003). Habitat: Deciduous riparian woodlands and lowland deciduous forest (alder, cottonwood, willow, aspen, oaks). Also found in urban parks and orchards. Low and mid-elevations. Nest primarily in dead trees (Marshall et al. 2003). In northeastern Oregon, the species most often uses quaking aspen larger than 6 DBH for nesting (Thomas et al. 1979). Threats and Risk Factors: Replacement of hardwood habitat with coniferous habitats (Marshall et al. 2003). Hairy Woodpecker Distribution: Widespread permanent resident from Alaska across Canada, south to Baja California across to the Gulf coast to south Florida, the Bahamas and west Panama. In Oregon it breeds in the east slope of the Cascades, Blue Mountains, and riparian areas across eastern Oregon and is a local and transient west of the Cascades (Marshall et al. 2003). Habitat: Dry and wet coniferous forests at low to mid-elevations. Also uses deciduous forest and riparian areas, especially if adjacent to coniferous forest. The species uses all ages of forest stands, though some authors report preference for older stands for nesting. Nest primarily in moderately decayed snags. Most nesting occurs in dead trees. In northeastern Oregon, ponderosa pine is the dominant species selected for nesting, although all tree species except grand fir were used. The species nests primarily in relatively open ponderosa pine stands with low basal area, low stem density, and open canopies. Snags most commonly selected for nesting ranged from DBH and were likely dead for less than 5 years (Bull et al. 1986). In burned forest habitats, these woodpeckers reach their highest densities in un-salvaged, recent (1-5 years) post-fire habitat with moderate to high densities of snags. Older burns do not support the high levels of wood-boring beetles that attract them to the recent burns (Cahall 2007, Cahall and Hayes 2008, Haggard and Gaines 2001, Kreisel and Stein 1999, Saab et al. 2007). Threats and Risk Factors: House sparrows usurping nests (NatureServe 2010). 31

32 Northern Flicker Distribution: Breeds from southeast Alaska, east to the west edge of the Great Plains, south to Mexico. In Oregon, it occurs across the state in appropriate habitats at low to moderate elevations (Marshall et al. 2003). Habitat: This species is a habitat generalist, though most abundant in open forests or forest edges. It uses coniferous and deciduous forest, riparian woodlands, and urban areas. Nests are in large snags (Marshall et al. 2003). In northeastern Oregon, northern flickers frequently nest in ponderosa pine, preferring snags greater than 20 inches DBH that are dead for less than 5 years. Nest areas typically contain low stem density, low basal area, and canopy closures less than 35 percent. Grassland areas adjacent to nest sites were preferred foraging areas (Bull et al. 1986). Within post-fire areas, the species is most abundant in areas with medium snag density. Population densities were significantly higher in unsalvaged areas and 5 or more years post-fire (Saab et al. 2007). Threats and Risk Factors: Loss of large soft to moderate snags (Marshall et al. 2003). Lewis s Woodpecker Distribution: Breeds from southwestern Canada and Montana, south to central California, Colorado, Arizona, and New Mexico. Winters from northern Oregon, south to northern Mexico and west Texas. In Oregon, it breeds in the east slope of the Cascades, Blue Mountains, and riparian areas across eastern Oregon. It is a local and transient west of the Cascades. Habitat: Source habitat includes oak woodlands, pine-oak habitat, old-forest in ponderosa pine vegetation types, cottonwood groves and burned forest (Wisdom et al. 2000). The species five major habitat types include ponderosa pine, oak-pine woodlands, cottonwood riparian forests, and areas burned by fire. Special needs consist of aerial insect populations for foraging, large soft or well-decayed snags for nesting, and relatively open canopy for flycatching (ODFW 2006). Thomas (1979) identified the minimum snag diameter suitable for Lewis woodpecker is 12 inches, while Saab and Vierling (2001) reported average snag size used by the species in conifer stands as about 18 inches DBH. Habitat suitability is 0.5 or greater when canopy closure is less than 50% and optimal (1.0) when canopy is less than 30%. Other components of suitable habitat include retention of one snag per acre greater than 12 inches DBH available shrub layer (Sousa 1983). In burned areas, the species is most abundant in areas with low snag density (<5 acre); most abundant in salvaged stands the first 4 years after fire; from 5-12 years post fire, no preference for salvaged stands (Haggard and Gaines 2001, Saab et al. 2007). Threats and Risk Factors: Loss of large pines and cottonwoods, loss of large, soft snags; fire suppression resulting in ponderosa pine being replaced by Douglas-fir and grand fir-white fir; and potential loss of remaining large pines due to fire. Reduced shrub cover due to fire suppression and grazing. Overgrazing in riparian habitats. Negative impacts from agricultural pesticides (Wisdom et al. 2000, NatureServe 2010). White-headed Woodpecker Distribution: Resident from southern interior British Columbia, though central Washington, northern Idaho, east and southwest Oregon, through the Sierras to southern California. In Oregon, it occurs on the east slope Cascades, Ochoco, Blue and Warner Mountains. It is rare west of the Cascades in the Siskiyou Mountains and Umpqua River basin (Marshall et al. 2003). Habitat: Source habitat is old forests in ponderosa pine cover types (Wisdom et al. 2000). Suitable stand conditions include canopy closure ranging from 10-40% with trees greater than 21 inches DBH at densities greater than 10 trees per acre (ODFW 2006). Mean size of nest trees reported from the eastern Cascades of Washington is 20 inches DBH (Buchanon et al. 2003), and inches in the eastern Cascades of Oregon 32

33 (Dixon 1995 in Marshall et al. 2003). Foraging consists of gleaning trees for insects and cone-feeding. Forage habitat is most commonly characterized as open pine stands with large trees. Availability of mature cone-producing pine trees is important in providing foraging habitat during the winter (Garrett et al. 1996). In post-fire areas, a mosaic of burn severities across the landscape provides habitat for white-headed woodpeckers by opening forest canopies in burned areas, while retaining decayed snags created before wildfire and live, cone-producing trees in unburned or low-severity burn areas. Threats and Risk Factors: Habitat degradation: Loss of large pines, loss of large snags; fire suppression resulting in ponderosa pine being replaced by Douglas-fir and grand fir-white fir; and potential loss of remaining large pines due to fire (Wisdom et al. 2000, NatureServe 2010). Williamson s Sapsucker Distribution: East slope of Cascades to the Rocky Mountains, from southern British Columbia south into Mexico. Winters from northern California south to Arizona, New Mexico, Texas and through Mexico. In Oregon it breeds east slope Cascades, Blue and Warner Mountains (Marshall et al. 2003). Habitat: Source habitats consist of open, late-seral stages of montane and lower montane forests (Douglasfir, western larch, grand fir and white fir, and ponderosa pine) and aspen and cottonwood stands (Wisdom et al. 2000). In northeastern Oregon, the species is found in grand fir and Douglas-fir forest types with high densities of snags. Nest sites were positively associated with density of large snags and negatively associated with large live tree density, slope and log density (Nielsen-Pincus and Garton 2007). Threats and Risk Factors: Reduction in late-seral forest habitat, reduction in numbers of large snags (Wisdom et al. 2000). Red-naped Sapsucker Distribution: Breeds from the Rocky Mountains west to the crest of the Cascades and the Sierras, from southern Canada south to Arizona, New Mexico, and west Texas. Winters southern US to central Mexico. In Oregon, it is widespread east of the Cascades (Marshall et al. 2003). Habitat: Breeding habitat is conifer and mixed conifer-hardwood forest. Typically found in riparian habitats, particularly aspen as well as in other hardwood stands, and less frequently found in mixed conifer forest. East of the Cascades they occur in young, mature, and old-growth forests, but are absent from very young stands. During non-breeding season they may occur in deciduous woods and urban areas. Nests occur in snags or dead limbs (Huff and Raley 1991, Marshall et al. 2003). In northeastern Oregon, the species was found to utilize both unmanaged old-growth as well as managed stands where snags of all size classes were used, but larger snags were preferred (Mannan and Meslow 1984). Level of use between unburned area and areas burned with low to moderate intensity was similar, but lower in areas of high intensity burn (Marshall et al. 2003). Threats and Risk Factors: Loss of large aspen snags and trees (Marshall et al. 2003). Pygmy Nuthatch Distribution: Western North America from southern British Columbia to northern Mexico, east to northwest Nebraska. In Oregon, it occurs from the east slope of the Cascades eastward to the Blue and Warner Mountains (Marshall et al. 2003). Habitat: Source habitat is old forests in ponderosa pine cover types. Nest sites occur in open ponderosa pine forests with large (mean ~18 DBH) trees and snags (Wisdom et al. 2000). In northeastern Oregon, areas 33

34 surrounding nest trees consisted predominately of ponderosa pine with canopy closure averaging about 27 percent (Nielson-Pincus 2007). Threats and Risk Factors: Habitat degradation: Loss of large pines, loss of large snags; fire suppression resulting in ponderosa pine being replaced by Douglas-fir and grand fir-white fir; and potential loss of remaining large pines due to fire (Wisdom et al. 2000). Red-breasted Nuthatch Distribution: Breeds from southeast Alaska across Canada, south to the mountains of the western US, the Great Lakes states, New England, and the Appalachians. Oregon: breeds and winters throughout forested areas of the state (Marshall et al. 2003). Habitat: Breeding habitat is conifer and mixed conifer-hardwood forest. East of the Cascades they occur in young, mature, and old-growth forests, but are absent from very young stands. During non-breeding season they may occur in deciduous woods and urban areas. Nests occur in snags or dead limbs (Huff and Raley 1991, Marshall et al. 2003). In northeastern Oregon, the species was found to utilize both unmanaged old-growth as well as managed stands where snags of all size classes were used, but larger snags were preferred (Mannan and Meslow 1984). Level of use between unburned area and areas burned with low to moderate intensity was similar, but lower in areas of high intensity burn (Marshall et al. 2003). Threats and Risk Factors: Loss of snags and structural diversity (Marshall et al. 2003). White-breasted Nuthatch Distribution: Widely distributed resident bird across North America. In the west, from southern Canada into northern Mexico, across southern Canada to the east coast from Prince Edward Island south to northern Florida. In Oregon, it occurs in low elevation forests west and east of the Cascades, interior valleys (Marshall et al. 2003). Habitat: Source habitat consists of old forests in ponderosa pine cover types, but also breeds in old forests of cottonwood-willow, Oregon white oak, and in unmanaged young forests of interior ponderosa pine. Nests occur more frequently in the cavities of live ponderosa pine than in snags. Suitable nest sites occur within the upper diameter classes of trees and snags, with nest tree diameter averaging 21 inches DBH (Wisdom et al. 2000). Threats and Risk Factors: Habitat degradation: Loss of large pines, loss of large snags; fire suppression resulting in ponderosa pine being replaced by Douglas-fir and grand fir-white fir; and potential loss of remaining large pines due to fire. Also, decline in old stands of aspen and cottonwood (Wisdom et al. 2000). Black-capped Chickadee Distribution: Resident from west-central Alaska and Canada, east to the Atlantic Coast, and south the northern California and North Carolina. In Oregon, primarily found in western Oregon and described as relatively uncommon in grand fir forests of the Blue Mountains (Marshall et al. 2003). Habitat: Nesting habitat in the Blue Mountains consists of quaking aspen and riparian hardwood sites. Minimum diameter of snags used for nesting is 6 inches DBH. Stand ages used for both reproduction and foraging range from young to late-old successional stages (Thomas et al. 1979). Threats and Risk Factors: Destruction and removal of riparian habitats (Marshall et al. 2003). 34

35 Chestnut-backed Chickadee Distribution: Breeds from southeast Alaska, western British Columbia, northern Idaho, and northwestern Montana, south through the Coast Ranges to southern California. It is a resident in Oregon along the coast and throughout most of the Coast Range and Cascades, and rare and locally uncommon in the eastern Blue and Wallowa mountains (Marshall et al. 2003). Habitat: Nesting habitats in northeastern Oregon include upland conifer with ponderosa pine, mixed conifer, grand fir, lodgepole pine, and subalpine fir. Minimum diameter of snags used for nesting is 6 inches DBH. Stand ages used for both reproduction and foraging range from young to late-old successional stages (Thomas et al. 1979). Threats and Risk Factors: Reductions in large snags and logs in habitats that have been managed under traditional silvicultural practices; possibly unsustainable conditions in late-seral stage montane forests that have transitioned from shade-intolerant to shade-tolerant species; degradation and loss of riparian habitat (Wisdom et al. 2000). Mountain Chickadee Distribution: Confined to middle to high elevations in Oregon, Washington and the Rocky Mountain states. In Oregon, it is a forest resident from west of the Cascades, east through the Blue Mountains (Marshall et al. 2003). Habitat: Nesting and foraging habitats in northeastern Oregon consist of upland conifer including ponderosa pine, mixed conifer, grand fir, lodgepole pine, and subalpine fir. Minimum diameter of snags used for nesting is 6 inches DBH. Stand ages used for both reproduction and foraging range from young to late-old successional stages (Thomas et al. 1979). Threats and Risk Factors: Loss of aspen and large conifer as nesting habitat (Li and Martin 1991). Conservation Status of PCE Species Conservation status is assessed at the global, regional, and state level by several sources. Table 10 summarizes conservation status at various levels. Species of highest concern consist of the white-headed and Lewis woodpeckers; both of which are classified as Sensitive in USFS Region 6 and by the Oregon Department of Fish and Wildlife. Black-backed and three-toed woodpeckers are also listed as Oregon State Sensitive, but show a lower degree of concern, ranked Vulnerable. 35

36 Table 10. Conservation Status of Cavity-nesting MIS. Species USFS Sensitive NatureServe Ranks 1 Global OR USFWS Birds of Conservation Concern 2 ODFW 3 Black-backed woodpecker G5 S3 Vulnerable Northern three-toed woodpecker Downy woodpecker G5 S4 Hairy woodpecker G5 S4 Northern flicker G5 S5 G5 S3 Vulnerable Lewis s woodpecker Yes G4 S2 S3 BCR 9, BCR 10 Critical White-headed woodpecker Yes G4 S2 S3 BCR 9, BCR 10 Critical Red-naped sapsucker G5 S4 Williamson s sapsucker G5 S4B S3N BCR 9, BCR 10 Pygmy nuthatch G5 S4 Red-breasted nuthatch G5 S5 White-breasted nuthatch G5 S4 Black-capped chickadee G5 S5 Chestnut-backed chickadee Mountain chickadee G5 S4 1 NatureServe Ranks: (NatureServe 2010) G5 or S5 Widespread, abundant, secure G4 or S4 Apparently secure G3 or S3 Vulnerable G2 or S2 Imperiled 2 Species of Concern in any BCR (Bird Conservation Region) Listed (USFWS 2008) 3 Oregon Department of Fish and Wildlife Sensitive Species ( Population Trend G5 Population trend data is obtained for PCE species mainly from two sources. Breeding bird survey data collected long-term along established survey routes provides indices for species population trends at the state level. A stable trend in Oregon is shown for 10 of the 15 PCE species, with long-term decreases shown for two species (northern flicker and mountain chickadee). Trend information is inconclusive for three-toed and white-headed woodpecker, likely due to low overall detection rates. It is not uncommon for some species to be under-represented in multi-species survey techniques since some species are more easily detected than others. Regional indices of species security are also assessed under the Partners in Flight program. Scores for Bird Conservation Region 10, which includes the WWNF, are shown in Table 11. Scores above 13 indicate a species may be of regional concern, and these include the black-backed woodpecker, Lewis woodpecker, pygmy nuthatch, red-naped sapsucker, white-headed woodpecker and Williamson s sapsucker. The combination of the state and regional indices shows stability and regional scores below the threshold of concern for 6 of the 15 species. Increased concern, but stable state trend is shown for 4 species, while decreasing trend and lower regional concern is shown for 2 species. Both white-headed woodpecker and Lewis woodpecker show the highest regional concern without available trend information at the state level. Three-toed woodpecker also contains no state trend information, but the regional score is below the threshold of concern. S5 36

37 Table 11. Population Trend Data for Cavity-nesting MIS. Species Breeding Bird Surveys 1 Partners in Flight Database 2 OR Reliability BCR 10 Black-backed woodpecker stable yellow 14 Downy woodpecker stable yellow 10 Hairy woodpecker stable blue 10 Lewis s woodpecker no trend red 18 Northern flicker decrease blue 13 Northern three-toed woodpecker No data 13 Pygmy nuthatch stable yellow 14 Red-breasted nuthatch stable blue 11 Red-naped sapsucker stable yellow 17 White-breasted nuthatch stable blue 8 White-headed woodpecker no trend red 18 Williamson s sapsucker stable blue 17 Black-backed chickadee stable blue 13 Chestnut-backed chickadee stable blue 11 Mountain chickadee decrease blue 12 1 Breeding Bird Survey (Sauer et al. 2011) - watch reliability ratings ( o Increase = significant (p<0.05) increase from o Decrease = significant (p<0.05) decrease from o Stable = yellow or blue reliability and no significant increase or decrease o No trend = red reliability and no significant increase or decrease Oregon - ( 2 Partners in Flight (PIF) Database ( Regional Combined Scores can range from 5 to 25. Regional Combined Score > 13 may be a species of Regional Concern (Panjabi et al. 2005). o Northern Rockies BCR 10 Habitat Trend at the Regional Level Wisdom et al. (2000) analyzed the current amount of available habitat in relation to historical availability for a range of species, producing a trend index by Ecological Reporting Unit (ERU). In the Blue Mountains, trend indices were reported for 8 of the 15 PCEs (Table 12). Decreasing or Strongly Decreasing trend is indicated for 6 of the 8 species. Strongest decreases are shown for species associated with mature ponderosa pine habitats with large snags (white-headed woodpecker, pygmy nuthatch) as well as single-story old forest, multi-storied Douglas-fir and western larch, and riparian cottonwood woodlands (Lewis woodpecker). Factors contributing to these declines include transition to shade-tolerant tree species, past timber harvest, and increased roading that allowed snag removal for firewood (Wisdom et al. 2000). Black-backed woodpecker habitats, consisting of a range of green and burned forest condition, have also decreased at the regional level due to past timber harvest, firewood removal, and fire suppression, The highest regional increase is shown for three-toed woodpecker, which is associated with late seral subalpine and montane conifer. Table 12. Long-term Regional Trend of PCE Habitats (Wisdom et al. 2000). ERU Relative Change & Species Group Trend Category Blue Mountains Black-backed woodpecker Decreasing Lewis s woodpecker Strongly decreasing 37

38 Species Northern threetoed woodpecker Pygmy nuthatch White-breasted nuthatch White-headed woodpecker Chestnutbacked chickadee Williamson s sapsucker Group ERU Relative Change & Trend Category Blue Mountains 100+ Strongly increasing Strongly decreasing Decreasing Strongly decreasing No change Decreasing R6 Decayed Wood Advisor (DecAID) The use of DecAID is a culmination of the most recent science and data available. As stated by Rose et al. (2001), DecAID is based on a thorough review of the literature, available research and inventory data, and expert judgment. Tolerance intervals are estimates of the percent of all individuals in the population that are within some specified range of values. In the case of the USFS Region 6 decayed wood advisor (DecAID) (Mellen-McLean et al. 2012), for example, they tell us what percent of pileated woodpeckers in a population use snags, up to or above certain diameters. Thus, an 80% tolerance level indicates 80% of the individuals in the population have a value for the parameter of interest (i.e. snag density) between 0 and the value for the 80% tolerance level. Or conversely, 20% of the individuals in the population have a value for the parameter of interest greater than the 80% level. The tolerance interval is the range between 2 tolerance levels. For example, the value for 80% is the level and the range of 0 to 80% is the interval. DecAID displays three tolerance levels (30%, 50% and 80%) for density and DBH class of snags in various vegetation condition groups, used by wildlife species. Ranges of snag densities by DBH class are provided as a synthesis of data from various studies. In the Sparta project area, vegetation condition groups consist of Ponderosa Pine/Douglas-fir and Eastside Mixed Conifer habitats and include both Large Tree and Small/Medium tree habitat types. The ranges of snag densities by DBH class for each vegetation group and habitat types are shown in Table 13. Wildlife snag use levels shown in Table 14 exclude those estimated for white-headed woodpecker because the data are from a declining population (Mellen-McLean et al. 2012). Table 13. Snag and Down Log Parameters and Tolerance Levels (source: DecAID). Habitat Type/ Structure Tolerance Levels Snag Densities Used by Wildlife by Diameter Class 80% > > 20 Ponderosa Pine/ > 10 50% Douglas-fir (Large) 1-8 > 20 30% > > > 10 80% > 20 Ponderosa Pine/ > 10 Douglas-fir 50% > 20 (Small/Medium) > 10 30% > 20 Mixed Conifer (Large) > 10 80% 4-18 > 20 38

39 Habitat Type/ Structure Mixed Conifer (Small/Medium) Tolerance Levels 50% 30% 80% 50% 30% Snag Densities Used by Wildlife by Diameter Class > > > > > > > > > > 20 To utilize DecAID data, snag numbers across the fire area needed to be modeled. The DecAID analysis area delineated to follow the Eagle Creek watershed boundary and be large enough to represent the variation in snag habitat and distribution from which the vegetation inventory data were collected, thus allowing a direct comparison between the analysis area and the vegetation inventory distribution histograms. Figures 1 through 4 display the existing condition of snag densities by size class (>10 dbh and >20 dbh) for the ponderosa pine/douglas fir (PPDF) habitat type and Eastside Mixed Conifer (EMC) habitat type within the Eagle analysis area. There is a total of 29,333 acres of PPDF and 20,986 acres of EMC within the analysis area. Figure 1. Comparison of current densities of snags >10 dbh and reference (historical) conditions within the PPDF habitat type in the analysis area 39

40 Figure 2. Comparison of current densities of snags >20 dbh and reference (historical) conditions within the PPDF habitat type in the analysis area Figure 3. Comparison of current densities of snags >10 dbh and reference (historical) conditions within the EMC habitat type in the analysis area 40

41 Figure 4. Comparison of current densities of snags >20 dbh and reference (historical) conditions within the EMC habitat type in the analysis area In 2015, the Eagle Fire burned approximately 8,796 acres of National Forest System managed lands to the north of the Sparta project area within the Eagle Creek watershed. The majority (95%) of the affected area burned at low or moderate severity. Of those acres that burned, 3,650 acres were in the ponderosa pine/douglas-fir (PPDF) habitat type and 3,444 acres were within the Eastside Mixed Conifer (EMC) habitat type. Analysis using a historic range of variability specific to the analysis area and weighted by the amount of each habitat type suggest that the Eagle fire has created high-density snag habitat that is above the natural range of variation for PPDF habitat in the analysis area. Currently 21% of the PPDF habitat in the analysis area has a high density of snags (snag density of >8 snags/acre that are 10 dbh or greater) compared to 12% historically (Figure 1). Within the EMC habitat type, 41% of the area currently has snag densities in excess of 12 snags/acre, compared to 34% historically (Figure 3). The Eagle fire caused a dramatic, short-term increase in snag numbers. Snag habitat occurring within the fire area is serving as intermittent habitat for most cavity excavators (Saab et al. 2004). The process of tree mortality and snag recruitment are balanced by the processes of snag decay and fall (Everett et al. 1999). Dahms (1949) found that 10 years post-fire, 50% of fire killed ponderosa pine snags remained standing but this declined to 22% standing after 22 years. It is estimated that about 75% of all snags may fall within 20 years (Keen 1929, Dahms 1949, Parks et al. 1999, and Everett et al. 1999). The effect of the Eagle fire is an immediate increase in snag habitat followed by a reduction in available habitat and a decrease in local populations as snags fall. Snag Habitat within the Sparta Project Area Snag density surveys were conducted within 23 stands distributed across the Eagle Creek-Paddy Creek and Little Eagle Creek subwatersheds during the summer of Sample results indicated an overall density of 4.73 snags per acre for snags 8 inches or greater in diameter and at least 10 feet in height. Density of snags greater than 21 inches averaged 1.32 per acre (Table 14). A summary of snags sampled by tree species and diameter class is provided in Table 14. Table 14 indicates that grand fir snags occur in higher densities than other species in the areas sampled for snags. This is because grand fir trees are susceptible to a greater variety of insects and pathogens than other 41