An Assessment of Spotted Owl Habitat on Non-federal Lands in Washington between 1996 and 2004

Forest Practices Rules for the Northern Spotted Owl (Strix occidentalis caurina) were adopted in May 1996. These rules, which apply to nonfederal lands, established 10 landscapes â€" known as Spotted Owl Special Emphasis Areas (SOSEAs) â€" wherein proposed harvest of suitable owl habitat would receive environmental review designed to provide a high level of protection. Under the rules, the level of habitat protection varied depending on whether habitat was located inside an owl management circle located inside or outside of SOSEAs or whether or not habitat lands were part of a Habitat Conservation Plan (HCP) approved by the U.S. Fish and Wildlife Service.

In 2004 the Department of Natural Resources and the Washington Department of Fish and Wildlife entered into a cooperative agreement to gather information that could be used by our agencies and the Forest Practices Board as part of a review of the Forest Practices Rules. Specific objectives of this study were to: 1) estimate the amount of suitable Spotted Owl habitat in 2004 on landscapes affected by state and private forest practices, 2) estimate the amount of Spotted Owl habitat harvested under the regulatory authority of the Forest Practices Rules between 1996 and 2004, and 3) determine the current status and net change (accounting for gains and losses) of Spotted Owl habitat on landscapes influenced by the Forest Practices Rules adopted in 1996.

The study area outside of SOSEAs included all forested lands within Status 1-3 (i.e. territorial) Spotted Owl management circles where at least 10% of the acres within a circle were under state or private ownership. In addition, all lands and all Status 1-3 owl circles inside SOSEAs were included in the study area. The study area included 450 owl circles and totaled 3,233,942 acres. The study area was divided into two sampling strata categories: 5 geographic zones (East Cascades, North Cascades, South Cascades, Olympics, and Southwest), and an updated GIS layer that mapped seral strata (early, mid, late, and �"other”).

We determined presence and absence of suitable Spotted Owl habitat at 1,514 randomly selected locations using orthophoto interpretation for early and �"other” strata and helicopter reconnaissance for mid and late seral strata. Helicopter classification accuracy rates were determined by ground visits to collect quantitative stand plot data at a subset of these same plots to determine whether or not the stand met the suitable habitat definitions of the Forest Practice Rules. These accuracy classification rates were used to adjust the helicopter data to more accurately estimate the amount of Spotted Owl habitat on the landscape in 2004.

The amount of harvested habitat from 1996-2004 was estimated first by calculating the total amount of harvest that occurred during this time period regardless of Spotted Owl habitat condition. We contracted with the U.S. Forest Service Pacific Northwest Forestry Sciences Laboratory in Corvallis, Oregon (FSL), to map both clear-cut and partial harvest areas that occurred between 1996 and 2004. Using this map information, our next step was to estimate the percentage of the harvested areas that also met Spotted Owl habitat criteria in 1996 (prior to harvest) as defined by the Forest Practices rules. This was done by modeling remote sensing information collected in 1996 by the Interagency Vegetation Mapping Project team to predict Spotted Owl habitat presence within the harvested areas identified by the change map produced by the U.S. Forest Service’s lab. Stand inventory data with known Spotted Owl habitat conditions obtained from the Washington Department of Natural Resources were used to develop these models.

The ground plot data used to calibrate the helicopter predictions for 2004 conditions were used to develop accuracy classification tables for the 1996 model predictions as well. We then compared the estimates of the amount of habitat existing in 2004 with the amount of harvested habitat from 1996-2004 to calculate a Relative Change Index (RCI) to assess how the amount of harvested habitat since rule adoption in 1996 related to the total amount of habitat remaining on the landscape in 2004. Data were summarized separately for federally approved Habitat Conservation Plans (HCP) and non-HCP landscapes. Data were also summarized separately for lands within SOSEA boundaries (both inside and outside Spotted Owl management circles) and for lands outside of SOSEAs. All lands outside of SOSEA boundaries were within the boundaries of owl management circles.

Study Area Summary

We estimated that there was about 816,300 total acres of Spotted Owl habitat on all land ownership categories in our study area in 2004. Most suitable owl habitat in 2004 (56%) occurred on federal lands, and lesser amounts were present on state-local lands (21%), private lands (22%) and tribal lands (1%). Approximately 75% of the habitat in the study area occurred on non-HCP lands. Approximately 172,000 total acres of forest were harvested on the study area from 1996-2004, most of which occurred on non-HCP lands (76%). The majority of the total harvest occurred on private (79%) and state-local (14%) lands.

An estimated 33% (56,400 acres) of the harvested lands also met Spotted Owl habitat conditions as defined by the Forest Practices Rules. Approximately 71% of the harvested habitat occurred on non-HCP lands. Most of the harvested Spotted Owl habitat was on private (77%) and state-local (15%) lands. We estimated an average RCI value of 6% (95% confidence Interval (CI) = 5% - 8%) of the maximum potential amount of habitat in 2004 was harvested during the 9 years following rule adoption in 1996. RCI values on the study area ranged from 4% in the Olympics to 32% in the Southwest zone.

Changes in non-HCP Spotted Owl Special Emphasis Areas

The majority of non-HCP acres within SOSEAs were on private (55%) and federal (43%) lands. We estimated that 277,200 acres of Spotted Owl habitat existed on non-HCP lands inside of SOEAs in 2004. The majority of the habitat acres on non-HCP lands within SOSEAs were either federal (64%) or private (35%). Most of the non-HCP SOSEA Spotted Owl habitat in 2004 (59%) occurred inside of Spotted Owl management circles. The percentage of the SOSEA landscape in 2004 that met Spotted Owl habitat definitions ranged from 31% in the East Cascades to 13% in the South Cascades. Overall, the percentage of non-HCP SOSEA landscapes meeting Spotted Owl habitat criteria was higher inside of circles (28%) compared to lands outside of owl management circles (18%).

We estimated that 30% (21,000 acres) of the total harvest inside of SOSEAs on non-HCP lands was in Spotted Owl habitat. An estimated 33% of the 21,000 acres of habitat harvested during 1996-2004 occurred inside of owl management circles. Most (~19,000 acres) of the non-HCP harvested habitat inside of SOSEAs was on private land. We estimated that an average of 4% (CI = 3% - 5%) of the Spotted Owl habitat on non-HCP lands, within owl management circles in SOSEA, was harvested from 1996-2004. In contrast, an average of 11% (CI = 9% - 13%) of Spotted Owl habitat in SOSEAs outside of owl circles was harvested during this same period. Overall, RCI values on non-HCP SOSEA lands ranged from 5% in the East Cascades to 10% in each of the westside study area zones.

Changes in Habitat Conservation Plan Landscapes

Habitat conditions and levels of harvest were somewhat different on HCP compared to non-HCP lands inside of SOSEAs. Private lands made up 46% of the non-HCP lands compared to 23% of the HCP landscape. State lands made up only 1% of the non-HCP lands compared to 77% of the HCP landscape. Most of the approximately 200,500 acres of Spotted Owl habitat in 2004 on HCP lands in our study area (74%) occurred inside of SOSEAs, compared to 45% on non-HCP SOSEA lands. The average percentage of the HCP landscape in 2004 meeting Spotted Owl habitat definitions was 22%, and ranged from a high of 28% in the East Cascades to a low of 14% in Southwest zone. Overall, the relative amount of HCP landscapes that met Spotted Owl habitat criteria was higher inside of circles (24%) compared to lands outside of owl management circles (20%).

Approximately 38% (16,100 acres) of the HCP landscape that was harvested from 1996-2004 met Spotted Owl habitat definitions, compared to 31% on non-HCP SOSEA lands. The amount of Spotted Owl habitat harvested on HCP lands relative to the total habitat on HCP lands did not differ from non-HCP lands, averaging 7% (95% CI = 6% - 8%). RCI values inside of circles did not differ from RCI values outside of circles within SOSEA landscapes. Overall RCI values on HCP lands ranged from 5% in the Olympics and South Cascades to 14% in the Southwest zone.

Changes in Owl Management Circles Outside of Spotted Owl Special Emphasis Areas

The majority of the non-HCP acres outside of SOSEAs were on federal (62%) and private (35%) lands. We estimated that 338,600 acres of Spotted Owl habitat existed inside Spotted Owl management circles on non-HCP lands outside of SOEAS in 2004. The majority of the non-HCP habitat outside of SOSEAs were on federal (83%) and private (14%) lands. The relative amount of Spotted Owl habitat in owl management circles outside of SOSEAs on non-HCP lands averaged 31% (CI = 27% - 34%) and ranged from 37% in the North Cascades (where federal lands comprised 94% of the landscape) to 7% in southwest Washington (where there was no federal lands within the study area).

We estimated that 33% (19,000 acres) of the total harvest inside owl management circles on non-HCP lands outside of SOSEAs was in Spotted Owl habitat. Most of this harvest (85%) occurred on private lands. Overall, RCI values averaged 5% (CI = 4% - 6%) and ranged from 1% in the North Cascades to 44% in the Southwest zone.

Analysis Considerations

Certain cautions should be considered, relative to our analyses, which may have influenced the results presented in this report. One underlying assumption inherent in this analysis was that our ability to accurately classify habitat was independent of land ownership. Due to access concerns onto private lands some of our data collection was restricted to public lands. We examined this concern in the report and concluded any potential bias was not significant and did not affect our overall conclusions.

Another caution is that our approach may have overestimated the amount of harvest and harvested habitat on federal lands. Approximately 72% of the total stand replacement harvest (~5,500 acres) estimated on federal lands was derived from applying a non-harvest correction factor (C _nharv, see page 33), even though the estimate of C _nharv was small (0.45 %). Conversely, stand replacement harvest attributed to C _nharv made up only 11% of the estimated harvest for state and private lands.

Estimates of habitat loss due to partial harvest in the East Cascades had a greater level of uncertainty than losses related to stand replacement harvest. We assumed that all partial harvest activity in the East Cascade zone was captured in the DNR FPA database and that partial change outside of the FPA database (representing ~5,000 acres) was not a result of forest practices. We also assumed that Spotted Owl habitat loss associated with approximately 40,000 acres of uneven-aged forest practices permits in western Washington was not significant. We assumed that most harvest of forests with Spotted Owl habitat attributes in western Washington would be clearcuts. Further analysis would be necessary to determine whether this assumption was valid.

Finally, 21 Status 1-3 Spotted Owl management circles were changed to Status 5 (unoccupied) during the 1996-2004 period. Twelve re-classed sites were in Spotted Owl Special Emphasis Areas, and 9 were outside Spotted Owl Special Emphasis Areas. In addition, 16 Status 1-3 spotted owl sites were new and added to the database during the 1996-2004 period. Most of these (n=12) were located in the East Cascade zone. Ten post 1996 Status 1-3 sites were overlapped SOSEA boundaries. Our study area and summary statistics for landscapes inside and outside of circles were based on the landscapes that were associated with Status 1- 3 owls as of 2004. As a result, the status of the landscape at the time of timber harvest in these areas may have been different then the status in 2004.

Conclusions and Recommendations

Quantifying the effects of the habitat loss we documented on regional Spotted Owl subpopulations in Washington was beyond the scope of this project. However, a number of conclusions can be derived regarding the potential effects of habitat loss. Recent Spotted Owl demographic studies have documented significant population declines in each of the study areas that overlap with our study area. Spotted Owls have large home ranges and use large amounts of structurally complex forest within those areas. State Forest Practices Rules identified 40% of the landscape as necessary to maintain the viability of an owl territory. With the possible exception of the East Cascade zone, our results indicate that the average landscape inside of owl management circles within most SOSEA landscapes were likely significantly below this threshold (Table 16, page 53). The percent of non-HCP landscapes (including all ownerships) in 2004, inside of owl management circles, that met Spotted Owl habitat criteria ranged from a low of 18% (95% confidence interval (CI) = 16% to 20%) in the South Cascades to a high of 34% (CI = 30% to 38%) in the East Cascades.

Our estimates of a 4% to 7% loss of habitat inside owl circles within SOSEAs between 1996 and 2004 (Table 32, page 81) magnifies the potential effect on those Spotted Owl sites that use habitat on non-federal lands. Loss of habitat in these landscapes is important because the Spotted Owl Special Emphasis Areas were identified in the state rules as strategic areas within the state where owls and habitat on non-federal lands contributes to the overall health of Washington’s population of owls. In addition, RCI values of SOSEA habitat loss outside of management circles were more than twice as high as the RCI values for lands inside of management circles. As a result, if this pattern continues over time, owl habitat inside of SOSEAs will become more and more restricted to only those landscapes inside of Status 1-3 owl circles.

This is not to say the habitat loss we documented is conclusively responsible for the observed Spotted Owl population declines. There is concern that Barred Owls and Spotted Owls may compete for resources, and that the former has a distinct advantage in this relationship that is now influencing the Spotted Owl population decline. The nature of the relationship between these two species is not clear, but the negative effects of a strong competitor like the Barred Owl would likely interact with the effects of habitat loss for Spotted Owls.

Given the results of our study and considering the ongoing decline of Spotted Owl populations we make the following recommendations:

1. Long-term landscape planning should be encouraged. Spotted Owl management on non-HCP lands appears to be largely driven by individual owl circle management. Within SOSEAs, we not only documented habitat loss within circles, but also estimated rates of habitat loss outside of circles that were approximately twice the rates inside circles. This pattern of habitat loss isolates habitat near the cores of Spotted Owls' home ranges and compromises the ability of the entire landscape to contribute to Spotted Owl conservation over time. While some habitat lands within SOSEAs are currently managed under habitat conservation plans, stronger regulatory approaches to conserving habitat at the landscape level may be needed if SOSEAs are to function as more than groups of occupied circles. As Spotted Owl populations decline and fewer circles are consistently occupied, the current structure of the Forest Practices Rules coupled with "decertification" of circles that are inconsistently occupied may result in further habitat loss within SOSEAs.
2. High quality, spatially accurate habitat maps should be developed - It is important to accurately identify the location and amount of Spotted Owl habitat in areas that are identified to contribute to the long-term conservation of Spotted Owls (e.g. SOSEAs). The sampling approach that we adopted to assess habitat abundance and change was necessitated by the lack of such maps. While these data may be available for some management circles, areas, or ownerships, they are neither common nor consistent. High-quality habitat maps based on the habitat definitions in the Forest Practices Rules are essential both for day-to-day rule implementation (i.e. review of Forest Practices Applications) and for policy evaluation. Additionally, habitat criteria and definitions should be periodically reviewed and updated to ensure maps are consistent with owl habitat requirements in the specific areas identified to support conservation.