Executive Summary

To improve conservation outcomes for biodiverse native Oregon white (Quercus garryana) oak ecosystems, the Intertwine Oak Prairie Working Group(OPWG) has developed a series of oak data and map products for the greater Portland-Vancouver metropolitan region. The OPWG encourages the use of this information by the community, planners, conservation practitioners, landowners, and others with the goal of protecting and restoring our native oak ecosystems and their rich cultural legacy.

Oak distribution data was compiled from field and computer mapping. Oak woodland patches were defined based on oak distribution, land cover, and planimetric data. Not every oak tree is reflected in the data, but we have high confidence in the overall landscape pattern and the locations of significant oak clusters. We also have high confidence that areas depicted as absent of oak do not harbor clusters of undocumented oak.

Scoring identifies the largest and least fragmented oak habitat, to help prioritize land acquisitions and locate restoration actions that improve connectivity. Other oak habitat excluded from high-scoring oak woodland patches also has value, and finer-scale analyses or those emphasizing different criteria may reveal other patterns and priorities. Future map updates will depend on available resources.

This “habitat core” version of oak woodland patches is comprised of the highest scoring oak woodland patches according to the methodology below. The habitat cores were used to model habitat connectivity. The two basic components for the habitat connectivity modeling were habitat cores and a resistance surface. The habitat cores acted as the sources and destinations for animal movement.

This product is best used for landscape-level planning for conservation as well as land use, development, transportation, utility, park, farm and forest planning, and project scoping. For individual site- or parcel-scale, this data should be used with caution and supplemented with field surveys, as needed.

This data has particular strengths and is built on assumptions which are important to keep in mind. For example, it is inappropriate to assume that Oregon white oak that do not receive high scores in this product are unimportant and of low conservation value. Individual oak trees can have high ecological and cultural significance and should be protected whenever possible.

Data Inputs to This Product

1. OakQuest 2019 Point Data, Intertwine (“OakQuest_2019”)

2. Regional Conservation Strategy 2011 Landcover, 5m resolution, Intertwine (“RCS_Landcover”)

3. EPA EnviroAtlas 2014 Landcover, 1m, Environmental Protection Agency (“EPA_Landcover”)

4. Buildings, Metro RLIS and Clark County, 2019 (“Buildings”)

5. Street Centerlines, Metro RLIS and Clark County, 2019 (“Streets_CL”)

6. Railroads, Metro RLIS and Clark County, 2019 (“Railroads”)

7. Oregon Urban Growth Boundaries, Oregon Department of Administrative Services Geospatial Enterprise Office, 2019 (“UGB”)

8. Washington Urban Growth Areas, Washington State Department of Ecology, 2019 (“UGA”)

Methodology

Habitat Cores were defined from Oak Woodland Patches. Methodology to Create Oak Woodland Patches is described first.

Methodology to Create Oak Woodland Patches

Landcover Base

1. “RCS_Landcover” was the beginning file for the base. This raster was at 5m resolution, and all raster operations that followed were at 5m resolution.

2. Where “EPA_Landcover” existed, it superseded “RCS_Landcover.” “EPA_Landcover” was more recent and higher resolution than “RCS_Landcover,” but covered less area and had fewer landcover categories.

3. Buildings superseded any landcover below in “RCS_Landcover” and “EPA_Landcover.”

4. Steps 1 through 3 created “Landcover_Base,” with the following six landcover categories: Water, Developed/Impervious Surfaces, Low Vegetation, Tree Cover, Agriculture, Soil/Barren/Sand Bars.

1. Oak Canopy

5. “OakQuest_2019” points were buffered by a 30-feet radius to create “Oak Canopy.” Data on canopy sizes of different “OakQuest_2019” points were not available. Therefore, a single, standard radius was assumed.

6. For “Oak Canopy” area calculations, this 30-feet radius superseded all landcover in “Landcover_Base.” This approach gave highest priority to “OakQuest_2019” data and the standard 30-feet radius used for area calculations. This also counted canopy that overhangs streets and other impervious surfaces.

Oak Woodland Patches

7. “Landcover_Base” and “Oak Canopy” were combined. “Oak Canopy” superseded all landcover below to create “Woodland_Patch_Base.”

8. A 1-acre radius moving window was applied to “Woodland_Patch_Base” to calculate tree canopy percentage per acre. Tree canopy that was oak or any other species was included in the calculation. The intent was to define woodland of any kind, which could then be further defined into Oak Woodland.

9. Area with 30% tree canopy or higher was selected and then recombined with tree canopy (of all species) from “Woodland_Patch_Base.” This aggregated some smaller tree canopy fragments into contiguous Oak Woodland areas. This also counted isolated oak trees as Oak Woodland which might have been smaller than 1-acre, and thus not large enough to be 30% tree canopy in a 1-acre moving window. 30% tree canopy was based on the minimum threshold for Oak Woodland noted in the Biodiversity Guide for the Greater Portland-Vancouver Region (2012).

10. “Streets_CL” centerlines, “Developed/Impervious Surfaces” landcover, and “Railroads” were patch breaks in Woodland. Woodland completely divided by those patch breaks was considered to not be contiguous across the patch break.

11. Contiguous Woodland within 164-feet (50m) from the points layer “OakQuest_2019” was designated as Oak Woodland, creating Oak Woodland Patches. Area beyond 164-feet (50m) was not defined to be Oak Woodland.

Scoring Methodology

12. Only Oak Woodland Patches were scored. Woodland Patches that were not defined as Oak Woodland Patches did not receive a score.

13. Oak Woodland Patches were scored with five indicators: (1) Patch Size based on Species Territory, (2) Patch Size based on Patch Size Range in the Study Area, (3) Total Oak Canopy within an Oak Woodland Patch, (4) Percent of Patch Bordered by Natural Land Cover, and (5) Percent of Oak Woodland within a 2km Radius.

14. Scores of 1 through 10, from lowest to highest score, were used for each indicator.

15. A multiplier was applied to each Ecological Indicator score to create a Summary Score for each Oak Woodland Patch. The multiplier gave different relative weighting to each Ecological Indicator.

16. Possible Summary Scores for each Oak Woodland Patch ranged from 100 to 1000.

17. Indicators and scoring were as follows:

INDICATOR AA

Multiplier: 15

Category: Patch Characteristic

Ecological Indicator: Patch Size based on Species Territory- Number of 8 ha (20 acre) Oak Woodland Patch units

Scoring:

• 1 (low)= <8 ha (<20 ac) Patch

• 4= 8 ha to <16 ha (20 to <40 ac) Patch

• 6= 16 to <49 ha (40 to <120 ac) Patch OR (3) closely associated Patches >8 ha (>20 ac) each OR (2) closely associated Patches >12 ha (>30 ac) each. (i.e., enough suitable habitat for 2-5 nuthatch pairs or acorn woodpecker colonies or for 1-3 female western gray squirrels)

• 8= 49 to <=162 ha (120 to <=400 ac) Patch OR (3) closely associated Patches >16 ha (>40 ac) each OR (2) closely associated Patches >24 ha (>60 ac) each. (i.e., enough suitable habitat for 6-20 nuthatch pairs or acorn woodpecker colonies or for 3-10 female western gray squirrels)

• 10 (high)= >162 ha (>400 ac) Patch OR (3) closely associated Patches >57 ha (>140 ac) each OR (2) closely associated Patches >85 ha (>210 ac) each. (i.e., enough suitable habitat for >20 nuthatch pairs or acorn woodpecker colonies or for >10 female western gray squirrels)

Indicator Notes:

• Scoring based on a combination of white-breasted nuthatch, acorn woodpecker, and female western gray squirrel territory size and western gray squirrel canopy gap behavior. Scoring levels are based on different combinations of 8 ha (20 acre) Oak Woodland Patch units.

• Closely Associated = <=164' between Oak Woodland Patch edges, without being separated by roads, paving, buildings, or railroads. (164' based on maximum canopy gap for western gray squirrel from Metro-PSU Connectivity Modeling).

INDICATOR BB

Multiplier: 30

Category: Patch Characteristic

Ecological Indicator: Patch Size based on Patch Size Range in the Study Area

Scoring:

• 1 (low)= Natural Breaks 1st

• 2= Natural Breaks 2nd

• 3= Natural Breaks 3rd

• 4= Natural Breaks 4th

• 5= Natural Breaks 5th

• 6= Natural Breaks 6th

• 7= Natural Breaks 7th

• 8= Natural Breaks 8th

• 9= Natural Breaks 9th

• 10 (high)= Natural Breaks 10th

Indicator Notes:

• Scoring breakpoints are based on the range and frequency of Oak Woodland Patch sizes in the full study area. A rounded Jenks Natural Breaks classification is used to create 10 groups for scoring.

INDICATOR CC

Multiplier: 20

Category: Patch Characteristic

Ecological Indicator: Total Oak Canopy within an Oak Woodland Patch

Scoring:

• 1 (low)= Natural Breaks 1st

• 2= Natural Breaks 2nd

• 3= Natural Breaks 3rd

• 4= Natural Breaks 4th

• 5= Natural Breaks 5th

• 6= Natural Breaks 6th

• 7= Natural Breaks 7th

• 8= Natural Breaks 8th

• 9= Natural Breaks 9th

• 10 (high)= Natural Breaks 10th

Indicator Notes:

• Oak Canopy area, based on total Oak Points, is totaled within each Oak Woodland Patch. Large Oak Woodland Patches with small amounts of Oak Canopy may score high under size indicators, but low in this indicator.

INDICATOR DD

Multiplier: 20

Category: Context Characteristic

Ecological Indicator: Percent of Oak Woodland Patch bordered by "Natural" landcover

Scoring:

• 1 (low)= 0 to <10%

• 2= 10 to <20%

• 3= 20 to <30%

• 4= 30 to <40%

• 5= 40 to <50%

• 6= 50 to <60%

• 7= 60 to <70%

• 8= 70 to <80%

• 9= 80 to <90%

• 10 (high)= 90-100%

Indicator Notes:

• Landcover is measured as the percent area from the edge of an Oak Woodland Patch to a distance of 100-feet according to the following categories:

  • "Natural"= Tree Cover, Water

  • Low Vegetation= Weighted at 75% Natural (i.e., if an Oak Patch is 100% surrounded by Low Vegetation, that would be weighted to 75% Natural, and receive a score of 8. Prairie landcover data for the study area does not exist. The weighting rationale is that data for low vegetation includes lawn and prairie, thus it is not as reliably "natural" as Tree Cover. This likely overweights lawn areas but tries to avoid underweighting prairie.)

  • Agriculture= Weighted at 50% Natural. (i.e., if an Oak Patch is 100% surrounded by Agriculture, that would be weighted to 50% Natural, and receive a score of 6.)

  • Developed/ Impervious/ Roads, Sandbars/Soil/ Barren= Not Contributing.

INDICATOR EE

Multiplier: 15

Category: Context Characteristic

Ecological Indicator: Percent area of Oak Woodland within 2km radius

Scoring:

• 1 (low)= 0 to <5%

• 2= 5 to <10%

• 3= 10 to <15%

• 4= 15 to <20%

• 5= 20 to <25%

• 6= 25 to <30%

• 7= 30 to <35%

• 8= 35 to <40%

• 9= 40 to <45%

• 10 (high)= 45 to 50%

Indicator Notes:

• Percent area within 2 km (1.25 miles) from centroid of Oak Woodland Patch that is defined as Oak Woodland Patch.

• Proximity to other Oak Woodland Patches counts regardless of other landcover in between, including streets or impervious surfaces.

• Scale goes up to 50%, and not 100%, because the maximum Oak Woodland percentage within any 2km radius within the study area is 48%.

Urban Scores

18. Oak Woodland Patches that intersected Urban Growth Boundaries (UGB’s) in Oregon and Urban Growth Areas (UGA’s) in Washington were selected to create this Urban Oak Woodland Patches data. The intent was to highlight presence and scoring of Oak Woodland Patches specifically in urban areas.

19. As with the previous Regional Conservation Strategy Habitat Prioritization Modeling, this approach used a consistent scoring across the entire region that could then be cropped to UGB/UGA areas, or any other focal area of interest. Indicator Scoring was set up so that there was still score differentiation of smaller oak patches in more developed areas to create meaningful scoring differences when cropped to UGB/UGA boundaries. This approach allowed for a single scoring method to operate across multiple scales. It avoided diluting the value of large habitat patches that remain within UGB/UGA boundaries. It also avoided multiple separate scoring systems with thresholds based on socio-political boundaries that are somewhat arbitrary from the perspective of ecological function.

Data Not Used in Scoring

1. Rare Species Occurrences

2. Soils

3. Wetlands

4. Floodplains

5. Forest interior versus forest edge

6. Patch edge to interior ratio

7. Historic vegetation

8. Historic fire regime (USDA Landfire data)

9. TNC Priority Willamette Valley Oak and Prairie Habitat Parcels

10. WMSWCD Oak Mapping

11. Metro Data Resource Center 2014 Coniferous vs Deciduous Landcover, 1m (DRC_Coniferous)

Methodology to Create Habitat Cores

In this project, Habitat Cores were defined from Oak Woodland Patches and included a subcategory for Urban Habitat Cores as described below.

Habitat Cores

1. Oak Woodland Patches and their scoring were the basic unit for developing Habitat Cores. At the scale of regional connectivity, the larger scale of Oak Woodland Patches (OWPs) was more appropriate than Oak Patches as source and destination habitat for regional oak habitat connectivity modeling. OWPs better match the scale of many oak woodland species territories.

2. OWP scores ranged from 100 to 920. This scoring range was grouped into 20 classes using a Jenks Natural Breaks Classification. This data clustering method calculates based on “natural breaks” in the data set distribution to minimize differences within groups and maximize differences between groups. The top 4 of 20 classes were selected, which equaled scores >= 666, and were designated as Habitat Cores. OWPs in this selection ranged from 85 acres to 456 acres. This step created 51 Habitat Cores.

The high scores of the selected OWPs demonstrated their importance for overall conservation prioritization and associated importance for inclusion in the connectivity analysis.

3. Some larger OWPs did not meet the >=666 scoring threshold above, as they may be somewhat isolated or have edges that abut less natural conditions. However, based on size of contiguous habitat, they may be potential sources and destinations for higher populations of oak woodland species movement. Therefore, to not exclude these more isolated, but potentially important OWPs, size was also used as a factor for Habitat Core selection. OWPs >= 100 acres that did not meet the >= 666 scoring threshold were added to the Habitat Cores category. OWPs in this selection ranged in scoring from 510 to 601. This step created an additional 21 Habitat Cores.

Based on the species-area relationship of island biogeography, larger habitat patches will likely have the greatest diversity of species and generally the highest populations of those species. This relationship has been demonstrated in rural and urban areas (Beninde et. al. 2015).

A threshold of >= 100 acres of oak woodland could provide habitat for at least 5 white-breasted nuthatch (Sitta carolinensis) pairs, 5 to 10 acorn woodpecker (Melanerpes formicivorus) colonies, or 2 female western gray squirrels (Sciurus grisesus).

4. The steps above resulted in 72 Habitat Cores.

Urban Habitat Cores

5. To more directly model relationships between urban areas and the highest scoring and largest habitat patches in less densely developed areas, an additional set of OWPs were incorporated as Habitat Cores. This approach is based on the project goals to inform opportunities in both rural and urban areas. Research grade observations from eBird and iNaturalist also demonstrated presence of species of interest such as white-breasted nuthatch (Sitta carolinensis) and acorn woodpecker (Melanerpes formicivorus) in urban sections of the study area.

6. OWPs within Urban Growth Boundaries (UGB’s) in Oregon and Urban Growth Areas (UGA’s) in Washington were defined as Urban OWPs, and their scores ranged from 100 to 830. To be selected as an Urban Habitat Core, the majority of the OWP had to be within the UGB/UGA, not just a small portion of an OWP edge. As with the steps above, this scoring range was grouped into 20 classes using a Jenks Natural Breaks Classification, and the top 4 of 20 classes were selected, which equaled scores >= 506. Seven of the 72 Habitat Cores in steps 1 through 4 above were also Urban Habitat Cores. This step added 17 additional Habitat Cores, ranging in size from 44 to 93 acres.

Final Habitat Cores

7. Any OWPs selected to be Habitat Cores that met scoring Indicator AA definition of “closely associated” were combined into a single Habitat Core for the purposes of connectivity modeling. This step merged 13 of the OWPs at this point in the process into 5 OWPs.

8. The steps above resulted in a total of 81 Habitat Cores.

9. The Habitat Cores from the steps above were reviewed by the Intertwine Oak Prairie Working Group (OPWG). This review process added Habitat Cores based on OPWG expert opinion. OPWG Habitat Core additions, and merges of those patches with other Habitat Cores from the steps above, following the “closely associated” definition, resulted in a total of 110 Habitat Cores used in the habitat connectivity modeling.

In Oak Woodland Patch scoring, all roads are breaks in OWPs. In the review of Habitat Cores, the OPWG identified OWPs separated by roads that had less of a functional role as a habitat separator. These roads included interior park roads, dead end roads with few houses, and driveways. Based on OPWG concern that several of these OWPs would be excluded from being Habitat Cores in the habitat connectivity modeling, the OPWG identified additional OWPs to be added as Habitat Cores.

OWPs that were not selected as Habitat Cores are still very relevant to the connectivity modeling as they are supportive of habitat connectivity between Habitat Cores.

Field Descriptions

Field	Description
CORE_ID	Unique ID for each Habitat Core.

Reference

Beninde J, Veith M, & Hochkirch N. (2015). Biodiversity in cities needs space: a meta-analysis of factors determining intra-urban biodiversity variation. Ecology Letters 18:581-592.

The Intertwine Alliance. (2012). Biodiversity Guide for the Greater Portland-Vancouver Region.Shiler A, editor. The Intertwine Alliance. Portland, Oregon.

Distribution Liability

The data contained herein are provided on an as-is, as-available basis without warranties of any kind, expressed or implied, including (but not limited to) warranties of merchantability, fitness for a particular purpose, and non-infringement. Intertwine Oak Prairie Working Group and Biohabitats expressly disclaim any warranty that the data are error-free or current as of the date supplied.