© Scott Warren Landscape-Level Forest Conservation Analysis: A Recent Case Study from Pennsylvania Dylan Jenkins and Tamara Gagnolet The Nature Conservancy.

© Scott Warren

Landscape-Level Forest Conservation Analysis:A Recent Case Study from PennsylvaniaDylan Jenkins and Tamara GagnoletThe Nature Conservancy – PA Chapter

West Branch Forest © George C. Gress / TNC

Overview

Justification and Context

Lessons Learned

Conducting the PA Forest Conservation Assessment

How the FCA is being used (and who is using it)

Lessons Learned

Justification and Context

– Ultimately about guiding the forest conservation community’s allocation of increasingly scarce resources

– Identify and prioritize forest landscapes to:• Keep them intact• Keep them functioning for high quality ecological and

economic values

– Provide a case study (2005-present) from PA

– Help to make your assessment more efficient

Lessons Learned

– Choose an easily understandable and replicable unit of analysis (e.g. watershed, habitat patch, or census tract) for which summarized statistics will be meaningful and informative.

– Assess conservation values and threats in separate analyses.

– Avoid using ownerships/parcels as basic unit of analysis; further, avoid using ownership altogether until the threats assessment portion of the analysis.

– Start as a “splitter” (rather than a “lumper”) in defining your unit of analysis, as it is easier to lump than split after analysis is conducted.

– Clearly communicate and thoroughly evaluate your assumptions and values (for example: emphasis on large habitat patch size or a focus on biodiversity) as you define and implement your methods.

– Design your analysis for replication; rely on available geospatial datasets for your study area (e.g. National Land Cover Database, USGS’s National Hydrography Dataset, ESRI’s Street Map USA, your state’s Natural Heritage Program data).

– Seek input from many stakeholders, but optimize (i.e., minimize) the number of partners involved in the actual analysis

Lessons Learned

– Choose ranking/selection criteria that are:• from readily available data for your study area (unless data collection is in your budget)• directly relevant to the purpose of your analysis (i.e. weed out peripheral information that

will only distract you from the main point)• directly informative for developing strategies or actions• intuitive to understand by a wide range of audiences

– Normalize the raw values of each ranking/selection criteria (for example, onto a scale of 1 to 10) for direct addition/combination.

– Assess quantitative and qualitative ranking/selection criteria separately.

– Beware of attempting to quantify characteristics that are more qualitative in nature (e.g. landscape context, threat urgency).

– Find that tricky balance between (a) including sufficient criteria for a useful and complete assessment, and (b) simplifying the criteria list to avoid complexity, redundancy, and facilitate understanding and buy-in from stakeholders.

Pennsylvania’s Forests (approx. 16.9 million acres)

Extent of Forest Fragmentation

Linear fragmenting features shown here include interstate highways, state and local roads, railroads, and utility rights-of-way.

Impacts of Forest Fragmentation

Creates inefficiencies for the production of economic and ecological values from forests…

– direct loss of interior habitat & biota

– introduction and spread of invasive species

– movement of some fauna away from linear feature

– decrease in nest pairing and success in forest birds

– increase in forest bird nest predation and cowbird parasitism

– interruption of movement and migration of terrestrial animals (e.g., fisher, Allegheny woodrat)

– creation of habitat for common edge and invasive species

– increased costs of harvesting and transporting wood products

However, like stand boundaries, charismatic human and natural features define areas of opportunity.

Extent of Forest Fragmentation

Aerial photograph (true-color, leaves-off) of Sproul State Forest, west of Lock Haven. From the photo, the area looks like a huge swath of unbroken forest…

PAMAP 2005

… but much linear infrastructure already exists on the ground (yellow lines);few “large,” intact blocks of fully functioning forest ecosystems remain in PA.

PAMAP 2005

Extent of Forest Fragmentation

Importance of Intact Forest Areas

Large, intact forest areas are vitally important for:

– area-sensitive species – interior forest species– ecosystem processes/services

• nutrient cycling• pollination• predator-prey interactions• water filtering• regulation of stream flow• carbon storage as biomass• natural disturbance regimes

– maintaining production of high quality economic values

© Tamara Gagnolet / TNC

Why does forest size matter?

© 2004 The Nature Conservancy. All rights reserved.

To maintain the full array of community types and species in Pennsylvania, some forest areas need to be large enough to support interior and area-sensitive species

and to withstand a range of natural disturbance events.

Pennsylvania ForestConservation Analysis

In 2006-2007, The Nature Conservancy and the Western Pennsylvania Conservancy began collaborating on a state-wide spatial analysis to:

– identify large, relatively intact forest areas

– assess the conservation value and threats to these forest areas

– prioritize the forest landscapes for allocating planning, protection, management resources where we will get the best return on investment

The analysis used readily available geospatial data to create a new dataset. The data sources included:

• USGS National Land Cover Dataset (2001)• ESRI Street Map USA (road data)• Railroads, mines, and other features downloaded

from PASDA (Pennsylvania Spatial Data Access)• National Hydrological Dataset (rivers and streams)• PA Natural Heritage Program Data

Pennsylvania ForestConservation Analysis

Pennsylvania ForestConservation Analysis:

Term Definitions

Landscape block: Basic unit of land bounded by Category A fragmenting features and composed of patches of forest and other land cover types.

Forest Patch: A component area of a Landscape Block consisting only of contiguous forest and bounded by both Category A and Category B fragmenting features and/or edges of dissimilar land cover types.

For the purposes of our spatial analysis, we divided fragmenting features into two groups to differentiate mapping boundaries:

Pennsylvania ForestConservation Analysis:

Term Definitions

Category A fragmenting features included interstate and state roads with speed limits greater than 25 mph, active railroads, and large rivers (greater than 5th order). These features set the landscape stage, in that they define boundaries of large units of land (“landscape blocks,” in our analysis).

Category B fragmenting features included local municipal and township roads with speed limits less than 25 mph, unpaved roads, utility rights-of-way, and inactive railroads. These features determine the quality and condition of forest habitats within landscape blocks.

We could only include features for which state-wide spatial datasets were available.

The highway is a boundary between two landscape blocks, and the smaller roads separate forest patches within each landscape block.

At this point in the project, we knew how the state was broken up into landscape blocks and forest patches. The next phase of the analysis was to evaluate the forest patches.

Cambria County © Jessica McPherson / WPC

Pennsylvania ForestConservation Analysis:

Term Illustration

“UBER STANDS”

Forest patches were evaluated based on several ecological parameters that were measurable with available spatial data:– Forest Condition

• patch size, interior forest, stand composition– Forest Biodiversity

• G1-G3 species; S1-S3 species; Biological Diversity Areas– Wetland/Aquatic Health

• stream miles, wetland acreage & diversity, floodplain acreage

All forest patches greater than 100 acres were assigned a score based on these metrics (26,000 unique units assessed!!)

The scores are weighted to emphasize large forest size and biodiversity.

Pennsylvania ForestConservation Analysis

Hypothesis: Socio-economic opportunities are based and dependent upon underlying health of ecological values, therefore assess ecological values first….

Forest Patch Score Calculation

Scores for Forest Patches (>100 acres)

The map shows the project datasets in their entirety: all the landscape blocks and all their 26,000 scored forest patches in PA.

Scores for Forest Patches(>100 acres)

– The Nature Conservancy’s and Western Pennsylvania Conservancy’s forest analysis relied on available spatial data and incorporated several key ecological parameters.

– That said, these parameters are not inclusive; other forested areas of biological significance and conservation importance certainly exist in Pennsylvania.

• For example, this forest analysis does not address smaller forested tracts which might possess important ecological values and functions at a more local / regional scale (e.g. source water protection or county natural areas).

– Because the analysis emphasized a landscape context, it is not a substitute for detailed, site-specific inventory and assessment…

KEA Report Card:

• Stand/property level inventory and assessment of ecological and economic condition

• All attributes are FIA compatible

• Therefore, attributes spatially scalable and comparable:

– Stand– Property– Patch– Block– Landscape– Ecoregion– State– Multi-state Region

Stand and Parcel Assessment

Selecting Priority Areas

To identify our priority forest areas, we selected forest patches with the highest scores that were within relatively intact landscape blocks.

This map shows TNC’s selected landscape blocks and the high-scoring forest patches within them.

The areas identified here do not represent all of the places in Pennsylvania that are important for the conservation of biological diversity.

1. Percent Private Ownership (“at risk” lands)– Significantly greater potential for parcelization and land

use/value fragmentation relative to public lands

2. Wind Development Potential:– Percent high wind (wind power classes 3 and above) on “at-risk”

lands (not PGC, DCNR, ANF) within forest patch (75%)– Proximity to existing transmission lines (25%)

3. Gas Development Potential (binary metric based on coarse geology data)

4. Projected Housing Development Pressure(percent change in rural housing density from 2000 to 2030 on private land within forest patches – proxy for likelihood of fragmentation)

Pennsylvania Forest Patch PrioritizationThreat Analysis and Metrics

All threat metrics assess existing and potential impact of fragmentation on conservation value of forest patches…

Threat Metric 1 of 4:Private Ownership

Threat Metric 2A of 4:Wind Development Potential

% High Wind on At-Risk Lands

Threat Metric 2B of 4:Wind Development Potential:

Proximity to Existing Electric Transmission Lines

Threat Metric 3 of 4:Gas Development Potential

Threat Metric 4 of 4:Housing Development Pressure

Composite Threat Score

Scatter Plot of Composite Conservation Valueand Threat Scores

Urgency of threats

Parcelization of forest patches

Feasibility of protection / conservation success

Presence of capable conservation partners (land trusts, woodland owners’ associations, etc.)

Conservation opportunity and existing relationships

Pennsylvania Forest Patch PrioritizationFeasibility Metrics

Linking FCA Prioritization with Strategies

Who is Using the FCA…

- Foundation of TNC-PA Strategic Plan (2008-2012)

- Conservation Action Planning (using forest blocks as targets)

- Landscape Protection Analyses (identifying priority landowners)

- Allocating resources for large-scale protection projects (like Elk County)

-  PA Wind & Wildlife Collaborative

- Informed our comments on PA DCNR Bureau of Forestry SFRP

- Guiding discussions with natural gas companies and DCNR TopoGeo

-  Integration into County Natural Heritage Inventories (WPC)

TNC Priority Forest Areas, Barrens Communities, & High Wind Speed Areas

Forest Landscape Protection Analysis

Forest Parcel Protection Analysis

Conclusions

- Successful assessments are empowering….…will allow your agency to make informed and consistent decisions on conservation opportunities and resource allocation

- Strongly recommend starting assessment with fragmentation analysis to identify landscape blocks and forest patches as basic unit of assessment and analysis.

- Conduct assessment in discrete components:- Fragmentation analysis to ID landscape blocks and forest patches- Conservation criteria: score and rank- Threat criteria: score and rank- Feasibility and other qualitative criteria

- Keep qualitative (e.g., feasibility) criteria for final stage of assessment

- Choose your dance partners carefully

- Take good notes, you will test and retest criteria and assumptions

Thank you.

Dylan Jenkins, CF Director of Forest ConservationTamara Gagnolet Conservation Planner

The Nature Conservancy in Pennsylvania

2101 N. Front St.Building 1, Suite 200Harrisburg, PA 17110

717-232-6001