2.12      Key Limiting Factors (Problems)

2.12.1      Ecosystem Diagnosis and Treatment (EDT) Modeling

EDT, or Ecosystem Diagnosis and Treatment, is a system for rating the quality, quantity, and diversity of habitat along a stream. The model uses a probe or indicator species, such as Coho or Chinook salmon, to identify the most significant problems in a stream and to identify reaches for protection and restoration. The methodology includes a conceptual framework for decision-making and a set of modeling tools with which to organize environmental information and rate the habitat elements with regard to the focal species. In effect, EDT describes how the fish would rate conditions in a stream based on our scientific understanding of their needs. The value of EDT is that it can identify the potential for a stream under a set of conditions such as those that occur now or those that might occur in the future. The result is a scientifically based assessment of conditions and a prioritization of restoration needs.

EDT was developed by Mobrand Biometrics, Inc. to provide a practical, science-based approach for developing and implementing watershed plans. It is a salmonid life history based procedure for rating the quality, quantity, and diversity of stream habitat. The model uses rating curves to relate habitat conditions to life stage survival and capacity. These life stages are then linked to form life history trajectories (or the path of a salmonid through space or a chosen or taken migratory course). Because habitat is described by reach and month, many potential trajectories can be formed. All successful trajectories are combined to form an overall estimate of capacity and productivity at a population level. The range of successful trajectories is a measure of life history diversity.

Each reach of a stream has a certain capacity or number of fish that can be supported for each life stage depending on the quantity of key habitat; a certain number of fish can spawn in the riffles while the pools can support a number of juveniles. Each pool or riffle has a quality that affects the survival of a life stage in that habitat. Quantity of habitat is thus measured as capacity. When capacity and survival over the course of a fish’s life history is integrated, an overall capacity for the species as a measure of the quantity of habitat can be derived. Overall survival is measured as the number of adult fish that return for each fish that spawns. This is termed productivity and is a measure of habitat quality.

The value of EDT is that it can identify the potential for a stream under a set of conditions such as those that occur now or those that might occur in the future. The result is a scientifically based assessment of conditions and a prioritization of restoration needs. Because each segment, or stream reach, is rated individually, we can systematically examine conditions along a stream from the perspective of the fish. In this way, we locate areas where conditions are particularly good or bad and identify things that need to be fixed. In particular, EDT identifies the “restoration potential” and the “protection value” of each reach. This helps us prioritize actions and focus them on areas with identified problems and where the potential for benefit is highest. 

The model provides the flexibility to incorporate the effects of up to 45 specific variables reported to affect fish survival. Functional relationships between conditions and rates are described in series of rule curves derived from an extensive review of the scientific literature. Effects of specific habitat conditions are used to scale stage-specific survival rates between normal ranges reported from empirical data.

EDT represents the state of the art in salmon habitat-fish modeling. It is in wide spread application for salmon recovery planning efforts throughout the Pacific Northwest. Many other approaches attempt to relate fish population characteristics to habitat conditions, based on a subset of the relationships incorporated into EDT has the flexibility of emulating a variety of alternative approaches depending on data inputs.

The following strengths and weaknesses of EDT are listed in the Lower Columbia Fish Recovery Subbasin Plan: Proposed Analytical Framework (The JD White Company, Inc. and SP Cramer & Associates, 2003):

Strengths of EDT

• Data input structure provides a systematic means of describing basin-wide habitat conditions, qualifying the quality of the input data, and identifying reach-specific limiting factors based on a comprehensive review of the state of salmon knowledge.
• Extensive documentation of underlying relationships and assumptions is available.
• Detailed representation of stream habitat conditions and effects of habitat on fish provides flexibility in representing factors of concern and projecting fish benefits of specific changes.
• Sender-based approach to estimating survival rates protects against unrealistic estimates of population productivity and capacity.
• Estimates of population response to habitat changes are robust even where specific inputs are uncertain.
• One of the few realistic alternatives for inferring historic, current, and future fish population characteristics where empirical estimates based on fish data are not available.
• Provides a means for estimating fish sensitivity to departures from Properly Functioning Conditions.
• Life cycle approach at the core of EDT facilities linkage with life cycle-based population viability approach for integrated analysis.

Weaknesses of EDT

• Highly mechanistic nature requires extensive data that is often unavailable.
• Model complexity can obscure transparency in underlying assumptions, which has led to characterization as a black box.
• Complex interactions of habitat effects on fish can bias projected fish response to change where habitat inputs are unrealistic.
• Incomplete inputs can produce very specific results that are difficult to corroborate without independent data.
• Does not provide explicit estimates of uncertainty in results based on input assumptions.
• Model accessibility is limited by system requirements and specialized expertise.
• Description of equilibrium population conditions does not allow for consideration of risk assessments based on random variables such as ocean conditions.
• Not developed for detailed evaluations of mainstem, estuary, and ocean limiting factors.

Much of the EDT modeling performed to date by the city of Portland has focused on coho salmon. Preliminary results for steelhead were recently completed. Additional work will continue including EDT analysis of other salmonid species, assessment of sources, and project effectiveness. The EDT analysis indicated that in a restored condition, Johnson Creek would probably operate differently than it does today. Many more successful population trajectories (as characterized within the EDT model) would begin from the lower sections of the creek. Also, a portion of the trajectories starting in upper reaches of Johnson Creek would rear in the middle sections, which would provide abundant habitat for juvenile rearing.

EDT Model results revealed the top five ranking of various survival attributes for three scenarios when set to a reference condition. These scenarios include successful trajectories, change in capacity, and productivity for coho salmon and are summarized by watershed section in Table 15. Additional modeling results are provided in Appendix E (PDF, 22 KB).

Table 15. EDT Model Results for Coho Trajectories, Capacity, and Productivity

Watershed Section	Successful Coho Trajectories when attribute set to reference condition (percent)	Change in Coho capacity when attribute set to reference condition (percent)	Change in Coho productivity when attribute is set to reference condition (percent)
Lower Johnson Creek	Habitat Diversity 23.5	Habitat Diversity 5.6	Temperature 9.3
	Sediment 5.5	Key Habitat 5.6	Channel Stability 6.5
	Channel Stability 2.5	Food 4.0	Habitat Diversity 2.7
	Flow 2.1	Temperature 1.2	Chemical Pollution 0.6
	Temperature 1.7	Sediment 0.8	Sediment 0.4
Middle Johnson Creek	Habitat Diversity 66.0	Habitat Diversity 14.3	Habitat Diversity 2.4
	Sediment 30.9	Food 9.0	Sediment 1.9
	Flow 17.0	Key Habitat 7.1	Temperature 1.5
	Temperature 10.3	Temperature 3.0	Channel Stability 1.0
	Channel Stability 5.2	Flow 2.6	Flow 0.7
Upper Johnson Creek	Sediment 66.7	Food 20.1	Sediment 9.8
	Habitat Diversity 30.2	Habitat Diversity 10.6	Flow 9.4
	Flow 11.1	Key Habitat 8.5	Food 7.7
	Channel Stability 1.6	Flow 4.8	Habitat Diversity 5.5
	Temp., Food., and Harassment 0.8	Sediment 3.6	Temperature 4.3
Crystal Springs Creek	Not complete as yet. See notes below.
Kelley Creek	Not complete as yet. See notes below.

Source: E-mail communication and meeting handouts from Chip McConnaha (Mobrand Biometrics and the Portland ESA Program).

EDT modeling and analyses continues on Crystal Springs Creek and Kelley Creek. The following preliminary conclusions were obtained from Chip McConnaha (Mobrand Biometrics) delivered via e-mail to Ali Young (Portland BES) on June 20, 2003:

There are three overarching habitat restoration priorities for Crystal Springs Creek including:

1)      Temperature. Need to reduce summer temperatures down to a reasonable level. Refugia may be present that would allow for some summer rearing of salmonids. A temperature budget should be performed to more completely understand the temperature regime.

2)      Access. Crystal Springs has numerous culverts, most of which inhibit if not block fishery access.

3)      Structure. Crystal Springs, like most of the Johnson Creek watershed, has very little structure. In fact, Crystal Springs Creek has appreciably less structure than Johnson Creek even accounting for the WPA work in Johnson Creek. Many sections are totally lined by concrete and wood is almost non-existent.

The overarching habitat restoration priorities for Kelley Creek include:

1)      Create larger contiguous habitat areas rather than isolated pockets of habitat. Connect habitat in Kelley Creek with the relatively good areas in Johnson Creek just above the Kelley Creek confluence.

2)      Protect and improve upper Kelley Creek tributaries for water quality (temperature and sediment).

For more information on restoration strategies for both Crystal Springs and Kelley Creek see Section 2.13.2 – Restoration.

Figure 10 (PDF, 79 KB) and Figure 11 (PDF, 37 KB) summarize Preliminary EDT Model results for Coho and changes in number of successful life history trajectories, total productivity, and total capacity within the Willamette and various subwatersheds of the Johnson Creek basin. 

The EDT Model was also run for steelhead. Preliminary results indicate that Steelhead are not doing as well as Coho, and although the overall degradation impacts/restoration potential patterns are similar, they are more exaggerated for steelhead.  Restoration efforts show the most potential for steelhead in the middle Johnson Creek segments. EDT Model results for steelhead trajectories, diversity, productivity, capacity, and abundance are summarized in Table 16.

Table 16. Preliminary EDT Model Results for Steelhead

	Number of Trajectories	Number of Sustainable Trajectories	Diversity  (percent)	Productivity (No. of fish)	Capacity (No. of fish)	Abundance (No. of fish)
Degraded	1232	0	0	0	0.50	0
Current	1232	39	3	2	41.3	20.6
Restored	1232	1227	100	17.75	411	388

Source: Portland ESA Program

Figure 12 (PDF, 80 KB) and Figure 13 (PDF, 82 KB) summarize Preliminary EDT Model results for winter steelhead and changes in number of successful life history trajectories, total productivity, and total capacity within the Willamette and various subwatersheds of the Johnson Creek basin. Note, that although the capacity is very small for Kelley Creek, the productivity increase in a restored condition is extremely important.

The EDT Modeling project identified the following key limiting factors that are most critical for coho salmon in the Johnson Creek watershed : 1) low habitat diversity due in part to a lack of wood; 2) simplified channel structure; 3) degraded banks including WPA rock work, bank grading, and channel lining; 4) degraded riparian areas; 5) high summer water temperatures; 6) excessive sedimentation; 7) lack of food (aquatic benthic macroinvertebrates); and 8) toxics. In addition, high bacteria levels throughout the watershed are considered problematic and should also be classified as a priority. The following section summarizes the key functions and processes that are associated with theses limiting factors.

Habitat diversity is a primary limiting factor in Johnson Creek resulting from channelization and confinement due to the WPA works and the lack of large woody debris.  Recommendations from researchers who developed the EDT Model include: a) wherever possible, removal of the WPA channelization should be prioritized; and b) large woody debris should be introduced through both new, anchored wood and through maturation of a healthy riparian area.

Simplified channel structure is caused by a lack of streamside complexity and a lack of microhabitats. Channels that are allowed to meander form off-channel complex habitats. Large wood and boulders promote shifts in flow changes and velocity currents that help support channel structure and roughness and riffle-pool sequences.

Degraded banks are not able to contain flows and withstand erosive forces. Vegetation and especially root systems play a key role in the integrity of stream banks. 

Degraded riparian areas are not able to promote channel compensation and integrity.  Healthy riparian areas provide structure and complex habitats and connectivity to uplands. Disturbance regimes can result in changes to vegetative classes.

EDT Model results also reveal the importance of restoring floodplain connectivity.  Particular emphasis should be placed on supporting the following Portland projects in Middle Johnson Creek: Kelley Creek Meanders, Alsop Brownwood, West Lents Restoration, East Lents Restoration including south of Foster and Springwater Wetland Complex Restoration projects. Lower Johnson Creek projects including Tideman Johnson/Errol Heights Restoration, Bell Station Flood Mitigation, and the Westmoreland Park Restoration project should also be a high priority (Portland ESA, 2002). 

Excessive sedimentation and high summer water temperatures limit production of coho salmon throughout Johnson Creek. Therefore, sediment and water temperature sources should be investigated and riparian buffers should be established to complete shading and provide natural biofiltration.

Lack of food in the Johnson Creek watershed is a function of the lack of habitat (overhanging vegetation and substrate structure) for aquatic species such as benthic macroinvertebrates. Additionally, high flow disturbances and poor water quality conditions including excessive sedimentation are contributing to a lack of food sources and availability.

Recent information on pesticides and other toxics that were not incorporated into the EDT Model indicate that water quality may be of greater importance, especially during storm runoff events and the potential for both chronic and acute toxicity levels for aquatic organisms throughout the watershed.

Both animal and human wastes cause high fecal coliform and E. coli bacteria levels.  A wide variety of animals utilize habitats throughout the Johnson Creek watershed - both native (wildlife) and domestic (pets and livestock). Human wastes can contribute to high bacteria levels through failing onsite septic systems and wastewater spills and overflows.