We conducted two multi-factor analyses of coho summer habitat for this assessment. The first was the Potential Coho Summer Habitat Analysis. As requested by the MidCoast Watersheds Council Tech Team, this analysis located stream reaches that were classified as "unconfined" (flowing through flat areas over 1.5 ha in size) and low-gradient (0 to 2 degrees, or 0 to 3.5% slope). The Functioning Coho Summer Habitat Analysis used the results of the potential habitat analysis, and added in several other factors from AHI and RBA surveys as described below.

The Potential Habitat Analysis was a sub-6th field analysis that provided data on specific stream reaches meeting the criteria of low gradient and "unconfined" as defined by DEM analysis. By contrast, the Functioning Habitat Analysis was a 6th field ranking.

There are many possible definitions of "functioning coho summer habitat." The MidCoast Watersheds Council requested we incorporate the following factors into this multi-factor analysis:

Factor	Effect of high value on ranking (+/-)
Length of potential habitat (unconfined,* low-gradient)	+
LWD frequency (pieces/100m)	+
Percent pools by area	+
Channel widths per pool	-
Length of riffle habitat with gravel substrate dominant	+
Length of riffle habitat with bedrock substrate dominant	-
Percent shading of stream channel	+
Rapid Bioassessment average coho/sq m	+

*see DEM analysis of stream confinement for details on definition of "unconfined"

We used the following data sources for the above factors:

Factor	Data source	Analysis link
Length of "unconfined" low-gradient streams	Digital elevation model (DEM)	derived_stream_confinement.htm der_stream_gradient.htm
LWD frequency (pieces/100m)	AHI data	ahi_lwd100.htm
Percent pools by area	AHI data	ahi_percent_pools.htm
Channel widths per pool	AHI data	ahi_channel_widths.htm
Length of riffle habitat with gravel substrate dominant	AHI data	ahi_gravel.htm
Length of riffle habitat with bedrock substrate dominant	AHI data	ahi_bedrock.htm
Percent shading of stream channel	AHI data	ahi_percent_shade.htm
Average juvenile coho/sq m	Rapid Bioassess-ment (RBA)	RBA1999.htm

Six of the factors in the analysis consist of AHI data. AHI data came from three separate sources -- USFS Region 6 tabular data, ODFW GIS data, and Lincoln District AHI data. For those factors that were analyzed as numeric means or percentages (LWD, % pools, channel widths per pool, and % shade), there were protocol differences between the different data sources. In addition, sixth fields differed in terms of data availability:

Data available for coho summer habitat multi-factor analysis	# of 6th fields
All data available	64
All data except % shade available	90
All data except % shade and RBA data	154

Because of differing protocols and data availability, we used a normalized ranking scale that was independent of the specific values for each parameter and the number of parameters available.

We used several steps to rank 6th fields for this multi-factor analysis:

For factors that consisted of lengths (i.e., length of unconfined, low-gradient streams; length of riffle habitat with bedrock substrate dominant, and length of riffle habitat with gravel substrate dominant), only two steps were needed:

1. Determine total length for each sixth field.
2. Determine sixth field watershed rankings, then normalize rankings to a scale of 1 to 100.

For the Rapid Bioassessment data, we used two steps:

1. Calculate average juvenile coho density as described in the Rapid Bioassessment section
2. Rank sixth fields for average density, then normalize the rankings to a scale of 1 to 100.

For the AHI factors (which originated from several different data sources), the procedure required more steps:

1. Determine average value for 6th field from each data source separately.
2. Determine ranking for each 6th field from each data source separately.
3. Normalize the ranking from each data source to a scale of 1 to 100.
4. Average the normalized rankings from all available data sources to get an "average normalized rank" for each sixth field watershed.

We averaged the normalized rankings across all parameters to get a final sixth field ranking for the multi-factor analysis (cohos_rnk in aqhab_sum_final.xls or aqhab_sum_final.zip). The final ranking uses a scale of 1 to 100.  Values for the final ranking ranged from 22 (best) to 83 (worst).

Specific methods for length factors For length of riffle habitat with gravel substrate dominant, we summed the lengths from all data sources and ranked the 6th fields from top (rank 1, greatest length of gravel-dominated riffles) to bottom (shortest length) based on the total length. We used the same procedure for the DEM-derived length of unconfined, low-gradient streams.

For length of bedrock habitat with gravel substrate dominant, we summed the lengths from all data sources and ranked the 6th fields from top (rank 1, shortest length of bedrock-dominated riffles) to bottom (longest length of bedrock-dominated riffles).

As described in step 6 above, we averaged all of the normalized individual factor rankings to get a single final Functioning Coho Summer Habitat Ranking for each sixth field watershed. Since all rankings were normalized to a scale of 1 to 100, this average is not biased by number of parameters included in the ranking.

However, it is important to be aware of the different levels of data availability in different sixth field watersheds. Only 64 of 217 sixth field watersheds had complete data for this ranking. To provide more data to the watershed council, we decided to rank those sixth fields that had AHI data but lacked shade or RBA data, so that a total of 154 sixth fields could be ranked. We recommend that users consider data availability when using the rankings.

It is important to note that all 6th fields with AHI data available were ranked in this analysis, even those which may be inaccessible to coho. Sixth fields without RBA data were ranked, as were 6th fields without % shade data.  We felt it was best to rank all 6th fields with AHI data, even those without RBA data, since lack of RBA data does not necessarily indicate lack of coho or coho access. Some 6th fields may not currently be populated with coho (or may not have been surveyed), but may offer good opportunities for restoration actions such as barrier removal that could re-introduce coho into those areas.

Only 6th fields with AHI data could be ranked in this analysis. Some 6th fields lacking AHI data may have good coho habitat functions. We recommend AHI surveys for areas not yet surveyed but offering good coho summer habitat potential, and we also recommend re-survey of areas surveyed several years ago to determine whether habitat has changed since the earlier survey.

We followed the above ranking procedures because we felt they provided the best interpretation of the data. However, many different ranking systems are equally defensible for an analysis of this type. If they wish, MCWC members will be able to re-rank 6th fields using alternative systems by manipulating the aquatic habitats summary shapefile aqhab_sum_final.zip.

As requested by MCWC, we used absolute lengths for analysis of potential habitat and substrates. Proportional lengths, rather than absolute lengths, provide certain advantages in tracking change over time. A discussion of proportional lengths versus absolute lengths is found in "Interpreting the results of aquatic habitat analyses: Absolute lengths versus proportion of surveyed lengths" above.

The rankings can help prioritize 6th fields for actions designed to improve coho summer habitat. The rankings should not be used alone for this purpose, but should be used in conjunction with other data, particularly field verification of suitable conditions.

Before using the rankings, we recommend careful review of the detailed methods for each individual analysis that entered into the multi-factor analysis.  All datasets have their limitations and proper uses, and many of these are discussed in the methods sections for the individual analyses.

The data collected in AHI surveys can change considerably over the course of a single year. Therefore, any future analyses of coho summer habitat should use the most recent AHI survey data. We recommend surveying new reaches that appear to offer high potential habitat value, as well as re-surveying critical reaches for which survey data are more than a couple of years old. Since AHI data from USFS and Lincoln District sources was not georeferenced, it was not possible to develop a site-specific, reach-by-reach analysis of functioning habitat that incorporated all of the available AHI data. However, it would be possible to locate some specific reaches that meet all of the Functioning Coho Summer Habitat criteria, by using the Aquatic Habitat Inventory data that originated from the ODFW GIS.  This analysis would be a logical next step for the Basin Planning Teams. For such a site-specific analysis, it will be particularly important to consider the age of the AHI data (survey date). Ground-truthing or re-survey of critical reaches is recommended, particularly if the AHI data in question are more than a year or two old.

Scale is a consideration in site-specific analyses such as the one described above. The ODFW data are entered on a 1:100k streams layer, while the DEM analysis is conducted at the 1:24k scale that is considered appropriate for watershed assessment at the 5th field level. This scale difference will need to be considered when conducted any site-specific analysis that uses both DEM and ODFW GIS data.

Please address questions or comments to:

Dr. Ralph Garono
Principal Investigator
Earth Design Consultants, Inc.
rgarono@earthdesign.com

Laura Brophy
Principal Investigator
Green Point Consulting
brophyl@peak.org | homepage