Monitoring Watershed Health[1]
State-Federal Monitoring Partnership
Sept 2002 Progress Overview
Steve Lanigan, Acting AREMP Module Lead
A group of Washington, Oregon, California, and federal employees have been meeting since November, 2001 to discuss how we can coordinate/integrate our different watershed health monitoring efforts. Our goal is to present agency “executives” with different options for implementing a cooperative monitoring effort. To date, we’ve learned about each agency’s monitoring program and objectives, and identified the following coordination focus areas.
· Sampling design
· Field attributes and protocols
· How to share data at several different scales (i.e., can we do this using a Decision Support Model?)
· Data management, data analysis, interpretation and reporting
This document provides an overview of what’s happening to address each of the focus areas, as well as other efforts that are related to watershed health monitoring. Please contact Steve Lanigan (slanigan@fs.fed.us, 503.808.2261) with edits/clarifications on anything in this overview, along with any other monitoring info (including websites) you are aware of, e.g., developing protocols for assessing landslides, statistical comparisons of different sampling designs, etc. that’s worth sharing with others. This document and links to other state and federal monitoring efforts can be found at http://www.reo.gov/monitoring/watershed/index.htm.
Why it’s of interest?
There are currently three basic sampling designs in use that share common features, but are based on answering different questions. Our goal is to integrate them together so we can use each other’s data for inference at a greater landscape scale.
· AREMP (Westside federal interagency monitoring program) - randomly select 6th field watersheds throughout the Northwest Forest Plan (NWFP) area. Within each selected watershed, randomly select (on 1:100,000 stream layer) and sample 4-12 sites
o Used for federal interagency effort to answer question: what is the status and trend of watershed conditions in the NWFP area?
o In-channel attributes, and riparian and upslope characteristics (derived from GIS layers) are aggregated together in a Decision Support Model (DSM) to describe watershed conditions.
· EMAP (monitoring program OR and WA (?) is using) – randomly select sample sites across the landscape (on 1:100,000 stream layer).
o Used by states to answer question: what is the status and trend in key parameters related to biological condition, habitat condition, and water quality for stream reaches?
· PIBO (Eastside federal monitoring program) - randomly select 6th field watersheds throughout the PacFish/InFish area. Within each selected watershed; physical habitat, aquatic biota, and riparian vegetation attributes are sampled within the lowest response reach (the “integrator reach”).
o Used by the USFS and BLM to answer the question: Is the management direction from PacFish and InFish resulting in improved conditions of stream and riparian areas?
o Baseline conditions are being assessed by comparing managed versus reference watersheds and trends described using a repeated measures design.
· Integrating National Forest large-scale monitoring efforts workshop.
o We’ll be holding a workshop Nov 18-19 in Portland to address several questions related to statistical costs/benefits of integrating AREMP, PIBO, and EMAP designs and protocols. Additional presentations/workshops are planned for 2003, including at the American Fisheries Society - Oregon Chapter meeting (February 27-28 in Eugene, OR) and the American Fisheries Society Western Division meeting (April 14 – 17 in San Diego, CA). Contact Steve Lanigan for more information.
· EMAP and AREMP
o Tony Olson and Phil Larsen (EPA) and Steve Lanigan and Chris Moyer (AREMP) are determining the statistical costs/benefits of sampling many 6th HUCs with one site (selected by random stratification) versus sampling many random sites within a 6th HUC (fewer HUCs would be sampled).
§ Use the EMAP sample sites to characterize 6th field watersheds (usually one per watershed).
§ Timeline: Present preliminary results for the Nov 18-19 workshop.
· Bob Gresswell (USGS-FRESC) is finishing a study to determine how many sites need to be sampled within a 6th field watershed to adequately characterize the watershed. . Study objectives are:
o How do individual indicators vary within and among, reach, segment, and subwatershed scales?
o How many samples will be needed to expand estimates obtained at sample sites to reach, segment, and subwatershed scales?
o What is the impact of alternative sampling strategies in subwatersheds?
o Timeline: Draft results due in Feb 2003.
· Bob Gresswell also has the following studies of interest underway. These should all help us decide on the best way to monitor watershed health.
o Develop methods and procedures to address the spatial distribution of indicators used as surrogates for those processes with inherently high variability in spatial distribution throughout the watershed, in order to increase the precision and reliability of assessing watershed condition.
§ This is an expected outcome of an assessment of 40 randomly selected third-order watersheds across western Oregon. Preliminary results expected by Feb 2003.
o Development and use of predictive models. Predictive models would rely on measures of key upslope indicators that have been shown to be empirically (or theoretically) related to key riparian or channel habitat or biotic attributes. This is an expected outcome of an assessment of 40 randomly selected third-order watersheds across western Oregon. Preliminary results expected by Feb 2003.
o Develop the relationships between management and watershed (or indicator) condition from watersheds where ACS management practices are being implemented, as a basis for adaptive management. Although we are not currently working on this objective specifically, we are involved in a paired-watershed study of current logging practices on private industrial forest lands. Our study incorporates multiple biological response variables (e.g., insects, amphibians, coastal cutthroat trout, steelhead, and coho salmon) at a variety of spatial scales from watershed to habitat unit. We are specifically interested in the relationship of fish movement to habitat use and the influence of disturbance (i.e., land management) on this relationship.
· Several people are using PIBO monitoring results to answer questions about needed sample sizes, sources of variability, and other topics related to determining watershed condition.
o Brett Roper (USFS National Aquatic Monitoring Coordinator) and others have assessed the observer and site variability associated with the PIBO monitoring effort. A paper is currently in press in the American Waters Resources Association that describes sample sizes needed to detect changes in watershed conditions. The results provide an indicator of the usefulness of each variable in large-scale monitoring assessments.
o Eric Archer (PIBO) and others have a Forest Service General Technical Report (in press) that uses the same quality assurance data to describe observer variability and the relation between observer and site variability for 15 physical habitat variables. The paper also attempts to describe the sources of variability within each method.
o Marc Coles-Ritchie (PIBO) and others have produced a similar General Technical Report (in review) that describes observer variability, the relation between observer and site variability, and sample sizes needed to detect changes for riparian vegetation methods described by Winward (2000).
o Nick Bouwes (EcoLogical Research) and Brett Roper (USFS) are using PIBO physical habitat, riparian vegetation, and general watershed characteristic data (from GIS) to define useful strata within the interior Columbia River basin. The expected outcome is a map of the strata and list of covariates that best explain watershed conditions.
· Brett Roper (USFS National Monitoring Coordinator), Jeff Kershner (PIBO coordinator), Steve Lanigan and Chris Moyer (AREMP) are determining the statistical costs/benefits of sampling many 6th HUCs with one site (arbitrarily defined as the integrator reach) versus sampling many random sites within a smaller number of 6th HUCs. The total number of sampled sites is the same.
o Look at signal to noise ratio for 3 different stream gradients (based on A, B, and C Rosgen stream types).
o How many watersheds need to be sampled to detect a change?
o How permanent do sites have to be (can they be off by several 100 meters)?
o Does the lowest response reach serves as an “index” of HUC stream network condition?
o How does watershed size affect analysis/results?
o Is there a way to integrate our monitoring program objectives (so that a common sample design is possible)?
Field attributes and protocols
· Field protocol comparison – Federal and state survey teams surveyed three stream sites in Mt. Hood National Forest in Oregon and three sites in Boise National Forest in Idaho during summer 2002. Heath Whitacre (USU graduate student) and Kirsten Gallo (AREMP) are analyzing the data.
o Objective 1: Compare consistency among observers using the same protocol.
o Objective 2: Determine which protocol method(s) most precisely measures each physical stream attribute through comparison of coefficients of variation between protocols.
o Objective 3: Integrate most precise physical attribute sampling methods into a sampling design that will most efficiently detect change in stream characteristics due to management.
o Timeline: Present initial results at Nov 18-19 workshop.
· State and Federal protocol comparisons
o Ongoing effort to identify what protocols are being used for each attribute.
§ Spreadsheet of comparisons is available from S.Lanigan.
· BPA has proposed establishing the following:
o Identify a small, core set of indicators that:
§ Encompass a range of biologically important factors.
§ BPA would ask (and pay) for in all of its watershed monitoring project contracts.
§ Would be measured with protocols compatible with those used by land management agencies monitoring in at least the same subbasin and hopefully across the basin. By compatible, it means the resulting data are either directly transferable (same protocols) or translatable (different protocols, but data are highly correlated [fuzzy curves?]).
§ Preferably could be summarized into a single metric of stream/watershed health (DSM?).
· Riparian and upslope vegetation monitoring
o Interagency Vegetation Mapping Project (IVMP)
§ Provides maps of existing vegetation, canopy cover, size, and cover type for the entire range of the Northern Spotted Owl using satellite imagery from Landsat Thematic Mapper (TM). More info is available on http://www.reo.gov/monitoring/lsog.
§ Is IVMP remote sensing data at a small enough scale to adequately characterize riparian and upslope vegetation for use in AREMPs Decision Support Model?
· AREMP is working with Remote Sensing Applications Center (RSAC) personnel to determine the accuracy of IVMP, based on aerial photo interpretation. Preliminary results expected in December 2002.
o Bob Gresswell is working with Lisa Ganio and Barb Schrader (OSU) to define relationships between riparian vegetation and fish distribution using field surveys and remote sensing.
§ In a related project Bob is developing continuous longitudinal patterns of riparian vegetation on several of the streams in the Coast Range and will be assessing the relationship to fish distribution.
o The PIBO staff is working with range managers and plant ecologists to assess the use of community type information for the monitoring of non-forested riparian areas.
§ Specifically, do we lose important information that is helpful in detecting changes when plant assemblages are grouped into community types?
· Western Regional Center for Monitoring and Assessment of Freshwater Ecosystems (Western Assessment Center)
o Chuck Hawkins (USU), Gretchen Hayslip (EPA) and several state (Rick Hafele, Oregon DEQ) and federal agencies (Rick Henderson, PIBO and Kevin Whalen, BLM) are developing the Western Assessment Center. The Center’s role will be to identify key research and technical assistance needs with the objective of strengthening Western biomonitoring monitoring programs
· The Northwest Biological Assessment Workgroup (NBAW)
o The NBAW has an annual meetings to promote better understanding of the biotic communities in freshwater aquatic ecosystems of the northwest. The NBAW also sponsors a series of workshops on the taxonomy of Pacific Northwest freshwater benthic invertebrates. Workshops are to augment training for scientists actively working on benthic macroinvertebrate taxonomy, and to seek consensus on standard levels of taxonomic effort to be applied in biomonitoring studies.
· http://www.epa.gov/r10earth/offices/oea/aqbioass.html
· Inventory and
Monitoring of Salmon Habitat in the Pacific Northwest - Directory and Synthesis
of Protocols for Management/Research and Volunteers in Washington,
Oregon, Idaho, Montana, and British Columbia
o This document reflects an effort to establish a consistent format for the collection of salmonid habitat data across the Pacific Northwest. More specifically, our objectives were to: 1) provide a synthesis of the salmon habitat protocols applicable to the Pacific Northwest, 2) recommend a subset of these protocols for use by volunteers and management/research personnel across the region, 3) link these protocols with specific types of habitat projects, 4) establish a Quality Assurance/Quality Control framework for the data derived from the use of these protocols, and 5) to the degree possible, identify the format and destination where the data is routinely sent.
· http://www.wa.gov/wdfw/hab/sshiap/dataptcl.htm
· StreamNet
This is a cooperative venture of the Pacific Northwest's fish and wildlife agencies and tribes and is administered by the Pacific States Marine Fisheries Commission. We provide data and data services in support of the region's Fish and Wildlife Program and other efforts to manage and restore the region's aquatic resources.
Assessing Watershed Health using a Decision Support Model (DSM)
Here's the basic scenario: State and federal partners are interested in finding a unified approach to analyzing monitoring data and describing the condition of aquatic resources. One possibility currently being discussed is the use of a Decision Support Model (DSM) to assess watershed health across a large landscape that includes state, private, and federal ownership. Theoretically, each agency could determine watershed health at whatever scale they are interested in, e.g., 6th field watersheds, Ecologically Significant Units (ESU), the Northwest Forest Plan area. Things are presently “complicated” because each agency may also be looking at different attributes (or measuring them differently than AREMP), and an integrated sample design still needs to be developed to allow inference at different scales.
AREMP is using a DSM to integrate all the collected watershed information into a watershed condition value between –1 and +1. By tracking the cumulative distribution of the watershed condition values over time, one can determine how watershed condition is responding to management plans. Use of a DSM also allows us to identify which attributes have the greatest affect on the watershed condition.
· A peer-reviewed DSM will be developed during winter 2002 for each of the eight aquatic provinces identified in the Northwest Forest Plan area.
§ Evaluation criteria curves (aka “fuzzy curves”) are used to assign a value between –1 and +1 to each attribute we collect in a watershed. This information is used by the DSM to derive an overall watershed condition value.
· We currently have evaluation criteria curves built for each of the 20 attributes we use to describe watershed condition, based on scientific literature.
· A peer-reviewed set of evaluation criteria curves will be developed during winter 2002 for each of the eight aquatic provinces identified in the Northwest Forest Plan area.
Sean Gordon (PhD student at OSU) is preparing a “white paper” to address some of the questions associated with DSM:
· Can we develop one model with state/federal agencies that everyone agrees to use, or should each agency develop their own model? What are advantages/disadvantages to each approach?
o Assuming one model is developed, would the model "penalize" someone for not using all the attributes? E.g., State agencies don't collect periphyton data. Will they get "penalized" for not entering periphyton data? Can the model be written so it doesn't penalize for some variables, but does for others (or is neutral about missing data)?
o What about a "core model" that uses a single agreed upon subset of variables, but allows each agency to add on additional fuzzy curves
· Evaluation Criteria Curves (aka “fuzzy curves”).
o What is the effect of using fuzzy curves derived from different protocols that measure things at different resolutions? E.g., Some methods may use a lazar level to measure habitat variables to nearest mm and other methods estimate to nearest m.
o What is the effect of using different fuzzy curves for a given attribute? Is it legitimate to say the resulting watershed conditions are comparable?
o How in the world would one keep track of numerous fuzzy curves (and refining them over time)?
GIS databases being developed by the Interorganizational Resource Information Coordinating Council (IRICC).
Data management, data analysis, interpretation and reporting
· Aquatic and Riparian Effectiveness Program (AREMP)
o Currently analyzing what corporate database is the “best fit” for our data.
§ NRIS (USFS database), ARIMs (BLM database), or separate “custom database.
§ We plan to migrate all our data to a corporate database, where it will be available to other agencies, in fall 2002.
o AREMP reports are posted on the web at http://www.reo.gov/monitoring/watershed/index.htm
§ Technical reports (2000 pre-pilot and 2001 pilot)
§ Annual summary reports (from 2001 on)
§ Protocol handbook
§ We also provide data summary reports to each field unit we collected data on. This information is available by contacting Chris Moyer (cmoyer@fs.fed.us, 541.750.7017). We plan to post summary data on our website beginning with our 2002 results.
o Our http://www.reo.gov/monitoring/watershed/index.htm website is being updated to include links to other state and federal monitoring efforts.
· Pacfish/Infish Biological Opinion (PIBO) related documents are posted on the web at http://www.fs.fed.us/biology/fishecology/emp/index.html
o PIBO Effectiveness Monitoring Plan
o Example of annual reports given to each forest / BLM office
o Sampling protocols
o Links to peer reviewed documents
o Forest / Resource Area specific reports, copies of the database and all other information is available by contacting Rick Henderson (rhenderson01@fs.fed.us) or Eric Archer (earcher@fs.fed.us)
· Washington Department of Ecology - Freshwater Monitoring
This agency conducts Washington statewide monitoring in river and stream waters quality, biological monitoring, lakes monitoring, nuisance aquatic plant monitoring, and provides evaluation and guidance services for volunteer monitoring groups. Monitoring information, including analytical and methods guidance documents, are at following web addresses:
o River water quality: www.ecy.wa.gov/programs/eap/fw_riv/rv_main.html
o Lake water quality: www.ecy.wa.gov/programs/eap/fw_lakes/lk_main.html
o Aquatic plants: www.ecy.wa.gov/programs/eap/lakes/aquaticplants/index.html
o Stream biology: www.ecy.wa.gov/programs/eap/fw_benth/fwb_intr.html
o Contact Rob Plotnikoff (rplo461@ecy.wa.gov) for more information.
· California North Coast Watershed Assessment Program
The North Coast Watershed Assessment Program, or NCWAP, is designed to meet four goals:
[1] C:\aremp\Federal & State Coordination\fy02\Overview_of_efforts_.doc