OBMEP ISRP response
I would like to thank the ISRP for the time and energy they have provide toward this review. The Confederated Tribes of the Colville Reservation agree with the ISRP that our integrated monitoring efforts in the Okanogan Basin are headed in the right direction. We also agree that the data set is “too short at the present time to reach many conclusions” limiting our ability to conduct thorough analyses. It is important to clarify that the OBMEP project is planned to continue as it has since 2005 with only minor changes under the Colville Tribes fish Accord (CHaMP (if approved) will represent additional data collection activities carried out by the same staff but under a separate BPA contract using Fish and Wildlife program or Biop related funding outside of the Columbia River fish accords). The purpose of this document is to address questions raised by the ISRP regarding the ongoing elements of OBMEP and the evolving local, regional, and basin-wide context within which it operates.
1. Purpose, Significance to Regional Programs, Technical Background, and Objectives
OBMEP was designed to adaptively inform the Okanogan habitat and fisheries management programs. Our primary objectives are to monitor habitat and natural production status & trends, and to evaluate the impacts of restoration efforts on ecosystem condition. OBMEP produces natural production data addressing Upper Columbia summer Chinook, sockeye, and steelhead, with an emphasis on steelhead. In addition, the project monitors habitat conditions that influence the performance of Okanogan fisheries and Okanogan ecosystems as a whole.
The Integrated Status and Effectiveness Monitoring Program (ISEMP), Pacific Northwest Aquatic Monitoring Partnership (PNAMP), Upper Columbia Salmon Recovery Board (UCSRB), NPCC, Bonneville Power Administration (BPA), and NOAA Fisheries (NMFS) utilize our raw data to answer questions of regional significance, and we try to make every effort to ensure our data are useful for this purpose. However, OBMEP by itself does not have the resources to conduct extensive statistical analyses or to compare our data with that of other programs used in the development of regional models without considerable support from other groups.
At the regional level OBMEP implements Monitoring RPAs associated with the FCRPS BiOp, and collaborates with the UCSRB, the Canadian Bi-Lateral Okanogan Technical Work Group (BOBTWG), and the Columbia River Inter-Tribal Commission to inform the management and recovery of Upper Columbia stocks. At the Columbia Basin level OBMEP works with PNAMP, ISEMP/CHaMP, and NMFS to address Basin-wide issues and to develop Basin-wide tools. The role of OBMEP is to lead habitat and natural production monitoring in the Okanogan Subbasin, and to coordinate with regional and basin-wide RM&E programs to maximize the efficiency of these collaborations.
The relationships between OBMEP and CHaMP are just being initiated so specific methodological decisions have not been made. At this time, it appears that ISEMP Project # 2003-017-00 will provide the command and control structures for CHaMP. Therefore they will make all the design decisions, conduct all the training, provide all the data management support, conduct all the statistical analyses, and produce all the reports related to CHaMP. The OBMEP expansion will only fund the field data collection and daily logistical support for CHaMP in the Okanogan River subbasin.
CHaMP will experimentally implement sampling designs and field techniques that differ from many Federal, Tribal (including OBMEP) and State programs. CHaMP was designed around a number of fish-habitat hypotheses that were used to set habitat monitoring priorities (Bouwes et al, 2010). However, the analytical and decision-making framework associated with these hypotheses is un-developed and untested, and the hypotheses brought forth by CHaMP are inconsistent with the Okanogan, Upper Columbia, and Columbia Basin management programs. For example, some of the key environmental attributes (e.g. embeddedness) used to analyze Okanogan limiting factors and prioritize the Upper Columbia habitat work schedule are excluded from CHaMP.
The OBMEP methods will remain unchanged for 2010-2011, but will include additional and parallel monitoring as recommended by CHaMP. CHaMP/ISEMP will provide the analytical resources needed to provide a comparison between the CHaMP approach and the more established methods that have been applied by OBMEP and across the Columbia Basin over the past decade (Appendix A). OBMEP participates in the PNAMP committees that oversee the design decisions that will guide the implementation of CHaMP. OBMEP staff will work with the PNAMP Habitat Status & Trends Committee to evaluate the results of the CHaMP project, and to adapt OBMEP if and when improved techniques become apparent and well integrated with the local, regional, and basin-wide ecosystem diagnostic and treatment systems.
Over the next 2-years all habitat, macro-invertebrate, water quality, snorkeling, and screw trap data collected by OBMEP will be processed into the STEM data bank (under a separate contract funded by NOAA Fisheries). Once these data are housed within the STEM database additional comparisons and statistical analysis comparing OBMEP, ISEMP and future CHaMP data will be possible. Once all these data are in one place and in one data format the process of developing cross walks to link these data will begin and the time needed for this step is unknown. We expect that realistically over the next 5-years both the existing OBMEP and new CHaMP programs will need to continue so we can make informed decisions that address local and regional monitoring needs. Over the next several years the region will benefit from a collaborative process of exploring the linkages and differences between these 2-habitat data collection efforts providing broader analyses of both fish and habitat interactions.
2. History: Accomplishments, Results, and Adaptive Management
Given that OBMEP is a Biop project our data collected for threatened Upper Columbia summer steelhead is collected in a manner to make it compatible with VSP and 5-year review criteria. We have worked hard to make sure that we remain compatible with ever changing guidance such as that put out by NOAA fisheries for RM&E and OBMEP is specifically called out as a specific component of this guidance. We will continue to have an open dialogue with NOAA fisheries to ensure our monitoring efforts remain compatible with the best science available. We share our data through various data networks such as stream-net, DART, USGS surface water monitoring program and others. OBMEP data is assimilated into the SOTR report prepared by CBFWA. We are actively engaged within PNAMP and the NPCC monitoring coordination activities in the Okanogan and across the Upper Columbia. Each of these activities or groups operates under an independent adaptive management paradigm that OBMEP contributes to and helps to inform (Figure 1).
Figure 1: Adaptive management Framework related to OBMEP and other monitoring efforts in the Okanogan River subbasin
Adaptive Management within the Upper Columbia ESU
Within the Upper Columbia ESU habitat actions and monitoring data are assembled to guide recovery efforts in an adaptive management framework (Figure 2). At the center of this framework are the upper Columbia monitoring projects including OBMEP. Data collected by OBMEP related to habitat and water quality are captured within this process so it can flow up to decision makers and down to people working on developing habitat actions. One of the primary reasons for using the EDT model for evaluation of our data is to make transparent linkages to actions that were undertaken to address limiting factors identified using this model during the recovery planning process. In addition to providing the data upon which the adaptive management framework is built the staff employed by OBMEP also engages at the Watershed Action Team and Regional Technical Team levels (Figure 2).
Figure 2: Adaptive management Framework used by the Upper Columbia Salmon Recovery Board.
Adaptive Management within the Okanogan River Basin
In addition, to being integrated with the Upper Columbia salmon recovery efforts the Colville Tribes and Okanogan Nation Alliance are working collaboratively to implement habitat, harvest, hatchery, and fisheries management actions across the board with Canada. Over the last 5-years habitat restoration projects have been proposed to a wide array of funding arenas and these habitat project proposals have utilized parts of the data collected under OBMEP. To make these linkages more explicate a reach assessment has been in development over the last 2-years to identify which tributary and main-stem reaches most in need of restoration and protection. The assessment is directly linked to the OBMEP database so it can be updated in the future rapidly. The outcome of this assessment tool will focus both protection and habitat recovery efforts funded under the Colville Tribes Fish Accords. In addition, we are currently working on completing a revised EDT3 model run utilizing empirical data collected by OBMEP that will update limiting factors for each of these reaches. With both assessment tools the Watershed Action Team in the Okanogan River basin will be posed to significantly update the implementation schedule used to prioritize targeted actions in the Okanogan River basin for the next 5-years.
Chief Joseph Hatchery Adaptive Management Plan
The Chief Joseph Hatchery Project (CJHP) adaptive Management plan calls for an annual project review that will incorporate considerable data archived into the OBMEP database. As part of the CJHP each year before decisions about harvest and brood stock management for the coming season have to be made, an Annual Project Review (APR) will be conducted through a workshop. The purpose of the APR is to implement the four steps of the In-Season Management Procedure (ISMP) described below.
The agenda of the APR follows the steps in the ISMP. The APR is a science driven process that informs policy decision makers. The end product of the APR is an action plan for the coming season. This action plan is completed at the end of the workshop and presented as a recommendation to decision makers. The APR participants will include habitat, harvest, and hatchery biologists. The workshop and subsequently adopted action plan constitute the all-H coordinated implementation of the CJHP.
Resource management goals directly affected by and relevant to the CJHP are to rebuild and maintain sustainable naturally spawning summer/fall Chinook populations in the Okanogan subbasin, and to provide harvest for tribal and non-tribal fishers. The long-term purpose of the CJHP is to contribute to harvest goals in a manner that is compatible with sustainable natural production (i.e., conservation). During the transition phase(s), the CJHP will be operated in a manner that may accelerate the re-population of Chinook throughout the Okanogan subbasin.
The Okanogan summer/fall Chinook population has been designated as a primary population by the co-managers. This population has two components, 1) those fish that spawn in the Okanogan River, and 2) fish that spawn in the main-stem Columbia River between Chief Joseph dam and Wells Dam.
For the first component, management of hatchery production and natural escapement will adhere to guidelines that minimize the genetic and ecological influence of hatchery fish on the naturally spawning population. These guidelines stipulate that the proportion of the natural spawning escapement composed of hatchery-origin fish must be less than 5%, unless the hatchery program is genetically integrated with the naturally spawning population, in which case this proportion must be less than half of the proportion of natural-origin fish in the hatchery broodstock (HSRG et al. 2004a).
Hatchery fish management for the second component will be less restrictive. Initially, no limitations will be placed on the proportion of hatchery fish spawning in this area. This is necessitated by the fact that water depth and clarity make it difficult to determine spawning numbers or composition (HOR or NOR). Studies will be undertaken to estimate the proportion of NOR fish passing Wells Dam that spawn in this section of the river.
There is uncertainty about the future natural production potential for Okanogan River summer/fall Chinook salmon. While the expectation is that investments in habitat improvements in the Okanogan will improve Chinook productivity and abundance, it is unknown when and to what extent those investments will be effective and these data will be provided by OBMEP. Future survival conditions in the main-stem Columbia, the estuary and the ocean also remain uncertain. Additionally, harvest management policies beyond the control of the Colville Tribe influence the viability of the Okanogan natural Chinook populations.
Because of this uncertainty and because of the annual variability in abundance of natural-origin adult returns, the hatchery program has been designed for flexible production and operations. This flexibility is reflected in the design and operation of the hatchery facilities and weirs and in a set of Decision Rules that determine the size of the hatchery program and the management of natural escapement abundance and composition (Table 1).
Table 1. CJHP Decision Rule set for Current, Transition and Long-term phases of the program
|
|
|
Transition
|
|
|
Natural Escapement
|
Current
|
Period 1
|
Period 2
|
Long Term
|
|
Minimum natural-origin (NOR) escapement
|
800
|
800
|
1,200
|
1,600
|
|
Proportion hatchery-origin spawners (pHOS)- maximum target for Integrated program
|
30%
|
30%
|
30%
|
30%
|
|
NOR escapement at which pHOS must be achieved
|
800
|
800
|
1,200
|
1,600
|
|
Maximum pHOS for Segregated Program
|
5%
|
5%
|
5%
|
5%
|
|
Minimum hatchery-origin spawners (HOS) + natural-origin spawners (NOS)
|
500
|
500
|
500
|
500
|
|
Hatchery Program
|
|
|
|
|
|
Proportion natural-origin brood (pNOB) target
|
100%
|
100%
|
75%
|
30%
|
|
Run size at which pNOB will be reached
|
800
|
800
|
1,200
|
1,600
|
|
NOR broodstock allocation
|
75%
|
75%
|
75%
|
30%
|
|
Adjust segregated program to reduce strays
|
Yes
|
Yes
|
Yes
|
Yes
|
|
Harvest
|
|
|
|
|
|
% of NOR in excess of NOS and NOB taken as harvest
|
10%
|
10%
|
10%
|
10%
|
The purpose of the Decision Rules is to assure that hatchery programs, terminal fisheries and weirs are managed to meet the guidelines for abundance, composition, and distribution of the natural spawning escapement. The ultimate goal of the Decision Rules is to increase the fitness of the natural population by maintaining a PNI > 0.67[1]. The Decision Rules are based on a set of key assumptions about our capability to accurately detect and respond to the annual abundance of natural-origin returns of Okanogan summer/fall Chinook to Wells Dam. The CJHP M&E plan identifies the information needed to update and apply the Decision Rules and describes how data will be collected to derive this information. Resource goals are expected to be met as a result of appropriate in-season management actions taken over time.
The keys to achieving resource goals over time are a) to assemble the most recent and relevant information and b) to use this information to operate fisheries, weirs and hatcheries consistent with the established guidelines each year. To this end, the Colville Tribe intends to implement a four-step, In-Season Management Procedure (ISMP) (Figure 3). This 4 step procedure is formalized in a database and a set of management tools that assure consistency and accountability. As shown in Figure 3, the dashed boxes represent three tools to be used to carry out the ISMP. These tools are all Excel spreadsheets that store and document data and assumptions and derive biological targets for the operation of terminal fisheries, weir and hatchery programs. The tools document the basis for these targets and establish expectations for all performance indicators. They also will help simplify the implementation process and document the rationale for the management actions taken each year. The management biologist responsible for implementation of in-season management will use these tools in preparation for the APR workshop, where analytical results will be presented and shared with all interested parties.
Figure 3. In-Season Management Procedure framework. The shaded box at the top represents the ISMP. The boxes below indicate management tools and reports used to support the procedure. For example, there is a long term plan supported in part by the AHA tool, which in turn informs Step 1 of the ISMP.
Step 1 Update Key Assumptions
The CJHP was developed from a set of key assumptions:
- The quality and quantity of Okanogan subbasin habitat which is updated by OBMEP using EDT
- Survival rates of fish migrating to and from the ocean
- Ocean survival
- Harvest rates in freshwater and ocean fisheries, and
- Effectiveness of weirs and live-capture techniques to remove hatchery fish from the spawning environment without killing substantial numbers of wild summer/fall Chinook
The first step in the ISMP is to update these assumptions each year based on data collected from monitoring and evaluation activities. This step ensures that the best available information and knowledge is applied to the in-season management process. The key assumptions driving the program will be reviewed each year at the Annual Program Review workshop. Conclusions from the workshop review will be captured, documented in the database and the results brought forward to Step 2 in the process.
Step 2 Review Decision Rules
After review and updating of the key assumptions is complete, the Decision rules may need to be altered to account for change in population policy status (e.g., ESA listing), collapse of the run, new science discoveries or other changes in salmonid management in the basin or the region (Figure 4).
Once the key assumptions and Decision Rules have been confirmed, Step 3 of the ISMP will be implemented. This will ensure that standards for the Okanogan Primary population are met and progress toward conservation and harvest goals is maintained.
Figure 4. Control panel for Step 2 in the In-Season Management Tool
Step 3- Update Stock Status Information.
In this step, the most recent stock status information will be entered into the database for both the hatchery and natural components of the run. Recent natural and hatchery escapement is entered by origin (hatchery vs. natural). The number of hatchery-origin spawners (HOR) from hatchery programs other than the Okanogan integrated program is entered separately as HOS (segregated). This information is used to determine pHOS for the integrated and segregated programs, as well as pNOS and pNOB. A PNI value is calculated from this data each year and a cumulative value tracked over a five-year period.
Step 4- Set Biological Target for the Coming Season
With updated stock status, managers will have the data needed to set biological targets (broodstock needs, harvest levels, weir catch and escapement) for the migration year based on run-size predictions. The run size prediction will be updated each week and entered into the analysis tool shown in Figure 5. The tool then outputs biological targets for the program, expected outcomes, and progress being made towards achieving pHOS, PNI and pNOB objectives over a five-year period.
To better refine in-season run size estimates at Wells Dam, a portion of each year’s juvenile outmigration will be PIT-tagged so that upon their return as adults, their migration timing and progress through the Columbia River mainstem dams can be calculated.
Figure 5. Example control panel for ISMP Step 4
The AHA models for the CJHP effort are built on the foundation of the EDT2 models that are populated by empirical data collected by OBMEP. The habitat data currently being collected by OBMEP is being rapidly assimilated in all aspects of the Colville Tribes fisheries management efforts. At this time, it is impossible to say with any certainty what or how the new CHaMP habitat monitoring will impact these adaptive management processes.
OBMEP has provided the Okanogan stakeholders with a centralized habitat and natural production monitoring program and data repository. Our accomplishments and results have been well stated and presented in our annual reports, and were not-reiterated in our proposal or in this response to the ISRP.
We agree with the ISRP that the time series of our data record is relatively short and has limited application at this time. However, OBMEP is well integrated into the Adaptive Management process in the Okanogan, and our continued efforts will play a significant role in both short- and long-term planning (Figure 1). OBMEP natural production information, including spawner, redd, and out-migrant data, represent the primary information source for Okanogan steelhead and fill critical data gaps for Chinook and sockeye salmon. Our ongoing collaborative studies with CRITFC in Omak Creek, and WDFW throughout the Okanogan River will allow us to incorporate additional information regarding parental origin, and kelt performance, age and growth, and origin is central to steelhead recovery in the Okanogan and our understanding of steelhead biology throughout the Upper Columbia. OBMEP disseminates natural production information to managers and scientists via our written reports. In addition, we publish our data to a number of information systems including those managed by ISEMP, StreamNet, DART, and various Federal and State agencies.
As part of our reporting services OBMEP data is post-processed to generate information that is brought to the Okanogan and Upper Columbia Adaptive Management frameworks. These frameworks bring together stock status assessments, limiting factors analysis, and integrative (All-H) planning. Long-term Adaptive Management plans for the Okanogan and Upper Columbia were constructed using the Ecosystem Diagnosis & Treatment (EDT) system (Subbasin and Recovery Plans) and the AHA model (Chief Joseph Hatchery Master Plan).
OBMEP participates in periodic program reviews including the Okanogan Annual Program Review (APR), Bi-Lateral Okanogan Basin Technical Work Group (BOBTWG), Upper Columbia Habitat Adaptive Management Science Conference, and the Upper Columbia Monitoring and Data Management Committee (MaDMaC). We incorporate EDT and AHA in these periodic planning processes to produce consistent scalable out-comes that can be utilized in both short- and long-term planning. It is too early to tell how CHaMP will become assimilated into the collective adaptive management framework. However, CHaMP has developed work plans showing that multiple agencies will be involved with training and integration of field data collection (Appendix A).
3. Project Relationships, Emerging Limiting Factors, and Tailored Questions for Type of Work (Hatchery, RME, Tagging)
We strongly support the use of consistent protocols throughout the Columbia River region. OBMEP was designed to consistently provide habitat data that can be statistically analyzed, and to provide the key inputs to EDT (and subsequently to AHA) needed to adaptively refine the Okanogan steelhead and Chinook limiting factors analysis. We are very concerned that the draft CHaMP protocols would lead to a habitat data set that is largely incompatible with existing EDT2/3 models thus leaving us without a way to present meaningful habitat data analysis to our local constituents, and without a system to consistently inform Okanogan and Upper Columbia habitat restoration efforts.
Therefore, we plan to maintain OBMEP in its current form in addition to the CHaMP methods until these major design and methodological differences can be worked out. We would strongly support a suggestion from the ISRP and NPCC as to the importance of linking CHaMP to the EDT3 model. By taking this position you would be clearing the way for specific regional studies to occur that would help illuminate understanding of fish-habitat relationships that could lead to constructive changes in both the future CHaMP protocols and the EDT3 model. Future planning efforts will be greatly enhanced by linking the EDT model with regional monitoring programs. The OBMEP proposal will update life history trajectories and make basin specific adjustments to the EDT3 model but wholesale fish-habitat relationship modifications would require a larger regional effort. We are committed to conducting EDT assessments in the Okanogan River subbasin as one way to achieve our goal of developing a more effective restoration effort within our sphere of influence. We hope the ISRP, NPCC and the entire region would share in our commitment and linkage to past Fish and Wildlife Program efforts by continuing to support the EDT model.
4. Deliverables, Work Elements, Metrics, and Methods
When OBMEP was first designed we followed the guidance provided by the Upper Columbia Strategy (Hillman 2004). We proposed 50 sites, with one panel sampled annually and the others sampled on five-year cycles, because that is what was used in the Oregon Coast coho studies (rotating panel design, see Urquhart and Kincaid 1999). This sample design for habitat, juvenile standing crop, water quality and macroinvertebrate surveys was developed to describe current status and to detect trends for a suite of indicators within the target population. While status is best monitored by using as many sites as possible representing the broadest geographical distribution of anadromous and resident fish populations, trends are monitored by repeated sampling of the same sites over time. The trend panel consists of 25 sites that are surveyed annually. The status panel consists of a different rotating panel of 25 sites sampled every year, and revisited every five years.
Table 2: Rotating panel design showing years when OBMEP habitat monitoring sites will be visited/revisited.
At that time, no one had conducted any analyses to determine the appropriate sample size. One objective of the ISEMP study was to determine, after five years of sampling, if the sample size and temporal design was appropriate. At the Upper Columbia Adaptive Management Workshop held in Wenatchee this past summer ISEMP presented information associated with variance decomposition which showed the value of a panel design, in particular the repeated visits across years. The variance decomposition confirms the repeatability of the metrics and protocols chosen showed residual variance terms that were small relative to the other terms. The analysis also confirms the metrics and protocols’ sensitivity to detect spatial and temporal patterns in that independent site and year terms are also resolved.
The sampling density of 25 sites in the Entiat is supported by preliminary Bayesian mixed model hierarchical regression analysis that shows subbasin wide, subwatershed and site scale trends in individual habitat metrics. The fixed panel of 25 annually repeated sites in the Entiat captures spatial and temporal habitat information without redundancy. That is, habitat characteristics have independent temporal trends at all 25 sites such that each site is capturing spatio-temporal information above and beyond that shared across the sites. It would appear that 25 sites within the Entiat represent a minimum sample size. The optimal sample size will be addressed with further analysis of the Wenatchee ISEMP habitat monitoring program, where an increased sampling density was employed to facilitate such investigations.
Today, the 50 sites that OBMEP samples can be thought of as 25 status and 25 trend sites or you could chose to look at as 50 sites per year, 25 of which may have some correlation in their trends. There is not an ISEMP analysis that says what the perfect number is, but when the Methow monitoring program (the last of the four populations in the Upper Columbia ESU) begins this year they will also implement habitat status and trend monitoring at 50 sites per year annually. Preliminary results and communications suggest that CHaMP will propose a 25 site sample size with a 3-year rotating panel design. However, this decision will be made by either ISEMP or the NWFSC prior to CHaMP site selection and data collection. It is this type of principle study design choices that could lead to major data integration conflicts and require a “move slowly” approach regarding the integration of the two (OBMEP and CHaMP) habitat monitoring programs. By moving slowly, we will be certain to minimize any unnecessary loss of data or information from the OBMEP habitat work to date.
We share the ISRP’s concerns regarding basic design aspects of CHaMP and how these programs might or might not be integrated. However, it is also impossible to say with any certainty what approach might be advisable at this juncture. OBMEP will work with the NOAA, NPCC, ISRP, PNAMP, RTT or any other interested statistician or research design experts regarding how best to evaluate and integrate these components. However, it is unlikely that this work could be completed prior to data being collected by CHaMP, and there is risk in changing protocols too quickly, without a local understanding of the benefits and risks of the new design. After both data sets have been integrated into STEM, which is scheduled to be complete in 2012, the data will be available for this work to efficiently commence and most of the design choices will have been made by CHaMP. If such a review were conducted it should be carried out by an independent third party that does not have a direct stake or bias associated with the outcome of the study. The recommendations could then be incorporated into the next ISRP review for these projects.
OBMEP data collected (past, present, and future) will be collected, shared, and analyzed by OBMEP and others. CHaMP data will be analyzed by CHaMP. Linking these programs was done in an attempt to collect field data efficiently and leverage existing personnel.
The OBMEP data and analysis routines are currently in a development phase. Habitat data will utilize the new EDT model (EDT version 3 is currently in development). During the development of the revised EDT model (funded independently through ICFI), ICFI took OBMEP habitat data to test and refine the model and develop tools specifically for evaluation of status and trend. We also collaborated on revisions to the life history trajectories for both summer steelhead and summer Chinook and revisions to the temperature survival rating curve. A revised EDT2 model run using OBMEP data was used during the planning for the Chief Joseph Hatchery Project as the base building blocks for habitat capacity and AHA modeling. We hope to have a completed summer steelhead EDT3 habitat and limiting factors report using OBMEP data collected through 2009 completed by early 2011. This report will be used to develop a revised implementation schedule for habitat actions in the Okanogan River basin within 2-years.
The map showing the specific locations for each of the adult summer steelhead data collection points is included in this response and labeled (Figure 6).
Figure 6: Summer Steelhead adult abundance data collection sites in the Okanogan River basin.
In the future, OBMEP fish data will likely be used to feed VSP criteria and Biop 5-years review but today specific uses and applications of our fish data are less tangible. We expect that these data will become increasingly important for steelhead management as the time series lengthens. At present the most important data we summarize is adult summer steelhead returns. We are fortunate that in the Okanogan River basin most of the habitats available to adult summer steelhead are in wadable streams that can be monitored visually. We have been able to get a near census count of adults returning to the Okanogan River basin annually since 2005. We are able to get complete counts at weirs located on the primary spawning areas of Omak, Bonaparte, and Inkaneep Creeks (Figure 6). These weirs are on small streams with no opportunities for avoidance. The data collected also indicates that summer steelhead do not enter these streams until just prior to spawning therefore adult enumeration at the weir is considered a surrogate for the number of spawners and has been verified by comparing redd survey estimates.
Since 2005, we have developed video monitoring sites at Zosel Dam, Antoine, and Ninemile creeks. Video monitoring allows us to count the exact number of adult summer steelhead moving above or below a fixed point. The Zosel Dam site provides an excellent break point for counting fish headed to Canada once you have subtracted the counts for Tonasket and Ninemile creeks and operates 24/7/365. The two tributary video arrays are installed in the early spring and removed after summer steelhead have completed there spawning activity as observed during redd surveys. These 2-tributary sites are located at the downstream end of the subwatershed extent. Observation from these video weirs shows that less than 1% of summer steelhead enter these small tributaries prior to spawning so, once again this has helped validate our assumption that adult counts equal spawner counts.
All areas not covered by either video or trap weirs are surveyed using redd survey methodology. The redd surveys are completed by reach areas and these reaches are independently evaluated to extrapolate the number of redds observed (assuming a 100% coconut and 1-ress per female) by the sex ratio collected at weir sites for the stream of the most similar size. Once the values are computed for each reach all reaches regardless of method used to collect the data are summed to get an Okanogan River population estimate of adult summer steelhead spawners. Occasionally, special situations occur that can make our estimates less accurate and in these cases we attempt to estimate the widest possible range of results and report these along with our best estimate based on our field experience rather than presenting a single number without confidence intervals.
We believe that our method of enumerating adult summer steelhead spawners in the Okanogan River uses a preponderance of the evidence approach to achieve a near census count. We consider this method to result in estimates as near to “truth” as is achievable in this basin and an estimate that is far more accurate then using a single method of data collection coupled to a random statistical design.
In addition we have been collection standing crop estimates and out-migrant estimates since 2005 and 2006 respectively. The standing crop estimates have lead to the development of a summer steelhead habitat carrying capacity estimator that has been used recently to evaluate different habitat restoration actions and made funding decisions related to the Colville Tribes fish accords. This tool will also be at the heart of a 5-years standing crop report that should be completed in the next 2-years. In addition, over the next couple of years we plan to begin producing annual out-migrant to adult and parr to adult indices so the status and trend of productivity can begin to be tracked. Once developed the specific data analysis methods will be incorporated into revised protocols for each data type we collect and it is our hope this work will be completed in the next 5-years.
Over the longer term, we plan to develop many tools that will link biological needs and criteria of salmon and steelhead into our habitat status and trend analysis. One tool has been under development for 2-years and this is our temperature reporting tool. This tool helps evaluate massive amounts of data in a way that it is linked to habitats used during specific life stages. The dimensions of time, space, species, and life history are used to establish criteria under which the habitat data are evaluated for status and trend analysis. For example, a reach of river where it is known that adult migration is delayed can be evaluated over the typical migration period on the basis of the number of days when the water temperature exceeded the threshold that delays migration. The status data set would be the actual hourly data but the trend is determined by the increase or decrease in the number of days of migration delay. This provides a much more meaningful and sensitive indicator as compared to making the trend the mean temperature for the year which has no real biological linkage. Another example would be in a spawning reach where warm summer temperatures have the ability to kill young steelhead before they emerge. At this site the time period would be changed to the period when this species eggs are typically in the incubation life history stage and you would measure the number of days when you exceed the LC50 for the species. If the number of days is increasing you have a major problem if the number of days is decreasing your moving away from a problem. You could then look at the status data over this same time period and evaluate it against state water quality standards and if you are below these levels it is unlikely that this represents a limiting factor.
We believe that habitat could be analyzed in a number of novel ways that could illuminate our understanding for fish and habitat relationships. However, realistically it might be years 10 to 15 of a 20+ year study design before most of these new analytical and automated reporting tools are completed. Many other monitoring studies continue to search for the perfect method for collecting monitoring data. However, we think that continuing to collect data using our original OBMEP design will provide a number of innovative analytical tools whose use and value will extend well beyond the confines of the Okanogan River subbasin and the Confederated Tribes of the Colville Reservation by the end of our planned 20+ year experiment.
The Okanogan (via OBMEP) has been proactive in the revision of its EDT model. We agree with the ISRP that Okanogan habitat and natural production monitoring offers a valuable opportunity to compare EDT model outputs natural production performance. EDT outputs represent long-term estimates of habitat potential, and are only sensitive to out of subbasin conditions in the most general way. However, within the Okanogan, EDT produces reach and diagnostic unit estimates of limiting factors and potential performance.
These diagnostic outputs represent management hypotheses that link long-term habitat conditions to long-term performance of a species and/or life stage. OBMEP will maintain and provide natural production data that can be used to compare against EDT model outputs, and to refine the model assumptions where possible. All Columbia Basin Subbasin Planning Teams have a seat in the EDT User Community, and OBMEP staff are currently members of the EDT Developer Network. These roles provide OBMEP and others with a venue to inform the collaborative development of the EDT system to ensure that it continues to meet the needs of the Columbia Basin co-managers.
ICF International maintains the EDT database and tools. They are ultimately responsible for participating in peer review, coordinating the Development Network, managing input from the User Community, and revising the EDT system as new information becomes available. The EDT User Community is a public-private partnership that has an extensive record of peer-reviewed literature. We agree that the Okanogan represents a valuable opportunity for model validation and improvement using OBMEP data, and welcome the use of our data for those purposes.
References
Bouwes, N., Bennett, S., and Weber, N. (2010) Tributary Habitat Monitoring Summary Report: A review of stream habitat sampling protocols used in the Colombia River Basin. Prepared for NOAA-Fisheries and Bonneville Power Authority. March 19, 2010. 25 Pages.
APPENDIX A
Regional Habitat Monitoring Support: Tools and Services Available to BPA Collaborators
June 28, 2010
Introduction
The Bonneville Power Administration (BPA) is working with regional fish management agencies to implement a tributary habitat action effectiveness strategy that includes monitoring the status and trends of fish habitat as required by the 2008 FCRPS Biological Opinion (BiOp) in at least one population per major population group. Efforts to standardize this work have been assisted by the Integrated Status and Effectiveness Monitoring Program (ISEMP), a BPA-funded and NOAA-sponsored program, which was tasked with developing regional habitat monitoring approaches through the development and field testing of recommended protocols, data capture, storage, and dissemination tools; and analysis. Given the recent pro-posed expansion in habitat sampling, some of these tools may be of interest to co-managers, contractors, and others wishing to implement habitat sampling in a consistent manner.
ISEMP has been coordinating and conducting the testing and development of habitat monitoring protocols and tools in several watersheds within the Upper Columbia, Salmon, and John Day subbasins since 2004. In cooperation with numerous agencies and contractors this work has enabled the side-by-side comparisons of numerous commonly employed habitat monitoring approaches and novel techniques across a diversity of Columbia River Basin ecoregions. These efforts recently culminated in the development of a core suite of habitat metrics and associated study design and data collection approach (a protocol). Importantly, this protocol is supported by data capture and data storage tools. The protocol is expected to undergo continued refinement as testing continues; however, we propose to coordinate this evolution such that changes are made globally across all participating projects, are based on data analysis, and occur at a rate that maintains a consistent core data set that can be fully supported by data capture, data storage, and data reduction tools.
The goal of this document is improved collaboration among habitat monitoring proponents in advance of the July BPA proposal deadline to support regional habitat monitoring. If interested, agencies may take advantage of the protocol and associated tools and study design support, with the goal of developing a large-scale regionally coordinated habitat monitoring program, one that is not hampered by gaps in data collection, independent and divergent habitat monitoring methods, disparate or hard-to-retrieve data, or redundant efforts.
The following are brief descriptions of the tools and services which could be provided to habitat monitoring collaborators. These tools and services fall into three categories: (1) sampling design support, (2) field sampling support, and (3) data management support. The extent to which these tools and services are pro-vided will be determined through collaborative agreements between ISEMP, BPA, and cooperating habitat monitoring contractors.
Sampling Design Support
GRTS Sample Design Documentation, Tools, and Application Guidance
A generalized random tessellation stratified (GRTS) survey design was recommended by Crawford and Rumsey (2009) for monitoring habitat status and trend in the Columbia River Basin. GRTS is a probabilistic sampling design that has been shown to be advantageous for generating habitat condition parameters with known statistical characteristics. Implementing a GRTS survey design correctly is critical to producing a final dataset with known statistical characteristics; however, doing so requires the implementation of strict procedures during the site evaluation and selection process. ISEMP has developed a GRTS Site Se-lection Protocol and Tool to support field crews with efficiently completing the process while strictly enforcing requirements. Standard data layers (described below) will be available to assist collaborators in developing GRTS sample draws that are consistent across the Columbia River Basin. Addition-ally, statistical support will be made available to assess the rigor of proposed GRTS sample draws over the three year duration of the current proposal cycle. The desktop tool, technical training, and future enhancements are available for use by collaborating agencies.
Standard GIS Layers to Support Sample Design and Data Reduction
The GRTS-based sampling approach is dependent upon compiling and processing GIS layers consistent with sampling decisions. For example, in order to select sites within a subbasin that have habitat available to steelhead and a gradient of less than 12%, a set of GIS data must be compiled that describes these areas (steelhead extent within a basin, barriers, and streams with a gradient less than 12%). ISEMP can assist cooperators in assimilating, managing, and processing GIS data to support sampling design needs as well as assisting groups that may have GIS support but are unfamiliar with this sampling process.
Budget Guidance
ISEMP has developed and administered budgets for numerous private and agency contractors who have conducted habitat monitoring using ISEMP habitat monitoring protocols through funds provided by BPA and is offering to share this experience with collaborating agencies. This direct experience in scoping personnel needs, equipment needs and costs, and other associated costs may be invaluable for BPA contractors who are new to this work. At the same time, the contribution of ISEMP’s experience in fitting the necessary monitoring work within BPA’s contracting requirements would benefit BPA by helping to improve the cost efficiency of this work and ease the contracting process.
GIS Data Management and Geoprocessing
Because regional monitoring programs cover large spatial scales, assimilating and managing spatial datasets in GIS, accounting for the geomorphic context of sampling, and performing watershed or subbasin scale analyses are important features of these programs. Collaborators will have support assimilating spatial datasets across multiple basins, including managing data and metadata within geodatabases and on web-accessible GIS servers. Also supported is the exploration and development of geospatial models, the use of remote sensing technologies to collect continuous GIS datasets, such as LIDAR and aerial photos, and integrating field-based tabular data within a geospatial context. The results and products from these developmental monitoring programs and assistance in implementing promising technologies will also be provided.
Field Sampling Support
Habitat Sampling Protocols
ISEMP has recently conducted a review of fish habitat requirements, interactions of processes that influence fish habitat, the spatial scales for the context of these interactions, and current monitoring programs (ISEMP 2010). This review culminated in the development of a protocol that we believe has the greatest probability of being comparable to other protocols and perhaps most relevant to salmonids. This protocol was designed to be applied across varying spatial contexts depending on the logistical constrains of the sites. In areas where GPS signals can be obtained, along with aerial photos, habitat units within reaches can be superimposed onto aerial photos with a map grade GPS. In situations, where a GPS signal is not obtainable units can be referenced to aerial photos and supplemented with on the ground measurements. Finally, in situations where a GPS signal and aerial photos cannot be obtained a stick and tape method can be employed to delineate the reach into habitat units. These approaches are described in the habitat protocol document (ISEMP 2010). ISEMP will be testing and refining these methods this summer and will provide a revised protocol in the fall of 2010 but no information in this has yet emerged.
Standardized Crew Training: Sampling and Data Capture Tool
Along with ISEMP’s experience in implementing habitat monitoring protocols comes several years of development of field crew training. Trained and experienced field crew leaders would be available to pro-vided coordinated, standardized training in application of the protocol. Such a standardized approach would promote crew efficiency and improve standardization across the region.
In addition to agency specific safety and other training, ISEMP would provide the following training to support cooperating agencies that implement the habitat protocol:
1. Three days of GIS training for cooperators’ GIS and sampling design coordinators to explain how to perform GRTS sample draws and allocate sample sites using a priori filters. By the end of this training, cooperators could expect to have sample draws completed, mapped, and listed in field-ready forms.
2. Three days of in-office training on data loggers, data flow management, and QA/QC.
3. Five days of in-field training for field crew leaders and technicians using the ISEMP protocol. Crews would be trained to measure habitat and record data with data loggers.
4. Five days of “as-needed” support throughout the field season for common questions involving data loggers and sampling methodologies.
5. Five days of post-season assistance with data capture and management issues.
As previously described, this protocol is based upon the statistical evaluation of numerous habitat protocols implemented throughout the Columbia River Basin. The protocol will be expected to evolve as standardized data are collected and analyzed from cooperating programs. Evolution of the protocol will be sup-ported through annual meetings of cooperating agencies and changes to the protocol will be implemented globally, thus enabling the necessary modifications to data capture and storage utilities. It is expected that some level of training effort will be required annually to maintain consistency among collaborators and to assist cooperating agencies in employing updated versions of data capture utilities.
Coordinated Habitat and Fish Sampling
Ultimately one of the goals of habitat monitoring is the identification of fish and habitat relationships, the identification of mechanisms through which habitat quality and quantity and natural processes contribute to freshwater productivity. As such, all habitat monitoring programs should capitalize to the extent possible, on existing fish monitoring activities and/or incorporate a fish monitoring component to supplement habitat monitoring information. In short, using fish vital rates (life stage specific survival and growth, and distribution as a function of habitat quality and availability) might constitute the most transferable and applicable suite of habitat characteristics. Methods and guidance are available from the ISEMP website (http://www.nwfsc.noaa.gov/research/divisions/cbd/mathbio/isemp/documents.cfm). At the very least, we urge cooperators to identify and summarize fish monitoring that coincides with proposed habitat sampling.
QA/QC crews to do repeat sampling across all participating watersheds
Repeated sampling of habitat monitoring sites within the same sampling season has proven to be an important component of GRTS-based, watershed-scale habitat monitoring. Repeat sampling assists with 1) quality assurance/quality control, 2) the assessment of crew variability as a component of variation, and 3) pro-viding improvements to temporal variability recognition (i.e., trend detection). Furthermore, repeat sampling will be important to ISEMP’s research goals of testing the performance of the recommended protocol across the Columbia Basin. To achieve these objectives, ISEMP proposes conducting repeat sampling visits for all watersheds in this program at 10 percent of all sampling sites during the same low-flow index period.
Data Collection Support
Database Schema, Data Dictionary, Meta Data Support and Tool
For a monitoring program at the scale of the Columbia River Basin to be successful, a robust data management system must be in place before initiating data collection. Monitoring habitat in 20 watersheds will generate a massive volume of data. Without properly documented metadata the context that provides meaning to these data would be inaccessible and most likely lost. To insure against that scenario, a system of data processing, storage, analysis, reporting, and distribution to meet the crucial needs of a large-scale monitoring program has been developed. These needs include: (a) documenting monitoring objectives, study design and intended analysis; (b) summarizing how, when, and where the monitoring data were collected, (c) supporting a range of analytical methods, such as hypothesis testing, time series analysis, structural equation modeling, and GIS support; and (d) adapting to changing requirements in the future.
Collaborating agencies will have the use of a data system that includes a centralized data warehouse and web-based data discovery tool (Status Trend and Effectiveness Monitoring, STEM Databank); data ex-change and loading procedures; a database schema that defines data storage format (Aquatic Resources Schema, ARS); metadata tools; data capture, validation, and summary tools (Automated Template Module, ATM); quality control and assurance procedures; and data stewards who support the system.
Field Data Capture Tools: Hand Held Loggers or ATMs and Paper Forms
Field crews need applications to support data capture, review, summarization, and reporting. A handheld data-logger program has been developed for the purpose of capturing data in the field. The current software application is transferable to any data logger device that uses.
CHaMP hypotheses are outlined into a conceptual diagram (Figure 1). The design of the sampling protocols and data structure will help to facilitate testing a large number of hypotheses over the next 10-20 years. The designers of CHaMP strongly believe that these hypotheses will illuminate our understanding of fish/habitat relationships. As these hypotheses are tested and our knowledge base grows OBMEP will slowly incorporate this knowledge into our existing monitoring program.
Windows 6.1 Mobile and has a screen size of 320x240 (support for other screen sizes will follow). The application includes some on-the-fly quality control by enforcing user established minimum and maximum values for each of the individual data entry fields. Drop down values minimize data entry errors and increase the speed of data entry. The program supports USB or Bluetooth GPS to enable quick documentation of sample coordinates and allows data to be immediately integrated into a GIS program providing instant functionality for mapping.
The habitat application runs a data check to identify any potentially missing data, before field crews leave the site. Once the user finishes the data check process, the data are stored in three locations within the logger to ensure multiple backups exist for the data. Users can define the time span for auto-saving the data based upon their preferences.
The data are stored in an XML file format, which is then processed through a conversion script. This process makes the data instantly compatible with the ARS Access database. Uploading data is a simple one-step electronic process, which avoids data entry errors that typically occur when transcribing paper data forms into a database. It also bypasses the data entry process, thus saving time and money.
While field-based handheld computers increase efficiency by reducing data entry effort and eliminating transcription errors, desktop applications have also been developed to perform quality assurance reviews, to derive metrics and summarize data, and to generate reports or submit data for archiving. Additionally, the desk-top applications provide a backup data entry mechanism in the event that data-loggers fail. The field-based and desktop applications have XML-based mechanisms to synchronize data. Thus transiting between data entry using handheld data-loggers and paper forms is not accompanied by a synchronization burden.
The Automated Template Module (ATM) supports the full breadth of the data capture workflow. This workflow includes documenting metadata about project and statistical design, entering survey event information and observations, performing quality assurance procedures, deriving metrics, and submitting data for archiving.
Currently, the ATM supports a wide range of data collection protocols including habitat, macro-invertebrate, water quality, snorkeling, screw trap, and PIT tagging protocols. Customized data entry forms can be generated to support the proposed habitat protocol, or any of the other protocols supported by the ATM. These forms would be set within the broader ATM tool and provide a solution that supports the full breadth of the data capture workflow. Data entry forms generated through the ATM are fully integrated within the previously described data system.
Figure 1: Conceptual diagram of hypotheses to be tested under the CHaMP project.
Help
Mitigation work
Reporting
Funding
Search
Log in