The Coded-Wire Tag (CWT) Recovery Project is an on-going data collection and data management program conducted by ODFW, WDFW, and PSMFC that supports a coastwide stock identification system for coded-wire tagged salmonid fish.  Within the Columbia Basin, the CWT is used extensively for identification of hatchery and wild anadromous salmonid stocks.  In particular, the tag recovery data are used to monitor the status of both threatened and endangered stocks.  In addition, the recovery data are used to assess a wide variety of studies designed to improve survival of hatchery-produced salmonids.  CWT recovery information also provides critical data for evaluating stock rebuilding programs sponsored by NWPCC's Fish and Wildlife Program.

Oregon and Washington carry out a coordinated sampling effort to collect CWTs from mature salmon and steelhead, which return to fishery (sport and commercial) and escapement areas (natural spawning grounds, hatcheries, and Bonneville Dam fishways) throughout the Columbia River Basin.  ODFW also samples the ocean commercial and recreational fisheries along the entire Oregon coast to recover CWTs from chinook and coho salmon.  Sampled heads of tagged fish are transported to tag recovery labs at Clackamas and Olympia where the CWTs are recovered and decoded.  The CWT recovery and catch/sample information is then forwarded to PSMFC's Regional Mark Processing Center where it is validated and made available to users via the on-line ‘Regional Mark Information System' (RMIS).

Fish managers, researchers, mitigation agencies and others use the CWT release and recovery data to evaluate a number of administrative, management and environmental effects on salmon and steelhead.  For example, the harvest management agencies combine CWT data with other data and information to estimate the effects of harvest regulation on populations of salmon and steelhead.  Others use CWT data to estimate the rates of escapement into the wild of a population of hatchery fish.  Others, including BPA, use CWT data to determine survival of different hatchery operations and the effectiveness of the hatchery programs they fund.  Others use CWT data to determine the effectiveness of specific hatchery or other management actions.

This CWT marking and recovery program is consistent with the Fish and Wildlife Program goals for monitoring and evaluation.  In addition to monitoring the status of both threatened and endangered stocks, CWT recovery data are used to assess a wide variety of studies designed to improve survival of hatchery-produced salmonids.  CWT recovery data also provide critical information for evaluating stock rebuilding programs under measures now sponsored by the 2000 FCRPS Biological Opinion.

I.  Overview of CWT Recovery Program Components

The large-scale CWT recovery programs of Oregon and Washington provide an essential service for all of the various State, Federal, and Tribal tagging programs in the Columbia Basin.  The goal of ODFW, WDFW, and USFWS's stock assessment tagging programs, for example, is to tag a statistically valid number of coho and chinook salmon from each hatchery such that survival rates plus ocean and freshwater distribution can be accurately documented.  These data allow for more accurate assessments of the proportion of wild and hatchery stocks in the Basin and further allow for valid statistical comparisons to be made among project groups.

Upon return to the fisheries, hatcheries, or spawning grounds, mature salmon and steelhead are typically sampled for CWT recovery purposes.  Fall chinook and steelhead are sampled visually to detect missing fins, which denote the presence of CWTs.  Conversely, the use of an electronic detector is required for spring and, summer chinook, and coho due to mass marking programs in effect for these stocks.  Biological data are collected in conjunction with the snout recovery.  Once the snouts are collected, they are transported to the Tag Recovery Lab where the CWTs are removed and then decoded.  The tag recovery data for all the individual tags recovered are recorded into the respective sampling agency's fisheries database where the data are collated and expanded based on the sampling rate of the various collection sites.  The recovery data are then sent to PSMFC for entry into the Regional Mark Information Center database (RMIS) for user access via the Internet

II.  Columbia Basin CWT Sampling Program

Oregon and Washington jointly share the task of sampling the Columbia River sport and commercial fisheries for CWT recovery purposes.  Sport and commercial fisheries target salmon and steelhead stocks throughout the lower 395 miles of the Columbia River stretching from the mouth at Buoy 10 upstream to the Priest Rapids Dam.  The primary mainstem sport fisheries occur below Bonneville Dam (including Buoy 10), at tributary mouths above Bonneville Dam, and at Hanford Reach on the upper Columbia River.  Tributary sport fisheries primarily occur from The Dalles Dam downstream.  Treaty Indian commercial fisheries operate between Bonneville and McNary dams and non-Indian commercial fisheries are limited to areas downstream of Bonneville Dam.  Additional sampling occurs for salmon and steelhead passing Bonneville Dam and returning to hatcheries and natural escapement areas.  Sampling in the mainstem Columbia River focuses primarily on stock compositions while sampling in tributaries focuses primarily on hatchery/wild compositions.

Salmonids encountered are examined for the presence of a CWT.  Fish containing a CWT will have their snout removed and be sampled for biological data.  Biological data will vary from project to project and may include length, weight, sex, skin color, other marks, and a scale sample.  Random sampling techniques are used to ensure that unbiased data are collected for the purpose of developing stock compositions for fish landed in fisheries and returning to escapement areas.  Sampling rates will vary depending on sampling location.  For Columbia River mainstem fisheries the goal is to sample a minimum 20% of the landed catch for the purpose of recovering CWTs.  The 20% sampling goal was adopted to ensure that CWT recoveries will be adequate for developing stock compositions for individual mixed stock fisheries and that stocks of low abundance will be adequately represented in resulting stock compositions.  In contrast, tributary fisheries are primarily single stock fisheries; therefore, the goal is to achieve a minimum 5% sample rate in tributary sport fisheries.  A 5% sampling goal is also in place for tributary escapement areas, including hatcheries and natural spawning locations.  The 5% sampling goal provides the data necessary to develop accurate catch estimates plus hatchery/ wild, stock, and age compositions.

1.  Columbia River Commercial Fisheries

Non-Indian commercial fisheries occur in the lower 140 miles of the Columbia River from the mouth at Buoy 10 upstream to Bonneville Dam, while treaty Indian fisheries occur in the 140 miles of the Columbia River between Bonneville and McNary Dams.  Columbia River non-Indian and Treaty Indian commercial salmon and steelhead fisheries may occur during February through October with the majority of the landings occurring during the mid-August through October time frame.  Seasons are set during the year based on expected run strength of various salmon and steelhead stocks.  

In recent years, the ESA has imposed severe constraints on mainstem non-Indian commercial fisheries and has greatly increased the need for precise stock accounting in fisheries. The advance of mass marking spring chinook has provided additional fishing opportunity in the spring, primarily during April and early May.  The BPA funded Select Area Fishery Enhancement (SAFE) Project has increased the time and area in which Columbia River non-Indian commercial fisheries occur in select areas.  Select Area fisheries occur from late February through October with the majority of the fisheries occurring during late April though early June and early September through October.

Salmonids landed in these commercial fisheries are generally sold to commercial fish buyers with the exception that some fish are sold directly to the general public.  Sampling of fish occurs only at commercial fish buying stations or processing plants because sampling of fish sold directly to the general public is unfeasible at this time.  Salmon and steelhead landed in commercial fisheries are sampled at fish processing plants and buying locations upon delivery of the catch to the commercial fish buyer.  

Sampling rates typically exceed 20% in most commercial fisheries because sampling occurs at fish buying stations or processing plants where large numbers of fish are delivered in a relatively short amount of time.  In recent years, increased fishing opportunities, in combination with funding restrictions, have impacted the agencies ability to maintain a 20% sampling rate for commercial fisheries sampled as part of this project.  In some cases the limited number of buyers, in conjunction with need to randomly sample the catch, results in sample rates that are significantly higher than 20%.

2.  Columbia River Sport Fisheries

The sport fishery located near the Columbia mouth is known as the Buoy 10 fishery and occurs during early August through mid-October.  Nearly all of the Buoy 10 catch is fall chinook and coho with a few steelhead being landed.  The fishery has been sampled since its resurgence in 1982.  Effort and catch is estimated on a weekly basis but is not part of the statistical creel program for the lower Columbia River sport fishery.  Effort is indexed by on ground trailer and rod counts at popular launch sites and bank angling locations.  Anglers are queried for success at boat ramps and bank fishing locations, but no on-water sampling occurs. In Washington, the Buoy 10 fishery is sampled using non-BPA funds.

The sport fishery on the lower Columbia River (below Bonneville Dam) occurs year round and targets different species or races throughout the year.  The fishery targets spring chinook during mid-February through mid-May, summer chinook during mid-May through July, fall chinook and coho during August through October, summer steelhead during mid-May through September, and winter steelhead during December through February.  In recent years, implementation of mass marking programs for spring and summer chinook and the adoption of selective fisheries has provided additional fishing opportunity during the April - July time frame.

Sport anglers encountered on the water, at bank fishing locations and at boat ramps or moorages will be queried regarding success in catching fish.  Boat and bank effort will be estimated by aerial ‘fly over' counts conducted over the lower Columbia River twice a week during February through October.  These data will be used as part of a statistical creel program that will estimate monthly effort and catch for lower Columbia River salmonid fisheries.  This fishery has been sampled as part of a statistical creel program since 1969.

Sampling of the mainstem sport fishery occurs at popular bank fishing and boat launch locations throughout the lower 146 miles of the Columbia River.  Sampling rates in the lower Columbia River and Buoy 10 sport fisheries have generally ranged between 15%-30% in recent years.  Achieving the 20% sampling goal on the lower Columbia River is a difficult task because the fishery extends over the 146 miles of the lower river and fish are landed throughout the day.

There are also some localized fisheries occurring in the Columbia River between Bonneville and McNary dams just below mainstem dams and at tributary river mouths.  Limited creel sampling of the salmonid sport fisheries in the mainstem Columbia River between Bonneville and McNary dams began in 1994.  Increased spring chinook returns in recent years have significantly increased fishing opportunities in this section of the Columbia River, primarily during the March through May time frame.  Ancillary sampling occurs in mainstem sport fisheries from Bonneville to McNary Dam and sampling rates are generally less than 20%.

The Hanford Reach fall chinook fishery occurs from mid-August through October.  Anglers are interviewed at boat ramps or bank fishing locations.  Trailer counts are conducted to estimate total effort.  Angler success data are applied to total effort estimates to estimate total catch.  The sampling rate for this fishery typically exceeds 20%.

Washington tributary spring chinook fisheries typically occur between April and June, and fall salmon fisheries primarily occur in September and October.  The fisheries occur on lower Columbia and Bonneville Pool tributaries and landed catch is sampled for the purpose of recovering CWTs.  Anglers are queried for success at boat ramps and bank fishing locations.  Sample rates generally do not exceed 10% except in the largest fisheries.  Bonneville Pool tributaries spring chinook fisheries are managed jointly between WDFW and Yakama Nation (YN) to meet hatchery escapement goals in addition to harvest sharing.  Creel programs exist in Oregon tributaries to the Columbia River but sampling efforts are not funded by this project.

3.  Selective Fisheries Sampling

Beginning in 1998, the adipose clip on hatchery coho was reassigned to be a mass mark for identifying hatchery fish.  As such, the majority of the coho now returning to the Columbia River are adipose marked without a CWT being applied; therefore, electronic equipment is required to detect the presence of a CWT. This situation greatly reduces the efficiency of the CWT sampling process and additional samplers are needed to maintain adequate sampling rates in fisheries and at escapement areas.  

Since 1998, sources for funding additional Oregon field sampling staff have been obtained via the Sport Fish Restoration Program (SFR).  Oregon State funds were also secured via the ODFW Fish Restoration Act to purchase electronic CWT detection wands.    Oregon has identified additional SFR and other state funding sources to match with available BPA and other federal funding to sample and monitor selective coho fisheries in both Columbia River and ocean locations.  Washington has purchased electronic CWT detection equipment and provides state funds to assist in the recovery of CWT's from mass marked coho salmon.

Beginning in 2001, the majority of the hatchery produced spring chinook returning to the Columbia River were mass marked with the adipose fin clip.  Selective sport and commercial fisheries targeting these mass marked spring chinook in the lower Columbia River occur primarily during March and April, with some fisheries occurring during the first half of May.  Selective spring chinook fisheries also occur in the mainstem Columbia River between Bonneville and McNary dams during March through May.  Beginning in 2002 selective sport fisheries were also adopted for summer chinook.  Summer chinook fisheries can occur during the mid-May through July time frame but have generally occurred during late June and July.  Level funding of the CWT program in recent years has impacted the states ability to sample these selective spring and summer chinook fisheries.   Additionally, funds typically used to sample other fisheries must be shifted to sample these newly emerging fisheries, thereby reducing the states ability maintain a 20% sampling rate in all mainstem fisheries.

4.  Columbia River Escapement Areas

Spring chinook, fall chinook, and coho are sampled at several Columbia River Basin hatcheries located downstream of Priest Rapids Dam during August through January.  Salmonids returning to hatcheries are sampled for the purposes of recovering CWT's and collecting biological data.  CWT recovery sampling at Washington and Oregon salmonid hatcheries is supported by a variety of funding sources with BPA funds from the CWT Recovery project providing a significant portion of the funding in Washington and only a minor portion of the funding in Oregon.  The majority of the funds used to hatchery returns are provided by non-BPA sources.  Washington hatcheries are sampled on most or all spawning days and the sample rate at hatcheries typically is 100% except for some fish that are passed upstream above the hatchery facility.  Most Oregon facilities are sample at the 100% rate for CWT recovery purposes.

Spring chinook, fall chinook, and coho returning to the Columbia River mainstem and tributaries from the mouth of the Columbia River upstream to Priest Rapids Dam during August through January are sampled for the purpose of recovering CWT's.  Peak counts (redd or live and dead fish) are used to estimate the total natural spawning populations.  In addition to sampling for CWTs and biological data, fish are separated according to stock based on skin color and/or external marks.  Fish counts are further divided into adults and jacks and by hatchery or wild origin.  CWT recovery sampling rates are typically less than 10% in most streams sampled.  Spawning ground surveys conducted on the Cowlitz and Lewis rivers are primarily funded by the local power companies.  Funding from this project supports only limited sampling in Oregon tributaries of the lower Columbia River.

Sampling also occurs at Bonneville Dam to determine stock composition of fall chinook and summer steelhead passing Bonneville Dam.  The steelhead data allows for the identification of Group A and Group B stocks that are necessary for stock status monitoring purposes in the Snake River Basin.  The goal of this project is to sample 1% of the total summer steelhead passing over Bonneville Dam for stock separation purposes.  Stock separation of fall chinook also occurs at Bonneville Dam.  Fish passing Bonneville Dam are identified as bright and tule stock based on skin color and these data are necessary for stock status monitoring purposes.

5.  Data Summarization and Analysis

CWT recovery data, fishery catch estimates, and escapement estimates in combination will be summarized and analyzed for the purpose of determining stock status of wild and natural salmon and steelhead populations in the Columbia River basin.  Data collected during the fishery sampling portion of this project will be used to estimate catch in commercial and sport fisheries and returns to escapement areas by species and/or stock.  These catch and return estimates plus catch and effort sampling data will be included in the PSMFC's CWT recovery database.  All fishery and escapement data will be delivered to PSMFC's Regional Mark Recovery Processing Center for merging into the RMIS database.  CWT recovery data will be used with total catch and escapement estimates to produce stock compositions for each fishery and escapement area and to estimate total returns by tag code.  The return by tag code estimates provide the basis for annually estimating the return to the Columbia River mouth for all major salmonid stocks, including stocks listed under the Endangered Species Act (ESA).

Annual abundance estimates plus fishery stock and age composition data are essential for monitoring the status of wild/natural and hatchery produced salmonid stocks in the Columbia River basin.  These data will be incorporated into run reconstruction databases and will be provided to the U.S. v. Oregon Technical Advisory Committee (TAC) for the purposes of tracking stock status of all major salmonid stocks returning to the Columbia River, including ESA-listed stocks.  The data will be made available to state agencies for use in developing Biological Assessments concerning proposed fisheries in the Columbia River basin and are commonly used by the NOAA Fisheries when completing Biological Opinions regarding fisheries in the Columbia River basin.  Forecasting and run reconstruction methodology is reviewed annually by the TAC and modified as needed.  Forecasts are provided to managers for use in planning fisheries for the upcoming year to protect depressed or listed salmonid stocks.

III.  Oregon Ocean Fisheries CWT Sampling

Oregon's ocean commercial troll and recreational fisheries target a multitude of regional and West Coast chinook and coho salmon stocks along the approximately 310 miles of the Oregon Coast in both state and federal offshore waters.  The evaluation of Columbia River salmonid stocks through BPA-funded CWT sampling is an essential component for determining stock composition, distribution, and survival characteristics of these important stocks.  

Recent inclusion of several Columbia River system stocks under the federal ESA and other critical stocks have increased the need for information for evaluating impacts in regional fisheries.  It is also needed to provide life history information to evaluate stock rebuilding strategies and management alternatives.

The ODFW Marine Resources Program (MRP) implements ocean salmon fisheries sampling and CWT collection program in close consultation with the Department's Inter-jurisdictional Fisheries Program in Portland.

This statistically-based and coordinated sampling project has been functioning since 1979 to collect critical data elements and CWTs from the ocean commercial troll and recreational ocean fisheries.  Project objectives are to:  (1) implement non-biased representative sampling at a minimum rate of 20% of landings by week, catch area (troll) and port (recreational), and species strata; (2) provide necessary CWT sampling and recovery data to evaluate stock contribution and distribution characteristics in Oregon's ocean fisheries; (3) provide information for evaluating stock survival rates; and (4) deliver collected data into PSMFC's RMIS database and make it available for regional and international salmon management forums to implement management strategies that meet harvest impact criteria for Columbia River basin stocks.

Seasonal port samplers are hired to collect CWT, effort, catch, and other biological data at coastal ports.  Salmon observed by samplers to have a fin clip or "mark" are electronically tested for the presence of a CWT and if the test is positive, have their snouts removed for later CWT extraction and decoding.

Funding provided by BPA represents only part of the overall federal/state support necessary to initiate and operate Oregon's yearly ocean salmonid CWT sampling program.  BPA supported about one third of the total OSM ocean sampling costs in 2004.

1.  Ocean Commercial Troll Fishery

Oregon's ocean commercial troll fishery has changed from historically targeting both coho and chinook to a primarily directed chinook fishery since the early 1990s.  Critical wild salmonid stock management and rebuilding needs for such stocks as Oregon's coastal wild coho have precipitated this change.  

Although ocean troll chinook regulations vary significantly by coastal area and year, with several ocean areas closed for part or all of the season, the 2004 ocean season was generally opened for most of the Oregon coast from mid-March through October.  Portions of the general chinook season between June and August have experienced rotating block closures over the past several years, to limit hooking mortality impacts on Oregon Coastal Natural (OCN) coho and keep chinook harvest on stocks such as Snake River fall chinook within baseline rates.  Additional but limited "late season" state water ocean troll fisheries take place during October, November, and December to harvest specific local chinook stocks.

Traditionally, the majority of ocean-caught chinook salmon were harvested in August.  In recent years, the harvest is much more evenly spread throughout the entire season.  Columbia River stocks are distributed over a wide time and area during the season.  In 2004, troll chinook were landed at about 90 traditional buying locations, primarily in Oregon's 12 major coastal ports.  

Oregon's ocean salmon fisheries are established by the PFMC and the state of Oregon in April (March openings are considered an in-season action under the prior year's regulations) each year.  Seasons are established on the basis of several factors including regional species (chinook and coho) stock status.  Columbia River basin stocks are important in setting these yearly harvest strategies as they include ESA-listed and other "critically" managed Columbia River chinook and coho populations.

2.  Ocean Recreational Fishery

Oregon's marine recreational salmon fishery has operated primarily for chinook salmon for the majority of the March through October seasons since the mid 1990's, due to catch quota restrictions on coho.  The recreational chinook fishery along most of the Oregon Coast opened in mid-March in 2004. Beginning in 1998 selective ocean fisheries for fin-clipped hatchery coho were adopted in ocean waters adjacent to the Columbia River, and since 1999 have also occurred off the central Oregon Coast.  

These coho seasons have been limited to the late June through September period off the Columbia, and from late June through August on the central Oregon Coast. These fisheries are monitored under specific operational salmon plans for impacts on wild stocks.  Primary tasks include tracking total fishing effort, dockside sampling , and at-sea observations.  Columbia River coho make up the vast majority of the harvested salmon in these hatchery only selective coho seasons.

Oregon's seasonal ocean recreational chinook catch is spread over eight months from March through October, with some exceptions for late season directed local fall chinook seasons.  Most of the chinook catch occurs between May and September.  

Generally, most of the chinook caught from the central coast to the Oregon / California border originated from California's Sacramento and San Joaquin basins, while chinook landed in Tillamook and Astoria have a higher proportion of Columbia basin chinook stocks.  In the past couple of seasons, recovered CWTs have indicated that Columbia basin chinook are making up a larger share of the landings throughout the Oregon Coast.        

IV.  CWT Extraction and Data Management/Analysis

Snouts are delivered to tag recovery labs in Clackamas or Olympia where the CWT is extracted and decoded.  The resulting tag code is entered and verified on a mainframe computer.  Associated fishery/recovery and biological data, collected when snouts are recovered, are uploaded to the mainframe computer and merged with previously entered CWT recovery data.  Based on program specific sampling rates, individual tag recoveries are increased by an expansion factor to estimate the total number of that particular tag present in a given fishery, hatchery, or natural escapement area.  

CWT recovery data are summarized to estimate the number of CWTs recovered for each tag code for each sampling program.  Throughout this process, the data are diligently error checked and errors corrected to ensure quality data.  The CWT recovery data are then forwarded to PSMFC and to managers for making in season fishery management decisions.

Summarized CWT data recoveries, fishery catch estimates, and estimated escapements for most Columbia River salmonid stocks are provided by several state and federal agencies for additional data analysis.  Data analysis includes run reconstruction of all major salmonid stocks.  Total returns are categorized by age and stock.  Included in total returns are fishery catches, escapement estimates for both hatchery and natural spawn fish, and dam counts.  Preseason run size forecasts are also developed annually.  Data are provided to the U.S. v. Oregon TAC on status of ESA listed stocks and is summarized annually in technical reports.  Annual stock assessment reports are produced and distributed to fish resource agencies throughout the basin.  All resultant databases are updated annually and are used in a variety of management forums.

The ocean recreational fishery is sampled at most major coastal ports including multiple charter boat business locations, and private boat fisherman at moorages, marinas, and launch ramp sites.  The evaluation of this angler and trip effort, expanded landed catch estimates by time and catch area, and CWT sampling are collectively used to evaluate Columbia River basin stock representation in both Oregon and regional fisheries for establishing appropriate management strategies.  These CWT data provide wider information for a variety of users through PSMFC's RMIS system.

V.  Regional Mark Processing Center's Role in Data Management

Once the CWTs are decoded and processed by ODFW and WDFW's tag recovery labs, the recovery and associated catch/sample data are reported to PSMFC's Regional Mark Processing Center.  The data are then subjected to another battery of error checks.  Upon validation, the recoveries are combined with the coastwide recoveries reported by other agencies.  Data users may then query the on-line ‘Regional Mark Information System' (RMIS) to obtain tag recovery data (reports or raw records) for research and harvest management analysis applications.

RMIS provides on-line access to all coastwide CWT data, including that for the Columbia Basin tagging studies.  Data sets include: 1) Release; 2) Recoveries; 3) Catch/Sample; 4) Location codes and 5) Data Descriptions.  The Mark Center also serves as the site for exchanging U.S. CWT data with Canada for Pacific Salmon Treaty purposes.  The CWT data can be accessed on PSMFC's computer via the following methods:

Telnet: telenet.psmfc.org
FTP: ftp.psmfc.org
WWW: http://www.rmis.org    or    http://www.psmfc.org/rmpc

Given the far ranging migratory behavior of salmon and steelhead trout, the benefits of PSMFC's central data repository for coastwide CWT release, recovery, and associated catch/sample data are self evident.  Prior to the establishment of the RMPC, harvest managers and researchers had to individually contact all agencies to gather tag recovery data.

The establishment of the RMPC's on-line "Regional Mark Information System" (RMIS) has also greatly enhanced timely access to CWT release and recovery data.  Data users can now specify either tag codes or specific location areas and download either standard recovery reports or individual data records in either "raw" or aggregated form.  

The RMPC also provides an essential focal point for the coastwide coordination of the usage of CWTs and other fin marks.  Regional agreements on marking and tagging are reached through consensus of the "Mark Committee" during the annual "Mark Meeting" each April.

The Mark Center's operations are consistent with purposes of the Anadromous Fish Conservation Act as the regional CWT database is supported by all marking agencies.  The cooperative reporting of CWT release and recovery data by all involved agencies, in turn, provides universal access to the data for all agencies.

The role of the Mark Center is unique on the west coast.  The regional CWT database is not duplicated in function or its entirety by other agencies.  As such, the operations of the RMPC are not in duplication of other projects funded by anadromous resources such as the Pacific Salmon Treaty, Anadromous (NOAA), Mitchell Act (NOAA), LSRCP, USFWS, or state funding.