- ID: 83
- State: Draft
- Owner: Carol Volk
- Collaborator(s): Andrew Hill, Andrew Murdoch, Arielle Gervasi, Arielle Gervasi, Boyd Bouwes, Braden Lott, Casey Justice, Chris James, Chris Jordan, Chris Moan, Christopher A. Beasley, Erin Morgan, Jacque Schei, James White, Jean Olson, Jesse Langdon, Keith van den Broek, Kristina Mc Nyset, Laurel Faurot, Matt Nahorniak, Nick Bouwes, Nick Weber, PAMELA NELLE, Seth White, Steve Bennett, Steve Fortney, Steve Rentmeester, Ted Sedell
- Spatial Design Category: Generalized Random Tessellation Stratified (GRTS)
- Sites in Design: No sites scheduled
- Has Location Privacy: No
- Data Repository: <none>
- Created by: Carol Volk
- Created: 1/30/2014
- Updated by: Carol Volk
- Updated: 4/30/2014
- Version History: v1.0 Draft (1/30/2014)
The details of this Sample Design, including all the parameters used to generate it, are included below. Sample designs must belong to a Study Plan.
Description
CHaMP is designed as a Columbia River basin-wide habitat status and trends monitoring program built around a single protocol with a programmatic approach to data collection and management (RM&E Workgroup 2010). CHaMP will result in the collection and analysis of systematic habitat status and trends information that will be used to assess basin-wide habitat conditions. When coupled with biological response indicators, this status and trends information will be used to evaluate habitat management strategies. This program will be integrated with ongoing Pacific Northwest Aquatic Monitoring Program (PNAMP) recovery planning efforts and will be part of the collaborative process across Columbia Basin fish management agencies and tribes and other state and federal agencies that are monitoring anadromous salmonids and/or their habitat. The implementation of CHaMP will characterize stream responses to watershed restoration and/or management actions in at least one population within each steelhead and spring Chinook Major Population Group (MPG) which have, or will have, “fish-in” and “fish-out” monitoring (identified in RPA 50.6), thereby meeting the requirements of RPA 56.3, RPA 57, and RPA 3. CHaMP was designed to deliver trends in habitat indicators and requires that monitoring occurs for three cycles of a sampling panel (see section 1.6), at least 9 years.
The target populations are the Upper Grande Ronde Chinook, Catherine Creek Chinook, and Upper Grande Ronde Steelhead. Three study designs were generated to meet the sampling needs of each population.
2011 Design Notes:
ChaMP’s Grande Ronde habitat monitoring program covers three distinct spatial domains, coordinating CHaMP, ODFW, and CRITFC habitat monitoring goals and interests. CRITFC coordinated the sampling for the Upper Grande Ronde’s and Catherine Creek Chinook spawning and rearing habitat. The sampling domain for ODFW (also CHaMP) covers the steelhead spawning and rearing domain, part of which overlaps the Chinook spawning area in both Upper Grande Ronde and Catherine Creek.
2011 Design summary:
Survey designs were created for each of the three major domains: Upper Grande Ronde and Catherine Creek chinook spawning and rearing domains and steelhead alone domain (i.e., the portion of the Grande Ronde steelhead domain that did not overlap with the Upper Grande Ronde and Catherine Creek domains).
Design Strata:
Upper Grande Ronde Steelhead Domain: Valley Class (Source/Transport/Depositional) + Ownership (public/private)
Upper Grande Ronde Chinook and Catherine Creek: No stratification
Legacy sites:
Upper Grande Ronde and Catherine Creek Designs: CRTFC surveyed the Upper Grande Ronde during 2010 using a GRTS design and a frame that CRTFC developed. These legacy sites were transferred to the 1:100 scale network used to develop the CBW master sample. Fifteen sites were targeted each year in each domain, 10 annual and 5 each year on a three year rotation. As we discussed the Upper Grande Ronde portion of the study, the target frame was reduced significantly because there were substantial private lands preventing sampling access.
Steelhead only domain: The target steelhead only domain was extended to cover 1-6th order wadeable, perennial streams. The standard CHaMP rotating panel design was adopted (15 annual sites, 10 rotating panel sites on a three year cycle).
Detailed Design Notes File: CRITFC.ODFW design goals.docx contains the discussion about developing both 2011 and 2012 designs.
2012 Design Notes:
1. The target population was expanded to include additional Chinook spawning areas in Beaver Creek, a tributary to the Upper Grande Ronde River. This stream is known to provide habitat for spawning and rearing Chinook salmon, but was excluded from the original sample frame due to access issues with the landowner. Access remains tentative, but it would be beneficial to survey sites in Beaver Creek on an opportunistic basis.
2. The target population will be expanded to include non-natal rearing habitat for Chinook, including portions of the mainstem upper Grande Ronde River between Beaver Creek and Five Points Creek and portions of various tributaries.
3. CHaMP’s panel sample sizes changed from 10 annual sites to 5; rotating panels 2 and 3 increased to 10 sites each.
4. ODFW has conducted steelhead spawning surveys; beginning in 2008, using a GRTS based rotating panel structure. This survey was not incorporated into the 2011 CHaMP design. CHaMP modified the selection of rotating panel 2 and 3 to incorporate legacy steelhead spawning sites. Details of the integration of the steelhead surveys into CHaMP’s design are covered in the site selection section.
2012 Design changes
Upper Grande Ronde Chinook:
Brief recap of the redesign goal for 2012:
· Expand the UGR chinook sampling domain
· Reduce original annual panel from 10 to 5 sites; assign the released annual sites to beef up rotating panels 2 and 3 to 10 sites each and distribute these 10 sites across the original UGR chinook domain and the expansion domain. Assign the 5 dropped annual sites to rotating panel 1.
Design layout treats the original 2011 and expanded domains as separate strata.
The allocation of number of RP 2 and 3 sites followed the desire to allocate the 10 sites in each panel without regard to stratification even though we found it necessary to stratify. We chose an allocation of 3 per panel in the original domain and 7 in the expansion domain. This follows the relative extent in each domain with one twist. Beaver Cr. was considered part of the original domain, but sites were deemed “access denied” so was excluded from the design. However, there is interest in having some sites fall in Beaver Cr. We incorporated Beaver Cr. into the expansion domain. There are 34 master sample sites in the original domain, 10 in Beaver Cr., and 61 in the expansion domain; total = 105 sites. The original domain is about 1/3 and the expansion + Beaver Cr about 2/3. The allocation we chose approximates this ratio.
Original domain: Keep the original design intact. For RP2 and RP3, select the first 3 sites in the original 2011 RP2 and RP3 set; assign the remainder to the ‘over’ or ‘extra’ pool. As indicated above, assign the five ‘released’ annual sites to RP1 pool.
Expanded domain: We rerandomized the expanded domain plus Beaver Cr. sites (71 sites total) to generate a spatially ordered balance within this stratum. Blocks of sites (16 sites per block) were examined for spatial balance with respect to both panels and ‘fit’ to the ODFW legacy sites (sampled in 2011 or planned for 2012 and 2013). The second block (site #17 in the ordered list) was selected for sample allocation, and sites were alternately assigned to RP2 and RP3. No legacy sites were incorporated into the selection in the expansion domain. It is possible that some sites will fall “too close” to legacy sites (1 km criterion?), or that Beaver Cr. sites will be access denied. Replace denied sites with the next on the list.
Strata updates for 2013: There were no changes to the Steelhead nor Chinook sampling designs in 2013. However, sampling by the Confederated Tribes of the Umatilla added several sites to cm.org to make sampling possible.
Strata updates for 2012: Expanded UGR domain treated as an additional stratum called Upper Grande Ronde Chinook Expansion.
GRTS Master Sample Input file: UGRr_chinook_domain_MS_legacy_coded_20120403.csv
R code: UGR.r, used strat_panel_function10.2.r
Output file: ugr_new_1_20120605.csv
Catherine Cr:
Design updates for 2012: Reduce original annual panel from 10 to 5 sites; assign the released annual sites to supplement rotating panels 2 and 3 to 10 sites each. Assign the 5 dropped annual sites to rotating panel 1.
Annual panel: 5 sites carried over from 2011; all were legacy sites sampled by CRITFC in 2010
Rotating panel 1: 12 sites coming from 5 “left over” annual sites and 7 rotating panel 1 sites
Rotating panel 2: 10 sites, selected from rerandomized list
Rotating panel 3: 10 sites, selected from rerandomized list.
GRTS master sample input file: Catherine_Cr_chinook_domain_MS_legacy_coded_20120403.csv
R code: Catherine.Creek.r; used strat_panel_function10.2.r
GRTS design output file: Cath_Cr_redesign_20120507.version2.xlsx
Steelhead only domain:
Candidate site lists were generated using Generalized Random Tesselation Stratification (GRTS) for two surveys in the Upper Grande Ronde River (UGRR) subbasin: steelhead spawning ground surveys (ST SGS) in 2008 and the Columbia Habitat Monitoring Program (CHaMP) in 2011. Both surveys are conducted by Oregon Dept. of Fish & Wildlife (ODFW) staff in LaGrande, OR. ODFW staff decided that sites from both studies should be melded into one sample set to maximize the number of locations with both ST SGS and CHaMP data. Ultimately the goal is to compare habitat conditions with steelhead use, both as adults and juveniles (another ODFW study currently underway).
ST SGS sites were broken into three panels designed to be used on a three year rotation. However, they were not stratified. Instead, they were given a use order during the initial GRTS draw in 2008, and sites were chosen in that order based on landowner permissions and other logistic constraints. A total of 30 sites per year were visited: 10 annual sites (selected from the first panel), 10 rotating panel sites (3 year rotation), and 10 “new” sites which were visited only once.
CHaMP’s rotating panel 2 and 3 designs were changed by incorporating the legacy ST SGS sites. A total of 20 ST SGS sites were transferred to CHaMP panels 2 and 3 to achieve the original ‘block’ (valley class by public/private) sample sizes as feasible. ST SGS annual sites were preferred over ST SGS panel sites for movement to CHaMP (because they have more fish data), and the supply of ST SGS annual sites was exhausted before the panel sites were chosen for CHaMP. ST SGS sites were chosen randomly if more sites than needed fell into a block. Those not randomly selected were designated as oversamples and randomly assigned a use order. A detailed description of the integration of the CHaMP and ST SGS
design file: ODFW Grande Ronde steelhead spawning with CHaMP.docx.
2013 Design Updates
There were minimal changes to the Grande Ronde design in 2013. Minor changes included the addition of a stratum to add 3 sites for sampling by the Confederated Tribes of the Umatilla Indian Reservation (CTUIR) and a shift of site CBW-360346 to the annual panel to also facilitate sampling by CTUIR.
Design Documentation Files:
No new design files were generated for 2013--changes to accommodate CTUIR sampling were made directly in champmonitoring design documentation.
2012 redesign:
Sample Design Parameters
- Has user sites
- 5 Panels
- Is stratified
Start Year
2011
Initiation Year
2013
Retirement Year
Study Plan
<none>
Data Repositories
<none>
Photos
<none>
Documents
<none>
Map of Sites
- Stratum
- Panel
- Occasion
Area of Inference
<none>
AOI Notes
<none>