E. Eric Knudsen, Ph.D.
Alaska Biological Science Center
U.S. Geological Survey
Anchorage, Alaska
September, 1996
Estimation of age composition for wild rainbow trout stocks in Southwest Alaska is a fundamental part of routine fish stock assessments studies. The Alaska Board of Fisheries' management plan for southwest Alaska rainbow trout (ADFG 1990) calls for maintenance of "historical size and age composition". Significant age information presently exists for wild rainbow trout in southwest Alaska (Faustini 1990; Minard and Dunaway 1991; Riffe 1994; and Lisac 1996) and fisheries staff, both state and federal, continue to add to this information base. Scales are the primary structure used for determining rainbow trout ages. They are preferred to otoliths because they are cheaper to collect and process and do not result in lethal collections as do otoliths.
Between- and within-reader variability associated with aging rainbow trout was reported by Coggins (1994) and a protocol for collection of scales, training readers, and processing age information will be offered by Fair (In Prep.). There remains, however, the question of whether scales and otoliths approximate the true age of southwest Alaska rainbow trout. Irving and Faustini (1994) reported differences between age compositions estimated from scales and otoliths. Because there are concerns that older fish may not be well-represented via routine scale analyses, an improved method for consistently estimating true age composition, that is both non-lethal and economic, is needed.
Techniques described by Ogle et al.(1994), or some modification of the same, may be applicable to rainbow trout. This temporal signature technique uses an individual's environmental growth history recorded in the scales to compare to the growth history from a master set of "known" aged fish. The premise of this technique is that the growth history recorded during the early years of a rainbow trout's life, a time when they can be reliably aged, can be used to classify older aged fish that show incomplete or indistinct growth history at the margins of their scales.
This study will determine whether the temporal signature technique will work on a population of rainbow trout from Southwestern Alaska. The study will indicate whether rainbow trout from a selected population can be more accurately aged by the use of this technique. If these hypotheses prove true, this technique can be explored on other populations needing accurate age structure to make management decisions.
1. Develop a master aging chronology for Kvichak River rainbow trout.
2. Estimate the age composition of Kvichak River rainbow trout using the master chronology.
The primary hypothesis to be tested is whether known age rainbow trout can be accurately aged by comparing their growth increment history to the master growth increment chronology developed from another group of known age fish from the same population. Once this has been demonstrated to be possible, ages of older fish can be calculated by first comparing the older individual's growth increment pattern for the early, reliable years to the master chronology and then subtracting the date of estimated birth from the date the scale was collected.
Other hypotheses to be tested include 1) whether the first annulus is reliably apparent, 2) the oldest age for which the growth increment pattern analysis can accurately assign age of an individual, 3) the number of individuals required per age class in the master chronology to accurately age other individuals, 4) the effect of age on the growth chronology, and 5) the minimum number of years required in the chronology.
The critical data for this project is a master chronology based on growth increment data from a sufficient number of individuals over a sufficient number of years. Growth increment patterns from additional individuals of known age are also required for validation of the technique.
We sought an archive of existing scales from a drainage from Southwestern Alaska. It was necessary to have samples from the same area for as many continuous years as possible. These samples needed to come from a single feeding population of fish to enhance the possibility that their growth increment patterns might be similar. It was also necessary to assume that the fish had the same environmental forces acting on their growth from year to year. The Kvichak River best met these criteria, with years of nearly continuous data containing several thousand samples. Scale samples were collected in most years by ADFG from rainbow trout captured with seines and hook and line fishing from late April to early May each year.
The scales and pertaining data were loaned to ABSC from ADFG, Dillingham Sportfish Division. The scales were pressed in acetate and most were visually aged by ADFG. Scales that were not aged by ADFG were left out of this study. Scales from no more than 200 fish are to be randomly selected from each year's data set to build the master chronology.
The pressed scale samples will be read using a Nikon Optiphot-2 microscope with a 4X lens. A video image of the imprint is projected on a monitor with a frame grabber. This image is accessed using an Optimas 4.2 bioscan program. The Optimas program is set up to measure from reference point to marked points and is calibrated using a stage micrometer. The scale focus (biological origin) is located and marked with a reference point. A transect line is then drawn from this reference point to the anterior margin of the scale. The annuli are found and marked along this transect. Annuli appear as dark bands of concentrated growth defined by areas of concentrated, broken circuli running forward from the posterior margin on one side of a scale around to the posterior margin on the other side.. The growth increment is measured at the center of the band with the most circuli breaking. The marked points are translated through the program into distances from the reference point located in the scale focus. From these distances, each year's growth on the scale can be measured to a thousandth of a millimeter. These data will then be exported to a spreadsheet for analysis.
Initial evaluations will include the influence of scale size and transect location on variation within and among different scales from the same fish. Results from the initial test will determine the number of scales to be used for each fish and the number of transects to be measured per scale. To test the reliability of the formation of the first annulus, we will also analyze the frequency distribution of the first growth increment.
The master chronology will be developed following the methods described by Ogle et al. (1994). The master chronology is based on average growth increments at each age which can be reliably aged. In the case of rainbow trout, reliably aged probably means younger than about 6 years old. The average annual growth of all reliably aged individuals over a series of years describes a template along which individuals to be aged are moved until their individual growth pattern matches the template. An error sums of squares is used to estimate the concordance between the individuals growth history and the master chronology template. Some work will be done to try to find simpler statistical approaches than used by Ogle et al. (1994), such as discriminant analysis.
Scale samples used for testing the effectiveness of the technique by against the master chronology template will be set aside from the archive scale set before the master chronology is developed. The aging and growth increment measuring methods will be the same for the evaluation group as for those used to make up the master chronology.
Expected Products/Study Endpoints
This project will demonstrate whether there is a more accurate, and non-lethal, way to age rainbow trout from Southwestern Alaska. This will give biologists a better tool for managing rainbow trout stocks. If, as the study progresses, it becomes apparent that the method does not work for rainbow trout, the study may be halted and progress to date summarized and reported, as appropriate.
Miscellany
Technology/Information Transfer:
This project is almost completely devoted to technology development in that it explores a possible non-lethal method for accurately aging rainbow trout. The technology will be transferred to fisheries managers via publication in the scientific literature and through direct consultation and/or outreach.
Personnel (see attached qualifications):
Name/ Role/ Responsibility:
E. Eric Knudsen Principal Investigator Study coordination, data analysis, writing
Brad Benter (FWS) Study implementation Image analysis, data management, writing
Cooperators/Partners:
U.S. Fish and Wildlife Service: Region 7 will provide, under an Interagency Agreement, a fisheries technician to conduct the majority of the image analysis and data management necessary for this project. Other FWS staff have been involved in conceptual design of the project. They will consult on experimental design, particularly on the statistical validity of newly developed chronology methods.
Alaska Department of Fish and Game: ADFG personnel have provided their 80,000-record scale analysis database. They have also provided Kvichak River rainbow trout scales for initial chronology development. ADFG personnel have been and will be involved in consultation on this project.
BRD Principal Contact: E. Eric Knudsen, Fisheries Research Team Leader, Alaska Biological Science Center, USGS - BRD, 1011 East Tudor Rd.., Anchorage, AK 99503, Phone: 907-786-3842, Fax: 907-786-3636, E-mail: Eric_Knudsen@usgs.gov.
Facilities/Equipment/Study Areas:
This work will be done in the laboratory and office, based on previously collected scales. It will entail many hours of biologist and technician time using an image analysis system to measure growth increments from rainbow trout scales. Subsequent data management, statistical analysis, and writing will be done in the office.
Legal And Policy-Sensitive Aspects:
There are no legal or policy-sensitive aspects to this project.
Budget/Personnel Requirements:
The funding required for this project ($10,000) will be used in an Interagency Agreement for the fisheries technician salary. All other project costs will be supported from the BRD fisheries research base budget.
Work And Reporting Schedule:
Task Completion Date
Review scale database and acquire scale samples from ADFG May, 1996
Establish image analysis protocols June, 1996
Build database of annual growth increments November, 1996
Completes statistical analysis of fitting individuals to chronology December, 1996
Complete draft manuscript February, 1997
Publication Fall, 1997
ADFG 1990. Southwest Alaska rainbow trout management plan. Division of Sport Fish, Anchorage.
Coggins, L. G. 1994. Precision of ages estimated from scales for rainbow trout in Bristol Bay, Alaska. Alaska Department of Fish and Game, Fisheries Data Series Report 94-26.
Fair, F. F. (In Prep.). Rainbow trout sampling and aging protocol. Alaska Department of Fish and Game. Fisheries Data Series Report.
Faustini, M.A. 1996. Status of rainbow trout in the Goodnews River, Togiak National Wildlife Refuge, Alaska. U.S. Fish and Wildlife Service.
Irving, D. B., and M.A. Faustini. 1994. Status of rainbow trout in the Goodnews River, Togiak National Wildlife Refuge, Alaska, 1988-1989. U.S. Fish and Wildlife Service, Alaska Fisheries Technical Report Number 24, King Salmon, Alaska.
Lisac, M. J. 1996. Length frequency, age distribution, and movement in Neguthlik and Ungalikthluk rivers, Togiak National Wildlife Refuge, Alaska, 1989-90.
Minard, R. E., and D. O. Dunaway. 1991. Compilation of age, weight, and length statistics for rainbow trout samples collected in southwest Alaska, 1954 through 1989. Alaska Department of Fish and Game. Fisheries Data Series Report 91-62.
Ogle, D. H., G. R. Spangler, and S. M. Shroyer. 1994. Determining fish age from temporal signatures in growth increments. Canadian Journal of Fisheries and Aquatic Sciences 51:1721-1727.
Riffe, R. 1994. Compilation of age, weight, and length statistics for rainbow trout samples collected in southwest Alaska, 1990-1993. Alaska Department of Fish and Game. Fisheries Data Series Report 94-17.
E. Eric Knudsen (Ph.D.). Fisheries Research Team Leader, Alaska Science Center
Training:
Bachelor of Science, Fisheries Science, University of Massachusetts, Amherst - 1974
Master of Science, Fisheries Science, Louisiana State University, Baton Rouge - 1976
Doctor of Philosophy, Wildlife and Fisheries Science, Louisiana State University, Baton Rouge - 1990
Relevant Work Experience:
Dr. Knudsen has 8 years of fisheries research experience, as well as over 12 years experience with restoration and management of Pacific salmon stocks, primarily in the Puget Sound and coastal Washington areas. He became Fisheries Research Team Leader at the Alaska Science Center in 1994 where he specializes in Pacific salmon ecology.
Representative Salmonid Publications:
Chapman, D. W., and E. E. Knudsen. 1980. Channelization and livestock impacts on salmonid habitat and biomass in western Washington. Transactions of the American Fisheries Society 109:357-363.
Knudsen, E. E., and S. J. Dilley. 1986. Effects of riprap bank reinforcement on juvenile salmonids in five western Washington streams. North American Journal of Fisheries Management 7:351-356.
Knudsen, E. E., J. M. Hiss, and D. P. Zajac. 1989. Quilcene spring chinook: a case study of limitations to Puget Sound spring chinook restoration. Proceedings of the Coho and Chinook Workshop, American Fisheries Society, North Pacific International Chapter.
Knudsen, E. E., D. P. Zajac, and B. Kenworthy. 1989. Quilcene National Fish Hatchery coded wire tagging results: coho - brood years 1974-81; fall chinook - 1972-75. U.S. Fish and Wildlife Service, Fisheries Assistance Office, Olympia, Washington.
Wampler, P. W., J.D. Anderson, and E. E. Knudsen. 1990. Evaluation of planting native hatchery coho fry into streams containing wild coho fry. U.S. Fish and Wildlife Service, Fisheries Assistance Office, Olympia, Washington.
Peters, R. J., E. E. Knudsen, J. C. Cederholm, W. S. Scarlett, and G. J. Pauley. 1993. Preliminary results of woody debris use by summer-rearing juvenile coho salmon (Oncorhynchus kisutch) in the Clearwater River, Washington. Pages 323-339 In L. Berg and P. W. Delaney, editors, Proceedings of the coho workshop, American Fisheries Society, Naniamo, British Columbia.
Step by Step Operating Procedures for Optimas 4.02
Turn computer on.
Open an Excel file that you want to download measurements into.
Scope: Nikon Optiphot-2
Power on.
Scope objective set on 4X plan
Green interference, and both neutral density .6 filters used.
Open field diaphragm to same size as image field of view.
Place turret condenser on O setting, with aperture diaphragm as small as possible.
Condenser fully raised.
Start with lowest lamp power and find optimal light while acquiring an image.
Make sure Green photo button on side is not pushed in.
Make sure silver bar on upper right side, marked with a T is in mid position.
Have camera mounted on scope using two inch adapter and .55x lens.
Camera should be mounted so that model and s/n are facing reader.
Camera: MTI ccd72
Power on.
Check cords.
Enhance, max- mid.
Gamma, adjust for contrast, start at .45.
Polarity, start at middle with switch to pos.
Bandwidth, ? Doesn't do much.
Stretch, off.
Black level, use preset and automatic.
Gain, switch on man. Dial on 150.
TV: Sony
Power on.
Input , RGB must be pushed.
Analog digital (ext sync) must be pushed.
The different settings on the right can be changed to the viewers liking.
Cords are labeled by color and position. Check these.
Check to see that camera in coming into TV.
Computer: compuadd
Optimas file found in ms-dos_6(C). Or shortcut icon.
Open Optimas and run Optimas.exe..
File, Open configuration, C:/Brad/plexi.cfg/open.
Place stage micrometer on scope and focus.
Go into data calibrate units, highlight Nikon_4X_Plexi, go to add/edit.
Highlight the word millimeters, go to calibrate.
When calibrate is selected you go to the TV. monitor where you have a green arrow. You place the point of this arrow on the middle of a millimeter line and push the left mouse button once. (You can use the keyboard arrow keys for fine adjustments.) You the move the green arrow to the farthest away millimeter bar center and push the left mouse button once, then the right button to return to the computer monitor. You then enter the number of millimeters you just measured. (Either 2 or 3.) This gives you a frame width of about 5.9mm. You then select OK . You should see that Nikon_4X_Plexi in now active. You then close this box and go to file/save configuration.
Go to data and select data collection.
Go to measurements in the collection box and select all data measurements.
Select PtDistanceToReference. And left click on box DDE and file next to this so that a small asterisk shows in each box. This allows you to use view box if you want to check your measurement against a stage micrometer. And shows that you will be exporting your data to a file. If you want to do some calibration checks then remove the asterisk from file.
Make sure the word single is selected under the nine measurement boxes on the upper right.
Now go to data and select data recorder. This is done only when you are ready to start making measurements.
Click on excel sheet and within a few second the program will go to the excel program you opened earlier. It is easiest to only have one file open at this time to limit problems.
You will then select which sheet from this file you want to export data to. If you use a short name for your file such as abc.xls it will be easier to see in the boxes they give you.
Select ok and then choose which box in the file you want to start exporting your data to. The data is recorded in one column from this point.
Then go to all measurements, select PtDistanceToReference and highlight. Select add and this will show PtDistanceToReference in the data to record box. Select done
Find image with scope.
Push the F9 key to acquire and freeze images, or use the little red F in the lower left corner of the screen and select and it will show a green A.
Move scale card on scope to place image where you want on screen.
Push the F9 key to freeze image.
Go back to bottom of computer screen and select the fourth box over that looks like a TV. screen, this is your full screen area of interest select and you will use this whenever you have problems with the image moving when you get the green arrow to close to the edge of the screen.
Now go to the data collection box and make sure the show reference box is checked.
Make sure the data export and data export boxes are without checks.
Now select the green arrow in the nine box group with one push of the left click.
You will now be on the TV. monitor where you will move the green arrow up next the the red reference point and left click once to highlight the point and once more to move the point. When the curser turns into a pencil with the reference point attached to it move the point to the upper left mark on the screen where the lines radiate from. Now you are in the correct place, hopefully without having moved the area of interest by getting to close to the edge, you right click once to drop the reference point and right click once more to go back to the computer screen.
Now put checks in the export data and extract data boxes.
You are now ready to measure and any time you right click from any of the nine boxes you download data so be careful or you will need to go back to data recorder and start that over where the data is wrong.
To do the measurements you select the upper left box with the three plus signs and a pencil.
This puts you on your image with a green arrow. The goal is to have the reference point on the focus of the scale and have the line on the screen bisecting the scale ant- posterior. You now find the annuli, put the point of the arrow exactly on the point of of the annuli and the transect and left click. You will start on annuli one and end on the edge of the scale, being a spring caught fish. After all annuli are left clicked on you will now right click and this will both download the measurements, and return you to the computer.
I then go to the excel file to make sure all is well. You need not do anything to the data recorder unless there is a mistake. It will continue to download in the next open column space.
Problems.
When you move the reference point or when ever you the green arrow box do the two right clicks in the same spot or you may have problems.
If you get an error box when right clicking try removing one of the asterisk from the file or DDE box.
When using the view box you need to have these asterisks in place .
Fisheries Project Home Page | Rainbow Trout Projects | Age Study Contents | Top of This Page
Last Reviewed: