VOL 34 NO 7 JuLY 2009

Fish News Legislative update Fisheries Journal Highlights FisheriesAmerican Fisheries Society • www.fi sheries.org American Fisheries Society • www.fi sheries.org calendar Job Center

Demonstration Flow Assessment and 2-D Modeling Making the Best use of Modeled Data: Multiple Approaches to Sensitivity Analysis of a Fish-Habitat Model

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 313 An innovative, new product from Smith-Root, the global leader in electrofishing technologies! PORTABLE ELECTROANESTHESIA SYSTEM Introducing the PES,TM a rugged, fully portable fish electroanesthesia unit for use in lab and field settings.

PORTABLE ELECTROANESTHESIA SYSTEM

PESTM CONTROLLER, ANESTHESIA TANK and ACCESSORIES

PORTABLE (for use in diverse field locations on multiple age classes - from juveniles to adults.) BATTERY OPERATED (Fully rechargeable) APPLICATION ACROSS WIDE RANGE OF SPECIES (particularly effective on warm-water fishes) VERY USEFUL IN FISH HATCHERY SETTINGS (to handle broodstock, to inoculate and spawn fish, etc) ALSO USEFUL IN MANAGEMENT APPLICATIONS (e.g. marking and tagging programs) ELIMINATES THE NEED FOR CHEMICALS (zero withdrawal period)

14014 NE Salmon Creek Ave.

Smith-Root Vancouver, WA USA 98686 314 Phone:Fisheries (360) 573-0202 • v o l 34 n o •7 Email: • j u l y 2009 [email protected] • w w w .f i s h e r i e s .o r g An innovative, new product from Smith-Root, the global leader in electrofishing technologies!

VOL 34 nO 7 PORTABLE ELECTROANESTHESIA SYSTEM juLy 2009 AMeRICANFisheries FISHeRIeS SOCIeTY • WWW.FISHeRIeS.ORG TM Introducing the PES, a rugged, fully portable fish EDITORIAL / SUBSCRIPTION / CIRCULATION OFFICES 5410 Grosvenor Lane, Suite 110 • Bethesda, MD 20814-2199 301/897-8616 • fax 301/897-8096 • main@fi sheries.org The American Fisheries Society (AFS), founded in 1870, electroanesthesia unit for use in lab and field settings. is the oldest and largest professional society representing fi sheries scientists. The AFS promotes scientifi c research and enlightened management of aquatic resources for optimum use and enjoyment by the public. It also encourages comprehensive education of fi sheries scientists and continuing on-the-job training. 320 AFS OFFICERS FISHERIES STAFF EDITORS

PReSIDenT SenIOR eDITOR ScIence eDITORS William G. Franzin Ghassan “Gus” N. Rassam Madeleine Hall-Arber Ken Ashley PReSIDenT eLecT DIRecTOR OF Doug Beard Donald C. Jackson PuBLIcATIOnS Ken Currens Aaron Lerner William E. Kelso FIRST Deirdre M. Kimball Contents VIce PReSIDenT Robert T. Lackey Wayne A. Hubert MANAGING eDITOR Beth Beard Dennis Lassuy Allen Rutherford SecOnD VIce PReSIDenT PRODucTIOn eDITOR BOOk ReVIeW COLUMN: COLUMN: William L. Fisher Cherie Worth eDITORS 316 PRESIDENT’S HOOK 340 GuEST DIREcTOR'S LINE Francis Juanes PAST PReSIDenT Ben Letcher A Look at AFS Members... The north American Freshwater Fish Mary C. Fabrizio Keith Nislow The recent AFS membership survey sheds Standard Sampling Project: Improving new light on our evolving membership, Fisheries communication eXecuTIVe DIRecTOR ABSTRAcT TRAnSLATIOn PORTABLE Ghassan“Gus” N.Rassam Pablo del Monte Luna along with their participation and priorities The need for standardization of freshwater ELECTROANESTHESIA for the Society. sampling methods has been recognized for Dues and fees for 2009 are: William G. Franzin years, and now a new comprehensive book SYSTEM $76 in North America ($88 elsewhere) for regular members, will serve as a crucial tool for helping to $19 in North America ($22 elsewhere) for student members, UPDAte: achieve this long-sought goal. and $38 ($44) retired members. Scott A. Bonar, Wayne A. Hubert, and Fees include $19 for Fisheries subscription. 318 LEGISLATION AND Nonmember and library subscription rates are $132 ($127). David W. Willis Price per copy: $3.50 member; $6 nonmember. POLICY Fisheries (ISSN 0363-2415) is published monthly by the Elden Hawkes, Jr. News: American Fisheries Society; 5410 Grosvenor Lane, Suite 110; Bethesda, MD 20814-2199 ©copyright 2009. Feature: 345 AFS uNITS Periodicals postage paid at Bethesda, Maryland, and at PESTM CONTROLLER, an additional mailing offi ce. A copy of Fisheries Guide for 320 FISH HaBItat JOUrNAL highLights: Authors is available from the editor or the AFS website, Demonstration Flow assessment and 352 TRANSAcTIONS OF ANESTHESIA TANK www.fi sheries.org. If requesting from the managing editor, 2-D Modeling: Perspectives Based on please enclose a stamped, self-addressed envelope with THE AMERIcAN FISHERIES and ACCESSORIES your request. Republication or systematic or multiple Instream Flow Studies and evaluation of reproduction of material in this publication is permitted only restoration Projects SOcIETY under consent or license from the American Fisheries Society. the methods presented by Railback and PUBLiCAtiONs Postmaster: Send address changes to Fisheries, American Kadvany (2008) may not give reproducible Fisheries Society; 5410 Grosvenor Lane, Suite 110; Bethesda, 353 BOOK REvIEW, MD 20814-2199. results, and demonstration flow assessments have significant drawbacks relative to NEW TITLES, FOR YOuR Fisheries is printed on 10% post-consumer two-dimensional hydraulic and habitat INFORMATION recycled paper with soy-based printing inks. modeling. Mark Gard AFs ANNUAL MeetiNg: 356 THE GREENING OF THE Feature: AFS’09 ANNuAL MEETING 330 FISH HaBItat PORTABLE (for use in diverse field locations on advertising Index Making the Best use of Modeled Data: CALeNDAr: 359 FISHERIES EvENTS multiple age classes - from juveniles to adults.) Emperor Aquatics, Inc...... 349 Multiple approaches to Sensitivity analysis of a Fish-Habitat Model ANNOUNCeMeNts: BATTERY OPERATED (Fully rechargeable) Floy Tag ...... 337 sensitivity analyses are essential tools in appropriately using modeled predictions to 360 JOB cENTER APPLICATION ACROSS WIDE RANGE OF SPECIES Halltech ...... 329 make management decisions. We compare three types of sensitivity analysis as applied (particularly effective on warm-water fishes) Hydroacoustic Technology, Inc. . . . 364 to an ecosystem model linking salmon and Lotek ...... 351 their habitats. VERY USEFUL IN FISH HATCHERY SETTINGS (to handle broodstock, to inoculate and E. Ashley Steel, Paul McElhany,Naomi J. spawn fish, etc) Northwest Marine Technology. . . . 363 Yoder, Michael D. Purser,Kevin Malone, Brad E. Thompson,Karen A. Avery, David ALSO USEFUL IN MANAGEMENT APPLICATIONS (e.g. marking and tagging programs) Onset ...... 319 Jensen, Greg Blair, Craig Busack, Mark D. Bowen,Joel Hubble, and Tom Kantz Oregon RFID ...... 357 ELIMINATES THE NEED FOR CHEMICALS (zero withdrawal period) O.S. Systems ...... 339 Smith-Root ...... 314 Sonotronics ...... 361 14014 NE Salmon Creek Ave. Tell advertisers you found them through cOVeR: Ambre Chaudoin, Matt Recsetar, and Olin Feuerbacher electrofi shing a stream in Vancouver, WA USA 98686 Fisheries! Southern Arizona. cReDIT: Scott A. Bonar Phone: (360) 573-0202 • Email: [email protected] Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 315 COLUMN: William G. Franzin AFS President Franzin can be PRESIDENT’S HOOK contacted at [email protected].

A Look at AFS Members...

AFS is a fairly large professional report and to follow up, Connelly and on meetings is that a majority of sur- society with a core of dedicated Brown will provide a detailed article vey respondents favored web-based members who pay dues annually or from their report in the August issue broadcasts of symposia from Annual have purchased life memberships of Fisheries. This column is to whet Meetings and the development of to obtain the benefits the Society your appetite for a closer look at their online professional development offers. I suggested what some of article; I think you will find it inter- opportunities. There is keen support these benefits are in my last column. esting. In the course of researching for mentoring in AFS, both through You have seen in Fisheries and on this Hook, I also solicited some data programs like the Hutton Scholarship the website that we are about to from Membership Coordinator Eva Program and mentoring of young adopt a new Strategic Plan to guide Przygodzki at AFS headquarters to professionals by established fisheries us over the next several years. The examine some trends in recruitment professionals. Only about a third of Strategic Plan required the Society and losses. I was interested in our respondents were fully aware of the to do a bit of research on ourselves; annual net recruitments and losses in mechanisms and roles of Chapters, for example, we collected informa- the different categories of member- Divisions, and Sections in AFS gov- tion from our Membership and ship over the readily available period ernance but a majority of surveyed Membership Concerns committees, of record. These are reported below members felt the Society was meet- President Fabrizio commissioned a after my précis of the membership ing their needs. Although a majority consultant to carry out a focus group survey. of survey respondents felt AFS was analysis of a selected group of AFS The 2008 survey was con- doing enough to recruit and keep members to get opinions about AFS ducted entirely electronically via new members, 30–40% thought we member services, and finally we com- an e-mail sent to 1,500 randomly could do more and a large major- missioned a membership survey. The selected members, of whom 55.7% ity thought we should do more to 2008 Membership Survey was similar responded. In the 11 years since involve AFS members in the Society. to surveys we conducted in 1997 1997, we as a society haven’t Finally, of nine future priorities of AFS, (Brown 1998) and in 2004 (Brown changed a great deal. We have a large proportions of survey respon- and Cooke 2005) but also different in slightly higher percentage of female dents felt that AFS should do more that a few new questions were asked. members, our mean age hasn’t public outreach and political advo- The questionnaire that formed the changed much, the core employ- cacy on behalf of aquatic resources. basis of the 2008 results was devel- ers remain the same, and we have As I mentioned above, look for the oped by a small group of members, more student members. Most of our detailed analyses of the questions in officers, and the executive director. members are fisheries biologists who the August issue of Fisheries. The 2008 survey, like the previous work more in freshwater than in I turn now to trends in member two, was conducted for us by the marine or estuarine ecosystems. More recruitment and losses in our various Human Dimensions Research Unit of us are attending Annual Meetings membership categories, annually and of Cornell University Department of but the meetings still attract a little over the last few years. The Society’s Natural Resources. The common ele- less than half the membership over a numbers grew during the middle ment in all these surveys was Tommy three-year period. In other words, in of the decade but have stabilized Brown, but this year Nancy Connelly any one year perhaps only 15-20% at around 8,700 members (Table 1 of the same unit led the study. In the of members attend the Annual and Figure 1). I found it interesting 2008 report, Connelly and Brown Meeting. Chapters are more impor- that annual recruitment in the period compared the outputs of the previous tant to most members than Divisions 1999–2003, although proportional to surveys with the findings in 2008 so or Sections, probably reflecting our membership numbers, was more than we could investigate some trends in large geographic distribution and the offset by annual losses on average various areas. What I report here will greater ability to participate locally in of 129 members per year (<1.5%; be some of the highlights of the 2008 a Chapter. A noteworthy observation Table 2 and Figure 2). If one looks

316 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g Table 1 and Figure 1. AFS membership by category, 2006–2008. at month-to-month gains and losses, we do gain new members at Annual Meetings but often lose many of them in the follow- ing year, meaning that each year shows the losses of attendees at the Annual Meeting that don’t maintain membership in AFS. Incrementally our numbers slowly have increased mainly by the addition of student members. In 2002–2003 the losses were declining and we entered a period of alternating gains and losses in 2006–2008 with an overall very small gain. We lost some regular members to the young profes- sional category in the first few years of making that category available but now it appears that a few young profession- als are leaving that category but not being picked up as regular members. This indicates that Table 2 and Figure 2. Annual membership gains and losses 1999–2003 and 2006–2008. some young professionals are leaving the Society when that benefit runs out. What does all this mean? Our core member- ship is quite constant and we get a few more new members per year than we lose but not enough to change our dynamic. How do we attract more of the two-thirds of fisheries profession- als in North America that are not AFS members? We offer even better member services, even better journals, have even better meetings, and actively recruit new members in the coming years. We can easily add international mem- bers with that extra outreach that our members suggest we should do. For example, recently, more than 30 young Mexican members References Research Unit Series 05-4. Department were recruited at the University of Natural Resources, Cornell of Mazatlan by Lourdes Rugge of Brown, T. L. 1998. AFS membership University, Ithaca, New York. Connelly, N. A., and T. L. Brown. 2008. the International Fisheries Section. survey shows Society efforts, services 2008 AFS membership study. Human Read more about us in the article on track. Fisheries 23(10): 6-11. Dimensions Research Unit Series 08-6. in the August issue of Fisheries. Brown, T. L., and S. J. Cooke. 2005. AFS Department of Natural Resources, membership study. Human Dimensions Cornell University, Ithaca, New York.

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 317 UPDAte: Elden Hawkes, Jr. LEGISLATION AND POLIcY AFS Policy coordinator Hawkes can be contacted at ehawkes@fi sheries.org.

Hearing on national Fish Habitat conservation Act Ocean conservation, education, and national Strategy On 16 June 2009, the House Committee on Natural for the 21st Century Act. Resources’ subcommittee for Insular Affairs, Oceans and On 18 June 2009, the House Committee on Natural Wildlife held a hearing to discuss the National Fish Habitat Resources’ subcommittee for Insular Affairs, Oceans and Conservation Act (NFHCA; H.R. 2565) and the Pacifi c Wildlife held a hearing to discuss the Ocean Conservation, Salmon Stronghold Conservation Act of 2009 (PSSCA; H.R. Education, and National Strategy for the 21st Century Act 2055). The hearing featured testimony from a diverse group (H.R. 21). This hearing also heard testimony from a diverse from federal, state, and non-governmental organization rep- group from federal, state, and non-governmental organiza- resentatives. All of the participants agreed that the NFHCA tion representatives. Kristen Fletcher of the Coastal States and the PSSCA both seek to achieve healthy habitats, Organization testifi ed that H.R. 21 has the components healthy fi sh populations, and healthy coastal ecosystems, as needed to reform ocean governance, improve the health of well as increased ecosystem services, aesthetic value, and our oceans and coasts, and create a national ocean policy. fi shing opportunities for the U.S. public. It is estimated that Fletcher further stated that the bill would help reorient they would also provide additional benefi ts to commercial government to formalize the practice of ecosystem-based and recreational fi sheries by contributing $10 billion annu- management and has the potential to build on the work ally to the nation’s economy. In regard to the PSSCA, many on the panel felt that it already done by existing regional ocean partnerships. would provide a unique cooperative forum for public and Margret Caldwell of the Center for Ocean Solutions private stakeholders to improve current salmon manage- stated that leadership similar to that which was demon- ment and conservation policies. The PSSCA will also assist strated in legislation like the Clean Water and Clean Air acts state governments to accelerate implementation of a is appropriate for the nation’s oceans. This would not only holistic, comprehensive salmon conservation and manage- provide a unifying framework for the existing sectoral laws ment approach that integrates habitat, harvest, hatcheries, and agencies, but it would also provide a national vision and hydropower. It was expressed that the PSSCA would and support system for states whose livelihoods depend on provide the focus and resources necessary to be successful healthy and resilient ocean systems. She concluded her testi- in the healthiest wild salmon rivers and will restore balance mony by stating that it is time for the United States to adopt to current efforts to recover depleted stocks and maintain a unifi ed approach to the management of one of its most healthy ones. critical resources and establish a national ocean policy. In regards to the NFHCA, many felt that the NFHCA Pietro Parravano of the Joint Ocean Commission Initiative provides a strong overarching framework to build on and testifi ed that passing H.R. 21 should provide a permanent further advance fi sh habitat protection conservation actions and comprehensive policy and implementation framework nationally. It was stated that the NFHCA will help fi sh every- for addressing ocean health. The policy should: where by supporting public-private efforts to protect and habitat naturally. • Authorize and fund implementation of comprehensive All of the panel participants urged the passage of both ocean governance legislation that creates a national the NFHCA and the PSSCA by Congress. It was added that ocean policy. the potential additional fi sheries losses are too great to not • Codify and strengthen the federal coordinating structure establish the cohesive framework that these bills would for implementing the national ocean policy. create. All members of the panel felt that both the NFHCA • Support regional marine and coastal management and and the PSSCA are entirely compatible and complementary. governance efforts. Some expressed a desire to work with the subcommittee and other stakeholders to integrate the purposes and goals He concluded by stating that with the passage of H.R. of the two bills if the subcommittee decides it would be 21, the benefi ts to be gained are enormous and are essen- more effi cient to merge them into a comprehensive national tial to revitalizing the economic and ecological health of our approach to fi sh habitat conservation. oceans, coasts, and Great Lakes.

318 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g Electronic Balloting on AFS Constitution and Rules Changes By Gwen White AFS Constitutional Consultant Gwen White can be contacted at [email protected].

We are testing a new mechanism for electronic review to facilitate broader and more timely participation of the AFS membership in updating the protocols that guide Society structure and function. We request your patience and welcome any suggestions on creating an effective electronic review process. The following changes were recommended by the Governing Board at their meeting on 7 March 2009. All three proposals are provided for your review at www.fi sheries.org. Comments can be submitted in the online forum by July 15 or e-mailed to the AFS Constitutional Consultant, Gwen White, [email protected]. Active Members may consider three amendments to the AFS Constitution and Rules for adoption by electronic ballot between August 1 and August 15. The three proposals address the following protocols:

a. Constitution Article IV and IX. Parliamentarian—Establishes the Constitutional Consultant as a nonvoting member of the Management Committee, creates an Apprentice position to the Constitutional Consultant, and a Past Constitutional Consultant’s Advisory Council. b. Constitution Articles III, VII, and IX. Appointment of committee and editorial board members—Allows the AFS President to delegate responsibility to appoint committee and editorial board members to the chair and journal editor, respectively. c. Rule 4.H. Fish Habitat Section—Upon petition from over 100 Active Members, the Governing Board recommends establishing a new Fish Habitat Section.

onset

$120 HOBO® Water Temp Pro v2 Loggers combine research-grade performance with exceptional value. Ideal for use in fresh or salt water, the HOBO® Water Temp Pro v2 provides a

Under number of key advantages: • Research-grade measurements • Waterproof to 120 meters (400 feet) • Easy to use graphing and analysis software study • Waterproof data shuttle for easy readout

Underwater Temperature Loggers: Considerations for Selection & Deployment Get tips on selecting & deploying underwater loggers To Learn more call Download our free white paper today 1-866-460-8979 at onsetcomp.com/fisheries HOBO visit us at onsetcomp.com Onset

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 319 Feature: Mark Gard Gard is a fish and wildlife biologist with Fish Habitat the U. S. Fish and Wildlife Service, Sacramento, California. He can be reached at [email protected]. Demonstration Flow Assessment and 2-D Modeling: Perspectives Based on Instream Flow Studies and Evaluation of Restoration Projects

ABSTRACT: Railsback and Kadvany Figure 1. Demonstration flow assessment replicate survey areas. (2008) make a compelling case for the use of demonstration flow assessments (DFA) for instream flow assessments as a more robust method than traditional one-dimensional habitat simulation techniques, such as the Physical Habitat Simulation System. But, based on experience with DFAs used for instream flow assessments and evaluation of stream restoration projects, the methods presented by Railsback and Kadvany (2008) may not give reproducible results, and DFAs have significant drawbacks relative to two-dimensional (2-D) hydraulic and habitat modeling. Application of the DFA methodology presented in Railsback and Kadvany (2008) to assess a stream restoration project on the Trinity River, California, did not give reproducible results, with substantial disparity between replicate surveys in the total quantity and spatial distribution of habitat. As an empirical Figure 2. An example of differences between initial and repeat surveys in the DFA effort on the two-dimensional habitat modeling Trinity River exemplified by holding, presmolt, and spawning habitat types in survey segment 1, method, DFAs have several drawbacks. 2006. Compared to 2-D models, DFAs require interpolation and extrapolation of results from the observed flows, and lack biological realism because they use categorical (binary) habitat suitability criteria. Further, 2-D modeling offers significant benefits overA DF in the context of adaptive management. Advantages of DFAs cited by Railsback and Kadvany (2008), namely the feasibility of assessing habitat with complex hydraulics, the ease in using mechanistic and theoretical conceptual models, and the ability to assess long reaches, also apply to 2-D models. Modification of the DFA methods presented in Railsback and Kadvany (2008) to give reproducible results may increase the cost of DFAs or reduce the length of stream that can be assessed with DFAs to the point where 2-D modeling may be a more cost-effective technique than DFAs.

320 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g Figure 3. Pit River site at 51 m3/s: A. Demonstration flow assessment results (from R2 Resources Consultants, Inc. 2003); Método de Evaluación de B. Two-dimensional hydraulic and habitat modeling results for deep/fast habitat (from Hardin-Davis, Inc. 2003); and Regímenes de Caudales y C. Two-dimensional hydraulic and habitat modeling results for adult rainbow trout (O. mykiss) rearing habitat (from Hardin-Davis, Inc. 2003). modelación 2-D: The DFA deep/fast habitat corresponds to adult rainbow trout habitat. Differences between A and B can be attributed to errors in DFA mapping, while differences between perspectivas sobre B and C can be attributed to the use of categorical versus continuous criteria. estudios de caudales A. ecológicos y evaluación de proyectos de restauración

RESUMEN: Railsback y Kadvany (2008) exponen un sólido caso en cuanto al método de Evaluación de Regímenes Caudales (ERC) aplicado a caudales ecológicos, como una herramienta más robusta que las técnicas tradicionales unidimensionales para simulación de hábitat, como por ejemplo el Sistema de Simulación de Hábitat Físico. Sin embargo, sobre la base de la experiencia con el ERC aplicado para evaluar caudales ecológicos y proyectos de restauración de caudales, los métodos presentados por Railsback y Kadvany (2008) pueden producir resultados B. no replicables, con diferencias sustanciales entre la réplica de los muestreos en cuanto a la cantidad total y la distribución espacial del hábitat. Como método de modelación empírica de hábitat bidimensional, el ERC tiene algunas desventajas. En comparación con los modelos 2-D, el ERC requiere de la interpolación y extrapolación de los resultados de los caudales observados y carece de realismo biológico ya que se fundamenta en el uso de categorías binarias como criterio para determinar cuán apropiado es un hábitat. Más aún, la modelación 2-D ofrece beneficios significativos en comparación al ERC en un contexto de manejo adapativo. C. Las ventajas del ERC mencionadas en Railsback y Kadvany (2008), que consisten en la viabilidad de evaluar el hábitat a partir de aspectos sobre hidráulica, la facilidad de utilizar modelos conceptuales mecanísticos y teóricos, y la habilidad de evaluar canales grandes también aplican a los modelos 2-D. La modificación de los métodos de ERC presentados por Railsback y Kadvany (2008) para generar resultados reproducibles, puede incrementar el costo de los ERC o reducir la longitud del caudal que sea objeto de evaluación, hasta un punto en el que la modelación 2-D puede resultar una técnica más económica.

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 321 Figure 4. Habitat availability curves for 1+ coho rearing in the lower Figure 5. Sacramento River fall-run Chinook salmon spawning Oak Grove Fork using 1-D, 2-D, and DFA methodologies. Adapted from habitat suitability criteria for velocity (from Gard 2006), along with McBain and Trush (2004). corresponding categorical criteria defined as velocities with a suitability greater than 0.5.

Figure 6. Pit River boulder-dominated riffle habitat at 51 3m /s: A Demonstration flow assessment results (from R2 Resources Consultants, Inc. 2003); and B Two-dimensional hydraulic and habitat modeling results for adult rainbow trout rearing habitat (from Hardin-Davis, Inc. 2003). The DFA and 2-D results are from different sites, because no boulder-dominated riffle sites were assessed with both DFA and 2-D. The results show areas of adult rainbow trout rearing habitat downstream of boulders which were captured with 2-D but were mostly not captured with DFA due to the minimum polygon size of 2.3 m2 for pocket water and 4.6 m2 for other mesohabitat types (R2 Resources Consultants, Inc. 2003). Specifically, Figure 6(A) only shows two polygons downstream of boulders, despite the presence of many boulders, while Figure 6(B) shows numerous areas of medium to high-quality habitat (yellow, orange and red colors) away from the channel banks, downstream of boulders.

322 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g INTRODUCTION 2. the need for many years of data; 3. the need to assume a linear relationship between the bio- In a recent feature, “Demonstration Flow Assessment: logical response variable and flow between each observed Judgment and Visual Observation in Instream Flow Studies,” flow; Railsback and Kadvany (2008) make a compelling case for the 4. the inability to extrapolate beyond the observed range of use of a judgment-based habitat evaluation technique, demon- flows; and stration flow assessments (DFA), to quantify the relationship between fish habitat and flow for instream flow assessments and 5. For spawning, the need for intermediate levels of spawn- they identify procedures for improving credibility and repro- ing—at low spawning levels, there will not be any redd ducibility of the technique. Notably, Railsback and Kadvany superimposition even with low habitat abundance, while (2008) cite many advantages of DFAs versus traditional one- at high spawning levels, the amount of superimposition dimensional habitat modeling, such as the Physical Habitat cannot be determined because individual redds can no Simulation (PHABSIM) System. The terminology for DFAs longer be identified. is still in flux and relate to a variety of habitat evaluation Demonstration flow assessments (CIFGS 2003) use direct techniques; DFAs are sometimes referred to as “expert habi- observation of river habitat conditions at several flows; at tat mapping,” “judgment based habitat mapping,” or “habi- each flow, polygons of habitat are delineated in the field. tat criteria mapping.” Recent DFAs in California have been Disadvantages of this approach are: used for instream flow assessments on the Pit (R2 Resources Consultants, Inc. 2003), Lower McCloud (Pacific Gas and 1. its potential subjectivity; Electric Company 2007) and Upper Tuolumne (McBain and 2. Limitations in the accuracy of delineation of the polygons, Trush, Inc. 2007) rivers, and for evaluation of stream restora- using the methods in CIFGS (2003); tion projects on the Trinity River (the subject of this article). 3. the need to have categorical (binary) habitat suitability These studies illustrate three important aspects of DFAs that criteria; should be considered: (1) the ability of the procedure out- lined in Railsback and Kadvany (2008) to produce a credible, 4. the need to assume a linear relationship between habitat legally defensible, and reproducible result; (2) the relative and flow between each observed flow; and performance of two-dimensional (2-D) hydraulic and habi- 5. the inability to extrapolate beyond the observed range of tat modeling (Gard 2006) versus DFAs; and (3) the range of flows. flows and actual effort needed for DFAs. Previous applica- The second disadvantage can be overcome, albeit at tions of judgment-based habitat evaluation techniques have increased cost, by surveying in the vertices of the polygons received criticism for their potential lack of reproducibility and making extensive depth and velocity measurements to (Arthington and Zalucki 1998; Young et al. 2004). This lack validate the polygon boundaries. The fourth disadvantage of reproducibility was demonstrated when replicate surveys can be overcome, again at substantially higher costs, by map- of the judgment-based Expert Panel Assessment Method pro- ping in habitat at a large number (on the order of 30) of duced disparate results (Swales and Harris 1995). Although flows. I agree with Railsback and Kadvany (2008) that DFAs can represent a substantial improvement over one-dimensional, REPRODUCIBILITY OF DEMONSTRATION FLOW or transect-based, habitat modeling, I conclude that 2-D hab- ASSESSMENTS: A Case Study from the itat modeling produces several advantages over either DFAs Trinity River or one-dimensional transect-based approaches that should be considered. The Trinity River, California, is the subject of a river res- The variety of alternative techniques available to evaluate toration effort focused on a 64 km reach from Lewiston Dam habitat can be broken down into three general categories: to the North Fork Trinity River. The goal of the restoration effort is to restore geofluvial processes which will increase 1. biological response correlations; the availability of aquatic habitats and result in increased 2. Demonstration flow assessments; and fish populations (U.S. Fish and Wildlife Service and Hoopa 3. habitat modeling (Annear et al. 2004). Valley Tribe 1999). Agencies and tribes working on recovery need a habitat evaluation method to assess progress toward Biological response correlations are generally developed increasing aquatic habitats. Goodman et al. (2007) evaluated by a linear regression of a biological response variable, such a DFA methodology and study design on the Trinity River as redd superimposition, versus flow. They can be used to similar to that presented by Railsback and Kadvany (2008). evaluate spawning habitat by examining the degree of redd In this case study, the DFA was evaluated by conducting rep- superimposition at different flows (Snider et al. 1996), or to licate surveys at three locations (Figure 1). The results of evaluate rearing habitat by examining juvenile production these surveys are presented herein. estimates at different flows (Hvidsten 1993). Disadvantages of this approach are: Methods 1. Difficulty in separating out effects of flows from year-to- year variation in escapement, survival, productivity, and Six clearly defined habitat types were developed that per- other factors; tain to species and lifestages of interest to the Trinity River

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 323 restoration effort (Table 1). Several species and lifestages with were calculated from the shapefiles using Xtools for ArcGIS similar habitat requirements were combined into aggregate desktop (ESRI 2006) and exported into Microsoft Excel for habitat types, so the types could be distinguished during field analyses. observations. The fry habitat type represented high quality The field effort (training and application) required 11 days rearing habitat for Chinook (Oncorhynchus tshawytscha) and to complete. Two days were spent on training and calibra- coho (O. kisutch) salmon with a total length (TL) less than tion. A total of 20.47 rkm of river habitat were mapped over 50 mm, as well as egg deposition and larval rearing habitat an additional 9 field days. In general, the survey crew con- for the foothill yellow-legged frog (Rana boylii) and western sisted of 3–4 members on 1 raft, with 2 individuals mapping toad (Bufo boreas). The presmolt habitat type corresponded habitats. Map generation prior to the field effort required an to high-quality rearing habitat for Chinook salmon from 50 to additional 20 staff days, while post-processing of the field 200 mm TL. The spawning habitat type represented high qual- data using ArcGIS took another 30 staff days, for a total cost ity spawning habitat for Chinook and coho. The holding habi- of $126,013 to conduct the DFA. tat type characterized the holding habitat for adult Chinook Reproducibility of the DFA methodology was assessed salmon. The ammocoete habitat type denoted lamprey rearing at three sites within the Trinity River, with a total length habitat. The large woody debris habitat type was applied as of 4.06 rkm. The sites were randomly selected to represent a surrogate for western pond turtle (Acitenemys marmorata) the range of habitat conditions in the Trinity River from habitat. Lewiston Dam to the North Fork Trinity River, and thus were The criteria for the salmonid components of the fry, not selected to be similar. At these sites the observation team presmolt, adult holding, and spawning habitat types were conducted two surveys. Although the observation team was defined from Trinity River habitat suitability curves to rep- not completely the same for the two surveys, some members resent high quality habitat (U.S. Fish and Wildlife Service were shared between repeat surveys. Surveys were completed and Hoopa Valley Tribe 1999; Yurok Tribe and U.S. Fish and at summer base flows, when the release from Lewiston Dam Wildlife Service, unpublished data). The foothill yellow- was 12.7 m3/s and discharge at the downstream boundary of legged frog and western toad require shallow, low-velocity the mapped reach ranged from 14.1 to 14.8 m3/s at the U. S. areas for egg deposition and larval development and were Geological Survey (USGS) gauging station above the conflu- included in the fry habitat type (Lind and Wilson 1996; ence with the North Fork Trinity River (USGS gage number Cavallo 1997; Lind 2004). The ammocoete habitat type was 11526400). Weather and daylight conditions did not differ defined as areas of low velocity (< 0.05 m/s), fine substrate, between the initial and repeat surveys. and deposition but excluding areas of anoxic sediments (Potter 1980; Wydoski and Whitney 2003). Large woody Results debris habitat type is important for the basking behavior of western pond turtles and was applied as a surrogate to In the Trinity River DFA surveys, discrepancies between describe their habitat (Reese and Welsh 1998). initial and repeat habitat areas varied by habitat type and Maps of the study areas were developed for field observa- survey segment (Table 2). An example of such discrepancies tions from 2001, 2005, and 2006 high resolution aerial pho- between initial and repeat surveys is presented in Figure 2. tography. The maps were developed in ArcGIS (version 9.2) The ammocoete habitat type produced both the largest and using the map book utility (ESRI 2006). Two sets of 1:207 smallest discrepancies between initial and repeat surveys. The to 1:276 scale maps were produced for each survey. During largest discrepancies were observed in survey segment 3 with field mapping, three habitat types were depicted on each set a repeat ratio (initial/repeat habitat area) of 11.8, whereas of maps to avoid errors associated with high levels of overlap the lowest discrepancies were observed in survey segment 2 among types (e.g., fry and presmolt habitat types). A team with a repeat ratio of 1.1. The repeat ratio in other habitat of qualified observers was selected from a subset of the agen- types varied from adult holding at 0.9 in survey segment 3 to cies involved in the restoration effort, including representa- fry at 2.7 in survey segment 2. tives of the California Department of Fish and Game, Hoopa Valley Tribe, U.S. Fish and Wildlife Service, and Yurok Tribe. Discussion Prior to field observations, areas were delineated on the field maps and the observers (a) were trained in the field meth- The rehabilitation actions on the Trinity River are designed ods and (b) calibrated their ocular estimates of depths and to promote geofluvial processes which will in turn increase water velocities to measured values. The survey methodology channel complexity and improve aquatic habitats (Barinaga included a three-step decision-making process for mapping 1996). Although specific sites on the Trinity River may be habitats. First, a crew member would identify an area that restored each year, the evolution of these sites with geoflu- potentially corresponded to a habitat type. The area would vial processes into their full habitat potential may occur over then be discussed among the group regarding habitat-type decades. Evaluation of the changes in aquatic habitat in the criteria until a group consensus was reached. Finally, a team Trinity River requires a method that is reproducible through member would delineate the habitat area on the map, con- time. The application of the scientific method requires obser- firming with the group about its placement and size. The sur- vations that are unbiased and reproducible. Without a means vey maps were scanned and rectified, and the habitat types of creating reproducible observations it is no longer science. were digitized into ArcGIS shapefiles (ESRI 2006). A final The DFA methodology as described above did not create quality-control step was performed by comparing habitat- reproducible results despite implementation of the methods type shapefiles against field maps. Habitat-type areas (m2) listed in Railsback and Kadvany (2008).

324 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g Table 1. Target species and life stage used by the DFA technique with associated habitat types and criteria applied to the Trinity River, 2006. Cover codes are as follows: NERA, non-emergent rooted aquatic vegetation; ER, exposed roots; SWD, small woody debris; LWD, large woody debris; H, grass or herbaceous; SS, shrub-scrub (primarily Salix spp. and Rubus spp.). Habitat Type Species Minimum Depth (m) Column Maximum Cover Substrate Habitat Size Velocity(m/s) Distance to Cover (m) Min. Max. Min. Max. In Water Out of Water Chinook and coho salmon fry <50 mm TL ER, H, NERA, ER, H, SS, Fry Foothills yellow legged 4.7 (m2) 0.12 0.61 0 0.12 0.61 frog eggs and larvae SS, SWD SWD

Western toad eggs and larvae ER, NERA, ER, SS, Presmolt Chinook >50 mm TL 4.7 (m2) 0.15 1.52 0 0.24 0.61 SWD SWD 50% Holding Adult salmonid holding 3.04 channel width Chinook salmon Spawning 9.3 (m2) 0.18 0.64 0.15 0.61 5-15 cm Coho salmon fines and Ammocoete Lamprey ammocoete 4.7 (m2) low 0.31 detritus

30 cm diameter LWD Western pond turtle LWD LWD 3 m length

Table 2. Summary of replicate surveys conducted to evaluate repeatability in the DFA method including survey length (rkm), habitat type area (m2), and ratio between surveys. A reproducible DFA would have ratios close to 1. Length Ammocoete Segment Survey Fry Holding Presmolt LWD Spawning (rkm) (m2) I 473 3,040 790 3,237 163 4,003

1 1.63 II 284 1,700 1,972 2,282 196 1,967 Ratio I/II 1.7 1.8 0.4 1.4 0.8 2.0 I 39 509 795 1,211 76 3,716 2 1.16 II 34 188 492 482 46 810 Ratio I/II 1.1 2.7 1.6 2.5 1.6 4.6 I 266 173 1,113 594 140 693 3 1.27 II 23 339 1,234 495 283 1,064 Ratio I/II 11.8 0.5 0.9 1.2 0.5 0.7

The habitat differences between initial and repeated mapped and what is digitized into ArcGIS due to rigorous surveys can be related to errors in judgment, field measure- GIS post-processing methods. ment, intergroup variation, or digitizing. Judgment error can Railsback and Kadvany (2008) suggested standardization be defined as imperfect detection of attributes which define of the DFA decision-making process through the implemen- habitats, such as over- or underestimating depths or veloci- tation of a conceptual model. This conceptual model lists ties. Field measurement error is a product of the difference steps in the decision process to help reduce judgment error between the actual size of the habitat and the mapped poly- associated with the technique. As observed in the Trinity gon. Another type of field measurement error could occur River, the conceptual model described by Railsback and when a habitat is drawn in the wrong location on the map. Kadvany (2008) is not sufficient to create reproducible data. Although Goodman et al. (2007) used a group that shared I suggest that creating reproducible DFAs will require addi- members between repeat surveys, an additional source of tional steps to reduce judgment and field measurement errors. error may be intergroup variation if the same observers could For example, judgment errors could be reduced by measuring not be assembled each time. The final and probably least habitat parameters rather than relying on ocular estimates. significant source of error is the difference between what is Field measurement errors could be reduced by implementing

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 325 more rigorous survey methodologies to accurately identify availability. As a result, I have been using a more sophis- habitat type boundaries. ticated approach of logistic regressions (Manly et al. 2002; Current applications of aquatic habitat assessment on the Ahmadi-Nedush et al. 2006) to develop habitat suitability Trinity River, subsequent to the results reported here, are criteria. employing several additional steps to improve the accuracy Railsback and Kadvany (2008) cite the following advan- and reproducibility of habitat assessment within the DFA tages of DFAs, versus traditional one-dimensional hydraulic framework (Damon Goodman, U.S. Fish and Wildlife service, and habitat modeling: Arcata, CA pers. comm.). This more rigorous DFA method- 1. assessment of habitat with complex hydraulics is more ology employs parameter measurement for all habitat crite- feasible, ria and incorporates a cover component. Furthermore, the boundaries of habitat types are delineated using tablet com- 2. DFA makes it easier to use mechanistic and theoretical puters, high accuracy GPS transceivers, and high resolution conceptual models, and aerial photographs. Evaluations of this survey methodology 3. DFA can assess long reaches. have lead to a more credible and reproducible DFA habitat assessment. This method will be applied to the Trinity River In contrast, 2-D modeling can handle complex hydrau- in 2009 to evaluate river restoration sites. The plan includes lics, including transverse flows, across-channel variation in evaluation of 2.5 km of bank rehabilitation sites at 5 flows water surface elevations, and flow contractions/expansions and an additional 4.5 km at summer base flow for $216,165. (Ghanem et al. 1996; Crowder and Diplas 2000; Pasternack Accordingly, this more rigorous DFA methodology will likely et al. 2004). To my knowledge, the only type of hydraulics more than double the costs or halve the length of stream that that would require the use of DFA would be situations where can be assessed, as compared with the methods presented in the gradient exceeds 10%; such circumstances violate the Railsback and Kadvany (2008). However, there are likely assumption of 2-D models that there is not a significant ver- to be tradeoffs of additional field time versus reduced post- tical component to velocities. Because DFAs are empirical processing effort for a more rigorous DFA, compared to the 2-D models, both DFAs and 2-D modeling can provide the methods presented in Railsback and Kadvany (2008). inputs (spatially explicit representations of habitat) needed for mechanistic and theoretical conceptual models. In fact, DFAS VERSUS TWO-DIMENSIONAL HYDRAULIC through geographic information system (GIS) post-process- AND HABITAT MODELING ing, complex aspects of habitat, such as: 1. Adjacent velocity criteria (Gard 2006), which addresses Demonstration flow assessments can properly be charac- the bioenergetic mechanism of transport of invertebrate terized as empirical two-dimensional habitat modeling, where drift from fast-water areas to adjacent slow-water areas habitat is empirically determined at a series of flows, rather (where fry and juvenile salmonids reside) via turbulent than being simulated for a range of flows. One of the few mixing; and cases I am aware of where the results of DFAs and 2-D mod- 2. Escape cover (Hardy et al. 2006), can be incorporated into eling can be directly compared is from the Pit River, which 2-D modeling, while it would be challenging to incorpo- had one study site where habitat was quantified with both rate these into DFAs. DFA (R2 Resources Consultants, Inc. 2003) and 2-D model- ing (Hardin-Davis, Inc. 2003). The substantial differences GIS post-processing can also be used to implement spa- between the results of the two methods (Figure 3) can likely tial resolution considerations mentioned by Railsback and be attributed to a combination of errors in DFA mapping and Kadvany (2008), such as eliminating habitat patches that are the use of categorical criteria for the DFA versus continu- smaller than a territory size for adult trout (very roughly 1 m2; ous criteria for the 2-D modeling. McBain and Trush (2004) Railsback et al. 2005) or a redd for spawning, for 2-D model- present comparisons of the numeric results of the DFA case ing. I have used 2-D models to examine two of the mechanis- study in Railsback and Kadvany (2008) versus 2-D modeling. tic conceptual models mentioned by Railsback and Kadvany Differences in the results of DFA and 2-D modeling (Figure (2008): drying redds (USFWS 2006a) and producing food 4) could possibly affect the decision-making of managers. (USFWS 2006b). In addition, Bowen et al. (2003) used 2-D Demonstration flow assessments by definition use categori- modeling with alternative habitat suitability criteria, such as cal criteria, while 2-D modeling typically uses continuous cri- measures of habitat diversity, one of the theoretical concep- teria (Figure 5). Categorical criteria are generally biologically tual models mentioned by Railsback and Kadvany (2008). unrealistic—they either (a) overestimate the habitat value Two-dimensional hydraulic and habitat modeling has been of marginal conditions if the categorical criteria are broadly used to assess long reaches; Hardy et. al. (2006) and Bowen et defined (e.g., setting suitability equal to one for any depths al. (2003) modeled a total of 9 and 26.3 km of river, respec- and velocities where the original Habitat Suitability Index tively, compared to 0.8 km mapped in the DFA discussed by value was greater than 0.1) or (b) completely discount the Railsback and Kadvany (2008). Recent advances in technol- habitat value of marginal conditions. The latter case would ogies for collecting bed topography data (which requires most be biologically unrealistic because it would not account for of the field effort) for 2-D modeling, including robotic total substantial fish use in areas that are deemed unsuitable from stations, survey-grade real-time kinematic GPS equipment, the categorical criteria. I agree with Railsback and Kadvany acoustic doppler current profilers, and echo sounders, have (2008) that many of the existing traditional PHABSIM cri- made it possible to collect the data for 2-D modeling over teria curves are suspect, largely because of biases from habitat long reaches fairly quickly. For example, an approximate esti-

326 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g mate of the cost (not including capital costs of equipment) ated cost, errors associated with extrapolation likely cannot of 2-D modeling for 1.9 km of the Lower McCloud River was be alleviated. For example, in the case study presented by $450,000 (Craig Addley, Entrix, pers. comm.), compared to Railsback and Kadvany (2008), sampling at flows in excess an estimated cost of $715,000 for a DFA of 4 km of the Lower of 9.2 m3/s was not conducted in part because observers McCloud River at 6 flows (Pacific Gas and Electric 2007). could not wade safely at higher flows. Assessment of habitat In addition, remote sensing techniques, such as photogram- at higher flows may be critical in systems where unimpaired metry and Light Detecting and Ranging (LIDAR), may hold high flows during snowmelt provide important benefits, such the promise of further increasing the length of river than as juvenile habitat associated with inundated floodplains can be assessed with 2-D models, once they become suffi- (Jeffres 2006). Historically, managers have confronted this ciently accurate (i.e., 0.03 m elevation) for purposes of simu- problem before. Empirical methods were first used for one- lating microhabitat. In contrast, the cost of a reproducible dimensional hydraulic and habitat assessments, but were DFA, including surveying the vertices of habitat polygons replaced by one-dimensional hydraulic and habitat modeling and extensive depth and velocity measurements to validate because of the time efficiency and extrapolative capability of the boundaries of habitat polygons, will likely substantially such models (Bovee and Milhous 1978). increase the cost of DFAs or reduce the length of river that I feel that 2-D modeling offers significant benefits over can be assessed by DFAs, when compared to the methods DFA for adaptive management. The following circumstances presented in Railsback and Kadvany (2008). would require that a DFA be repeated, while 2-D modeling Fundamentally, there is a trade-off in costs for DFAs ver- could just be used as is with new sets of habitat suitability sus 2-D modeling. A much more extensive data collection criteria: effort is needed for 2-D modeling to accurately characterize the bed topography of a river reach, but DFAs require that 1. habitat needs to be evaluated for additional species, such data be collected multiple times for a given river reach, i.e., as because of an Endangered Species Act listing; at each flow to be assessed. In addition, for 2-D modeling, 2. New information becomes available on habitat require- bed topography and substrate and cover mapping data can ments, requiring a change in the habitat suitability crite- be collected at a very low flow, with the only data needed ria used to quantify habitat; or at high flow being water surface elevations at the up- and downstream ends of the site and flow, and edge velocities for 3. biological verification shows errors in the habitat suit- validation purposes. The cost of DFAs or length of river that ability criteria. can be assessed by DFAs is sensitive to the minimum polygon In contrast, situations where habitat requirements are poorly size used. The use of too large a polygon size can result in understood lend themselves well to the use of 2-D model- errors in mapping (Figure 6), while a small minimum poly- ing in an adaptive management framework. Preliminary cri- gon size could substantially increase the cost of a DFA or teria could be developed through a Delphi analysis (Bovee decrease the length of river assessed by a DFA, particularly 1986), and as additional information is developed on habitat for complex habitats. The minimum polygon size selected for requirements, resulting revised criteria could be used with use in a DFA should correspond to the relevant scale for the the 2-D modeling. Two-dimensional hydraulic and habitat smallest life stage/species assessed by a DFA, for example, on modeling can be used to simulate the amount of habitat for the order of 0.1 m2 for trout fry. One aspect of the minimum alternative designs prior to construction, a cost-effective polygon size that still needs to be defined is whether the rel- method to conduct adaptive management for stream restora- evant scale for adult trout would include an entire feeding tion projects (Gard 2006). In contrast, DFA, as an empirical station, comprised of both a slow velocity at the fish location method, cannot be used to quantify the amount of habitat and faster adjacent velocities for food delivery, in which case expected with alternative stream restoration project designs the minimum polygon size for adult trout (only including the prior to construction. DFA also lacks the ability to do sen- slow velocity at the fish location) would be smaller than 1 m2. sitivity analyses or construct uncertainty bounds related to Railsback and Kadvany (2008) note that assessment of habitat suitability criteria. habitat at flows other than the flows mapped during the DFA requires interpolation between, or extrapolation from, results CONCLUSIONS from the observed flows. I view this as a significant disadvan- tage of DFA, and note that two-dimensional hydraulic and Difficulties with the DFA method presented in Railsback habitat modeling does not require interpolation or extrapo- and Kadvany (2008), as applied on the Trinity River, include lation, because hydraulic parameters are quantified at each the lack of reproducibility, which was demonstrated by the flow to compute habitat. The flow dependence of DFA is disparity between replicate survey results. The large disparity also a practical limitation because there are many challenges between initial and repeated surveys limits the utility of this to gathering data, scheduling crews, and getting controlled method unless substantial improvements can be made. Given steady flows. Practitioners of DFA have found that this is the additional effort that is needed to obtain reproducible DFA a substantial challenge and constraint (Scott McBain, pers. results, 2-D modeling potentially does not require more effort comm. McBain and Trush, Inc., Arcata,CA); 2-D modeling than DFA, but yields more useful information. In addition, is able to avoid most of these challenges and constraints. 2-D modeling can allow for additional species/life stages to be Although errors associated with interpolation can be reduced considered after the fieldwork has been completed, allows for by mapping additional flows, albeit with an increased associ- more spatial based assessments of habitat conditions such as

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 327 distance to cover, and can provide a fine scale resolution over Ghanem, A., P. Steffler, F. Hicks, and C. Katopodis. 1996. a wide range of flows not feasible with DFAs. Two-dimensional hydraulic simulation of physical habitat conditions in flowing streams. Regulated Rivers: Research ACKNOWLEDGEMENTS and Management 12:185-200. Goodman, D. H., A. C. Martin, P. P. Petros, and J. Klochak. Many thanks to D. Goodman for providing copious infor- 2007. Judgement based habitat mapping on the Trinity River, mation on the Trinity River DFA evaluation and reviewing 2006. Draft version 11/15/2007. Report produced in collabo- the manuscript. I also thank S. Gallagher, P. Zedonis, D. ration with Hoopa Valley Tribe, Trinity River Restoration Program and the Yurok Tribal Fisheries Department. U. S. Threloff, D. Cox, R. Vadas, C. Chamberlain, S. McBain and Fish and Wildlife Service, Arcata, California. three anonymous reviewers who contributed insightful com- Hardin-Davis, Inc. 2003. Pit River 2-dimensional instream ments that greatly improved the manuscript. flow study draft final report. Prepared by Hardin-Davis, Inc., Corvallis, Oregon, for Pacific Gas and Electric Company, July 2003. REFERENCES Hardy, T. B., R. C. Addley, and E. Saraeva. 2006. Evaluation of interim instream flow needs in the Klamath River, phase II Ahmadi-Nedush, B., A. St-Hilaire, M. Bérubé, É. Robichaud, final report. Report prepared by Institute for Natural Systems N. Thiémonge, and B. Bobée. 2006. A review of statistical Engineering, Logan, Utah, for the U. S. Department of the methods for the evaluation of aquatic habitat suitability for Interior, 31 July 2006. instream flow assessment. River Research and Applications Hvidsten, H. B. 1993. High winter discharge after regulation 22:503-523. increases production of Atlantic salmon, Salmo salar, smolts Annear, T., and 15 others. 2004. Instream flows for riverine in the River Orkla, Norway. Pages 175-177 in R. J. Gibson resource stewardship, revised edition. Instream Flow Council, and R. E. Cutting, eds. Production of juvenile Atlantic Cheyenne, Wyoming. salmon, Salmo salar, in natural waters. Special Publication Arthington, A. H., and J. M. Zalucki. 1998. Comparative of the Canadian Journal of Fisheries and Aquatic Sciences evaluation of environmental flow assessment techniques: 118. review of methods. Land and Water Resources Research and Jeffres, C. A. 2006. Ephemeral floodplain habitats provide best Development Corporation, Canberra, Australia. growth conditions for juvenile Chinook salmon in a California Barinaga, M. 1996. A recipe for river recovery? Science river. M. S. Thesis. University of California, Davis. 273:1648-1650. Lind, A. J. 2004. Foothill yellow-legged frog (Rana boylii) envi- Bovee, K. D. 1986. Development and evaluation of habitat ronmental documentation for monitoring and risk analysis. suitability criteria for use in the Instream Flow Incremental USDA Forest Service, Pacific Southwest Research Station, Methodology. Instream Flow Information Paper 21. U. S. Redwood Sciences Laboratory, Arcata, California. Fish and Wildlife Service Biological Report 86(7). Lind, A. J., and R. A. Wilson. 1996. The effects of a dam on Bovee, K. D., and R. Milhous. 1978. Hydraulic simulation in breeding habitat and egg survival of the foothill yellow-legged instream flow studies: theory and techniques. Instream Flow frog (Rana boylii) in northwestern California. Herpetological Information Paper 5. U. S. Fish and Wildlife Service FWS/ Review 27:62-67. OBS-78/33. Manly, B. F. J., L. L. McDonald, D. L. Thomas, T. L. McDonald Bowen, Z. H., K. D. Bovee, and T. J. Waddle. 2003. Effects of and W. P. Erickson. 2002. Resource selection by animals: sta- flow regulation on shallow-water habitat dynamics and flood- tistical design and analysis for field studies, second edition. Kluwer Academic Publishers, Dordrecht, The Netherlands. plain connectivity. Transactions of the American Fisheries Society 132:809-823. McBain and Trush, Inc. 2004. Synthesis of geomorphic, veg- etation, and instream flow studies, Oak Grove Fork project, Cavallo, B. J. 1997. Floodplain habitat heterogeneity and the Oak Grove Fork and Clackamas River upstream from North distribution, abundance, and behavior of fishes and amphib- Fork Reservoir. Prepared by McBain and Trush, Inc., Arcata, ians in the Middle Fork Flathead River Basin, Montana. California, for Portland General Electric, 31 March 2004. M.Sc. thesis, University of Montana, Missoula. _____. 2007. Tuolumne River flow evaluation from CIFGS (Clackamas Instream Flow/Geomorphology Subgroup). O’Shaughnessy Dam to Early Intake. Proposed study plan 2003. Estimating salmonid habitat availability in the lower and methods. Prepared by McBain and Trush, Inc., Arcata, Oak Grove Fork using expert habitat mapping, summary of California, for San Francisco Public Utilities Commission, methods and preliminary results. Report prepared by McBain 12 July 2007. and Trush Inc., Arcata, California, for Clackamas Instream Pacific Gas and Electric Company. 2007. Pacific Gas and Flow/Geomorphology Subgroup, 5 March 2003. Electric Company McCloud-Pit hydroelectric project, FERC Crowder, D. W., and P. Diplas. 2000. Using two-dimensional project 2106. Licensee’s revised relicensing study plan. Pacific hydrodynamic models at scales of ecological importance Gas and Electric Company, San Francisco, California, 4 May Journal of Hydrology 230:172-191. 2007. ESRI (Environmental Systems Research Institute). 2006. Pasternack, G. B., C. L. Wang, and J. E. Merz. 2004. ArcGIS 9.2. ESRI, Redlands, California. Application of a 2D hydrodynamic model to design of reach- Gard, M. 2006. Changes in salmon spawning and rearing habi- scale spawning gravel replenishment on the Mokelumne tat associated with river channel restoration. International River, California. River Research and Applications 20:202- Journal of River Basin Management 4:201-211. 225.

328 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g Potter, I. C. 1980. Ecology of larval and metamorphosing lam- A. J. D. Ferguson, ed. The ecological basis for river manage- preys. Canadian Journal of Fisheries and Aquatic Sciences ment. John Wiley & Sons, Inc., New York. 37:1641-1657. USFWS (U. S. Fish and Wildlife Service). 2006a. Relationships Railsback, S. F., B. Harvey, S. K. Jackson, and R. H. between flow fluctuations and redd dewatering and juve- Lamberson. 2005. inSTREAM: the individual-based nile stranding for Chinook salmon and steelhead in the stream trout research and environmental assessment Sacramento River between Keswick Dam and Battle Creek. model. Humboldt State University and Redwood Sciences U.S. Fish and Wildlife Service, Sacramento, California. Laboratory, Arcata, California. _____. 2006b. Flow-habitat relationships for macroinverte- Railsback, S. F., and J. Kadvany. 2008. Demonstration flow brates in the Sacramento River between Keswick Dam and assessment: judgment and visual observation in instream Battle Creek. U.S. Fish and Wildlife Service, Sacramento, flow studies. Fisheries 33(5):217-227. California. Reese, D. A., and H. H. Welsh Jr. 1998. Habitat use by west- U.S. Fish and Wildlife Service and Hoopa Valley Tribe. 1999. ern pond turtles in the Trinity River, California. Journal of Trinity River flow evaluation final report. Prepared by U. Wildlife Management 62:842-853. S. Fish and Wildlife Service, Arcata, California, and Hoopa R2 Resource Consultants, Inc. 2003. Pit River habitat mapping: Valley Tribe, Hoopa, California for the Secretary of the U.S. results of the August 2002 demonstration flow study. Final Department of the Interior, June 1999. report prepared by R2 Resources Consultants, Inc, Redmond, Washington, for Pacific Gas and Electric Company, 5 May Wydoski, R. S., and R. R. Whitney. 2003. Inland fishes of 2003. Washington, 2nd ed., rev. and expanded edition. University Snider, B., K. Vyverberg and S. Whiteman. 1996. Chinook of Washington Press, Seattle. salmon redd survey lower American River fall 1994. Young, W. J., B. C. Chessman, W. D. Erskine, T. A. Raadik, California Department of Fish and Game, Environmental D. J. Wimbush, J. Tilleard, A. J. Jakeman, I. Varley, and Services Division, Stream Flow and Habitat Evaluation T. J. Verhoeven. 2004. Improving expert panel assessments Program, Sacramento. through the use of a composite river condition index—the Swales, S., and J. H. Harris. 1995. The expert panel assessment case of the rivers affected by the Snowy Mountains hydro- method (EPAM): a new tool for determining environmental electric scheme, Australia. River Research and Applications flows in regulated rivers. Pages 125-134 in D. M. Harper, and 20:733-750.

Halltech

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 329 Feature: Fish Habitat

E. Ashley Steel, Paul McElhany, Naomi J. Yoder, Michael D. Purser, Making the Best Use of Modeled Data: Kevin Malone, Brad E. Thompson, Multiple Approaches to Sensitivity Karen A. Avery, David Jensen, Greg Blair, Craig Busack, Mark D. Bowen, Analysis of a Fish-Habitat Model Joel Hubble, and Tom Kantz

ABSTRACT: Fisheries management has become increasingly Steel is the team lead for the Landscape Ecology dependent on large and complex models; however, tools and and Recovery Science Team at NOAA’s Northwest technologies for model evaluation have lagged behind model Fisheries Science Center (NWFSC) in Seattle development and application. Sensitivity analyses can test the degree and can be contacted at Ashley.Steel@noaa. to which particular model inputs or internal parameters affect model gov. McElhany is a research ecologist in the outputs and are recommended in model construction, calibration, Conservation Biology Division, NWFSC. Yoder and assessment. We describe three parallel sensitivity analyses of the Ecosystem Diagnosis and Treatment (EDT) model and draw combined completed the project as a fisheries biologist conclusions. The details of how each agency conducted and utilized with the U.S. Bureau of Reclamation in Denver, sensitivity analyses are outlined and the trade-offs between simpler and Colorado. Purser is a senior scientist with more intensive sensitivity analyses are described. Combined insights Snohomish County in Everett, Washington. Malone on the EDT model include identification of input parameters to which is an independent fisheries consultant in Port the model is surprisingly insensitive and quantification of prediction Orchard, Washington. Thompson completed the intervals. We conclude that known uncertainties in input data and project as a research scientist with the Washington internal parameters lead to large prediction intervals around estimates Department of Fish and Wildlife in Olympia and is of population abundance and productivity but that the identification of now the fisheries division manager for the U.S. Fish high priority reaches for restoration and protection is relatively robust and Wildlife Service in Lacey, Washington. Avery to these uncertainties. We recommend that sensitivity analyses are is a fisheries biologist in the Conservation Biology applied to all models used in fisheries management. Division, NWFSC. Jensen is a private statistical consultant with Jensen Consulting in Eugene, Oregon. Blair is a fisheries consultant with ICF Haciendo mejor uso de los datos: Jones & Stokes in Vashon, Washington. Busack is the chief fish scientist for Washington Department múltiples enfoques al análisis de of Fish and Wildlife in Olympia. Bowen is a fisheries sensibilidad de un modelo biologist with the U.S. Bureau of Reclamation in Denver, Colorado. Hubble is a fisheries biologist de hábitat en peces with the U.S. Bureau of Reclamation in Yakima, Washington. Kantz is special project biologist with Resumen: El manejo de pesquerías se ha vuelto cada vez más Pierce County in Tacoma, Washington. dependiente de modelos grandes y complejos; sin embargo, las herramientas y tecnologías para evaluarlos se han mantenido rezagados con respecto a su desarrollo y aplicación. Los análisis de sensibilidad, Introduction por un lado, pueden probar a qué grado los parámetros o entradas de un modelo en particular afectan las salidas y, por otro, se recomienda Fisheries management has become increasingly aplicarlos durante la construcción, calibración y evaluación del dependent on large and complex models. Such mod- modelo. Aquí se describen en paralelo tres análisis de sensibilidad els are developed to help natural resource managers sobre el modelo de Tratamiento y Diagnóstico de Ecosistemas (TDE) address complex issues by providing the estimates y se llega a conclusiones comunes. Se destacan los detalles de cómo las of ecosystem parameters or biological response that agencias condujeron los análisis de sensibilidad y se describe el balance are necessary for making fisheries and habitat deci- entre sus formas simples y complejas. Una introspección combinada sions. The challenge for scientists and managers is to de estos análisis aplicados al TDE incluye la identificación de aquellos parámetros de entrada a los que el modelo es particularmente insensible develop models that enable informed and transparent y la cuantificación de intervalos de predicción. Se concluye que tanto decisions based on available scientific information, incertidumbre conocida en los datos de entrada como los parámetros which is generally imperfect. The EcoPath/EcoSim internos dan como resultado amplios intervalos de predicción con framework, for example, has been used to study and respecto a las estimaciones de abundancia poblacional y productividad, understand foodwebs in hundreds of marine systems pero la priorización de la restauración y protección es relativamente (Pauly et al. 2000). ALFISH is a spatially explicit, robusta a dicha incertidumbre. Se recomienda que los análisis de age-structured model to explore fish density dynam- sensibilidad se apliquen a todos los modelos de manejo de pesquerías. ics that has been applied in the Everglades (Gaff et

330 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g al. 2004). The Atlantis modeling framework, developed in and make recommendations about the use of sensitivity anal- Australia, has been employed to support integrated oceans yses for large and complex models. management (Smith et al. 2007). These models are diverse but share a dependence on large numbers of parameters with Sensitivity Analysis complex interactions. Advances in computing and mathematics have enabled Sensitivity analyses test the ability of inputs (data input such models to grow increasingly complex; however, tools into the model such as empirical observations), parameters (estimated relationships within the model), and model com- and technologies for model evaluation have lagged behind ponents (smaller sub-models or sets of parameters within the model development and application. Sensitivity analyses pro- larger model) to affect model outputs and are a standard tech- vide objective criteria with which to evaluate model output, nique used in model construction, calibration, and assessment allowing users to explore how uncertainties in inputs and in (Saltelli et al. 2000a). Sensitivity analyses can aid decision- parameter values propagate through the model. The results makers by highlighting those inputs and components with the of sensitivity analyses can be used to provide more detailed largest influence on the outputs.T hey are recommended when model outputs, e.g., confidence intervals or precision esti- model output is used for decision making (Haefner 2005; ISAB mates, to refine the model structure by removal of unnecessary 2001) and often provide insights into how the model arrived at parameters, and to improve the way modeled results are used a particular prediction and into potential biases in predictions. to make decisions. A further goal of sensitivity analyses is often to reduce the We employed a variety of sensitivity analysis techniques uncertainty in model output. In this case, a sensitivity analysis in the evaluation of a complex model used throughout the can identify those input factors or model parameters that, if Pacific Northwest for salmon recovery planning, Ecosystem measured more precisely, would provide the greatest reduction Diagnosis and Treatment (EDT). Sensitivity analyses were in model output uncertainty. completed by three public agencies and provide an ideal A useful byproduct of sensitivity analyses can be the esti- opportunity for exploring the appropriate use of modeled mation of the distribution of modeled outputs, often estimated predictions in salmon recovery and freshwater restoration through Monte Carlo simulations, which would result from planning, in specific, and in environmental management in incorporation of particular uncertainties. In this article, we general. In this article, we explain sensitivity analyses and refer to the percentile bounds, e.g., 90%, of that output distri- describe the EDT model before presenting three different bution as a prediction interval; these bounds describe the range sensitivity analyses of the EDT model (Table 1). In conclu- of predictions that the model produces when a set of known sion, we make explicit the trade-offs between these three uncertainties is incorporated. As a practical matter, using a types of sensitivity analysis, synthesize findings about the distribution of inputs rather than a single-point estimate illus- EDT model, draw conclusions about multi-agency analyses, trates that a given model will produce a range of predictions

Table 1. Comparison of EDT sensitivity analyses completed by three agencies.

Agency U.S. Bureau of Reclamation Washington Department of Fish and Wildlife NOAA Fisheries

Stream attributes, habitat survival rules, benchmarks, adult parameters, Parameters varied Stream attributes Stream attributes all other internal parameters, Region (species) Naches River (summer steelhead); All Puget Sound watersheds East Fork Lewis River (fall Chinook), Naches river (spring Chinook); Snoqualmie River (fall Chinook), American River (spring Chinook) Germany Creek (coho), and Washougal River (winter steelhead)

All parameters Scale of sensitivity One at a time simultaneously; groups of All parameters simultaneously; analysis related parameters groups of related parameters

Type of alternate Systematic Monte Carlo Monte Carlo parameter value selection

Hydrologic regime natural, Hydrologic regime regulated, Habitat inputs, Least sensitive parameters all flow attributes, Macroinvertebrates internal habitat survival Habitat-off-channel relationship parameters

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 331 when all feasible inputs are considered. When modeled output one parameter or in any one input impacts model output but is then used to estimate the effects of an on-the-ground action they ignore interactions among parameters. Often the results of or group of actions, a decision-maker can make a better deci- an OAT sensitivity analysis can be calculated analytically but sion by considering the full range of plausible model outputs. computer simulations may be necessary for complex models. Consider an example application. In the Puget Sound Global analyses generally are more difficult to produce; they Salmon Recovery Plan Executive Summary (Shared Strategy evaluate output uncertainty when all (or many) input factors Development Committee 2007), EDT was used to predict are altered simultaneously. Variance partitioning methods take that dam removal will improve North Fork Nooksack River this a step further, producing an estimate of the proportion of Chinook salmon (Oncorhynchus tshawytscha) abundance by the output variance attributable to each input factor. These 30.8%. Managers tasked with deciding whether to invest in methods are computationally demanding for complex models dam removal or other suites of habitat restoration actions (Saltelli et al. 2000a) and have rarely, if ever, been applied to would benefit from knowing (a) the range of modeled abun- ecological models used in a management context. dances given uncertainties in user input and internal model Note that sensitivity analyses, in general, cannot estimate parameters, (b) whether and how often the model would pre- model accuracy (the distance between modeled output and dict a population decline post-dam-removal if uncertainty in the truth) but, rather, model precision (similarity of repeated all model parameters were considered, and (c) the likelihood model runs given a set of uncertainties which might include that the model might have predicted an even more dramatic model structure, parameters, or inputs). In this article, we improvement in fish population performance. For example, will report on three different and independent OAT analyses should the value of 30.8% be interpreted to mean a range from completed by the U.S. Department of the Interior/Bureau of 30.1 to 40.2% or from –10 to 70%? Reclamation (Reclamation), the Washington Department of There are a variety of methods for sensitivity analysis Fish and Wildlife, and the National Oceanic and Atmospheric (Saltelli et al. 2000a). Local analyses, which alter factors “one Administration Fisheries Service (NOAA Fisheries) as well as at a time” (OAT), are the simplest to produce because all global analyses completed by NOAA Fisheries. Because of the parameters, except those specifically being evaluated, are held complexity of the EDT model, no agency evaluated whether constant. OAT analyses can determine how uncertainty in any the overall model structure was realistic nor did any agency

Figure 1. Results of the variance partitioning for the East Fork Lewis River, Washington, fall Chinook salmon abundance (without harvest). The size of each pie wedge represents the proportion of total model output uncertainty (total effects) attributable to that group of parameters. Parameter groupings identified in white are internal parameters that do not vary from basin to basin. Solid shading indicates parameter groupings which model users can modify when running the model in a particular basin and for a particular population.

Trajectory Seed 4% Habitat Integration Type Rules 3% 10% 5% Benchmark Population Description 15% 0% Food Multiplier Juvenile Age 1% 2% Pattern Life stage 1% Life History 0% Monthly Pattern Data 1%

Reach Data Global 14%

Adult Age 44%

332 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g evaluate the sensitivity of modeled output to variations in ties. EDT is predominantly a deterministic model. There is no model structure. uncertainty in the relationships between fish habitat and fish production nor in population descriptors such as migration Ecosystem Diagnosis and Treatment timing, fecundity, and ocean survival. An electronic library of EDT documentation is available from ICF Jones & Stokes EDT is a system for rating the quality, quantity, and diver- (http://www.jonesandstokes.com/index.php?option=com_co sity of stream habitat, relative to the species-specific needs ntent&task=view&id=488&Itemid=784). of salmon (Lichatowich and Mobrand 1995; Mobrand et al. 1997; Blair et al. in press). EDT has been applied in nearly Testing the extremes every salmon-bearing watershed in Washington and many in Oregon. In Puget Sound and in the Columbia Basin, the model Bureau of Reclamation biologists conducted a sensitivity has been widely used to assist in setting recovery goals (e.g., analysis of the EDT model to determine how the EDT equilib- Lower Columbia Fish Recovery Board 2004; Northwest Power rium abundance output for Chinook salmon and for steelhead and Conservation Council 2005) for threatened and endan- (O. mykiss) varies as a function of stream attributes (Yoder gered salmonids and to predict consequences of proposed et al. 2006). EDT was one model in a chain of hydrological, management actions (e.g., Shared Strategy Development physical, and biological models employed by Reclamation to Committee 2007). The EDT model compares estimated fish evaluate water storage and flow management options in the performance (e.g., population size or productivity) among cur- Yakima River, Washington. By exploring a wide range of input rent, historical, and potential habitat conditions. To do this, values (that included both plausible and implausible values the EDT model requires thousands of data inputs and contains for flow and habitat alterations), Reclamation analyses pro- well over 1,000 internal parameters. Data inputs refer to those vided guidance for managers in how to use EDT output more parameters which model users can modify when running the effectively and in defining priorities for future data collec- model in a particular basin and for a particular population. tion. Reclamation expected that productivity and abundance Internal parameters are those that do not vary from basin to estimates would be sensitive to changes in stream attribute basin (Figure 1). As described in the introduction, the EDT inputs; they aimed to discern the set of stream attributes to model is not unique in its complexity or size; we report the which the abundance and productivity estimates were most results of sensitivity analyses on the EDT model as a case study and least sensitive. of how to conduct and interpret sensitivity analyses of large Systematically varying model input values over a range ecosystem models as well as to provide specific guidance on and observing the amount of change in model output from using EDT output in management decisions. a given baseline is one form of model sensitivity analysis The EDT model consists of four major components (Blair (Haefner 2005). Reclamation analyses systematically varied et al. in press): EDT habitat inputs (e.g., maximum temperature, bed scour, etc.) for each river reach. These habitat attributes are typi- 1. a reach level stream and environmental description. cally input into EDT on a 0–4 scale. Reclamation started The reach level environmental description is based on 46 with the current set of attribute inputs and varied each attri- stream attributes (e.g., flow, stream temperature, macroin- bute input, independently, by plus or minus 1, and then reran vertebrate abundance, and habitat types). Users are also the model. They then varied each attribute input by plus and asked to enter a “level of proof” rating for each stream minus 2, 3 and 4, working within the limits of possible input attribute which describes the general method used to values, which are 0–4 for most habitat attributes. Varying the estimate that parameter (e.g., empirical data or expert input by ±4 quantified the sensitivity of the model to extreme opinion). changes in input attributes. Non-focal habitat attributes and 2. a set of “rules” that relate the environmental condi- all other parameters were held at their point estimates (Yoder tion to life stage survival and capacity (as defined for et al. 2006). the Beverton-Holt equation). The species-specific rules Sensitivity ratings for most habitat inputs were calculated are rating curves for each salmonid species that link the using the following equation (after Haefner 2005): stream attributes to life stage survival or density. The  (A0 − A1 ) habitat rules reduce the predicted species-specific survival   and maximum density to reflect local conditions in the  A1  stream. S = Pavg0 − Pavg1 3. biological data on target species such as adult and juvenile where: age structure, run timing, sex ratio, fecundity, and ocean

survival. A0 is the original EDT output point estimate; 4. integration of the reach by life stage estimates to the A1 is the new EDT output point estimate given a change population level based on a disaggregated Beverton-Holt in attribute value; function (Beverton and Holt 1957; Moussalli and Hilborn 1986). Pavg0 is the mean initial input attribute value over all reaches; The outputs of an EDT analysis are the productivity and capacity parameters of a Beverton-Holt function, equilibrium Pavg1 is the mean changed input attribute value over abundance, and reach-level restoration and protection priori- all reaches;

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 333 empirical data exists. One can then expect that the precision A similar equation was used for attribute responses on differ- of modeled output will also be basin-specific. ent numerical scales. This concern led WDFW to initiate a Monte Carlo simu- lation study to develop basin-specific estimates of model pre- Results: Using sensitivity analysis results to inform cision based on the quality of the stream input data (Busack decision-making and Thompson 2006). To identify the plausible range of stream attribute values, WDFW developed a survey and que- Reclamation found that model outputs for both steelhead ried eight biologists with EDT experience to elicit ranges of and Chinook salmon populations in the Yakima River were plausible stream attribute scores based on both the reported most sensitive to variations in the following stream inputs: stream attribute score and level of proof rating. For example, alkalinity, embeddedness, fine sediment, habitat backwater the survey asked “If a reach had a score of 2 for turbidity, with pools, habitat primary pools, miscellaneous toxins, temper- a level of proof of 3, what do you think the highest and lowest ature-daily maximum, temperature-daily minimum, and tur- likely true value for turbidity might be?” WDFW then created bidity. Reclamation also identified stream inputs to which triangular probability distributions, using the reported value model results for all three focal populations in these river as the mid-point and the survey’s average range as the end- segments were relatively insensitive (Table 1), including a points, for each combination of attribute, level of proof, and large number of flow parameters. Because of the non-linear stream attribute score. mathematics within the model, sensitivity analyses were nec- The probability distributions were sampled in two ways. essary to understand these relationships. In the first analysis, all stream attribute parameters were The relative insensitivity of EDT predictions of fish abun- varied simultaneously and in the second analysis, each dance to flow attributes provided valuable information about of nine stream attribute parameter groupings were varied EDT usage in the evaluation of a proposed water storage and sequentially. In each analysis, 500 independent input stream flow management project on the Yakima River. The rela- attribute data sets were created for each of 10 Puget Sound tive insensitivity of EDT to direct flow attributes reflects the watersheds supporting 15 Chinook salmon populations. Each model assumption that flow has a small physiological effect on of these 100,000 input data sets (2 analyses X 10 watersheds fish survival and that impacts of flow manifest themselves in X 10 parameter groups X 500 iterations) represent plausible other physical attributes. One of the most obvious potential stream attribute input data sets, given the stream attribute impacts of the proposed water storage project would be the values and levels of proof recorded in the original EDT runs change in flow volumes and flow patterns. However, because for reach watershed (Busack and Thompson 2006). flow effects almost all other physical stream attributes, the impact of a real-world change in flow on habitat conditions Results: Using sensitivity analysis results to inform (e.g., stream wetted area, bed scour, embeddedness, and hab- decision-making itat-type distribution) would need to be determined indepen- dently and entered into the model as changes to non-flow Using the first WDFW approach, prediction intervals attributes. Although Reclamation’s conclusions about the flow (range of 95% of Monte Carlo outputs) for Puget Sound attributes are applicable only to the Yakima basin, these con- Chinook salmon populations generally had coefficients of clusions are supported by research in other basins (Northwest variation of 3.5%, 7%, and 10.5% for productivity, capacity, Hydraulic Consultants 2005). Subsequently, Reclamation and equilibrium abundance estimates respectively. WDFW developed support models to calculate the values for those concluded that, given the plausible range of stream-attribute EDT inputs (i.e., habitat and temperature attributes) most inputs, variability in EDT output was generally low, although influenced by changes in stream flow. Furthermore, based on model results with larger prediction intervals were observed these findings, Reclamation facilitated the development of for some basins or populations (Busack and Thompson 2006). a customized version of EDT to address the major ways that These results suggested that uncertainty in stream-attribute flow regimes can be altered by land use practices (Lestelle et factors alone does not produce large uncertainties in model al. 2006). output. One conclusion from this result is that small errors in stream-attribute inputs will not propagate into large errors Surveying users to estimate uncertainty in habitat ratings in model output, so long as model structure and other inputs such as fecundity, smolt-to-adult survival rates, etc., remain Washington Department of Fish and Wildlife (WDFW) constant. Another possible explanation for these results is is responsible for co-managing 14 evolutionarily significant that some relationships between habitat inputs and popula- units (ESUs) of salmon including over 100 individual popula- tion outputs are not captured in the model. tions. WDFW has used EDT extensively to assist in the devel- Using the second approach, in which they varied subsets opment of recovery goals and plans throughout Washington. of habitat parameters sequentially, WDFW identified a set Because EDT relies on detailed stream habitat input, there is of stream attributes which, across multiple watersheds, pro- concern that uncertainty in these stream habitat input vari- vided the largest contribution to the variance of model out- ables might lead to large amounts of uncertainty in model put for Puget Sound Chinook salmon populations. Generally, output. The accuracy and precision of stream habitat data Chinook salmon capacity estimates were most sensitive to vary widely from stream attribute to stream attribute and uncertainty in ratings for the habitat type group (e.g., relative from basin to basin depending on data collection methods amounts of primary pool, tailout, small cobble riffle) while and on the means of estimating model input values where no productivity estimates were most sensitive to uncertainty in

334 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g channel stability and sediment ratings. The precision of model and reach-specific restoration and prioritization actions are estimates could best be improved by increasing accuracy and often used by state, tribal, and local groups in the processes of precision of attributes in these groups. Of these, only sedi- developing recovery plans and prioritized lists of restoration ment was also identified in the Reclamation analysis (Yoder and protection actions. Therefore, NOAA needs the best et al. 2006). The one stream attribute that consistently had possible understanding of the precision and accuracy of EDT little or no impact on model output was macroinvertebrate output in order to evaluate whether model output is appropri- diversity (Busack and Thompson 2006). Therefore, if the ately used to inform these decisions. management objective is to increase precision in EDT model The goals of the NOAA analyses were to quantify a more outputs and habitat-sampling funds are limited, WDFW rec- complete prediction interval, by incorporating uncertainty in ommended not increasing the sampling intensity for macro- non-habitat inputs and in internal parameters, and to iden- invertebrates, an expensive and time-consuming effort. Of tify which input parameters are responsible for the major- course, there may be many important reasons to sample mac- ity of output variability (McElhany et al. in press). NOAA roinvertebrates other than to improve EDT precision, as we Fisheries designed a global sensitivity analysis (Sobol 1993; have strong evidence from other sources that macroinverte- Saltelli et al. 2000a, 2000b) that incorporated known or esti- brates are a good indicator of salmon habitat quality (Quinn mated uncertainty on nearly every parameter in the EDT 2005). model. In the first step, a potential distribution was defined WDFW was also able to investigate the sensitivity of for each input and internal model parameter. Distributions for rankings of reaches for restoration or protection to habitat stream attributes were defined using WDFW’s survey results inputs. For Puget Sound watersheds, the top restoration and (Busack and Thompson 2006). Distributions for other param- protection rankings were relatively stable within basins. In eters were defined during collaborative meetings (including particular, uncertainty in stream attribute input data caused scientists from Snohomish County, Pierce County, NOAA little shift in the predictions of where the best habitat resto- Fisheries, WDFW, Reclamation, and Mobrand, Jones & ration opportunities exist for Chinook salmon within each Stokes) at which measurement errors, empirical data, simu- watershed. They concluded that managers could be confident lation studies, and expert opinion were considered. To esti- that modeled high priority reaches were consistent despite mate one prediction interval, the model was run at least 250 uncertainties in stream attribute input data. Restoration and times. For each parameter in each run, a value was randomly protection rankings for reaches originally ranked as fourth and independently selected from the distributions defined or lower, however, were not as robust to the uncertainties above. Using this approach, all possible combinations of in stream attribute inputs (Busack and Thompson 2006). parameters, both internal and user-defined, had a possibility Therefore, WDFW concluded that rankings for these mid- of being included in the simulation study (McElhany et al. priority reaches should be used with caution in decision- making.

BEYOND Figure 2. Histogram of the 894 successful Monte Carlo simulations of abundance (Neq) for the East Fork Lewis fall HABITAT: Chinook salmon population, calculated without harvest. The solid bars represent abundance estimates of 2000 or greater. Note that 1,000 Monte Carlo simulations were attempted but 106 resulted in unusable parameter combinations. SENSITIVITY ANALYSES OF ALL PARAMETERS 200 NOAA Fisheries is 180 responsible for over- seeing the develop- 160 ment of recovery plans for most marine and 140 anadromous species 120 listed as threatened or endangered under the 100 Endangered Species 80 Act (ESA). NOAA Fisheries also contrib- 60 utes information to 40 the management of Number of Simulations freshwater and estua- 20 rine habitats and to the prioritization of 0 freshwater and estua- 0 500

rine restoration and 1000 1500 2000 2750 3500 4000 4500 5000 5500 6000 6500 7000 protection actions. Population parameters Abundance modeled using EDT

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 335 in press). The resulting prediction interval therefore reflects relationship as expected, or another factor is limiting model the distribution of model outputs given known and estimated output and until this bottleneck is removed the parameter of uncertainties in the stream attributes (from WDFW), the interest can have no effect. The variance partitioning pro- population descriptors, and the internal model parameters. vided initially unexpected results. Much of the emphasis of In conducting the global sensitivity analysis, NOAA EDT critique and evaluation has been on the use of expert Fisheries was also able to estimate the effect of uncertainty in opinion to generate stream attributes and on the formulation any one parameter, or group of parameters, on the final pre- of the rules or equations linking stream attributes to survival diction distribution. The total prediction interval was par- (ISAB 2001). It might have been expected that the uncer- titioned using an approach that is analogous to an analysis tainty in stream attributes would have a relatively large effect of variance (ANOVA) to identify the contribution of each on the size of the final prediction interval. As shown in an parameter or set of parameters to the total output uncertainty example using East Fork Lewis River fall Chinook salmon (Sobol 1993; Satelli et al. 2000a). Because of the large number (Figure 1), uncertainty in stream attribute inputs had a small of parameters, NOAA applied a hierarchical approach, vary- relative impact on the size of the prediction interval. Other ing sets of parameters (e.g., all stream attributes) together. parameter groups, such as the adult parameters (e.g., fecun- Once the most sensitive groups of parameters were identified, dity, ocean survival) and the benchmarks, contributed most the analysis focused on determining which specific param- to the variability in the model prediction. eters in these sensitive groups contributed most to prediction Variance partitioning results were consistent across all uncertainty (McElhany et al. in press). three basins for which NOAA ran the sensitivity analysis (McElhany et al. in press). Although there is variation among Results: Using sensitivity analysis results to inform populations in which parameter groups have the greatest decision-making sensitivities, stream attributes have consistently lower sensi- tivities than other groups of parameters (McElhany et al. in NOAA created prediction intervals for multiple popu- press). This should not be interpreted to mean that instream lations. For example, the 80% prediction interval for fall habitat is not important to salmon or that stream attribute Chinook salmon abundance (equilibrium abundance of a parameters do not affect EDT results. The importance of Beverton-Holt model) under current conditions without the adult parameter group should not be surprising as adult harvest in the East Fork Lewis River, Washington, using the parameters include ocean survival, fecundity, sex ratio, and global sensitivity analysis, ranged from 35 to 2,274 fish with age structure. This group of parameters defines the reproduc- a mean predicted abundance of 941 (median = 635). In other tive potential of the species and needs to be captured with words, given uncertainties in the input data and internal high precision if estimates of fish production from any model parameters, model output for this scenario was between about are to reflect observed data. 35 and 2,274 fish in 80% of EDT simulations (McElhany et al. in press). Conclusions: In addition to providing a more informed estimate of the Synthesizing multiple approaches mean model prediction, the analysis also provides managers with a distribution of model predictions (Figure 2). If one needed to know whether the population prediction exceeds, Trade-offs between types of sensitivity analyses for example, 2,000 fish, inferences based on the mean esti- mate would suggest not. In contrast, inferences using the pre- Each of the sensitivity analyses described above took a dif- diction intervals indicate that, given the input data, model ferent mathematical approach to questions about the impacts parameters, and the uncertainty distributions, there is a 17% of model inputs on model outputs and about model precision. chance that EDT will produce an estimate greater than 2,000 The Reclamation approach, simply varying an input and fish (Figure 2). In this way, prediction distributions for popu- observing the response in model output, provided invaluable lation abundance, productivity, and capacity can provide insights to the first question. As a result of their sensitivity increased information for model-supported decision-making analyses, they were able to avoid misuse of the EDT model, (Steel et al. in press). in isolation, for predicting the fisheries response to a range Assessments of the reach priorities for restoration and of proposed management actions. Simple sensitivity analyses protection suggest that these may also shift as uncertainty is such as these have limitations. They cannot identify interac- incorporated into the modeling process but that the suite of tions between uncertainties in multiple inputs or between high priority restoration reaches is relatively robust to uncer- inputs and internal parameters. As well, they provide only tainties in inputs (McElhany et al. in press). These results crude estimates of model precision, a range rather than a dis- are consistent with those of WDFW (Busack and Thompson tribution of potential outputs. Reclamation’s systematic sen- 2006). sitivity analysis represent a large amount of work because of Like the other agencies, NOAA was also able to identify the large number of inputs considered. However, often these groups of parameters to which the model appears less sensi- systematic variations in inputs and internal parameters can tive, at least for the basins examined (Table 1). When a model be conducted with a minimum of computer programming or is not sensitive to variation in particular parameters, it may specialized software. Failure to conduct such simple sensitiv- be because those parameters are not important for estimat- ity analyses in model development and before using modeled ing the output of interest, the parameter is already estimated predictions in management decisions severely reduces the with a high degree of precision, the model is not capturing a value of modeling for decision making.

336 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g Using a Monte Carlo approach to sensitivity analysis adds Value of multi-agencies collaborations two levels of complexity to the sensitivity analysis but pro- vides more robust information. The first added complexity The three analyses described here were conducted in par- is the necessity of quantifying not only the range but also allel by three public agencies with differing goals and respon- sibilities. Clear communication between agencies about the distribution of possible values for the inputs and inter- how the EDT model would be used, about how sensitivity nal parameters of interest. WDFW used an opinion survey analyses were conducted, and about interpretation of sensi- to estimate the range of plausible values around each esti- tivity analyses has greatly improved each individual analy- mated stream attribute. In the future, empirical data from sis. For example, the NOAA analyses were able to build on pilot studies or other regions could also be used to estimate WDFW’s opinion survey of habitat input distributions. And, these input distributions. The second complexity is simply sharing of information was required to come to a consensus the Monte Carlo sampling routine that requires programming about the appropriate distribution of plausible parameters or software support. The added value of the Monte Carlo for the remaining inputs and internal parameters. Since the final value of sensitivity analyses depends on agreement of approach is the quantification of a distribution of plausible these distributions, such consensus building was essential. outputs that can be used to quantify particular risks of inter- For example, if one agency felt that the uncertainty around est (Figure 2). parameter X varied between 5 and 20 while another agency Variance partitioning is particularly important for complex believed that the parameter uncertainty only varied between models because it quantifies not only independent impacts 16.5 and 17.5, it would be difficult to design one sensitivity of particular parameters or groups of parameters but inter- analysis that would be considered valid by both agencies. In actions between parameters. Variance partitioning allowed addition, these complex collaborations enabled all available NOAA to compare the total impacts of suites of parameters empirical data to be brought to bear on the discussion. on model outputs and to identify those parameters or groups Each agency has a different mandate for natural resource management and therefore, different objectives for sensitivity of parameters whose uncertainty had the greatest impact analysis. WDFW, for example, calculated customized predic- on model output. Variance partitioning methods require tion intervals for every watershed in Puget Sound for which customized programming and more advanced mathematics EDT analyses have been completed (Busack and Thompson (Sobol 1993). Their value in complex ecological models is 2006). They are now able to provide guidance on the best just beginning to be explored. opportunities to reduce model output variance in each Puget

Floy Tag

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 337 Sound watershed, if additional funding were available for conclusions, it suggests that that conclusion is robust to the habitat research. By working together, each agency was also architecture and assumptions of any one model. Where com- able to benefit from the combined results of the three sen- peting models are unavailable, sensitivity analyses can suggest sitivity analyses and to synthesize the results into one set of whether model output is robust to particular assumptions. conclusions about appropriate uses of the EDT model. Third, where possible, models must be evaluated with respect to both precision and accuracy. It is frustrating that Using sensitivity analyses to draw conclusions about EDT years of analyses on one model can only produce information about model precision yet this is a common situation. Models By combining results from the three sensitivity analyses, are developed to assist decision-makers in exactly the sort of we are able to draw several robust conclusions about the EDT situations for which data to quantify predictive accuracy are model given known and unavoidable uncertainties. First, the unavailable. In these circumstances, it is important to main- model is less sensitive to uncertainties in stream attributes tain skepticism until empirical data can be collected. than might have been expected. Investments in increased Finally, models should not be used in management deci- data collection on habitat may not yield increased preci- sions without at least some sensitivity analyses. Sensitivity sion in EDT output. Using the EDT model in isolation to analyses provide essential information when using modeled predict the effects of some environmental changes, such as predictions in a decision-making context. They enable an impacts of flow alterations, can provide misleading results, assessment of the probability of making a wrong conclusion especially where strong interactions with other stream attri- as a result of uncertainties in the data used to run the model butes are expected and not explicitly included in the model. and of uncertainties in the parameters used to build the Sensitivity analyses have guided the development of new model. Sensitivity analyses can be used to focus research and models in the Yakima Basin and can continue to guide the data collection efforts on parameters that will result in bet- collection of data to improve model precision. Second, the ter decision-making and, hopefully, increased cost efficiency model is most sensitive to parameters that describe adult pop- and improved fish populations. Sensitivity analyses are par- ulations and reproductive potential. The NOAA sensitivity ticularly important for complex models where the relation- analyses suggest that most of the uncertainty in model predic- ships between inputs and outputs are not transparent and for tions of salmon abundance and productivity does not derive models that rely on inputs or internal parameters that are from uncertainty about habitat condition but from uncer- tainty in the other parameters—even knowing stream attri- known to be uncertain. Models are often also used to pro- butes perfectly, most of the model uncertainty would remain. vide evidence for a specific thesis or as a rationale for specific Research can provide refined estimates of adult population actions. In these cases, a sensitivity analysis serves to provide parameters or rules linking specific stream attributes to sur- transparency to the analysis and can be included as an impor- vival that would improve model precision. Third, as expected tant element of evidence building (Saltelli et al. 2000b). As for a complex ecosystem model, prediction intervals for the increasingly complex management decisions demand increas- EDT-based estimates of abundance, productivity, and capac- ingly complex models, sensitivity analyses become essential ity are large. The size of the prediction intervals suggests that tools in appropriately using modeled predictions in decision- decisions based on point estimates for these fish performance making. metrics alone are risky. Fourth, the highest priority reaches for restoration and protection are relatively robust to known Acknowledgements parameter uncertainties We would like to acknowledge the support of the Northwest Implications for the use of models in fisheries management Fisheries Science Center information technology depart- ment in conducting these analyses. Bruce Watson and Jesse The use of models in fisheries management has a long his- Schwartz at ICF Jones & Stokes provided reviews of early tory and models will continue to play an integral role in fish- manuscript drafts. Lars Mobrand at ICF Jones & Stokes pro- eries and fish habitat management for the foreseeable future. vided insight and expertise that contributed greatly to these The combination of these three sensitivity analyses and an analyses. George Pess and Mike Ford of NOAA Fisheries investigation of the implications of their combined results provided constructive reviews as did the Science Editor. provide an opportunity for general guidelines on the use of We also thank Katherine Zehfuss of Science Applications models in fisheries management. International Corporation for her contributions to the First, managers can improve their use of modeled data by Reclamation analysis. requesting prediction intervals (and if possible confidence intervals). Distributions of model output provide increased References information on the risks of achieving or failing to achieve particular thresholds. Confidence or prediction intervals are Beverton, R. J. H., and S. J. Holt. 1957. On the dynamics of particularly important where modeled output is passed on as exploited fish populations. Fishery Investigations 2: 1-533. input to another model. Blair, G. R., L. C. Lestelle, and L. E. Mobrand. In press. The Second, since most models provide predictions that are Ecosystem Diagnosis and Treatment Model: a tool for assess- imperfect, the use of multiple independent models can ing salmonid performance potential based on habitat con- provide a stronger basis on which to base decisions (ISAB ditions. In E. Knudson and H. Michael, eds. Pacific salmon 2001). When multiple independent models provide similar environment and life history models: advancing science for

338 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g sustainable salmon in the future. American Fisheries Society, Saltelli, A., K. Chan, and E. M. Scott (editors). 2000a. Bethesda, Maryland. Sensitivity analysis. John Wiley and Sons, Ltd., Chichester, Busack, C. A., and B. E. Thompson. 2006. Improving salmon UK. habitat productivity models for setting escapement goals Saltelli, A., S. Tarantola, and F. Campolongo. 2000b. Sensitivity and prioritizing habitat recovery funding. Final report to the analysis as an ingredient of modeling. Statistical Science Pacific Salmon Commission, Project SF-2006-H-27, Dec 15. 15(4):377-395. Washington Department of Fish and Wildlife, Olympia. Shared Strategy Development Committee. 2007. Puget Gaff, H., J. Chick, J. Trexler, D. DeAngelis, L. Gross, and Sound salmon recovery plan. Shared Strategy Development R. Salinas. 2004. Evaluation of and insights from ALFISH: a Committee, Seattle, Washington. Available at: www.nwr. spatially explicit, landscape-level simulation of fish popula- noaa.gov/Salmon-Recovery-Planning/Recovery-Domains/ tions in the Everglades. Hydrobiologia 520:73-87. Puget-Sound/PS-Recovery-Plan.cfm (Accessed 14 June Haefner, J. W. 2005. Modeling biological systems: principles 2007). and applications. 2nd Edition. Springer, Inc., New York. Sobol, I. M. 1993. Sensitivity estimates for nonlinear mathe- Independent Science Advisory Board (ISAB). 2001. Model matical models. Mathematical Modelling and Computational synthesis report: an analysis of decision support tools used in Experiment 1:407-414. Columbia River basin salmon management. ISAB, Portland, Smith, A.D.M., E.J. Fulton, A.J. Hobday, D.C. Smith, and Oregon. Available at: www.nwcouncil.org/library/isab/ P. Shoulder. 2007. Scientific tools to support the practical isab2001-1.pdf (Accessed 15 Nov 2007). implementation of ecosystem-based fisheries management. Lestelle, L. C., B. Watson, and G. Blair. 2006. Species-habitat ICES Journal of Marine Science 64: 633-639. rules: Supporting documentation for updated flow rules for application in EDT. Mobrand, Jones & Stokes, Vashon, Steel, E. A., T. J. Beechie, M. Ruckelshaus, A. H. Fullerton, Washington. Available at: www.jonesandstokes.com/index. P. McElhany, and P. Roni. In press. Mind the gap: uncer- php?option=com_docman&task=doc_download&gid=121 tainty and model communication between managers and sci- (Accessed 15 Nov 2007). entists. In E. Knudson and H. Michael, eds. Pacific salmon Lichatowich, J. A., and L. E. Mobrand. 1995. Analysis of Chinook environment and life history models: advancing science for salmon in the Columbia River from an ecosystem perspec- sustainable salmon in the future. American Fisheries Society, tive, Mobrand Biometrics. Bonneville Power Administration, Bethesda, Maryland. Portland, Oregon. Available at: http://pisces.bpa.gov/release/ Yoder, N. J., M. D. Bowen, K. P. Zehfuss, J. Hubble, B. documents/documentviewer.aspx?doc=25105-2 (Accessed 15 Watson, B G. R. Blair. 2006. Systematic sensitivity analysis Nov 2007). of the EDT model applied to selected fish populations in the Lower Columbia Fish Recovery Board. 2004. Lower Columbia salmon recovery and fish and wildlife plan. Lower Columbia Fish Recovery Board, Longview, Washington. Available at: www.lcfrb.gen.wa.us/December%20Final%20%20Plans/ lower_columbia_salmon_recovery_a.htm (Accessed 14 June 14, 2007). McElhany, P., E. A. Steel, D. W. Jensen, K. Avery, N. Yoder, B. Thompson, and C. Busack. In press. Dealing with uncer- tainty in ecosystem models: lessons from a complex salmon model. Ecological Applications. Mobrand, L. E., J. A. Lichatowich, L. C. Lestelle, and T. S. Vogel. 1997. An approach to describing ecosystem perfor- mance “through the eyes of salmon.” Canadian Journal of Fisheries and Aquatic Sciences 54: 2964-2973. Moussalli, E., and R. Hilborn. 1986. Optimal stock size and har- vest rate in multistage life history models. Canadian Journal of Fisheries and Aquatic Sciences 43(1):135-141. Northwest Hydraulic Consultants. 2005. Final report— Instream flow assessment pilot project. Prepared for Shared Strategy for Puget Sound, USA. Available at: www.shared- salmonstrategy.org/resources.htm#water_quantity (Accessed 14 June 2007). Northwest Power and Conservation Council. 2005. Subbasin plans in Columbia River basin fish and wildlife program. Northwest Power and Conservation Council, Portland, Oregon. Available at: www.nwcouncil.org/library/2000/2000-19/ FullReport.pdf (Accessed 15 Nov 2007). Quinn, T. 2005. The behavior and ecology of Pacific salmon O.S. Systems and trout. American Fisheries Society and University of Washington Press, Seattle.

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 339 Column: Guest Director's line The North American Freshwater Fish Standard Sampling Project: Improving Fisheries Communication

Introduction Scott A. Bonar, Wayne A. Hubert, and David W. Willis Bonar is with the U.S. Geological Survey, Arizona Cooperative Fish A city was lost because gear was not and Wildlife Research Unit, University of Arizona, Tucson, and can be standardized. In 1904, Baltimore burned. Known as the Great Fire, the conflagra- contacted at [email protected]. Hubert is with the U.S. Geological tion burned 56 ha of the downtown Survey, Wyoming Cooperative Fish and Wildlife Research Unit, University business district and 1,526 structures, of Wyoming, Laramie. Willis is at the Department of Wildlife and Fisheries resulting in over $150 million in property Sciences, South Dakota State University, Brookings. damage (Frederick 2004; Welden 2004). The massive destruction probably could have been prevented. There was plenty dard medical procedures and compared Standardization of routine fisheries of water available and hundreds of fire- to data distributions that summarize data collection provides several ben- fighters from surrounding cities arrived to results from the standard procedures efits. Data collected over a sequence of help, but they could do nothing because applied to many other individuals and years can be compared. Because fish the couplings of their fire hoses would lead to indices of health for an individual. populations often take years to respond not fit onto the city’s hydrants. Although If these diagnostic procedures were not to management actions, standardized efforts to standardize hose couplings standardized, we would be unable to time-series data can be used to assess the had been emphasized for over 30 years, evaluate even the most basic data about response of fish populations to manage- 600 different coupling sizes were in use our health. In fact, if standardization was ment actions. at the time. People were rightly angered not used in countless other facets of our Data can be compared across large with the fire-fighting profession because society, our lives would be much more geographic regions when standard data standardization had not yet been accom- difficult. collection methods are applied. Increasing plished. Because of this fire, standardiza- For data collection purposes, standard- numbers of studies at landscape or tion of fire couplings across the United ization means that data are collected in regional scales are being conducted in States moved forward at a rapid pace, a single defined manner so comparisons fisheries science. Standardized sampling leading to heightened safety in North can be made. Although routine data can provide comparable samples to American cities. collection has been standardized in many assess the effects of regulations, invasive Rarely does failure to standardize have other disciplines, for freshwater fish sam- species, habitat improvements, or other such dramatic consequences. However, pling it has not. Before now, most data management strategies over large geo- standardization of measurements, collections for freshwater fishes in North graphic regions. Standardized data can industrial processes, languages, and data America were standardized only by state also be useful for measuring large-scale collection has greatly advanced human or provincial agencies (Bonar and Hubert effects, such as global climate change, on development (Nesmith 1985). We are 2002). population structure, body condition, or often unaware of standardization, yet it Many of us have tried to compile fish- abundance of fish. is all around us, affecting our everyday eries data for region-wide assessments or Standardization can encourage data lives. It includes business letter formats, over long periods of time. Often it is very sharing and improve communications time zones, computer components, and difficult. Waters are sampled with differ- among fisheries professionals and the shoe sizes. The list of standardized items ent gears and reporting of effort is not general public. In addition, standardiza- is seemingly endless. In addition, data consistent, making comparisons almost tion can promote cooperation among collection and summarization for many impossible. Data requiring thousands of biologists in different countries who disciplines are standardized, including hours to collect have highly diminished speak various languages. medicine (Beers and Berkow 1999), utility because they cannot be compared Other factors can also be enhanced meteorology (Lockhart 2003; Schiesl to data from other waters to provide by using standard sampling techniques. 2003), geology (Assaad et al. 2004), and assessments. Reflect on the lost informa- Standard techniques to monitor the water chemistry (Eaton et al. 2005). For tion that this has cost our profession over abundance of rare or endangered spe- example, cholesterol, body temperature, many decades because data collection cies can assist in their conservation. and blood pressure are measured by stan- was not standardized. Standardization can allow collected data

340 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g Detroit Public Library Scott A. Bonar to be compared to regional or range- wide averages to assess fish in individual Scott A. Bonar National Weather Service populations relative to an array of other populations. An initiative of the Fisheries Management Section of American Fisheries Society (AFS) provided move- ment toward standardization of fresh- water fisheries sampling methods in North America. The North American standardized sampling project involved contributions by at least 284 fisher- ies biologists from 107 agencies and organizations in Canada, Mexico, and the United States, as well as several countries outside of North America. The development of these procedures was led by the Fisheries Management Section Standardization of methods and techniques has been responsible for huge advancements in of AFS in partnership with the U.S. Fish medicine, computer technology, industry, and data analysis. and Wildlife Service, U.S. Geological Survey (Cooperative Fish and Wildlife Research Units and National Biological and agencies that provided various forms to recommend a small set of methods Information Infrastructure program), U.S. of assistance. The project collaborators that could effectively sample coldwater Bureau of Reclamation, U.S. Bureau of decided that standard methods could or warmwater fishes in primary fresh- Land Management, National Park Service, best be reported in a book made avail- water habitats—small lakes and ponds, National Fish and Wildlife Foundation, Arizona Game and Fish Department, able to individuals who would like guid- reservoirs, large natural lakes, wadable and the AFS Fisheries Information and ance on sampling methods for particular streams, or large rivers. Technology and Education Sections. environments and fish species and who Forty-seven authors, each a recog- recognize the long-term value of using nized expert in sampling fish in specific Project History standardized sampling methods. aquatic environments, were selected for The resulting book, Standard Methods this project by the editors and mem- In the 1980s and 1990s, work to for Sampling North American Freshwater bers of the AFS Fisheries Management move toward standardized fish sampling Fishes, identifies sampling methods for Section. Most of the authors had pub- procedures in the United States was general population assessments and lished extensively on sampling techniques conducted by the Fisheries Techniques monitoring of fishes in various water or had been responsible for designing Standardization Committee of Fisheries body types in North America. The book standard sampling procedures for their Management Section. This effort resulted also provides a means whereby existing state, province, or region. Experts were in a compilation of methods being used and future data can be compiled into obtained from as wide a geographic across North America (Gabelhouse et al. databases to enable comparisons of sam- range as possible across Canada, Mexico, 1992); however, selection of a subset of pling statistics among populations. The and the United States, and represented a methods for standardized sampling of standardized sampling methods enable wide variety of federal, state, and provin- fish in ponds, lakes, streams, or rivers was a standard set of indices of population cial agencies, non-government organiza- not completed. structure and abundance to be computed tions, and universities. At the 2004 AFS Annual Meeting, and compared among populations, To develop sampling methods, authors a unanimous vote by attendees at the across regions, and over time. Previous were provided with the information col- Fisheries Management Section meeting reviews of fish sampling methods have lected by the AFS Fisheries Management initiated an effort to extend the work of typically been exhaustive, providing the Section in the 1980s on sampling tech- the Fisheries Techniques Standardization largest variety of sampling methods pos- niques used across the United States. In Committee and develop standard sible for a given species or type of water addition, dozens of fish-sampling proto- sampling procedures. The effort was body, and allowing many choices (e.g., cols used by federal, state, and provincial facilitated by a grant from the Fisheries Murphy and Willis 1996). The purpose agencies were provided to authors, and Management Section, the 10 collaborat- of the standard sampling book was not these techniques were used as a spring- ing agencies listed above, and other enti- to review all methods available to sample board for selecting standard sampling ties, as well as a multitude of biologists fish populations in all water bodies, but techniques applicable across North

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 341 America. Guidelines were provided to the authors to insure that representing over 4,000 fish populations in 43 states and their proposed standard methods had the following attributes. provinces across North America. These data were obtained • Standard-sampling methods recommended would be the through mail surveys of all federal, state, and provincial most common and statistically valid of those used by con- agencies in North America. servation agencies across North America. Recommendation • Computation of the standard indices are described, as well of common methods will make standardization easier, both as how to calculate needed sample sizes to provide precise politically and logistically. Exceptions may occur if a common estimates of the indices. References to other books and method is considerably flawed. book chapters (e.g., Brown and Austen 1996; Guy and • The book would differ from fisheries techniques books that Brown 2007) providing more detail regarding statistical present a variety of methods, including those that are experi- analyses are included. Additionally, examples of databases mental or “cutting-edge.” Authors were asked to resist the for storage and management of standardized data are tendency to promote “latest cutting-edge methods” in lieu included. of commonly used methods. • Recommended procedures for converting non-standard • The simplest methods possible were chosen because com- data to standard data are included. plex methods are much less likely to be standardized. • Procedures for preventing transfer of exotic or invasive • Standard-sampling methods will be updated in the future organisms among water bodies while using standard fish much like Standard Methods for the Examination of Water sampling methods are described. Quality and Wastewater (Eaton et al. 2005) to reflect • Methods for standardizing electrofishing power are advances in fisheries science. presented. • Standard habitat measuring techniques (e.g., Armantrout Standard sampling methods were reviewed and designed 1982; Bain and Stevenson 1999) would not be included, with leadership by the AFS in collaboration with most fed- except those used to standardize fish-sampling gear, such as eral, state, and provincial agencies in North America, but it is conductivity, Secchi depth, and water color. important to note that these standard methods are recom- • Standard sampling methods focused on large-juvenile and mendations. In no way are agencies or biologists required adult life stages that are easily sampled with standard gears. • Standard sampling methods were designed to minimize fish to use the methods. The standard methods described in this mortality, so methods such as toxicants and explosives were book are provided to those who would like guidance when not included. initiating standard monitoring programs, those who would like • Standard sampling methods should allow for fish species guidance as to how to gather data that can be compared to identification and measurement of lengths and weights if North American or regional averages, or those who would like possible. to modify existing programs to more standardized programs. Those who are satisified with local standardized sampling Results plans already in place are not pressured to change. However, those who decide to use the proposed standard sampling Standard fish-sampling methods were recommended in out- methods in their monitoring or assessment programs will find line form by authors and the recommendations were reviewed that they will possess powerful tools to assess fish populations by 54 fisheries professionals from 33 federal, state, and provin- and identify responses to fisheries management actions. cial agencies as well as the 3 co-editors of the book. Outlines of Standard Sampling Methods for North American Fishes recommended methods were turned into text, reviewed by the identifies sampling methods for general or routine fish popula- editors, and constructive criticism was provided to the chapter tion assessments. These methods provide a means to facilitate authors. The second draft of the text of each chapter was then comparison with data collected across North America and sent for anonymous review by 2–3 specialists. These special- over time. However, it is recognized that standard sampling ists included project sponsors, other sampling experts, and methods may not be appropriate when conducting specialized European experts in fish sampling. The resulting book will be research or management studies. Consequently, techniques published by the AFS books division in the summer of 2009. other than those identified in the standard sampling book, and Features of the book include: specifically designed for a particular study, may be appropriate. • Descriptions of specific gears and procedures for sampling either coldwater or warmwater fishes in ponds, natural Future directions lakes, reservoirs, small streams, or large rivers. • North American and regional averages and ranges were Even before the publication of the book, natural resource reported for indices of length structure, body condition, fish agencies in several states and provinces were moving to adopt growth, and catch per unit effort for common game and these methods. Some directions for future development that nongame fishes collected using the standard techniques. will allow these methods to provide even more information are These averages and ranges were calculated from data apparent.

342 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g Validation of Methods—A continu- ing challenge for fisheries scientists has been to determine the extent to which sampling data, collected with all of their biases and inconsistencies, represent populations. Use of standard sampling removes considerable variation in the data collected. However, bias still occurs. Validation is the process of defining the extent to which a sampling method rep- resents the true population. Validating the standard methods presented in the book presents a rich area of research that could improve our ability to inter- Kevin Pope pret sampling data. Validation of some of the standard methods is already underway, with Ontario researchers test- ing the standard gill-netting methods in Canadian lakes (Sandstrom et al. 2008). Boat electrofishing to capture warmwater fish in standing waters. Making standardized databases available via Internet—Averages, percentiles and ranges were presented in the book in tabular form for ecoregions and across the ranges of individual spe- cies. The organization follows Carlander (1969, 1977, 1997), whose efforts to compile data were of substantial value to fisheries biologists. However, to increase the usefulness of these sum- maries and to update them more easily, new data need to be incorporated into an interactive database that can be accessed via the Internet. This database could be updated rapidly with the col- lection of new data, could be a node at various agency websites, and could provide customized averages and ranges

for comparison with a push of a button. David Willis Development of an interactive database is a challenge for the future. Updating the standard sampling methods—Typically AFS reference books, such as Fisheries Techniques Setting a fyke net to capture warmwater fish in standing waters. (Murphy and Willis 1996) and Inland Fisheries Management in North America (Kohler and Hubert 1999), are updated compare among them between editions, mechanism is needed where comments approximately every 10 years. We antici- will advance techniques, contribute to on the methods could be cataloged and pate that there will be a need to update their ease of use, and reach a wider used when writing future editions. In this reference book and add to the data audience. Most standard methods addition, translation of the book into summaries in a similar fashion. manuals are updated on a regular basis. Spanish and French would facilitate Undoubtedly, the information in the For example, Standard Methods for communication of the information to first edition of the standardized sam- the Examination of Water Quality and additional fisheries professionals in the pling book will need to be improved in Wastewater is now in its 21st edition Americas and allow for greater data the future. Updating methods at regular (Eaton et al. 2005). A website or other sharing and communication. intervals, while preserving the ability to

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 343 Acknowledgements Society, Portland, Oregon. The Hague techniques used by conservation agen- Publishing, Billings, Montana. cies in the U.S. and Canada. Fisheries We thank the authors who have Assaad, F., P. E. LaMoreaux, and T. H. Techniques Standardization Committee described the standard sampling pro- Hughes, 2004. Field methods for Report. Fisheries Management Section, tocols: M. S. Allen, P. Bailey, D. Beard, geologists and hydrologists. Springer- American Fisheries Society, Bethesda, D. A. Beauchamp, K. I. Bonvechio, J. Verlag, New York. Maryland. Boxrucker, M. Brouder, P. J. Braaten, D. Bain, M. B., and N. J. Stevenson. 1999. Guy, C. S., and M. L. Brown (editors). Britton, S. Bryan, P. Budy, S. Contreras- Aquatic habitat assessment: common 2007. Analysis and interpretation of Balderas, R. A. Curry, A. Dolloff, J. B. methods. American Fisheries Society, freshwater fisheries data. American Dunham, C. S. Guy, D. P. Herzog, P. Bethesda, Maryland. Fisheries Society, Bethesda, Maryland. J. Howell, R. M. Hughes, A. C. Iles, Beers, M. H, and R. Berkow (editors). Kohler, C. C., and W. A. Hubert (edi- S. Jacks, R. Kinnunen, N. P. Lester, 1999. The Merck manual, 17th edi- tors). 1999. Inland fisheries manage- C. Luecke, A.J. Loftus, J. Lyons, M. tion. Merck Research Laboratories, ment in North America, 2nd edition. McMaster, N. Mercado-Silva, L.E. Whitehouse Station, New Jersey. American Fisheries Society, Bethesda, Miranda, R. Neumann, D. L. Parrish, Bonar, S. A., and W. A. Hubert. 2002. Maryland. C. P. Paukert, J. T. Peterson, J. Pitlo, Standard sampling of inland fish: Lockhart, T. J., 2003. Atmospheric K. Pope, M. C. Quist, C. Rabeni, Benefits, challenges, and a call for measurements. Pages 691-720 in T. action. Fisheries 27(3):10-16. R. S. Rogers, A. E. Rosenberger, R. D. Potter, and B. R. Colman, editors. Brown, M. L., and D. J Austen. 1996. Handbook of weather, climate, and Schneidervin, S. Sharon, S. Smith, G. Data management and statistical tech- water: dynamics, climate, physical Thiede, R. F. Thurow, R. A. Whaley, and niques. Pages 17-62 in B. R. Murphy meteorology, weather systems, and D. Zafft. This project would not have and D. W. Willis, editors. Fisheries measurements. Wiley, Hoboken, New been possible without the support of techniques, 2nd edition. American Jersey. C. Ault, P. Barrett, H. Bolton, and M. Fisheries Society, Bethesda, Maryland. Murphy, B. R., and D. W. Willis (edi- Parker of the U.S. Fish and Wildlife Carlander, K. D. 1969. Handbook of tors). 1996. Fisheries techniques, 2nd Service; J. de Vos of the Arizona Game freshwater fishery biology. Volume 1. edition. American Fisheries Society, and Fish Department; W. J. Fleming, A. Iowa State University Press, Ames. Bethesda, Maryland. Ostroff, D. Beard, B. Shanks, and B. K. _____. 1977. Handbook of freshwater Nesmith, A. 1985. A long arduous march Williams of the U.S. Geological Survey; fishery biology. Volume 2. Iowa State toward standardization. Smithsonian A. Loftus of Loftus Consulting; J. Kosa University Press, Ames. 15(12):176-194. formerly of the U.S. Bureau of Land _____. 1997. Handbook of freshwater Sandstrom, S., N. Lester, G. Morgan, Management, now with the U.S. Fish fisheries biology. Volume 3. Iowa State T. Haxton. 2008. Efficacy of the and Wildlife Service; J. Wullschleger University Press, Ames. proposed North American gill net stan- and L. Norris of the National Park Eaton, A. A., L. S. Clescerl, E. W. Rice dard for monitoring fish communities Service; M. McKinstry of the U.S. and A. E. Greenberg. 2005. Standard in Ontario, Canada. Abstract, 138th Bureau of Reclamation; M. Snyder methods for the examination of Annual Meeting of the American and others with the National Fish and water and wastewater, 21st edition. Fisheries Society, Ottawa, Ontario, Wildlife Federation; G. Rassam, execu- Jointly published by the American Canada. American Fisheries Society, tive director, AFS; and members of the Public Health Association (APHA), Bethesda, Maryland. Fisheries Management, Education, and Washington, D.C; American Water Schiesl, J. W., 2003. Instrument develop- Fisheries Information and Technology Works Association (AWWA), Denver, ment. Pages 691-720 in T. D. Potter, Sections of the AFS. Colorado; and Water Environment and B. R. Colman, B.R., eds. Handbook Federation (WEF), Alexandria, Virginia. of weather, climate, and water: References Frederick, G. E. 2004. The Great dynamics, climate, physical meteorol- Baltimore Fire, February 7, 1904: Fire ogy, weather systems, and measure- Armantrout, N. B. (editor). 1982. Department operations. Fire Museum ments. Wiley, Hoboken, New Jersey. Acquisition and utilization of aquatic of Maryland, Lutherville. Welden, G. R. 2004. No reason to burn: habitat inventory information. Gabelhouse, D. and 20 coauthors. the story of the Great Baltimore Fire. Western Division American Fisheries 1992. Fish sampling and data analysis Fire Museum of Maryland, Lutherville.

344 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g News: AFS UNITS

Workshop student receives 1-on-1 instruction Keynote speaker Ray Troll and Program Chair Workshop students inspect a boat and trailer from a USGS expert for suturing a fish (after Rich Grost model each other’s fish-art. for invasive hitchhikers. graduating from bananas) during the Fisheries Telemetry workshop.

Oregon Chapter: Holds annual meeting in Bend impacts to aquatic resources was planned and moderated by a Photos: www.RichardGrost.com panel of undergraduate students led by Chapter Student Liaison The Oregon Chapter annual meeting in February 2009 in Kristle Warren as a project for an Oregon State University (OSU) Bend drew more than 450 professionals, including a record 60+ problem-solving class. An even less conventional (but packed) students, to participate in “Sharing Our Passion” for fish and fish- session highlighted unique and inspiring images of fish, includ- eries. The theme recognized that as professionals we’re expected ing underwater photography and video with photographer to be pragmatic and practical—but we can, and often should, narrations. also be passionate. Passionate about the resources we manage, Two day-long workshops preceded the meeting and educated the species and habitats we study, the methods we develop and some 50 members. “Fisheries Telemetry,” organized by Shaun employ, the information we assemble, and the interpretations Clements, John Beeman, and Carl Schreck, had students suturing we devise. But mostly, passionate about that lowest common bananas first, then fish. The workshop was supported by the denominator we all share. Intentional or not, the AFS serves as a “support group” for professional fish-o-philes—a place we can commiserate in good times and bad, when other workers, friends, even family no longer appreciate the prolific use of our favorite 4-letter “f-word” (fish, of course). This meeting cel- ebrated that root connection and evoked passion like few others. Fish-art guru Ray Troll set the perfect tone with his entertain- ing and inspirational presentation “Fish Worship—How I Became a Scientific Surrealist,” explaining how fish and science have infused his entire life and career with passion. His creative, color- ful art has spread internationally, celebrating fish and their myriad relation to humans. One oddity revealed: the classic skull and fish design “Spawn til You Die” has become a favorite fashion Scholarship Awards Chair Scott Heppell with winners Mac Barr, Autumn among demented antagonists in B-rate horror movies. Troll par- Smith, and Matt Sloat. ticipated throughout the meeting, presenting his work devising and directing community art projects, discussing his explorations of the vicious fishes of the Amazon, and unleashing his musical madness by leading some hard-driving, fish-inspired “sub-aqua- punk-bush-rock” during the Fishhead Jam Session. Technical sessions accommodated 150 presentations and 36 posters, on subjects as diverse as aquaculture advances, climate and conservation, cyanobacteria and fish dynamics, dam remov- als, fish passage facilities, estuary and marine research, permit- ting and statistics, education and outreach, lampreys, bull trout, sea turtles, and, of course, zebra mussels. Notably, one session addressed revision of the Chapter’s white paper regarding marine Ticket hawkers Christy Fellas, Melissa Ann Ocana, Lindsey Arnold, and reserves, and another special session on synthesizing pesticide Heather Reiff.

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 345 Oregon Department of Bill Wingfield Memorial Award in Fish Fish and Wildlife (ODFW), Culture: Tony Amandi (on behalf of OSU, U.S. Geological the ODFW Fish Health Team), for long- Survey (USGS), and term contributions to fish health and vendors including AMIRIX production; Systems (VEMCO) Fishery Team of the Year: Tim Hoffnagle, and Lotek Wireless. Don Hair, Joseph, Deb Eddy, Sally Equipment displays Gee, and Fred Monzyk; for exemplary helped students visualize research on artificial propagation and how to configure field recovery of threatened Snake River stations to effectively spring/summer Chinook; and monitor tagged fish. Broken Oar: Tim Shibahara and Chris Bill Brignon, Heidi Andersen (accepted by Jason Dunham), Brett Blandon, The second workshop, Karchesky, for surviving a “watch this!” and Londi Tomaro receive Best Paper and Poster Award recipients. “Identification and test-drive where the manufacturer Control of Aquatic swamped his own jet boat at the base Invasive Species—How to Not Be a Vector,” of Willamette Falls. was organized by Jill Hardiman and sup- From a record number of scholarship ported by USGS, U.S. Fish and Wildlife th applications, scholarships were awarded Service, the 100 Meridian Team, Oregon to: Kari Rosch, Mount Hood Community State Marine Board, Sea Grant, U.S. Forest College (AA candidate); Autumn Smith, Service, Bureau of Land Management, and OSU (BS candidate); Mac Barr, OSU Portland State University. It featured a mock (masters candidate); and Ben Clemens, boat and trailer inspection, but first, stu- OSU (Ph.D. candidate). The inaugural and dents were familiarized with the scoundrels uniquely prestigious Carl Bond Scholarship most recently invading area waters and the was awarded to Matt Sloat, OSU (Ph.D. methods most useful in monitoring and candidate), in support of his work on life controlling them. Throughout the meeting, Tim Shibahara and Chris Karchesky receive the history evolution and anadromy in Pacific vendors of telemetry equipment, water Broken Oar Award from Stephanie Burchfield. salmonids. Scott Heppell and his commit- quality instruments, environmental services, tee did an outstanding job managing the and other products and services displayed scholarship process and evaluating many along a deck overlooking the snowmelt- worthy applicants, most of whom we laden Deschutes River. hope to see again next year. Professional awards were managed Best Student Paper honors went to expertly by Mindy Simmons and the William Brignon (OSU) and Heidi Andersen Awards Committee, resulting in presenta- (OSU, runner-up), while Best Student tion of the following: Poster honors went to Brett Blundon (OSU) Award of Merit: Central Oregon and Londi Tomaro (OSU, runner-up). Mike Flyfishers, for sustained dedication to Hudson did his usual excellent job planning conservation work benefiting native for and presenting these important awards. The Student-Mentor Mixer was packed Casey Baldwin and the custom art drawn by fish; Ray Troll just for ORAFS. Award of Merit: Sunriver Anglers Club, with some 60 students making rapid-fire, for sustained dedication to conservation 8-minute “speed-dating” rounds to meet work benefiting native fish; and query all 40 mentors. Despite the hec- Award of Merit: Leesa Cobb, for sus- tic pace, students raved about the direct tained dedication to marine conserva- exposure this event provides to a diversity tion work; of career options, guidance, and even Award of Merit: Port Orford Resources direct job opportunities. Mentors were Team, for sustained dedication to mostly hoarse at the end. Christy Fellas was community-based marine conservation the catalyst with critical assistance from work; Kristle Warren and Shivonne Nesbitt. Award of Merit: Bianca Streif and At the business meeting, President Neil Janine Castro, for exemplary service in Ward announced the renewal of a legisla- planning the 2008 Western Division tive liaison position to serve as our eyes and Blane Bellerud fondles the original Joe Tomelleri AFS workshop: “Stream Restoration: ears in the capitol and keep the Chapter burbot that he won in the live auction. Integrating Practical Approaches”; abreast of legislative issues and opportuni-

346 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g ties. Dave Ward updated us on Western Division AFS projects, Doug Young received his past president remembrance (a legacy baby blue T-shirt of his ordering), and External Director Jeremiah Osborne- Gowey updated members on committee activities. Standing committees convened for strategy sessions and there was renewed passion evident, especially within the Marine Habitat, Natural Production, Legislative, and Education and Outreach committees. A glimpse of the Fishhead Jam Session featuring Mike Faler on bass and Ray Troll on guitar and The raffle and auction, organized by lead vocals. Terry Smith, featured a huge assortment of desirable stuff including an original anadromous fish restoration project for the upper Deschutes River Basin. Thanks Joe Tomelleri burbot painting, custom Don Ratliff and Jim Bartlett, for the tour photo prints by members Richard Grost and the chili dogs! and Mary Edwards, outdoor gear, boating Overall, the 2009 meeting fulfilled adventures, and custom art by Ray Troll expectations as the Oregon Chapter’s (including a piece drawn on the tablecloth largest and most important techni- at dinner). Students deftly worked the cal and professional networking event. floor selling raffle tickets and supporting Thanks are due to all who attended, the auction. Enough passion ensued to assisted, presented, moderated, donated, raise over $10,000 toward future scholar- and especially the ExCom for a year of ships and other Chapter activities. behind-the-scenes planning: President The Jam Session, organized and Neil Ward, Past President Doug Young, grounded by bass-man Mike Faler, grew President Elect Rich Grost, Vice President Kitty Griswold models the custom fish-shirt into an amazingly choreographed mass Christy Fellas, Secretary Treasurer Shaun being drawn by Ray Troll as one of the auction of some 30 member musicians tuning, Clements, Internal Director Martyne prizes. crooning, and generally synergizing in Reesman, and External Director Jeremiah various combinations until literally booted Osborne-Gowey. This offstage in the wee hours. The audio crew, and incom- feast included classic bluegrass, folk, ing President Elect R&B, rock, and of course the custom Demian Ebert, are fish-fare of Ray Troll. One member used already planning the dance floor to shake her newly-won the next Oregon hip-booties. “Lumpsuckers of love, love, Chapter meeting love,” oh yeah. 24–26 February Finally, some 50 members participated 2010 at the Hilton in in a post-meeting free-lunch tour of Eugene. Portland General Electric’s Round-Butte —Rich Grost fish passage project. This project, well under construction, is one of the largest ExCom: Doug Young, Neil Ward, Rich Grost, Martyne Reesman, Christy and most complex fish passage facilities Fellas, and Shaun Clements (not present: Jeremiah Osborne-Gowey). ever built and a key element to the major

Samuel Chan and his undergraduate class (Dan Udell, Ben Cate, Emma Garner, Shane Smith, Kristle Warren, Cameron King, and Trygve Kaalaas) planned and moderated the “Pesticide Impacts to Aquatic Resources” symposium.

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 347 the group on the second day. Sixteen quet, Jack Swearingen of Tennessee presentations were given in 4 technical Wildlife Resources Agency received the sessions, with 12 presentations by stu- Outstanding Fisheries Scientist Award dents. Southern Division President Cecil for his work on the state’s muskellunge Jennings attended (despite inclement fishery. Amy Wales of Tennessee Valley weather conditions in Georgia) and Authority was recognized with a Service addressed the Chapter during its busi- Award for her long tenure as the ness meeting. The AFS 2009 planning Chapter’s secretary treasurer. Johnathan committee also met on the morning Davis of Tennessee Technological of the first day to continue working University received the Best Student on logistics for the Parent Society’s Paper Award for his presentation on upcoming meeting in Nashville. At monitoring strategies for the endan- the business meeting, Jim Habera gered duskytail darter, co-authored by Amy Wales receives a Service Award for her (president elect) and Jason Heneger his faculty advisor, Brad Cook. After the long tenure as secretary-treasurer. (secretary treasurer) were elected, and banquet, lively musical entertainment join Hayden Mattingly (president) and was provided by the folk-bluegrass trio Tennessee Chapter Frank Fiss (past president) as Chapter “2nd Nature,” along with student-led Holds annual meeting at state park officers. The Chapter voted to make games on the breaks. The meeting The Tennessee Chapter held its a donation to the pending paddlefish adjourned with a pledge to get every- annual meeting 3–4 March 2009 book, and set aside $1,500 to help one involved with preparations for the at Montgomery Bell State Park. in-state students travel to the AFS AFS 2009 Nashville meeting. Approximately 55 people attended, 2009 Nashville meeting. At the ban- with an additional 30 students joining —Hayden Mattingly

Dakota Chapter Holds annual meeting in Bismarck Despite a meager participation from South Dakota Game Fish and Parks because of out-of-state travel restrictions, the Dakota Chapter meeting was held in Bismarck on 23–25 February. Approximately 70 participants mingled at the socials, attended the sessions, participated in business meeting, and recognized award recipients. Students from South Dakota State University (SDSU) dominated the awards given by the Chapter. North Central Division President Mark Porath and Society President Bill Franzin each participated in the meeting. Three SDSU students tied for the Best Student Paper award. Sauger Scholarship winners from left, Ryan Andvik, Bethany Landon Pierce presented “Evaluating Stocking Success of Galster, Nick Peterson, and Andrew Wuestewald. Paddlefish in Lake Francis Case, SD: a work in progress.” His

Bethany Galster, right, receives the Schmulback Clark Williams, right, of the Lake Region Will Sayler, right, receives the Distinguished Memorial Scholarship from Dakota Chapter Anglers Association accepts the Aquatic Professional Service Award from Dakota President Randy Hiltner. Resource Conservation Award from Dakota Chapter President Randy Hiltner. Chapter President Randy Hiltner.

348 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g South Dakota State University students who won best paper and best poster awards, with their advisors. From left: Mike Brown, Andy Jansen, Landon Pierce, Travis Schaeffer, Justin Van De Hey, Brian Graeb, Nick Peterson, and Dave Willis (photo credit Dan James). co-authors were Brian Graeb and and Andrew Wuestewald from SDSU Jason Sorenson. Justin VanDeHey received Kriel scholarships. Named won for his presentation “Non-Lethal after long-time Dakota Chapter Sampling of Walleye and Yellow member Al Kriel, and funded by his Perch for Stable Isotope Analysis: family, the award pays for full mem- A Comparison of Three Tissues,” bership in the American Fisheries co-authored by Mark Fincel and Society. Steve Chipps. Travis W. Schaeffer Will Sayler received the presented, “Compensatory Growth Distinguished Professional Service and Metabolic Responses of Female award and the Aquatic Resource Yellow Perch Subjected to Symmetric Conservation Award was given to Feed: Fast Cycles,” with coauthors the Lake Region Anglers Association Daniel Spengler, Casey Schoenebeck, (North Dakota). Mike Brown, and Steve Chipps. New Chapter officers were elected Two SDSU students tied for the at the meeting. Greg Simpson (sec- Best Student Poster award. Andy retary/treasurer), Brian Fletcher (vice Jansen won for his poster “Effect of president), and Jeff Hendrickson a Simulated Cold-Front on Hatching (president elect), join incoming President Mike Barnes. North Central Division President Elect Mark Success of Yellow Perch Eggs.” Porath attended the meeting, and as this Brian Graeb and Dave Willis were —Mike Barnes photo shows, raked it in during the raffle! co-authors. Nick Peterson won for his poster titled, “Determination of Bluegill Size and Age at Maturity in Southeastern South Dakota Impoundments,” co-authored by Justin VanDeHey and Dave Willis. Bethany Galster from South Dakota State University was the winner of the Schmulbach Memorial Scholarship. Named after James C. Schmulbach, this competi- tive scholarship is awarded by the Chapter to an eligible junior or senior undergraduate studying fisheries in either North or South Dakota. Sauger scholarships (travel awards) went to South Dakota State University students Bobbi Adams, Emperor Aquatics, Inc. Ryan Andvik, Nick Peterson, and Andrew Wuestewald. Will French

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 349 Michigan Chapter

This is because the fellowship experi- leading to better management practices. ence is intended to engage the fellow Because one of the MDNR’s goals is to in a project that’s separate from their maintain resources for future generations, graduate research and coursework, my role as Fenske Fellow on this project and the fellowship committee wants to will aid in establishing management ensure that the fellow can manage the practices for sustaining lake sturgeon in The Fenske Fellowship: fellowship project in addition to their the future. A Graduate Student Opportunity in other responsibilities. For my fellowship, As a Fenske fellow, I had the opportu- Fisheries Management I needed to identify two mentors—one nity to attend several Management Team By Amy M. Schueller from a management agency and one from meetings for the Fisheries Division. These Schueller is a doctoral student at the MSU—and work with those mentors to meetings engage top administrators of the Department of Fisheries and Wildlife at develop a project. Thus, the fellowship Fisheries Division in group decision-making Michigan State University and can be was also meant to provide experience processes. One of the main meeting tools contacted at [email protected]. working within a management agency, they use are submitted issue statements, When I started graduate school at along with mentors, on a specific manage- which are documents that include key Michigan State University (MSU), it was ment-related project, as well as expose the pieces of information and requests for with the idea that someday, I would like student to how agencies are structured decisions related to this information. Issue either to become a professor or work and function. statements can be submitted by anyone for an agency as a researcher. Like many When it came time for me to pick the in the division to request a Management fisheries professionals, my fascination project that I wanted to work on, I picked Team decision on, for example, whether or with fisheries management stems from a project that would benefit the state not a project will receive money, a protocol childhood experiences fishing, camp- as well as provide me an opportunity to or regulation will be changed, how fish ing, and hunting. During my time at play an active role in influencing fisheries populations will be managed in the face MSU, I have participated in numerous management decisions. I also picked a of a new disease, or if a bill in the state activities common to the graduate student project which provided me the opportu- legislature will be supported. This system experience. One opportunity that I didn’t nity to write a document that included allows for everyone in the division to have anticipate was applying for and being internal, external, and stakeholder reviews. their voice heard on the topics that are selected as the first Janice Lee Fenske In 2005, I attended a meeting led by important to them and the stakehold- Excellence in Fisheries Management Gary Whelan from the MDNR on cur- ers they represent. I was impressed with Fellow for 2007–2008. By taking on this rent lake sturgeon research projects. At the MDNR’s ability to work together and fellowship, I gained insight into how an the meeting, the lake sturgeon commit- come to a consensus even when those agency functions, how important human tee was tasked with updating the state’s involved didn’t agree on all aspects of a dimensions are to fisheries management, Lake Sturgeon Rehabilitation Strategy topic. As part of the process to update the how to effectively communicate in group (LSRS). I later recalled this meeting when LSRS, the LSRS will need to be put forth settings, create timelines, balance duties, I was choosing a project for the Fenske as an issue statement and agenda item and plan meetings. My experience as a Fellowship application and contacted Gary for a Management Team meeting. Thus, Fenske Fellow has helped me to grow to inquire about the status of this project. attending Management Team meetings both personally and professionally, and He stated the updates hadn’t progressed has given me insight into how issue state- has given me the opportunity to influence much and felt the LSRS needed some- ments should be posed to be clear and fisheries management as a student, and one to take a central leadership role to concise and move forward with a decision. I wanted to share my experiences with move the project forward. I jumped at As a graduate student researcher other students. this opportunity and elected to work with focused on sturgeon biology, it is easy The Fenske Fellowship was created Gary and Dan Hayes (my advisor) from to get somewhat disconnected from the by an endowment to the Department of MSU on the project. human dimensions of resource manage- Fisheries and Wildlife at MSU honoring As Fenske fellow, my role in this ment. Lake sturgeon rehabilitation affects Janice Lee Fenske, the first female biologist process was to coordinate Michigan’s many diverse stakeholder groups such in the Michigan Department of Natural sturgeon researchers in order to take their as Sturgeon for Tomorrow and Native Resources (MDNR) Fisheries Division. individual expertise and combine it into American tribes. Before becoming the The fellowship provides individuals from a coherent update of Michigan’s LSRS. Fenske Fellow, I thought working with under-represented groups with experience This update will be based on knowledge stakeholders would be relatively easy, until working with a management agency on acquired since the strategy was first com- I found that stakeholders often do not see a fisheries-related project. Besides typical piled in 1997. The new LSRS will benefit eye-to-eye when considering how natural application materials like a CV, project lake sturgeon in Michigan by redefining resources should be managed. Some proposals, and reference letters, applying goals and objectives for their populations stakeholders want lake sturgeon for recre- requires a discussion of time constraints. based on more complete knowledge, ational harvest, others want lake sturgeon

350 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g Fenske Fellowship Selection Committee with Amy Schueller and her mentors. Schueller with her mentors Dan Hayes and Gary Whelan. for cultural and ritualistic reasons, and still ment actions were discussed, allowed toward progress on pressing fisheries others want lake sturgeon for ecosystem for efficient decision making. Having a management issues. rehabilitation. Juggling all of these needs prepared framework for decision making, Although I have officially finished my and wants is no small feat; it takes people, like an issue statement, for committee tenure as a Fenske Fellow, the update to money and political management skills members to comment on and change the Lake Sturgeon Rehabilitation Strategy is ongoing and I plan to continue my which are rarely taught in the classroom. was much more efficient than starting involvement. The completed update of For the most part, these critical people completely from scratch. This is just one the Lake Sturgeon Rehabilitation Strategy skills are learned on the job. I am fortu- of the beneficial lessons that I took away nate to have gained this insight while still will help guide lake sturgeon manage- from Management Team meetings. pursuing my graduate studies. ment efforts in Michigan, and maintain Through my involvement in the Fenske One of the biggest lessons that I took a sustainable population in the future for Fellowship, I feel fortunate to have had the away from this experience is that prepara- all stakeholders. It is rewarding to know opportunity to influence fisheries manage- tion, planning, and communication are that I was a part of these efforts, and essential to efficient and effective decision ment as a graduate student. I was able to these incredible experiences are something making. We worked with a lake sturgeon be involved and take a central leadership I would not have had if it were not for committee to update the LSRS. Keeping role during the decision making process the Fenske Fellowship. Finally, I encour- these members informed of the progress thus far, for lake sturgeon management age students to search out these types of the project was crucial to keeping the in the state of Michigan. The Fenske of opportunities in order to influence management decisions because this project moving forward. Also, effective Fellowship is a unique opportunity which was definitely a professional learning preparation for the committee meeting, fosters student participation with a variety experience. where goals, objectives, and manage- of professionals, who work together

Lotek

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 351 vOLUMe 138 issUe 3 JOUrNAL highLights: MAy 2009 TRANSAcTIONS OF THE AMERIcAN FISHERIES SOcIETY

To subscribe to AFS journals go to www.fi sheries.org and click on Publications/Journals.

Survival, and Habitat use of naturally Reared and G. Leibowitz, Joan P. Baker, M. Robbins Church, Jana E. Hatchery Steelhead Fry in Streams: effects of an Compton, Bruce A. Miller, Michael A. Cairns, Bruce P. enriched Hatchery Rearing environment. Christopher P. Hansen, and Henry R. La Vigne, pages 564-580. Tatara, Stephen C. Riley, and Julie A. Scheurer, pages 441- 457. Do Blue crab Spawning Sanctuaries in north carolina Protect the Spawning Stock? David B. Eggleston, Red Snapper Larval Transport in the Northern Gulf Geoffrey W. Bell, and Steven P. Searcy, pages 581-592. of Mexico. Donald R. Johnson, Harriet M. Perry, Joanne Lyczkowski-Shultz, and David Hanisko, pages 458-470. An evaluation of Data entry error and Proofi ng Methods for Fisheries Data. Christopher L. Johnson, Hydroacoustic Target Strength Distributions of Gabriel M. Temple, Todd N. Pearsons, and Timothy D. Alewives in a net-cage compared with Field Surveys: Webster, pages 593-601. Deciphering Target Strength Distributions and effect on Density estimates. Thomas E. Brooking and Lars G. The Genetic Legacy of Stocking Muskellunge in a Rudstam, pages 471-486. Northern Minnesota Lake. Loren M. Miller, Steven W. Mero, and Jerry A. Younk, pages 602-615. The effect of Sample Size on the Stability of Principal components Analysis of Truss-Based Fish Marine Distribution, Life History Traits, and the Morphometrics. Patrick M. Kocovsky, Jean V. Adams, and Accumulation of Polychlorinated Biphenyls in chinook Charles R. Bronte, pages 487-496. Salmon from Puget Sound, Washington. Sandra M. Spatial Distribution, energetic Status, and Food Habits O’Neill and James E. West, pages 616-632. of eastern Bering Sea Age-0 Walleye Pollock. Jamal H. juvenile chinook Salmon Summer Microhabitat Moss, Edward V. Farley Jr., Angela M. Feldmann, and James Availability, use, and Selection in a central Idaho N. Ianelli, pages 497-505. Wilderness Stream. Dean E. Holecek, Kara J. Cromwell, Linking Habitat Quality with Trophic Performance of and Brian P. Kennedy, pages 633-644. Steelhead along Forest Gradients in the South Fork [Note] Recent and Historical Spatial Distributions of Trinity River Watershed, California. Sarah G. McCarthy, juvenile Rockfi sh Species in Rocky Intertidal Tide Jeffrey J. Duda, John M. Emlen, Garth R. Hodgson, and Pools, with emphasis on Black Rockfi sh. Rebecca S. David A. Beauchamp, pages 506-521. Studebaker, Karah N. Cox, and Timothy J. Mulligan, pages Reproductive Abnormalities in Trout from Western U.S. 645-651. National Parks. Adam R. Schwindt, Michael L. Kent, Luke Balancing Aquatic Habitat Fragmentation and control K. Ackerman, Staci L. Massey Simonich, Dixon H. Landers, of Invasive Species: enhancing Selective Fish Passage Tamara Blett, and Carl B. Schreck, pages 522-531. at Sea Lamprey Control Barriers. T. C. Pratt, L. M. Steelhead Life History on california’s central coast: O’Connor, A. G. Hallett, R. L. McLaughlin, C. Katopodis, D. Insights from a State-Dependent Model. William H. B. Hayes, and R. A. Bergstedt, pages 652-665. Satterthwaite, Michael P. Beakes, Erin M. Collins, David R. Swank, Joseph E. Merz, Robert G. Titus, Susan M. Sogard, Population Dynamics of the Smalleye Shiner, an and Marc Mangel, pages 532-548. Imperiled cyprinid Fish endemic to the Brazos River, Texas. Bart W. Durham and Gene R. Wilde, pages 666-674. Seasonal Patterns of Abundance, Growth, and Site Fidelity of juvenile Steelhead in a Small coastal Sablefi sh Predation on juvenile Pacifi c Salmon in the California Stream. Susan M. Sogard, Thomas H. Williams, Coastal Marine Waters of Southeast Alaska in 1999. M. and Heidi Fish, pages 549-563. V. Sturdevant, M. F. Sigler, and J. A. Orsi, pages 675-691. Modeling Stream network-Scale Variation in coho [Note] effect of Hook Removal on Recapture Rates of Salmon Overwinter Survival and Smolt Size. Joseph 27 Species of Angler-Caught Fish in Australia. Gene R. L. Ebersole, Mike E. Colvin, Parker J. Wigington Jr., Scott Wilde and William Sawynok, pages 692-697.

352 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g Publications Book Review, New Titles, For Your Information

Skeena River Fish and Their Habitat Allen S. Gottesfeld and Ken A. Rabnett, Ecotrust, Portland, Oregon 2008, ISBN: 978-0-9779332-5-9, 352 p. $29.95, paperback The Skeena River is the second largest river system in British The authors have done an excellent job of locating and Columbia, Canada, draining 54,432 km2 of the north-central summarizing the information on Skeena fisheries hidden in region of the province. It has a limited fish fauna (about 32 spe- government reports and memoranda. Of the more than 783 cies) but supports commercially and culturally important popula- citations in the book, 689 are gray literature. Although this makes tions of Pacific salmon (Oncorhynchus spp.) as well as trout, charr, the Skeena seem scientifically neglected, in fact the river has a whitefishes, sturgeon, burbot, and eulachon. Although parts the distinguished history as a site of Pacific salmon research. Studies basin have suffered some habitat degradation from logging, min- by scientists of the Fisheries Research Board of Canada between ing, ranching, and urban development, most fisheries habitat has the 1940s and the 1970s, particularly at Lakelse Lake and Babine been little altered. The Skeena could, therefore, provide a habitat Lake, helped establish the foundation of salmon management in reference for more developed basins in the Pacific Northwest. As the Pacific Northwest. In the past few decades, however, most yet, there has been no comprehensive synthesis of information new information has come from intermittent monitoring and on Skeena River habitat and fishery resources that would allow management surveys. managers and researchers to take advantage of this opportunity. Although the book presents a lot of information that most In Skeena River Fish and Their Habitat, Gottesfeld and Rabnett individuals would otherwise have difficulty accessing, it also has take the first steps toward such a synthesis. a number of important limitations. There is no obvious logic to the selection of the 18 sub-basins, except that they divide up The book appears to be an enlarged and updated version of a the catchment. Each sub-basin is discussed as though it was report the authors prepared for the Skeena Fisheries Commission an independent unit rather than part of a whole. This criticism (Gottesfeld et al. 2002), an Aboriginal organization that focuses applies equally to the sub-sections of each main section. For on fisheries management, science, and conservation in the example, descriptions of geology, vegetation, water quality, etc., Skeena River basin. Ecotrust, an environmental non-governmental are not presented in the context of Pacific salmon needs, which organization, assisted with publication of the book as part of its is the focus of the book. Nor do the authors make any attempt collaboration with the Wild Salmon Center in Portland, Oregon, to locate the Skeena in relation to other larger river systems. The to document the distribution and status of Pacific salmon popula- book is a description of data on the Skeena River and nothing tion throughout their range. more. Finally, perhaps because the book is an expanded report, I The book is organized into two main sections—an overview found the prose dry and uninteresting. of physical, ecological and fisheries characteristics of the whole —Michael Healey Skeena basin, followed by more detailed descriptions of 18 sub- Professor Emeritus, basins. The basin and each sub-basin are described in terms of Institute for Resources Environment and Sustainability environment (physical characteristics, water quality, vegetation), University of British Columbia fish resources, fisheries, and human development. The Skeena Current Address: Peachland, British Columbia drains a diverse geography from dry interior plateau to wet [email protected] coastal forest and includes 7 of British Columbia’s 14 biogeocli- matic zones. The 18 sub-basins reflect this biogeoclimatic diversity Reference in only a very general way. Skeena River Fish and Their Habitat Gottesfeld, A. S., K. A. Rabnett, and P. E. Hall. 2002. also deals primarily with Pacific salmon species, including steel- Conserving Skeena fish populations and their habitat. Skeena head (O. mykiss). Other freshwater species are covered briefly in Fisheries Commission, Hazelton, British Columbia. Available the first section and barely mentioned again. This is a reflection of online at: www.skeenawatersheds.com/downloads/Skeena%20 the limited data on species other than salmon and steelhead. WFSP%2012%20(low%20res).pdf.

tial interest to AFS members: Lessios on Biology, Assessment, and New Titles divergence of marine organisms across Management of North Pacific Annual Review of Ecology, the Panama isthmus, Pyke on the biology Rockfishes. Edited by J. Heifetz, J. Evolution and Systematics volume 39. of mosquitofish, Crowder et al. on the Dicosimo, A. J. Gharrett, M. S. Love, V. M. Edited by D. J. Futuyma, H. B. Shaffer, and impacts of marine fisheries in an ecosys- O’Connell, and R. D. Stanley. Alaska Sea D. Simberloff. Annual Reviews, Palo Alto. tem-based context, and Schwartz on the Grant College Program, Fairbanks. 2007. 2008. 674 pp. $84 (cloth and online); performance of the Endangered Species 550 pp. $50.00 (cloth); seagrant.uaf.edu/ www.annualreviews.org. Act. This year’s volume includes more bookstore/pubs/AK-SG-07-01.html. This annual compilation of reviews this color and better quality graphics than ever This book represents the proceedings year includes various papers of poten- before. of the 23rd Lowell Wakefield Fisheries

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 353 Symposium held in 2005 as part of the oregonstate.edu/dept/press/c-d/CatchingtheEbb. Bethesda, Maryland. 2008. 484 pp. $69 ($48.30 AFS Annual Meeting in Anchorage. The html. for AFS members; paper); www.afsbooks. volume includes 28 papers that cover life org/54063p.html. The Conservation Professional’s Guide to history, population structure and specia- Working with People. By S. A. Bonar. Island If Fish Could Scream: An Angler’s Search tion, age and growth, fishery manage- Press, Washington, DC. 2007. 198 pp. $50 for the Future of Fly Fishing. By P. Schullery. ment, habitat requirements, and stock (cloth); $25 (paper); www.islandpress.org/book- Stackpole Books, Mechanicsburg, Pennsylvania. assessments. store/details.php?prod_id=121. 2008. 208 pp. $24.95 (cloth); www.stackpole- books.com/cgi-bin/stackpolebooks.storefront. Dietary Supplements for the Health and Cutthroat: Native Trout of the Quality of Cultured Fish. Edited by H. An Island Out of Time: A Memoir of Smith West—Second edition. By P. Trotter. Nakagawa, M. Sato, and D. M. Gatlin. CABI, Island in the Chesapeake. By T. Horton. University of California Press, Berkeley. Cambridge, Massachusetts. 2007. 244 pp. W.W. Norton, New York. 2008. 319 pp. 2007. 547 pp. $34.95 (cloth); www. $140.00 (cloth); www.cabi.org/bk_BookDisplay. $15.95 (paper); www.wwnorton.com/catalog/ ucpress.edu/books/pages/11059.php. asp?PID=1995. spring08/033146.htm. Twenty years after the publication Environmental Risk Assessment of of the first edition of this book comes a Marine Ornamental Shrimp: Biology, Genetically Modified Organisms. Volume 3. Aquaculture and Conservation. By R. Calado. thoroughly revised and updated second Methodologies for Transgenic Fish. Edited by Wiley-Blackwell, Ames, Iowa. 2008. 263 pp. edition. The book is intended primarily A. R. Kapuscinski, K. R. Hayes, S. Li, and G. Dana. $199.99 (cloth); www.wiley.com/WileyCDA/ for anglers and naturalists but contains CABI, Cambridge, Massachusetts. 2007. 304 pp. WileyTitle/productCd-1405170867.html. an extensive bibliography for those $150.00 (cloth. www.cabi.org/bk_BookDisplay.as who might want to use it as a reference p?SubjectArea=&Subject=&PID=205 Metal Contamination in Aquatic book. The text is well illustrated with Environments: Science and Lateral Fish Diseases. Edited by J. C. Eiras, H. Segner, Management. By S. N. Luoma and P. S. photographs, line drawings, and maps. It T. Wahli, and B. G. Kapoor. Science Publishers, Rainbow. Cambridge University Press, New contains 15 chapters, 3 of which are intro- Enfield, New Hampshire. 2008. 1340 pp., 2 vol- York. 2008. 573 pp. $120.00 (cloth); www. ductory, while the others focus on extant umes. $149.00 (cloth); www.scipub.net/fisheries/ cambridge.org/uk/catalogue/catalogue. (and extinct) subspecies. fish-disease.html. asp?isbn=978052186057

Fish Forever: The Definitive Guide to Otolith Atlas for the Western For Your Understanding, Selecting and Preparing Mediterranean, North and Central Eastern Healthy, Delicious, and Environmentally Atlantic. By V. M. Tuset, A. Lombarte, and C. Sustainable Seafood. By P. Johnson. Wiley, A. Assis. CSIC-Scientia Marina 72-Suppl. 1, Information Hoboken, New Jersey. 2007. 438 pp. $34.95 Barcelona. 2008. 203 pp. €59.44 (cloth); www. (cloth); www.wiley.com/WileyCDA/WileyTitle/ icm.csic.es/scimar/index.php/secId/6/IdArt/3723/ Advances in Fisheries Bioengineering. productCd-076458779X.html. Edited by S. V. Amaral, D. Mathur, and E. P. Taft. (free pdf version available). American Fisheries Society, Bethesda, Maryland. The Fishermen’s Frontier. By D. F. Arnold. Pollution of Lakes and Rivers: A 2008. 239 pp. $69 ($48.30 for AFS members); University of Washington Press, Seattle. 2008. www.afsbooks.org/54061p.html. 267 pp. $35 (cloth); www.washington.edu/ Paleoenvironmental Perspective, Second uwpress/search/books/ARNFIS.html. edition. By J. P. Smol. Blackwell, Malden, Alaska’s Fishing Communities: Harvesting Massachusetts. 2008. 383 pp. $64.99 (paper); the Future. Edited by P. Cullenberg. Alaska Sea Fishers’ Knowledge in Fisheries Science www.wiley.com/WileyCDA/WileyTitle/pro- Grant College Program, Fairbanks. 2007. 106 pp. and Management. Edited by N. Haggan, B. ductCd-1405159138.html. $10 (paper); seagrant.uaf.edu/bookstore/pubs/ Neis, and I. G. Baird. UNESCO, Paris. 2007. 437 AK-SG-07-02.html. pp. €30 (paper); publishing.unesco.org/details. Red Summer: The Danger, Madness, and aspx?Code_Livre=4521. Exaltation of Salmon Fishing in a Remote Anaesthetic and Sedative Techniques for Alaskan Village. By B. Carter. Scribner, New Aquatic Animals. By L. G. Ross, B. Ross, and B. Freshwater Fishes of Texas: A Field Guide. York. 2008. 234 pp. $25.00 (cloth); www.simon- Ross. Blackwell Publishing, Oxford. 2008. 222 Edited by C. Thomas, T. H. Bonner, and B. G. says.com/content/book.cfm?tab=1&pid=616775 pp. $139.99 (cloth); www.wiley.com/WileyCDA/ Whiteside. Texas A&M University Press, College &er=9780743297066. WileyTitle/productCd-1405149388.html. Station. 2007. 202 pp. $23.00 (paper); www. tamu.edu/upress/BOOKS/2007/thomas.htm. Reproductive Biology of Crustaceans: Animal Intelligence: From Individual Case Studies of Decapod Crustaceans. Edited to Social Cognition. By Z. Reznikova. Great Lakes for Sale: From Whitecaps to by E. Mente. Science Publishers, Enfield, New Cambridge University Press, New York. 2007. Bottlecaps. By D. Dempsey. University of Hampshire. 2008. 548 pp. $125.00 (cloth); 472 pp. $130.00 (cloth); $60.00 (paper); Michigan Press, Ann Arbor. 2008. 107 pp. www.scipub.net/biology/reproductive-biology- www.cambridge.org/catalogue/catalogue. $22.95 (paper); www.press.umich.edu/titleDetail- crustaceans.html. asp?isbn=9780521532020. Desc.do?id=269167. The Riverscape and the River. By S. M. Haslam. Catching the Ebb: Drift-fishing for a Life in Grenadiers of the World Oceans: Biology, Cambridge University Press, New York. 2008. 404 Cook Inlet. By B. Bender. Oregon State University Stock Assessment, and Fisheries. Edited by A. pp. $130.00 (cloth); www.cambridge.org/us/cata- Press, Corvallis. 2008. 296 pp. $22.95 (paper); Orlov and T. Iwamoto. American Fisheries Society, logue/catalogue.asp?isbn=9780521839785.

354 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 355 AFS Annual Meeting: The Greening of the AFS’09 Annual Meeting

to bring the AFS meetings to We’re proud the South, where the diversity of our freshwater fishes is unrivaled in the country. World-class fishing opportunities abound nearby, or you can join the field trip to the Duck River, which harbors one of the most diverse aquatic faunas in the country, if you’re more inclined to non-game fishes. As an organization responsible for helping to conserve this diversity, each year we look for ways to further reduce our impact on aquatic life. This year, we’ve focused our attention in reducing the amount of waste that is caused by the meeting.

In Nashville, you won’t receive:

• An author index in your printed meeting book and program guide. • A CD or memory stick with all of the abstracts. • An expensive tote bag that will be stashed in the back of your closet. • Drink tickets for soft drinks, beer, and wine

In Nashville, you will receive:

• A lighter meeting book and program guide sans author index (which saved 20,000 pages of paper when we printed the book). While author indices are handy for finding talks given by your

356 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g favorite fi shery biologists (or your own talk!), we • Good food and drinks. Southern hospitality is have decided to eliminate it to save paper and ink. legendary, and we’re going to ensure that you Instead, you will receive… experience the best of it by providing food and • Instructions on how to download and search beverages. There will be no drink tickets in Nashville the program and abstract fi les. These fi les will for soft drinks, beer, or wine at the four main be posted on the AFS website before the meetings, socials. The kegs, which are reusable, will also save where it will be a snap to electronically search for on packaging waste from bottles or cans. In case your favorite speakers (or poster authors) and read you have trouble remembering the great time you their abstracts. You can create a schedule of talks you want to hear prior to arriving at the meeting, or had in Nashville, we’re also giving you coasters you can easily download the program and abstract for your drinks, made from 100% post-consumer fi les to your laptop computer to search for speakers recycled tires. The main social will be held at Smiley and strategize during the meeting. The program Hollow, a working farm that has been in the Smiley and abstract fi les will also be uploaded onto the Family since the 1800s. Much of the produce that computers in the Internet Café to facilitate speaker you will eat that night is grown locally—on the farm searches. itself. • Reusable and recyclable bags. Conferences have been green for years by handing out canvas Finally, don’t forget to think about sustainability when bags that can be used at the grocery store. Our making your personal decisions on travel to the meeting. bags are made of recycled materials that can also In this time of tight travel restrictions, try to carpool to the be used to carry lunch, plant lettuce and herbs for portable gardens, or carry fi eld gear. If you’re already meeting or share hotel space. We have set up a discussion overwhelmed with conference bags, you can drop board on the AFS Facebook site so that you can list or these off along with your milk jugs at a recycling look for rides to the meeting. We look forward to seeing center. y’all there!

Oregon RFID Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 357 International Governance of Fisheries Ecosystems: Learning from the Past, Finding Solutions for the Future Michael G. Schechter, Nancy J. Leonard, and William W. Taylor, editors x

TO ORDER:

Online: www.afsbooks.org

Mail: 458 pages American Fisheries Society List price: $69 c/o Books International AFS member price: P.O. Box 605 $48 Herndon, VA 20172 Item number: 550.56P Phone: 703/661-1570

Published July 2008 Fax: 703/996-1010

Fisheries experts increasingly acknowledge the importance of globalization on local, national, and international fisheries. This book brings together fisheries and governance experts from across the globe who present case studies on a broad spectrum of the internationally shared fisheries that inhabit diverse freshwater and marine ecosystem types. Case studies provide the biological background of the fisheries resource, including status and threats to the resource and its ecosystem. The case studies review the evolution and current governance institutions of the fisheries resource, with particular focus on international or global institutions. Each study concludes with an evaluation of the effectiveness of the current fisheries governance institutions, and recommendations for changes.

358 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g To submit upcoming events for inclusion on the AFS Web site Calendar, send CALeNDAr: event name, dates, city, state/province, web address, and contact information FISHERIES EvENTS to cworth@fi sheries.org. (If space is available, events will also be printed in Fisheries magazine.)

More events listed at www.fi sheries.org, click "Who We Are," click "calendar"

jul 20-24 Sixth International Fisheries Observer and Monitoring conference Portland, Maine www.st.nmfs.noaa.gov/iformc209 jul 22-27 early Life History Section's 33rd Annual Larval Fish conference and American Society of Ichthyologists and Herpetologists conference Portland, Oregon www.dce.k-state.edu/conf/jointmeeting Jul 27-30 Waterpower XVI Spokane, Washington www.waterpowerconference.com Aug 24-28 4th International Otolith Symposium Monterey, California [email protected] Aug 30-Sep 3 American Fisheries Society 139th Annual Meeting Nashville, Tennessee www.fi sheries.org/afs09 Sep 16-19 World Fishing exhibition 2009 Vigo, Spain www.worldfi shingexhibition.com Sep 21-25 International council for the exploration of the Sea Annual Science conference Berlin, Germany www.ices.dk Sep 25-28 Combined Australian entomological Society’s 40th AGM & Scientifi c Conference / Society of Australian Systematic Biologists / 9th Invertebrate Biodiversity and conservation / Australian coral Reef Society conference Darwin, Australia www.evolutionbiodiversity2009.org Sep 28-Oct 2 Australian Society for Limnology Congress: Water in a Dry Land: Sustaining Arid Zone Rivers and Wetlands Alice Springs, Northern Territory www.asl.org.au Oct 15-17 cAnSee eighth Biennial conference—ecological economics: Prosperity for a Sustainable Society Vancouver, British Columbia, Canada www.cansee.org Oct 22-24 Western Division of the AFS 2nd Annual Student Colloquium Fort Collins, Colorado Nate Cathcart: [email protected] or http://welcome.warnercnr.colostate.edu/afs-home/ index.php Oct 25-30 Sixth International Symposium on Sturgeon Wuhan, Hubei Province, China www.iss6.org nov 2-4 International Symposium on Integrating Offshore Renewable energy System and Aquaculture Newport, Rhode Island http://seaagrant.gso.uri.edu/baird/2009 Dec 9-12 Fourth Shanghai International Fisheries and Seafood expo Shanghai, China www.gehuaexpo.com 2010 Mar 1-5 Aquaculture 2010 San Diego, California www.was.org

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 359 EMPLOYERS: To list a job opening on the AFS Online Job Center submit a position Announcements: description, job title, agency/company, city, state, responsibilities, qualifications, salary, closing date, and contact information (maximum 150 words) to jobs@fisheries. Job Center org. Online job announcements will be billed at $350 for 150 word increments. Please send billing information. Listings are free (150 words or less) for organizations with Associate, Official, and Sustaining memberships, and for Individual members, who are faculty members, hiring graduate assistants. If space is available, jobs may also be printed in Fisheries magazine, free of additional charge.

Fishery Management Biologist, Finfish Resources, increase constituents understanding of specific program Point No Point Treaty Council, Kingston, Washington. issues and initiatives. Duties also include development Salary: $52,770 to $63,248 of new media such as video and audio segments. Closing: Until filled. Assist in developing constituent relationships and Responsibilities: Assist member tribes natural resource preparing executives for outreach events. Coordinate directors in managing harvest of salmon, steelhead, communications messaging and activities across the and marine fish resources. Responsibilities include stock office. assessment, fisheries management planning, providing Qualifications: Experience working on or leading advice on the regulation of fisheries, and working projects and programs, writing and editing cooperatively with other tribes and state/federal communications materials, and conveying natural agencies. Coordinate with other Treaty Council staff as resource information to diverse audiences. College part of an integrated fisheries management program degree. that includes hatchery management planning, salmon Contact: Dan Farrow, 301/713-2325, reference job recovery, and habitat protection/restoration. announcement number NMFS-HC-2009-0015. See Qualifications: Desire M.S. with study emphasis www.usajobs.gov/. in at least two of the following areas: fish biology, population assessment, statistical procedures, Ph.D. Research Assistantship, Department of Ecology, applied research, natural resource economics, and Montana State University. natural resources management. Two or more years Salary: Stipend of $18,000 per year with substantial of professional experience required. Actual fisheries tuition waiver. management experience is desirable as is knowledge Closing: 1 August 2009. of U.S. vs. Washington and treaty fishing rights. Self- Responsibilities: Investigate the role of biodiversity in starter with demonstrable communication skills, ability infectious disease risk using salmonid whirling disease to prioritize and willingness to take tasks to completion. of western streams as the model system. Collaborate Contact: PNPTC, attn HR, 7999 NE Salish Lane, with engineering and molecular population genetics Kingston, Washington 98346; [email protected]; faculty. Direct involvement in both field surveys and 360/297-6544. laboratory experiments relating the tubificid community to disease risk in salmonids. Communications Specialist (2 positions), NOAA, Qualifications: B.S. or M.S. degree in ecology, zoology, Fisheries, Office of Habitat Conservation. biology, or a related field. Strong quantitative and Salary: $60,900–133,500. writing skills. Prior experience with field methods Closing: 10 July 2009. used in stream ecology and benthic macroinvertebrate Responsibilities: Develop and implement strategies for collection and identification and/or disease ecology raising the visibility and improving the understanding preferred. of the mission of OHC with a variety of audiences, Contact: Submit cover letter, resume, copies of including NMFS and NOAA leadership, Capitol Hill, transcripts and GRE scores, and names and telephone media working with NOAA Public Affairs, state and numbers of 3 references to by either e-mail or regular federal partners, and other stakeholders. Support the mail to Billie L. Kerans, [email protected], process to define objectives and key messages for 406/994-3725, www.montana.edu/wwwbi/staff/kerans/ communication efforts and assist in developing an kerans.html. annual strategic communications plan. Responsible for Postdoctoral Scientist, Department of Fish Wildlife, development of written promotional and information Colorado State University, Fort Collins. materials including press releases, brochures, annual Salary: Commensurate with qualifications and accomplishment reports, and web site content to experience.

360 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g Closing: 24 August 2009. of transcripts and GRE scores to: Jesse Trushenski, Responsibilities: Study climate change effects on Southern Illinois University Fisheries Illinois Aquaculture native cutthroat trout. Model persistence of native Center, 1125 Lincoln Drive, Room 173 Carbondale, cutthroat trout given impending changes to interacting Illinois 62901-6511; [email protected]; 618/536-7761 aquatic and terrestrial factors caused by climate change. Work involves climate modeling, data analysis, Ph.D. Assistantship in Mapping Ecosystem and developing a Bayesian Belief Network decision- Services, Virginia Polytechnic Institute. support tool. Salary: $22,000–24,000 per year plus tuition. Qualifications: Earned Ph.D. in aquatic or fisheries Closing: 31 July 2009. ecology, experience in research on ecology of stream Responsibilities: Participate in a multidisciplinary fishes, substantial experience with modeling, ability effort to examine where/when biological conservation to create and manage large databases, skill in using enhances delivery of aquatic ecosystem services. statistics and computers, strong record of publishing Participate in conceptual-model development for papers in high-quality refereed journals. Desired and spatial analyses of relations among conservation experience with climate models and fish conservation. practices, biodiversity, delivery of ecosystem services, Contact: E-mail cover letter, resume with four and human well being in a U.S. river basin. Perform references, transcript copies, and publications to Kurt project data analysis and report writing, while Fausch, [email protected]. Department completing Ph.D. coursework. of Fish, Wildlife, and Conservation Biology, CSU, Fort Collins, CO 80523. Qualifications: M.S. in landscape ecology, ecological economics, conservation biology, geography, or related discipline. Commitment to multidisciplinary research, Ph.D. Research Assistantship, Fisheries and Illinois demonstrated scientific productivity, including peer- Aquaculture Center, Department of Zoology, Southern reviewed publications, strong statistical skills experience Illinois University, Carbondale. with large geo-spatial datasets, excellent writing skills. Salary: $1,426 per month plus full tuition waiver. Contact: Send letter of interest, resume, GRE scores, Closing: 1 July 2009. Responsibilities: Conduct research evaluating the dynamics of fatty acid profile change and accumulation of persistent organic pollutants in hybrid striped bass, with the goal of identifying nutritional strategies to minimize risk and maximize value of cultured seafood. Address the controversial use of marine-derived feedstuffs i.e., fish oil and fish meal, as vectors for environmental contaminants that can accumulate in fish tissues along with nutritionally beneficial long-chain polyunsaturated fatty acids. See http://fisheries.siu.edu. Start date: Fall 2009. Qualifications: M.S. in aquaculture, aquatic toxicology, or related field. Must meet admission requirements for the graduate school and Department of Zoology at SIUC at www.science.siu.edu/zoology/ programsgraduate.html. Contact: Submit a letter of interest, resume and CV, contact information for three references, copies Sonotronics

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 361 names of three references to Paul Angermeier, Qualifications: Admission requires a B.S. degree Department of Fisheries and Wildlife Sciences, Virginia in aquaculture, fisheries, agricultural economics, Tech, Blacksburg, Virginia 24061-0321; 540/231-4501; or a related field, a minimum GPA last 2 years of [email protected]. 3.0 and GRE score of 1,000 verbal quantitative. Minimum TOEFL score of 550 paper based or M.S. Graduate Research Assistantship, Aquaculture equivalent for international students. Strong and Fisheries Center, University of Arkansas. quantitative statistics, mathematics skills, and Salary: First year—$17,800. Second year—$18,800. computer proficiency required. Maintenance of Closing: 1 August 2009 or until filled. large data set is desired. Responsibilities: Use store scanner data to analyze Contact: Complete forms at below link and mail market trends and retail pricing issues for catfish, hardcopy to Dey Aqua and Fish Center, UAPB 1200 crawfish, clam, and shrimp, and use household-based North University Drive, Mail Slot 4912, Pine Bluff, scanned data to analyze consumer behavior. Arkansas 71601. For questions see www.uaex.edu/aqfi, [email protected].

7 JULY

362 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g How to Tag a Shrimp (Once) that Molts 30 Times Faster growing strains of disease-resistant shrimp are providing a significant boost to the aquaculture industry. Developing these strains is easier if the performance of various families, reared together under identical conditions, can be easily identified for evaluation.

Tagging shrimp and other crustaceans is complicated by an exoskeleton which can be shed as many as 30 times as the animal grows to a marketable size. Hawaii’s Oceanic Institute’s Shrimp Program, with assistance of experts from Washington Dept. of Fish and Wildlife, set out to find the right tag for the job. Trials were conducted using NMT’s Visible Implant Elastomer (VIE). VIE is injected as a liquid under clear or translucent tissue, and it then cures to a pliable solid that remains externally visible. The tags are available in 10 colors, of which six are fluorescent for enhanced visibility and detection. VIE was injected into the muscle of the sixth abdominal segment in juvenile shrimp ranging from 0.2 g to 3.9 g, and adults weighing from 23 g to 47 g. After 10 to 14 weeks, tag retention in shrimp tagged as juveniles was 99.9% while 100% of the tags were retained in those tagged as adults1. The utility of VIE has since been proven with many species of shrimp2 and other crustaceans (see www.nmt.us for references). The tags are typically retained well through molting and have little effect on growth or survival.

Researchers around the world use VIE to evaluate the performance of various strains of shrimp3. By varying tag locations and colors, VIE provides the many unique codes required to identify test groups. Please contact us if we can help with your research.

1Godin, D.M. et al. 1996. Aquaculture 139:243-248. 2Brown, J. H. et al. 2003. Aquaculture Research 34(1):49-54. 3Goyard, E. et al. 2002. Aquaculture 204(3-4):461-468.

Top: VIE Tags in a posterior abdominal section of a pandalid shrimp. Center: VIE is injected into the sixth abdominal segment of a juvenile white shrimp. (Photo courtesy of S. Arce.) Bottom: An NMT Air Driven Elastomer Injection System can be used to tag 500 shrimp per hour.

Northwest Marine Technology, Inc. www.nmt.us Shaw Island, Washington, USA Northwest Marine Technology

Corporate Office Biological Services 360.468.3375 [email protected] 360.596.9400 [email protected]

Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g 363 Hydroacoustic Technology, Inc. 364 Fisheries • v o l 34 n o 7 • j u l y 2009 • w w w .f i s h e r i e s .o r g