fisheries · fisheries • vol 38 no 2 • february 2013• 49 contents fisheries vol 38 no 2 feb...

Fisheries

Partnering in a Global Network

Best Student Presentation/Poster Process

Best Practices for Managing Your Data

Computer-Facilitated Species Identification

Ecological Networks in Fisheries

American Fisheries Society • www.fisheries.orgVOL 38 NO 2FEB 2013

03632415(2013)38(2)

Robert L. Stephenson, John H. Annala, Jeffrey A. Runge, and Madeleine Hall-Arber, editors

The Gulf of Maine (GOM) is arguably one of the best studied marine ecosystems in the world. Interest in its physical environment, fisheries, and Canada/USA boundary have resulted in considerable research attention for more than a century. The GOM is also highly managed by two nations with a commitment to implementing an ecosystem approach to management.

The papers in this book review the management and policy tools and approaches required to implement integrated policy and management in the GOM; synthesize the current ecological and oceanographic understanding of the GOM, and the social, economic, and cultural interactions within the gulf; assess anthropogenic and external in- fluences on the gulf ecosystem; and examine the science required to observe and predict changes in the GOM ecosystem, along with strategies to implement an ecosystem approach to management.

Advancing an Ecosystem Approach in the Gulf of Maine

415 pages, paperList price: $79.00AFS Member price: $55.00Item Number: 540.79PPublished October 2012

TO ORDER:Online: www.afsbooks.org

American Fisheries Societyc/o Books InternationalP.O. Box 605Herndon, VA 20172Phone: 703-661-1570Fax: 703-996-1010

Fisheries • Vol 38 No 2 • February 2013• www.fisheries.org 49

Contents

Fisheries VOL 38 NO 2 FEB 2013

President’s Hook51 Think Globally, Act Globally With over 9,000 members representing 64 nations, AFS is in an ideal position to address global fisheries issues and, coordinating closely with the other members of the World Council, ensure that the best scientific and management approaches are being applied.

John Boreman — AFS President

Guest Director’s Line88 Revisiting the Process for Qualifying for the AFS/ Sea Grant Best Student Presentation and Poster AwardsAn important mechanism to highlight the students in our profession and the future leaders of AFS.

Melissa R. Wuellner and Shannon Fisher

COLUMNS

52 How to Manage Data to Enhance Their Potential for Synthesis, Preservation, Sharing, and Reuse—A Great Lakes Case StudyInvesting in data management is necessary for sharing, documenting, and preserving data that are collected and reported during the scientific research process.

Tracy L. Kolb, E. Agnes Blukacz-Richards, Andrew M. Muir, Randall M. Claramunt, Marten A. Koops, William W. Taylor, Trent M. Sutton, Michael T. Arts, and Ed Bissel

65 SuperIDR: A Tool for Fish Identification and Information RetrievalSuperIDR is a software package for computer-facilitated fish identification and access to ecological and life history information that will prove useful to students and practicing fisheries professionals.

Uma Murthy, Edward A. Fox, Yinlin Chen, Eric M. Haller-man, Donald J. Orth, Ricardo da S. Torres, Lin Tzy Li, Nádia P. Kozievitch, Felipe S. P. Andrade, Tiago R. C. Falcao, and Evandro Ramos

FEATURES

Cover: Painting - Lake Whitefish. Credit: Sarah Gilbert.

83 AFS 2013 Little Rock: Preparing for the Challenges Ahead

LITTLE ROCK MEETING UPDATE

93 Transactions of the American Fisheries Society, Volume 142, Number 1, January 2013

JOURNAL HIGHLIGHTS

87 The Montana State University Student Subunit of the American Fisheries Society (MSUAFS)

Ben Galloway

UNIT NEWS

96 February 2013 Jobs

ANNOUNCEMENTS

95 Fisheries Events

CALENDAR

84The volume of sand that moved into neighborhoods extensively damaging homes during Hurricane Sandy gives evidence of the historic and natural processes that would otherwise occur to replenish or build barrier islands.

76 Ecological Networks in Fisheries: Predicting the Future?Villy Christensen’s 2012 Twin Cities Plenary Talk—a chron-icle of ecological networks (food webs) in fisheries and his personal journey.

Villy Christensen

PLENARY SESSIONS

84 Beyond Biology: Can We Reconcile the Land–Water– Humankind Interface?

Larry Dominguez

HURRICANE SANDY RETROSPECTIVE


MEMBERSHIP TYPE/DUES (Includes print Fisheries and online Membership Directory)

Developing countries I (Includes online Fisheries only): N/A NORTH AMERICA; _____$10 OTHERDeveloping countries II: N/A NORTH AMERICA; _____$35 OTHERRegular: _____$80 NORTH AMERICA; _____$95 OTHERStudent (includes online journals): _____$20 NORTH AMERICA; _____$30 OTHERYoung professional (year graduated): _____$40 NORTH AMERICA; _____$50 OTHERRetired (regular members upon retirement at age 65 or older): _____$40 NORTH AMERICA; _____$50 OTHERLife (Fisheries and 1 journal): _____$1, 737 NORTH AMERICA; _____$1737 OTHERLife (Fisheries only, 2 installments, payable over 2 years): _____$1,200 NORTH AMERICA; _____$1,200 OTHER: $1,200Life (Fisheries only, 2 installments, payable over 1 year): _____ $1,000 NORTH AMERICA; _____$1,000 OTHER

JOURNAL SUBSCRIPTIONS (Optional)

Transactions of the American Fisheries Society: _____$25 ONLINE ONLY; _____$55 NORTH AMERICA PRINT; _____$65 OTHER PRINT North American Journal of Fisheries Management: _____$25 ONLINE ONLY; _____$55 NORTH AMERICA PRINT; _____$65 OTHER PRINT North American Journal of Aquaculture: _____$25 ONLINE ONLY; _____$45 NORTH AMERICA PRINT; _____$54 OTHER PRINT Journal of Aquatic Animal Health: _____$25 ONLINE ONLY; _____$45 NORTH AMERICA PRINT; _____$54 OTHER PRINT Fisheries InfoBase: ____$25 ONLINE ONLY

Recruited by an AFS member? yes noName

EMPLOYERIndustryAcademiaFederal gov’tState/provincial gov’tOther

FisheriesAmerican Fisheries Society • www.fisheries.org

EDITORIAL / SUBSCRIPTION / CIRCULATION OFFICES5410 Grosvenor Lane, Suite 110•Bethesda, MD 20814-2199(301) 897-8616 • fax (301) 897-8096 • [email protected]

The American Fisheries Society (AFS), founded in 1870, is the oldest and largest professional society representing fisheries scientists. The AFS promotes scientific research and enlightened management of aquatic resources for optimum use and enjoyment by the public. It also encourages comprehensive education of fisheries scientists and continuing on-the-job training.

Fisheries (ISSN 0363-2415) is published monthly by the American Fisheries Society; 5410 Grosvenor Lane, Suite 110; Bethesda, MD 20814-2199 © copyright 2013. Periodicals postage paid at Bethesda, Maryland, and at an additional mailing office. A copy of Fisheries Guide for Authors is available from the editor or the AFS website, www.fisheries.org. If requesting from the managing editor, please enclose a stamped, self-addressed envelope with your request. Republication or systematic or multiple reproduction of material in this publication is permitted only under consent or license from the American Fisheries Society. Postmaster: Send address changes to Fisheries, American Fisheries Society; 5410 Grosvenor Lane, Suite 110; Bethesda, MD 20814-2199. Fisheries is printed on 10% post-consumer recycled paper with soy-based printing inks.

2013 AFS MEMBERSHIP APPLICATIONAMERICAN FISHERIES SOCIETY • 5410 GROSVENOR LANE • SUITE 110 • BETHESDA, MD 20814-2199

(301) 897-8616 x203 OR x224 • FAX (301) 897-8096 • WWW.FISHERIES .ORG

PAID:

NAME

Address

City

State/Province ZIP/Postal Code

CountryPlease provide (for AFS use only)Phone

Fax

E-mail

PAYMENTPlease make checks payable to American Fisheries Society in U.S. currency drawn on a U.S. bank, or pay by VISA, MasterCard, or American Express.

_____Check

_____American Express

Account #______________________________________

Exp. Date _____________

Signature ______________________________________

_____VISA

_____MasterCard

All memberships are for a calendar year. New member applications received Janu-ary 1 through August 31 are processed for full membership that calendar year (back issues are sent). Applications received September 1 or later are pro-cessed for full membership beginning January 1 of the following year.

AFS OFFICERSPRESIDENTJohn Boreman

PRESIDENT ELECTRobert Hughes

FIRST VICE PRESIDENTDonna L. Parrish

SECOND VICE PRESIDENTRon Essig

PAST PRESIDENTWilliam L. Fisher

EXECUTIVE DIRECTORGhassan “Gus” N. Rassam

FISHERIES STAFFSENIOR EDITORGhassan “Gus” N. Rassam

DIRECTOR OF PUBLICATIONSAaron Lerner

MANAGING EDITORSarah Fox

EDITORSSCIENCE EDITORSMarilyn “Guppy” Blair Jim BowkerHoward I. BrowmanMason BryantSteven R. ChippsSteven CookeKen CurrensAndy DanylchukMichael R. DonaldsonAndrew H. FayramStephen FriedLarry M. GigliottiMadeleine Hall-ArborAlf HaukenesJeffrey E. HillDeirdre M. Kimball

DUES AND FEES FOR 2013 ARE:$80 in North America ($95 elsewhere) for regular members, $20 in North America ($30 elsewhere) for student members, and $40 ($50 elsewhere) for retired members.

Fees include $19 for Fisheries subscription.

Nonmember and library subscription rates are $174.

Jim LongDaniel McGarveyRoar SandoddenJeff SchaefferJesse TrushenskiUsha Varanasi Jack E. WilliamsJeffrey Williams

BOOK REVIEW EDITORFrancis Juanes

ABSTRACT TRANSLATIONPablo del Monte Luna


The 1993 American Fisheries Society (AFS) annual meet-ing was held in Portland, Oregon, and was the first AFS govern-ing board meeting I attended as a member (I was president of the Marine Fisheries Section). Carlos Fetterolf, AFS president at the time, chaired the meeting. Every time someone in the meeting used the term “national” in reference to the society-wide level, Carlos demanded a 25-cent donation to the meeting kitty. His intent was to get people out of the habit of using the term and into the habit of recognizing that AFS members repre-sent more nations than just the United States. Nowadays I rarely hear the term national used at AFS meetings and, when I do, I am tempted to ask for a quarter.

The AFS is indeed an international society. My latest count indicates that AFS membership now extends to 63 nations out-side of the United States. We now have a Canadian Aquatic Re-sources Section and a Mexico chapter in the Western Division. We also have a very active International Fisheries Section that provides the AFS with a means of reaching out to our interna-tional members, keeping them informed of AFS activities, and making them feel welcome when they attend our meetings.

One of the major international roles that AFS plays is as a member of the World Council of Fisheries Societies (WCFS), which is composed of fisheries and aquatic science professional societies in North America, Asia, Europe, and Australia. Re-cently, the Mexican Fisheries Society and Canadian Aquatic Resources Section joined the World Council, representing fish-eries professionals in their respective countries. Our executive director, Gus Rassam, serves as the World Council’s executive secretary, and AFS member Doug Beard is its current presi-dent. The mission of the WCFS is to promote international co-operation in fisheries science, conservation, and management. This past year, the WCFS sponsored the 6th World Fisheries Congress in Edinburgh, UK, which had a large number of AFS members in attendance; the 7th will be held in 2016 in Pusan, Korea.

Reasons for increasing AFS involvement in global fisheries science, management, and conservation issues, such as those being addressed through the World Council, should be obvious. Problems that our members are addressing in North America are common throughout the world. Prominent global fisheries is-sues include the role of aquaculture as a source of much-needed protein, especially in underdeveloped regions, and its impact on local ecosystems. Increased trade among nations has led to a concurrent increase in introductions of invasive aquatic species. Identification, treatment, and prevention of fish and shellfish diseases in aquaculture and sea farming operations, as well as in the wild, have become a global concern. Impacts of alternative modes of generating electricity, such as offshore wind farms, on aquatic ecosystems and fishing practices have also become an

increasing cause for con-cern. Overharvest of fish-ery stocks is still occurring throughout the world, re-quiring closer international cooperation in develop-ment of stock assessment techniques, genetics-based methods of stock identifi-cation, economic analyses of import and export mar-kets, and ecosystem-based approaches to fisheries management and conservation. Finally, global-scale environ-mental issues affecting fisheries resources (e.g., climate change, ocean acidification, eutrophication of inland and coastal waters, and chemical contamination) are drawing increased attention from our world leaders. With over 9,000 members represent-ing 64 nations, the AFS is in an ideal position to address these global issues and, coordinating closely with the other members of the World Council, ensure that the best scientific and man-agement approaches are being applied.

This is the message I intend to deliver at the annual meeting of the Japanese Society of Fisheries Science (JSFS) meeting in Tokyo next March and the Fisheries Society of the British Isles (FSBI) meeting in Glasgow in July 2013. Through the World Council, the AFS, JSFS, and FSBI have had an exchange pro-gram in which the societies’ leaders are invited to address each other’s annual meetings. At the AFS 2012 annual meeting in St. Paul, Shuichi Satoh, representing the JSFS, and Ian Win-field, representing the FSBI, spoke at our business meeting. In addition, the AFS, JSFS, and FSBI have had a tradition of co-sponsorship of symposia at their respective scientific meetings. However, the AFS and other members of the World Council need to do more than have their society officers deliver a few words of salutation at business meetings and cosponsor sym-posia.

Ideally, cooperation among professional fisheries societies should involve exchange programs that accommodate visiting scientists and managers or student internships. The societies could assist in locating opportunities within their respective spheres of influence, screening candidates, and seeking finan-cial assistance. Garnering financial assistance is most often the bottleneck for these opportunities and will likely not get easier in the current global economic climate. Short of cosponsoring professional and student exchange programs with our fellow societies, increased sharing of information related to the science and management of fisheries can probably be undertaken with

COLUMNPresident’s Hook

Continued on page 94

AFS President Boreman may be contacted at: [email protected]

Think Globally, Act GloballyJohn Boreman, President


How to Manage Data to Enhance Their Potential for Synthesis, Preservation, Sharing, and Reuse—A Great Lakes Case Study

ABSTRACT: Proper data management (applying coordinated standards and structures to data collection, maintenance, re-trieval, and documentation) is essential for complex projects to ensure data accuracy and accessibility. In this article, we used a recent project evaluating changes in Lake Whitefish (Core-gonus clupeaformis) growth, condition, and recruitment in the Great Lakes as a case study to illustrate how thoughtful data management approaches can enhance and improve research. Data management best practices described include dedicating personnel to data curation, setting data standards, building a relational database, managing data updates, checking for and trapping errors, extracting data, documenting data sets, and coordinating with project collaborators. The data management actions taken ultimately resulted in a rich body of scientific publication and a robust database available for future studies.

Tracy L. KolbMichigan Department of Natural Resources, Charlevoix Research Station, 96 Grant Street, Charlevoix, MI 49720. E-mail: [email protected]

E. Agnes Blukacz-RichardsGreat Lakes Laboratory for Fisheries and Aquatic Sciences, Fisheries and Oceans Canada, 867 Lakeshore Rd., Burlington, ON L7R 4A6, Canada

Andrew M. MuirGreat Lakes Fish Commission, 2100 Commonwealth Blvd., Suite 100, Ann Arbor MI, 48105, and Michigan State University, 14 Natural Resources Building, East Lansing, MI 48824

Randall M. ClaramuntMichigan Department of Natural Resources, Charlevoix Research Station, 96 Grant St., Charlevoix, MI 49720

Marten A. KoopsGreat Lakes Laboratory for Fisheries and Aquatic Sciences, Fisheries and Oceans Canada, 867 Lakeshore Rd., Burlington, ON L7R 4A6, Canada

William W. TaylorMichigan State University, 14 Natural Resources Building, East Lansing, MI 48824

Trent M. SuttonUniversity of Alaska Fairbanks, School of Fisheries and Ocean Sciences, 905 N. Koyukuk Dr., Fairbanks, AK 99775

Michael T. ArtsEnvironment Canada, 867 Lakeshore Rd., Burlington, ON L7R 4A6, Can-ada

Ed BisselMichigan State University, 14 Natural Resources Building, East Lansing, MI 48824

Cómo manejar datos para incrementar el potencial para su síntesis, preserva-ción, intercambio y reutilización –los Grandes Lagos como caso de estudioRESUMEN: en proyectos complejos, un manejo ap-ropiado de datos (aplicación coordinada de estándares y estructuras a recolección, mantenimiento, recuperación y documentación) resulta esencial para asegurar la precisión y accesibilidad de los mismos. En la presente contribu-ción se utiliza un proyecto de evaluación de los cambios en el crecimiento, condición y reclutamiento del coregono en los Grandes Lagos, como caso de estudio para ilustrar cómo un manejo adecuado de datos puede incrementar y mejorar la investigación. Las mejores prácticas en cuanto a manejo de datos incluyen: dedicar personal a la curación de datos, fijar estándares en los datos, construcción de una base de datos relacional, manejo de actualización de datos, revisión y filtro de errores en los datos, extracción de datos, documentación de bases de datos y coordinación con co-laboradores del proyecto. Las acciones de manejo de datos que se tomaron resultaron en la producción de un cuerpo importante de publicaciones y en una base de datos robusta, disponible para investigaciones futuras. Los recursos inver-tidos en el manejo de datos permitieron que este proyecto sirviera de modelo para tomar los primeros pasos hacia el objetivo común de compartir, documentar y preservar datos que son recolectados y reportados durante el proceso de una investigación científica.

FEATURE

Investing in data management allowed this project to serve as a model for taking the first steps toward a common goal of shar-ing, documenting, and preserving data that are collected and reported during the scientific research process.

CONTEMPORARY FISHERIES RESEARCH NEEDS THOUGHTFUL DATA MANAGE-MENT PRACTICES

Data are the infrastructure of science, and modern scientific architecture has become increasingly complex. This trajectory can be partly explained by the preference of granting agencies toward projects that address broad-scale research questions; partly by advances in computing and communications technol-ogy that allow the scientific community to work with larger


data sets that transcend conventional spatial, temporal, and disciplinary boundaries (Lélé and Norgaard 2005; Wake 2008; Carpenter et al. 2009); and partly by advances in computing that have allowed data-intensive science (Newman et al. 2003) and modeling projects that rely on previously collected data to increase in frequency and magnitude (Kelling et al. 2009; Borg-man 2010).

Any research project with multiple objectives or one that combines the expertise of multiple principal investigators—or even one that simply combines data from multiple institutions—will have the capacity to generate immense quantities of varied information, require the assimilation of previously acquired data, or both. This raises a variety of logistical complexities with regard to quality control, security, and accessibility of data and, as such, these projects can benefit greatly from formal data management strategies for entry, update, storage, validation, ac-cess, annotation, provenance (i.e., information regarding the or-igins, identification, ownership and structure of a data set), and archiving (McDonald et al. 2007; Brunt and Michener 2009; Kelling et al. 2009). Recognizing this, many funding agencies now require that all prospective grantees address data manage-ment as part of the project application (National Institutes of Health 2003; U.S. Fish and Wildlife Service 2006; National Oceanic and Atmospheric Administration [NOAA] 2010; Na-tional Science Foundation 2010).

Unfortunately, modern ecological data management prac-tices have not evolved as quickly as their data sets (Katz et al. 2007; McDonald et al. 2007; Barnas and Katz 2010; Hook et al. 2010). Data management is an often underrecognized and underutilized tool (Michener and Jones 2011). The majority of scientists still manage data through spreadsheet entry, indi-vidualized post-entry error checking and manual grouping, or extraction of data for analysis (Porter and Ramsey 2002; Borg-man et al. 2007; Nelson 2009) A recent survey of ecologists found that they felt that their own institutions lacked planning, technology, and funding for data management in the short term (during the project) and long term (post-project) and did not adequately provide training in data management (Tenopir et al. 2011). Heterogeneity in the practices and quality of data man-agement limits data reuse, data sharing, and data integration and does not facilitate standardization of data collection methods or support economic efficiency given current fiscal climates.

A fundamental disconnect occurs between the broadly based, complex, interdisciplinary, and collaborative projects requiring data that are accessible, electronic, decipherable, er-ror-free, and reusable and the heterogeneous and idiosyncratic data sets that are routinely being generated from the thousands of fisheries researchers collecting data in the course of their work. Fisheries managers and scientists must embrace the need to recognize data management as a critical step toward organiz-ing their discipline and resolving this tension.

THE STATES OF FISHERIES DATA MANAGEMENT AND PEER-REVIEWED LITERATURE

Scientific data collection and compilation can occur at dif-fering spatial scales, and the larger the scale, the more necessary it is to commit resources to data collection and management. Some examples of larger scale regional fisheries database ef-forts include FishMAP, a Great Lakes fish migration passage and knowledge database; GLATOS Web, a Great Lakes acous-tic telemetry database; the Multistate Aquatic Resources In-formation System (MARIS); the National Fish Habitat Action Plan; the Pacific Northwest Salmon Habitat Restoration Project Tracking Database; StreamNet, which compiles and dissemi-nates fish data from state, tribal, and federal agencies in the Pacific Northwest; and the Fisheries Information Networks (FINs), which are regional, cooperative, state, and federal data integration and management programs for the Pacific Region (PACFIN), the Atlantic Region (Atlantic Coastal Cooperative Statistics Program [ACCSP]), the Gulf of Mexico (GulfFIN), and Alaska (AKFIN; e.g., Beard et al. 1998; Katz et al. 2007; MARIS 2008; McLaughlin et al. 2010; Wang et al. 2011). In addition, there are regional fisheries databases housed at NOAA Fisheries Service Science Centers, which have long histories of managing data (NOAA 2011). At a smaller scale than these regional efforts, there are the data management endeavors of individual state agencies, coordinating groups of multi-affili-ated fisheries researchers, university fisheries research teams, and the many individual projects that require the construction of databases during the course of their research (e.g., Watson and Kura 2006; Katz et al. 2007; Heidorn 2008; Frimpong and Angermeier 2009).

The current state of “how to manage fisheries-specific databases” in the peer-reviewed literature can be summarized as follows: the regional efforts listed above have multiple per-sonnel dedicated to behind-the-scenes data management, using very sophisticated practices, but detailed descriptions of their specific efforts have not been documented in the peer-reviewed fisheries literature (e.g., K. Barnas, Pacific Northwest Salmon Habitat Restoration Project Tracking Database; D. Donald-son, GulfFIN; D. Infante, National Fish Habitat Action Plan; W. Kinney, StreamNet; C. Kruger, GLATOS Web; A. Loftus, MARIS; E. Martino, ACCSP; R. McLaughlin, FishMAP, per-sonal communication). There are also countless textbooks on the structural mechanics of database design (Hernandez 2003; Pratt and Adamski 2007; Ling Liu and Özsu 2009), which tend to ignore the specialized needs of the scientific field. Finally, there are specific fisheries projects that required construction of a database for which the results of the findings have been published, but details of the data management plans have not (Watson and Kura 2006; Katz et al. 2007; Frimpong and An-germeier 2009; Whiteed et al. 2012). Very few generalized de-scriptions detailing both the technical and practical aspects of managing data generated by typical collaborative research proj-ects are available to fisheries professionals to use as a resource (McLaughlin et al. 2001; Baker and Stocks 2007).


For individuals or teams of fisheries scientists collecting data independent from regional database efforts, formal data management guidance is not readily available as a resource, yet project-specific data management plans are increasingly re-quired as a prerequisite for research grant applications. There-fore, the purpose of this article is to provide a synthesis of data management best practices for the typical fisheries investiga-tor to serve as kind of a broad proxy for grant application and research plans while underscoring the added value that can be accrued by using these best practices. These best practices sup-port data integrity throughout the project and position data to be reusable by future users by ensuring that they are accessible, electronic, decipherable, and error-free.

We used a recent collaboration among federal, state, and university fisheries researchers as a case study to highlight how data management works. Although data management practices are generally masked during the publication process, the authors feel that they are a fundamentally important part of scientific inquiry and communication and therefore should be subject to the same rigorous evaluations and discussions in the primary lit-erature as other scientific methods. Not all fisheries projects will require all of the steps described subsequently, but we hope that this article serves as a guide for researchers asking themselves at the start of a new endeavor, “To what extent and how should data management be implemented for this project?”

LAKE WHITEFISH CASE STUDY BACKGROUND

In 2004, two teams of scientists from Purdue University and Fisheries and Oceans–Canada independently submitted pre-proposals to the Great Lakes Fishery Trust requesting grant money to study Lake Whitefish (Coregonus clupeaformis) re-cruitment dynamics in the Great Lakes (Sutton et al. 2007). In a rare occurrence, the reviewers felt that the projects were similar enough to suggest that the two teams collaborate (M. Coscarelli, personal communication). Recognizing that choos-ing not to collaborate meant competing as two similar projects vying for a limited pool of money, the groups merged, submit-ted a full proposal for one project that addressed two different sets of potential Lake Whitefish recruitment impediments, and received funding for 3 years, where yearly funding depended on the success of the collaboration. The researchers convened as a group to discuss issues associated with data management (Table 1). The result of that discussion was agreement that the expanded project and conditional nature of the funding required implementing formal data management practices and a decision was made to allocate project resources to obtain a dedicated data curator as a permanent member of the research team.

DATA MANAGEMENT BEST PRACTICES

Selecting a Data Curator

A data curator is responsible for the technical and practical aspects of data management throughout a research project—al-though for large, complex projects, data curation is often done

by a team of individuals, which may include subject-matter ex-perts, data users, information technology staff, computer pro-grammers, and a metadata librarian (Lord et al. 2004; Cragin et al. 2008; Akmon et al. 2011). A curator’s major responsibilities are to incorporate, organize, document, and retrieve data that they curate (Heidorn 2008; Witt 2009; Witt et al. 2009). The curator adds value to the research project by checking, verify-ing, and correcting data sets, as well as by providing software tools for data access, manipulation, and assimilation of any previously collected data, if required (Research Information Network [RIN] 2008; Cragin et al. 2010). Data curators apply rigorous procedures to ensure that the data sets they manage meet quality standards in relation to the structure and format of the data themselves (examples given in the following sections), ultimately contributing value by making data more discover-able and easier to access for potential reuse. A dedicated cura-tor combines the benefits of expertise available to researchers in disciplines with centralized data repositories with the agility and advantages of localized data storage and management (RIN 2008). Though formal training in database design and manage-ment is ideal, a data curator need not be a professional database developer or computer programmer; he or she can simply be someone who has experience and is comfortable managing data. Our data curator was a postdoctoral researcher with experience managing modest-sized (<1 million records) databases obtained during past research projects.

Establishing Data Requirements

Before any Lake Whitefish project data collection occurred, our curator’s job was to determine what data were going to be

TABLE 1. Discussion items that help identify the data management needs for a collaborative research project.

Given that we want to store all project data together, does a single member of the research team have the skill set and time to manage data for the entire project? Do we know someone reliable but outside of the research team who could curate the data?

How much data will we be collecting? What is the maximum size of our data set?

Once we have collected data, will housing them require multiple tables? Can we use “flat file” (single data table) organization or do we need a re-lational database?

How complicated will data entry be? How many different people will be entering data, at how many different locations? The more complex data entry, the greater the probability of errors and the more dedicated error oversight required.

If multiple PIs are working on separate parts of the project, how important is it that their data be able to interact? Do the PIs need to combine data to answer research questions? If so, properly defining relationships among data is critical.

Does our grantor require data management or data sharing as part of our grant stipulations? Will our data be shared beyond the PIs?

If we need to use a relational database, how much will it change through time? How many researchers will need to access identical data simultane-ously but separately? Will version control be critical to ensure that everyone is accessing the same data?

Are our data unique and can they be reproduced? Will we want to draw from these data sets for future studies? Is it worth the investment to pre-serve our data?


collected and by whom and who would be responsible for post-collection processing of data (Hernandez 2003; Pratt and Adam-ski 2007; Ling Liu and Özsu 2009). Three universities, two U.S. natural resources agencies, three Canadian government agen-cies, one tribal resource agency, and various commercial fishing operators worked together to collect data at 13 sites across lakes Michigan, Superior, and Erie from 2004 to 2006 (Figure 1). Adult, juvenile, and larval Lake Whitefish were sampled using gill and trap nets, beach seines, and plankton nets, respectively. Sampling effort parameters (e.g., date and location) and envi-ronmental data (e.g., water temperature) were collected during each sampling event. Biological data (e.g., length and weight) were collected on each life stage of the Lake Whitefish and their prey. Subsequent laboratory analyses resulted in the generation of further biological and physiological data (e.g., food hab-its, proximate body composition, and fatty acid composition). These collections resulted in a data set with more than 250,000 records. To ensure that data collection was standardized, our curator held initial meetings with individual project collabora-tors and collective meetings with the research group. During the individual meetings, the curator asked the collaborators the following questions:

• How are data defined; what formats will these data take (e.g., numbers, pictures, acoustic records, physi-cal specimens, etc.); what are the units of measure-ment associated with numerical data; which data are textual? What information about data collection will be archived (e.g., sampling effort data such as weather, collection gear, sampling crew names, etc.)? How many records will be generated seasonally and over the entire project?

• How will data be captured or created (e.g., research vessel, fish tagging, moored buoy, online surveys, etc.)?

• What are the spatial and temporal coverage of data col-lections?

• Once data are collected, will they be postprocessed? If so, where will they be sent and what processes will occur?

• What are the timelines for data collection and post-processing? Are data being collected all at once or throughout a season? Are data being generated and recorded continually or in batches?

• How soon after processing or collection will data be sent to the curator for input into a database and how soon after input will data be needed for analysis? Will data be transferred all at once or in batches?

• How do data relate to other data (e.g., will a sampling event be related to multiple fish caught during that event, or will multiple stomach contents be related back to an individual fish)?

An initial meeting with the entire research team allowed for the development and documentation of a predefined set of standards for coding categorical data, such as sampling loca-tions and Linnaean names of fish and invertebrate species (we recommend using standard Integrated Taxonomic Information System codes) and to determine how spatial and environmental

Figure 1. Schematic of the complexity involved in collecting and processing data for a Great Lakes Fishery Trust–funded Lake Whitefish project (Sutton et al. 2007). Lines indicate data exchange between entities. Solid lines represent data exchange between collector and processor. Dashed lines rep-resent data exchange between a collector or processor and the data curator. N.B., schematic organizes data collectors in rows for formatting reasons and does not imply any type of hierarchy.


data would be captured and classed. For example, collaborators agreed to standardize classification of all nets as trap, gill, seine, or plankton. These initial meetings helped the collaborators to improve their understanding about the scope of the project and facilitated standardizing field sampling methods. Although it seems intuitive that collaborative projects would function this way (i.e., with or without a dedicated curator), the lack of a dedicated person accountable for bringing data management is-sues to the group and forcing standardization at the onset of the project often results in a situation whereby issues, sometimes uncorrectable, are overlooked until much later in the research process, increasing the time required to identify and correct er-rors (McLaughlin et al. 2001; Wallis et al. 2008).

Creating and Populating the Database

After becoming familiar with project data and collabora-tors’ collection procedures, a data model was created. The data model and associated entity relationship diagrams identified all attributes (i.e., data elements) in each table and defined how tables related to each other through key fields. Our model called for seven primary data tables and 11 lookup tables (Figure 2, Table 2). After the data model was created, we developed a means for data storage (i.e., relational database) using Microsoft Office Access software. (Special note on spatial data: There are two options when working with spatial data: (1) using an aspa-tial, tabular, relational database where records have a unique identifier, which can be linked to geocoordinates in a second, separate spatial database; or (2) working exclusively within a single spatially enabled database [supports spatial data types]. In the second case, each individual record’s spatial geometry is stored as an attribute and the database is integrated with the

spatial software. We used the first option and stored only the latitude and longitude of our sampling events, importing those spatial attributes into ArcGIS when we needed to make maps or wanted to do specialized spatial analyses.)

Although we used Microsoft Access to develop our data-base, many relational database management software programs (RDBMSs) are available to researchers (Table 3).When select-ing an RDBMS, researchers should consider the advantages and disadvantages of the price, required operating system, compat-ibility with other software programs, user accessibility, level of technical expertise, anticipated upgrade costs (time and money), and constraints imposed by the quantity of data to be managed.

Most scientists use spreadsheet software to store their data (e.g., Excel), rather than an RDBMS (Borgman et al. 2006). Though both spreadsheets and RDBMSs organize data in rows (in data storage language these are called “records”) and col-umns (“fields”), spreadsheets store data in individual tables as “flat files,” meaning that tables are not linked, whereas RD-BMSs store data across multiple, interrelated tables, with the expectation that the user will primarily want to work with data across multiple tables simultaneously. Storing data using linked tables is the foundation of the relational database.

If data can be stored in a single table, a relational database is not necessary. If multiple tables are required to store data, cre-ating a relational database using an RDBMS is the best option, because relational databases have rules that maximize data in-tegrity across tables (Hernandez 2003). Generally, data integrity rules include the following: (1) tables that are constructed prop-erly and efficiently (i.e., each table represents a single entity,

every column in each table is comprised of distinct fields, fields are not repeated within a table, and each record is iden-tified with a unique value called a “pri-mary key” used for linking data among tables); and (2) data integrity (validity) is imposed at the record, table, and relation-ship levels (i.e., every table has a column for the key field and keys are used to create re-lationships among tables).

An experienced curator is able to har-ness the strengths of the relational da-tabase model and

Figure 2. Data model—design of the Lake Whitefish recruitment database indicating the primary data tables, relation-ships between tables, and fields used as keys in the relationships. Additionally, 11 lookup tables (not included in figure) standardized the entry of location, gear, weather conditions, fish species, life stage, sex, maturity, invertebrate species, fatty acid, age structure, and prey species data (see Table 2). The ∞ symbol represents a one-to-many relationship between table IDs. For example, one sampling ID in the sampling table can relate to more than one fish or invertebrate ID in the fish and invertebrates tables, or one fish ID in the fish table can relate to more than one lipid, age, stomach, or reproduction ID in their respective tables.


software by taking advantage of the built-in procedural logic that relates information among tables, allowing users to focus solely on using declarative logic to extract data; for instance, when combining data from multiple spreadsheets the user has to manually relate data among different tables, whereas in the relational database the user simply has to indicate data for ex-traction because relationships between tables are predefined. The chances are lessened that data will become useless if knowledge of those relationships is ever lost because the re-lationships among data are required to be declared explicitly. Using an RDBMS, data tables can adapt to changing sampling designs and protocols without necessitating structural changes so that new data can easily be incorporated in the future. RD-BMSs also offer several advantages related to data integrity and quality compared to spreadsheets. Properties of atomicity, consistency, isolation, and durability (ACID) describe the vari-ous mechanisms that the underlying RDBMS software uses to ensure data integrity between transactions (Haerder and Reuter 1983). Though spreadsheets optimize flexibility and ease of use by pairing data storage with visualization, RDBMSs optimize data integrity through ACID principles (Haerder and Reuter 1983; Hernandez 2003).

One foundation of ACID principles is the key field (pre-viously mentioned), which is defined as a unique identifier that links data across tables. Keys provide the quickest way to retrieve data when searching or sorting and make it easy to

summarize data from multiple tables. Keys can assume multiple formats as long as they are unique. The simplest format for a key is an autonumber, where each new record is assigned a sequential number starting at 1. In the Lake Whitefish recruitment database, the key for identifying each individual Lake Whitefish was a con-catenation of sampling date, sampling loca-tion, life stage, and fish ID (e.g., 05_06_2007_ElkRapids_AD_001); this allowed the key itself to convey mean-ing and to function as more than just a serial number.

Finally, RDBMSs offer several addi-tional advantages over spreadsheets, includ-ing the ability to store,

manage, and analyze data sets of considerably larger size. RD-BMSs run so efficiently because they only retrieve data required through a user-specified query, whereas spreadsheets load the entire data set into memory when the spreadsheet file is opened. In addition, the ability to partition a database into multiple files across multiple hard disks can reduce disk contention (bottle-necks caused by multiple processes accessing the same location on disk at the same time), making large and complex databases easier to work with. Additionally, RDBMSs use indexing to speed up which query results are returned for large data sets by reducing the number of records that must be scanned to return the desired result (Ling Liu and Özsu 2009).

Version Control

Sampling and postprocessing of samples collected for the Lake Whitefish project occurred over 3 years; therefore, the co-ordination of data submittal and updates to the database was done semiannually by the data curator. Every time new data were uploaded or existing data were corrected, a new version of the database was created.

It is critical that the curator exert control over the perpetua-tion of multiple versions of a single database. If version control is not implemented, different versions of files, related files at different locations, and information cross-referenced among files are all subject to the viral phenomenon of cascading repli-

TABLE 2. Database table names (columns) and non-key fields (rows) for the Lake Whitefish recruitment database. Italics indicate fields linked to lookup tables.

Sampling Fish Invertebrates Lipidsb,c Agea,b,c Stomachb, c Reproductionc

Date Species Species Total lipids Age Stomach weight Gonad weight

Time Life stage Length Fatty acid type Structure used to determine age

Prey species Egg diameter

Location Length Weight Fatty acid concentration

Prey frequency Egg weight

Latitude Weight Sex Prey weight Sperm velocity

Longitude Sex Biomass Sperm tail length

Depth Maturity Density Sperm cell volume

Gear Liver weight Milt volume

Tow speed Body condition Mean spermatocrit

Tow distance

VFI1

Ambient conditions

Protein2

Energy2

Moisture2

1 Visceral fat index; 2 in muscle tissue.Alphabetic superscripts delineate data collected for life stages or groups as follows:a larval Lake Whitefish.b age-0 juvenile Lake Whitefish.c adult Lake Whitefish.


cates being distributed, utilized, and reproduced. How version control is implemented depends on whether there are single or multiple users and whether versions across space and time need to be synchronized (Van den Eynden et al. 2011). Our suggested best practices for a version control strategy are listed in Table 4.

One of the most elegant advantages of using relational data-bases is that they can be programmed to be as self-documenting as the users require. “Self-documenting” refers to the process in which data transactions are logged along with their identify-ing features, such as who ordered the transaction, a time/date stamp identifying when the transaction occurred, and the na-ture of the specific transaction. We designed our database to be self-documenting in the sense that all changes to data were recorded in the database itself, so users could query which data were updated, by whom, and when. Each time the database went through major updates, a description of what occurred was pro-vided to users along with the new version. We implemented versioning control using an FTP site, because the site offered security and ease of distribution with minimal upfront program-ming. At the end of the project, our curator ensured that each collaborator was provided with a final version of the database that contained not only verified data but also previous versions of the database with records of any updates that occurred during the course of the project. This allowed the collaborators to audit revision history and recover deleted information if necessary.

Quality Control and Standardization

One of our curator’s most valuable contributions to the Lake Whitefish recruitment project was quality control and standardization of data. Initially quality control involved coor-dinating the group to use standardized naming conventions and code lists during field collection and data processing. A second-

ary layer of quality control operated inherent in the relational design model and embedded integrity rules—eliminating duplicates, re-ducing redundancy, increasing consistency, and maximizing accuracy of data (Hernandez 2003; Borer et al. 2009; Hook et al. 2010). The creation of formal data definitions disallowed data of one type to be entered as another type (e.g., wind direction was stored as text so it was impossible to accidentally store it as a number). Our curator also placed checks on data ranges and coding through lookup tables for gear types, sampling sites, and fish identi-fication codes and used input masks to ensure that data were formatted consistently.

One data management best practice is to audit, capture, and intercept suspect values as close to the source of the collection as possible because reworking data long after the actual sampling is complete is more time consuming and can result in unusable samples. To verify that data were accurate, our curator performed simple reviews of data by periodically check-

ing for missing values; verifying that data were in their proper columns; scanning for impossible values; and generating simple statistics such as frequencies, means, ranges, and clusters to de-tect errors or anomalous values (e.g., the lengths of all age-0 juveniles should fall between 20 and 70 mm; the latitude and longitude for Little Bay de Noc in northern Lake Michigan can-not map to southern Lake Michigan, etc.). Our curator also es-tablished more sophisticated error traps by communicating with project collaborators to identify what errors were most common to their data contributions. For instance, because Lake White-fish eat small diet items, if a diet item exceeded 20% of the total length of the predator, the sample was flagged. Sampling event data and individual fish biological data were entered into the database at different times and in different places, so our curator created a query to verify that the total number of fish recorded during a sampling event was equal to the number of biological records for individual fish associated with that sampling effort.

Our curator also verified that collaborators were not violat-ing their own sampling protocol. For instance, the collabora-tors agreed that when sampling adult female Lake Whitefish for fecundity, they would only collect eggs from females that were “green” (i.e., with mature eggs that were still attached to the skein) and not from females that were “running or spent” (i.e., eggs that were free flowing or had already been deposited). After data collection, egg samples were sent to one laboratory and female fish were sent to another for analysis. Upon receipt of data, the curator could use simple queries to easily match up all of the samples to ensure that fecundity was not estimated for fish that fell outside of the maturity requirements.

Data Extraction Scientists may not have been comfortable extracting data

from a relational database, yet they still needed to be able to

TABLE 3. Comparison of available technologies to manage data.a Structured query lan-guage (SQL) is a programming language designed for working with relational databases. Other considerations include operating system or integration with other clients (desktop or geographic information system software).

Concerns/needs Spreadsheet—basic

Database—intermediate

Database—advanced

Desktop or server based

Desktop Desktop Server based

Spatially enabled No No Yes

Security No Low High

Multiuser data entry No No Yes

Size of data set Limited Limited Unlimited

Web-based No No Yes

Examples Excel Access, Microsoft SQL Express, SQLLite

SQL Server, Oracle, MySQL

Cost Low Low High (although some open-source RDBMSs are available)

Level of programming experience needed

None Little to none Expert

a This table is meant to indicate the general advantages/disadvantages of the different tiers of technologies to manage data, and the characterizations do not hold true in all cases.


easily extract data for analysis. One of the greatest benefits of having a data curator was that collaborators could simply send an e-mail or make a call and ask for data to be assembled and formatted in whatever way was needed and the curator could deliver those data quickly and efficiently. For more general-ized data selections, the curator set up standardized forms in the database with checkboxes that allowed collaborators to select data without assistance. As an overseer of data extractions from the database, our curator could ensure that two collaborators ac-cessing the same data were doing so identically, decreasing the likelihood that different conclusions might be reached because different data were selected for analysis.

Because most statistical software requires data in a flat for-mat, it might seem counterintuitive to take the time to create a database only to extract and flatten data for analysis. However, having taken the time to standardize, assemble, and properly structure data, there is no end to the various combinations that the curator can provide to the collaborators, and extracting data in any flat format takes mere seconds. Multiple statistical soft-ware packages allow selection of data within a database via structured query language (SQL); for example, users of R (R Development Core Team 2011) have several CRAN packages available that retrieve the results from relational databases as entire data frames (R Development Core Team 2011).

Initially, most collaborators felt uneasy allowing a data cu-rator to develop and manage a database because they thought that it might limit their influence over extraction and analysis processes, thereby increasing the distance between themselves and their data. Consequently, our curator ensured that all col-laborators had open access to the database and served as teacher and advisor for those who wanted to learn how to extract their own data, while allowing those who were familiar with data-bases the freedom to extract on their own. In fact, most of the collaborators eventually ended up extracting their own data as the project matured and they became familiar with the database structure and operation.

Data Documentation and Archival

Ultimately, the value of data are enhanced, not exhausted, by their subsequent publication and use (RIN 2008; Whitlock 2011). If data are not properly documented, no one outside of the original collectors will be able to use them properly; and because memories fade, eventually even the data originator may have trouble recalling important information relevant to a data set (Akmon et al. 2011). Broadly, “documentation” (descriptive information about data sets, also called “metadata”) includes the following components: what data are; when they were col-lected; how they were collected; geographic scope of the proj-ect; contact information of collectors; directions for citation; any information relevant to interpretation (e.g., processing that occurred, confounding factors, how missing data were handled, quality assessment, projection information, etc.); and individual definitions for each data field (see Table 5 for an example of data documentation for a single table of the Lake Whitefish da-tabase).

Multiple standards provide models for data documenta-tion; the most comprehensive and broadly applicable are the Federal Geographic Data Committee Content Standard for Digital Geospatial Metadata (FGDC-STD-001-1998) and the International Organization for Standardization standards (ISO 9001:2011). For the Lake Whitefish project, our curator used Federal Geographic Data Committee standards to create a for-mal data dictionary, which was provided to each collaborator at the conclusion of the project. An object linking and embedding reference to the data dictionary was embedded in the data set so that it can continue to be accessed and interpretable into the future. For broadest access, it is best to archive data using open-source formats rather than proprietary formats when possible.

Scientific Value-Added Aspects of Data Management

As is becoming more common in research, field and labo-ratory samples for the Lake Whitefish recruitment project were obtained by multiple collectors at sites separated by substan-tial geographic distances. In consultation with the data curator, researchers were able to efficiently merge field and labora-tory data contained in the relational database and effectively extract them to investigate complex relationships and identify mechanisms related to the effects of declines in Lake Whitefish growth and condition on recruitment potential of populations across the Great Lakes.

For example, all sampling events were recorded in the sam-pling table where each event had its own unique ID. Then, fish caught during a sampling event were stored in their own fish table, where every fish had its own unique ID and the ID of its sampling event. When lipid analyses were done on an individual fish, lipid data were stored in their own lipid table along with the ID from the fish the lipids were extracted from. Thus, even though these pieces of information were being stored in sepa-rate tables, the relational database, which linked the IDs among tables, allowed analyses to be performed across all fields with-out the onerous manual linking required if spreadsheets were

TABLE 4. Version control strategies—best practices.

Identify a single location for the storage of versions (in the Lake Whitefish case a secure FTP site).

Decide how many versions to simultaneously maintain (in the Lake White-fish case, one version).

Uniquely identify versions using a meaningful naming convention, which should include the status of version (e.g., draft, working, final).

Record changes made to each version and maintain old versions for back-ups.

If applicable, manage any merging of entries or edits by multiple users.

Police users so that multiple working versions are not being developed in parallel.

Set permissions to read and write data to the database.

Develop formal procedures for destruction of any master files.

Properly document all version control procedures.


used. Using the links among the sampling, fish and lipids tables, Muir et al. (2010) found that female Lake Whitefish in poorer physiological condition had a tendency to produce age-0 juve-niles with poorer body composition at some sites, but this pat-tern was not evident across all sites. In the same manner, using the linkages among three other tables (sampling, fish, and stom-ach), Claramunt et al. (2010a) were able to partially explain this spatial pattern of juvenile condition by showing that early life history survival was likely dependent on favorable growth early in development, which allows an earlier ontogenetic diet shift to emergent spring macro-invertebrates, demonstrating that the link between parental and juvenile physiological condition was influenced by early life growth rates.

The male contribution to Lake Whitefish reproduction and recruitment potential was explored by Blukacz et al. (2010) using linkages among the fish, reproduction, and lipids tables to show that male fish in better condition tended to produce higher quality sperm, suggesting that males are not irrelevant to Lake Whitefish recruitment potential. In addition, the linkages between the sampling and fish tables were used by Claramunt et al. (2010b) to relate larval fish densities to several abiotic and biotic factors, including adult stock size, abiotic conditions

during incubation, and spring productivity. It was pre-cisely the extracting and combining data using linked tables that enabled the research team to efficiently ad-dress more complex and related questions and provide a more thorough understanding of Lake Whitefish re-cruitment potential.

Follow-on components of the Lake Whitefish recruitment project were also able to benefit from ef-forts to create and curate the Lake Whitefish relational database. For example, some members of the Lake Whitefish recruitment project team secured additional funding to analyze stable isotopes from tissue samples archived during the original project sampling. The proj-ect database made it straightforward to match the stable isotope data to the original project data through the ad-dition of a stable isotope table. Information queried from linkages among the fish, lipids, and new stable isotope tables was used to address questions about the connection between Lake Whitefish condition and prey quality (Fagan et al. 2012) and the use of C:N ratios to predict lipid content (Fagan et al. 2011).

DATA CURATION BEYOND THE SINGLE RESEARCH TEAM

It is the sum of all of the actions our curator took to ensure proper data management throughout the life cycle of the project (setting initial standards, coordi-nating data transfer, building a relational database, managing data updates, error checking/trapping, data extraction, data documentation, teaching, coordinating and communicating with project collaborators) that ul-timately resulted in a rich body of scientific publication

and a robust database available for future study (Brenden et al. 2010). In our experience, the efforts toward effective data man-agement more than justified the time and expense.

But does data management for a single project have ben-efits beyond that project? Does it behoove a granting agency to impose proper data management if the collaborators are already going through the formal publication process? We suggest that it does, because a well-designed and defined database is the equivalent of formal documentation of scientific methods. In a time of restricted financial resources, grantors who want to maximize the scope of their investments would be well served by minimizing the redundancy in data collection that occurs when data sets are lost through lack of proper management and, by extension, archiving.

Devoting resources to data management has benefits even beyond increasing the consistency and accuracy of individual project data. Relational database approaches facilitate inte-gration of information from multiple sources, affording more robust, scientifically defensible decision-making capabilities (McLaughlin et al. 2001; Baker et al. 2005; Baker and Stocks 2007). Effectively documented and structured data sets encour-age data sharing and communication among collaborators by

TABLE 5. Parameter documentation for sampling table in the Lake Whitefish database.

Sampling table

Column Definition Format Example

Date Date of sampling mm/dd/yyyy 06/25/2005

Time Time (military) gear was deployed

hh:mm 13:40

Location N, Naubinway; BBdN, Big Bay de Noc; ER, Elk Rapids; BH, Bailey’s Harbor; L, Ludington; S, Saugatuck, PP, Point Peelee; WP, Whitefish Point; RB, Row-ley’s Bay; M, Muskegon; MC, Manitowec; FI, Fisherman’s Island; BB, Brimley Bay

Text N

Latitude Stores the latitude of the sam-pling site in decimal degreesa

Number (double)b 42.64606

Longitude Stores the longitude of the sampling site in decimal de-grees

Number (double) −86.22633

Water temp Water temperature (°C) Number (double) 10

Air temp Air temperature (°C) Number (double) 12

Depth Stores the depth of the sam-pling station in meters.

Number (double) 2.3333

Gear TN, trap net; GN, gill net; S, seine; NN, Neuston net; G, grab; SL, sled; MN, mysis net; H, hy-droacoustics

Text GN

Tow speed The speed (m/s) at which the sampling gear was towed

Number (double) 1.5

Tow distance The distance (m) over which the sampling gear was towed

Number (double) 822

Comments Any comments related to the sampling event

Memo PI was arrested by Saugatuck police

a Our data were projected in the World Geodetic System (WGS84)—this information belongs in the metadata or data dictionary that describes each sampling parameter in detail. b Double means that the number is noninteger.


motivating them to make explicit all of the nuances of their data (McLaughlin et al. 2001; Porter and Ramsey 2002; Birnholtz and Bietz 2003). Databases can serve as storage for unique or irreplaceable records that can only be properly preserved for reuse though effective documentation and management (Brunt 1994; Borgman et al. 2007; Heidorn 2008). Only well-managed and documented data allow for reproduction of research where checks and balances operate at the most fundamental level (Parr 2007; Heidorn 2008; Borgman 2010). Others believe that be-cause most research is publicly funded, data belong to society at large, and best practices should be used when managing those data for preservation and reuse (Costello 2009; Guttmacher et al. 2009; Borgman 2010). Effectively managed data allow for repurposing, thereby saving money that might otherwise be used for redundant collections (Hale et al. 2003; Carlson and Anderson 2007; Heidorn 2008). Finally, properly documented and organized data have unlimited potential for reuse by pro-viding archival material to address future problems, thereby advancing science in ways possibly unforeseen by the original collectors (Postel et al. 2002; Nelson 2009; Borgman 2010).

One contentious issue surrounding data reuse is reluctance by researchers to share data beyond the original collaborators or close colleagues. Secrecy in guarding research has been part of scientific culture throughout history, and recent articles ex-ploring the data sharing attitudes find scientists overwhelmingly unwilling to freely share data within and among their own com-munity (Blumenthal et al. 1997; Campbell et al. 2002; Blumen-thal et al. 2006; Vogeli et al. 2006; Haas 2011; Tenopir et al. 2011), where willingness to share data is positively correlated with the ease of extraction and relationship to requestor (Witt et al. 2009; Cragin et al. 2010). In some sense, curators negate certain issues surrounding resistance to sharing that have to do with expending time and energy to prepare data, but address-ing the underlying scientific-professional reward structure that does not reward sharing remains outside their scope of influence (McDade et al. 2011).

Issues surrounding ownership and security also determine the extent to which data are shared (Beard et al. 1998). When research projects are funded by federal government agencies, philanthropic organizations, or private industries, grantor-specific stipulations often influence how data will be retained and disseminated (Fishbein 1991) as well as being subject to the Freedom of Information Act (5 U.S.C. 552). One simple solution to data sharing and ownership issues is a data shar-ing agreement. Data sharing agreements should be specific to each project and should include intended level of exposure (e.g., within the group only, within the field only, publicly accessi-ble), level of control applied to data outflow, whether an em-bargo period will be applied to data availability, and how data will be recognized when being used by others. In our case, our data sharing agreement stipulated that data would flow freely among principal investigators (PIs) and that each PI could de-cide to share or not share their portion of the data beyond the original collaborators at their discretion.

Building and managing databases can be challenging, espe-cially if long-term data management is underfunded. Granting institutions may recognize the benefits of requiring data sets as deliverables but may also be loath to become their ultimate rest-ing place. One field that is taking on the challenge of long-term digital curation is library science. University libraries are creat-ing institutional repositories as part of a larger technology and service structure that can contribute resources and expertise in data curation (Cragin et al. 2010). Data centers (open-standard, interoperable, nonproprietary web services) are also becoming widely established (Baker and Bowker 2007; Costello 2009). The lure of data centers is that by providing open or semi-open access to data, they act as a dual facilitator for finding and stor-ing data and, as of yet, no one repository has been established as the mainstay for fisheries data. Examples of established open-access ecological data repositories are the Long Term Ecologi-cal Research Network, DataONE, and MARIS (Baker et al. 2000; MARIS 2008; Michener et al. 2012). All three provide a framework for assimilation and management of disparate data sets with tools for data discovery and guidance on data manage-ment. The NOAA also has its own sophisticated internal data centers, whose services, as far as we were able to ascertain, are not available to non-NOAA researchers.

Not everyone is sold on the idea of depositing their data in open-access repositories though. Tenopir et al. (2011) found that only 15% of ecologists expressed a willingness to place their data into an open-access repository, and the majority expressed different conditions for doing so, including the fol-lowing: opportunities to collaborate (80%), mandatory reprints provided (75%), coauthorship (65%), results of analyses not disseminated without data providers’ approval (46%), legal per-missions obtained (40%), and monetary reimbursement (28%). Not included in the survey was an embargo period allowing PIs the first chance to publish on data, but we assume that would also be a consideration of data providers. To mitigate issues specifically related to recognition, formal and consistent cita-tion of databases will need to become more common in our field (NOAA 2012).

The concept of depositing data in an open-access repository was so foreign to the Lake Whitefish project team that we never seriously considered using one for our data set. This decision resulted in data that can now only be obtained through com-munication with a PI personally. We realize that our decision not to use a repository undermines a key message of this article, which is that data will not remain accessible without a plan for their preservation (Uhlir 2010), but the decision also accurately depicts the state of existing data preservation practices of most scientists in our position and field. We believe that whatever the future of institutional repositories and open-access data centers, they will continue to stay underutilized if they cannot support existing data practices specific to each scientific field and ad-equately mitigate the cultural issues associated with data shar-ing and recognition.

Given the shift toward large collaborative projects, we predict that formalized data management will become a more


integral part of research and will require explicit allocation of funds and recognition of professional productivity (McDade et al. 2011). We recommend that resource commitment associated with data management be estimated at the outset of the project so that dedicated resources can be requested from funding agen-cies for proper data curation. We also exhort funding agencies to support these requests for additional resources by realizing the benefits they provide in ensuring data availability for future research. Furthermore, an actual data set offered as a product of externally funded research is perhaps one of the most con-crete and useful deliverables that can be produced as return on investment.

Organization is an emergent property for any complex sys-tem, and efforts like the Lake Whitefish database are necessary as first steps in developing greater information organization within the fisheries research community. Looking beyond the development of a single database ultimately probes at a number of underlying systemic issues relating to large-scale informa-tion leveraging, in particular, resistance to sharing data, how to preserve and use historical data sets, the general lack of meth-odological standardization, and assessing whether the creation of these large-scale data endeavors yields returns enough to jus-tify their investment in resources. Ultimately, the fisheries com-munity should continue to examine ways to improve efficiency (reduce fragmentation) in research, reduce the duplication of effort in data collection, and spearhead efforts to coordinate data standards at a national level in order to adequately transfer sci-entific information. This can only be accomplished if we take the first steps toward a common goal of sharing, documenting, and preserving data that are collected and reported.

ACKNOWLEDGMENTS

We thank E. Volkman, C. Benoit, A. Charlton, R. Cripe, G. Fodor, J. Hoffmeister, V. Lee, A. McAlexander, R. Mollen-hauer, S. Shaw, D. Rajchel, B. Williston, and W. Zak for their assistance in the field and the laboratory. Thanks also to M. Ebener and D. Tagerson for assistance with the Lake Superior samples, C. Krause on Lake Erie, and L. Barbeau, D. Frazier, D. Hickey, K. King, T. King, P. Jensen, P. Peeters, B. Peterson, and J. Peterson for Lake Whitefish collections in Lake Michi-gan. We thank D. Clapp, E. Grove, K. Porath, T. Pattison, D. Wiefreich, J. Witzig, and two anonymous reviewers for their thoughtful comments that helped improve this article. Funding for this project was provided by the Great Lakes Fishery Trust, project number 2004.570, Department of Forestry and Natural Resources at Purdue University, Department of Fisheries and Oceans (to M. Koops) and Environment Canada (to M. Arts).

REFERENCESAkmon, D., A. Zimmerman, M. Daniels, and M. Hedstrom. 2011. The

application of archival concepts to a data-intensive environment: working with scientists to understand data management and pres-ervation needs. Archival Science 11:1–20.

Baker, K. S., B. J. Benson, D. L. Henshaw, D. Blodgett, J. H. Porter, and S. G. Stafford. 2000. Evolution of a multisite network in-formation system: the LTER information management paradigm.

Bioscience 50(11):963–978. Baker, K. S., and G. C. Bowker. 2007. Information ecology: open

system environment for data, memories and knowing. Journal of Intelligent Information Systems 29:127–144.

Baker, K. S., S. J. Jackson, and J. R. Wanetick. 2005. Strategies sup-porting heterogeneous data and interdisciplinary collaboration: towards an ocean informatics environment. Pages 1–10 in Pro-ceedings of the 38th Hawaii International Conference on System Sciences. (HICSS) 2005, 3–6 January, Big Island, Hawaii. IEEE, New Brunswick, New Jersey.

Baker, K. S., and K. I. Stocks. 2007. Building environmental informa-tion systems: myths and interdisciplinary lessons. Pages 253b–253b in System Sciences, 2007. Hawaii International Conference on System Sciences (HICSS) 2007. IEEE, New Brunswick, New Jersey.

Barnas, K., and S. L. Katz. 2010. The challenges of tracking habitat at various spatial scales. Fisheries 35(5):232–241.

Beard, D. T., D. Austen, S. J. Brady, M. E. Costello, H. G. Drewes, C. H. Young-Dubovsky, C. H. Flather, C. L. Gengerke, A. J. Loftus, and M. J. Mac. 1998. The multi-state aquatic resources informa-tion system. Fisheries 18(2):14–18.

Birnholtz, J. P., and M. J. Bietz. 2003. Data at work: supporting sharing in science and engineering. in Proceedings of the 2003 Interna-tional ACM SIGGROUP Conference on Supporting Group Work, Sanibel Island, Florida. 9(12):339–348.

Blukacz, E. A., M. A. Koops, T. M. Sutton, M. T. Arts, J. D. Fitzsim-mons, A. M. Muir, R. M. Claramunt, T. B. Johnson, R. E. Kin-nunen, M. P. Ebner, C. Suski, and G. Burness. 2010. Linking Lake Whitefish (Coregonus clupeaformis) condition with male gamete quality and quantity. Journal of Great Lakes Research 36(1):78–83.

Blumenthal, D., E. G. Campbell, M. S. Anderson, N. Causino, and K. S. Seashore. 1997. Withholding research results in academic life science: evidence from a national survey. Journal of American Medical Association 277(15):1224–1228.

Blumenthal, D., E. G. Campbell, M. Gokhale, R. Yucel, B. Clarridge, S. Hilgartner, and N. A. Holtzman. 2006. Data withholding in genetics and other life sciences: prevalences and predictors. Aca-demic Medicine 81(2):137–145.

Borer, E. T., E. W. Seabloom, M. B. Jones, and M. Schildhauer. 2009. Some simple guidelines for effective data management. Bulletin of the Ecological Society of America 90(2):205–214.

Borgman, C. L. 2010. Research data: who will share what, with whom, when, and why? Paper presented at the China–North America Li-brary Conference, Beijing, China.

Borgman, C. L., J. C. Wallis, and N. Enyedy. 2006. Building digital libraries for scientific data: an exploratory study of data practices in habitat ecology. Paper presented at the 10th European Confer-ence on Research and Advanced Technology for Digital Libraries, September 17–22, Alicante, Spain.

Borgman, C. L., J. C. Wallis, and N. Enyedy. 2007. Little science confronts the data deluge: habitat ecology, embedded sensor net-works and digital libraries. International Journal of Digital Librar-ies 7(1–2):17–30.

Brenden, T. O., M. P. Ebner, and T. M. Sutton. 2010. Special Issue on Assessing the Health of Lake Whitefish Populations in the Lau-rentian Great Lakes. Journal of Great Lakes Research 36(1):1–142.

Brunt, J. W. 1994. Research data management in ecology: a practical approach for long-term projects. Pages 272–275 in J. C. French and H. Hinterberger, editors. Seventh International Working Con-ference Scientific and Statistical Database Management. IEEE Computer Society Press, Washington, D.C.

Brunt, J. W., and W. K. Michener. 2009. The resource discovery ini-


tiative for field stations: enhancing data management at North American biological field stations. Bioscience 59(6):482–487.

Campbell, E. G., B. R. Clarridge, M. Gokhale, L. Birenbaum, S. Hil-gartner, N. A. Holtzman, and D. Blumenthal. 2002. Data with-holding in academic genetics: evidence from a national survey. Journal of the American Medical Association 287(4):473–479.

Carlson, S., and B. Anderson. 2007. What are data? The many kinds of data and their implications for re-use. Journal of Computer-Mediated Communication 12(2):635–650.

Carpenter, S. R., H. A. Mooney, J. Agard, D. Capistrano, R. S. De-Fries, S. Díaz, T. Dietz, A. K. Duraiappah, A. Oteng-Yeboah, H. M. Pereira, C. Perrings, W. V. Reid, J. Sarukhan, R. J. Scholes, and A. Whyte. 2009. Science for managing ecosystem services: beyond the Millennium Ecosystem Assessment. Proceedings from the National Academy of Sciences 106(5):1305–1312.

Claramunt, R. M., A. M. Muir, J. Johnson, and T. M. Sutton. 2010a. Spatio-temporal trends in the food habits of age-0 Lake Whitefish. Journal of Great Lakes Research 36(1):66–72.

Claramunt, R. M., A. M. Muir, T. M. Sutton, P. J. Peeters, M. P. Ebner, J. D. Fizsimons, and M. A. Koops. 2010b. Measures of larval Lake Whitefish length and abundance as early predictors of year-class strength in Lake Michigan. Journal of Great Lakes Research 36(1):84–91.

Costello, M. J. 2009. Motivating online publication of data. Bioscience 59(5):418–427.

Cragin, M. H., C. L. Palmer, J. R. Carlson, and M. Witt. 2010. Data sharing, small science and institutional repositories. Philosophical Transactions of the Royal Society 368:4023–4038.

Cragin, M. H., C. L. Palmer, and T. C. Chao. 2008. Relating data practices, types, and curation functions: an empirically derived framework. Proceedings of the American Society for Information Science and Technology 47(1):1–2.

Fagan, K. A., M. A. Koops, M. T. Arts, and M. Power. 2011. Assess-ing the utility of C:N ratios for predicting lipid content in fishes. Canadian Journal of Fisheries and Aquatic Sciences 68:378–385.

Fagan, K. A., M. A. Koops, M. T. Arts, T. M. Sutton, and M. Power. 2012. Lake Whitefish feeding habits and condition in Lake Michi-gan. Advances in Limnology 63:399–415.

Federal Geographic Data Committee. 1998. FGDC-STD-001-1998. Content standard for digital geospatial metadata (revised June 1998). Federal Geographic Data Committee, Washington, D.C. Available: http://[email protected]. (January, 2013).

Fishbein, E. A. 1991. Ownership of research data. Journal of Academic Medicine 66(3):129–133.

Freedom of Information Act (FOIA). 5 United States Code, Section 552. Department of Defense Publication 5400.7, Part 286 of Chapter 32. Code of Federal Regulations.

Frimpong, E. A., and P. L. Angermeir. 2009. Fish traits: a database of ecological and life history traits of freshwater fishes of the United States. Fisheries 34(10):487–495.

Guttmacher, A. E., E. G. Nabel, and F. S. Collins. 2009. Why data-sharing policies matter. Proceedings of the National Academy of Sciences 106(40):168–194.

Haas, M. R. 2011. To share or not to share? Professional norms, refer-ence groups, and information withholding among life scientists. Organization Science 21(4):873–891.

Haerder T., and A. Reuter. 1983. Principles of transaction-oriented da-tabase recovery. ACM Computing Surveys 15(4):287–317.

Hale, S. S., A. H. Miglarese, M. P. Bradley, T. J. Belton, L. D. Coo-per, M. T. Frame, C. A. Friel, L. M. Harwell, R. E. King, W. K. Michner, D. T. Nicolson, and B. G. Peterjohn. 2003. Managing troubled data: coastal data partnerships smooth data integration. Environmental Monitoring and Assessment 81:133–148.

Heidorn, P. B. 2008. Shedding light on the dark data in the long tail of

science. Library Trends 57:280–299. Hernandez, M. J. 2003. Database design for mere mortals: a hand’s on

guide to relational database design. Addison-Wesley, New York.Hook, L. A., S. K. S. Vannan, T. W. Beaty, R. B. Cook, and B. E.

Wilson. 2010. Best practices for preparing environmental datasets to share and archive. Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee.

International Organization for Standardization. 2011. Certification for high standards of quality management, ISO 9001:2011. Interna-tional Organization for Standardization, Geneva, Switzerland.

Katz, S. L., K. Barnas, R. Hicks, J. Cowan, and R. Jenkinson. 2007. Freshwater habitat restoration actions in the Pacific Northwest: a decade’s investment in habitat improvement. Restoration Ecology 15(3):494–505.

Kelling, S., W. Hochachka, D. Fink, M. Riedewald, and R. Caruana. 2009. Data-intensive science: a new paradigm for biodiversity studies. Bioscience 59(7):613–620.

Lélé, S., and R. B. Norgaard. 2005. Practicing interdisciplinarity. Bio-science 55(11):967–975.

Ling Liu, M. and T. Özsu, editors. 2009. Encyclopedia of database systems. Springer, New York.

Lord, P., A. Macdonald, L. Lyon, and D. Giarretta. 2004. From data deluge to data curation. Pages 371–357 in Proceedings of the UK e-Science All Hands Meeting.

(MARIS) The Multistate Aquatic Resource Information System. 2008. About MARIS. Available: http://www.marisdata.org. (January, 2013).

McDade, L. A., D. R. Maddison, R. Guralick, H. A. Piowar, M. L. Jameson, K. M. Helgen, P. S. Herendeen, A. Hill, and M. L. Vis. 2011. Biology needs a modern assessment system for professional productivity. Bioscience 61:619–625.

McDonald, L. L., R. Bilby, P. A. Bisson, C. C. Coutant, J. M. Epifanio, D. Goodman, S. Hanna, N. Huntly, E. Merrill, B. Riddell, W. Liss, E. J. Loudenslager, D. P Phillipp, W. Smoker, R. R. Whit-ney, and R. N. Williams. 2007. Research, monitoring, and evalua-tion of fish and wildlife restoration projects in the Columbia River Basin. Fisheries 32(12):582–590.

McLaughlin, R. L., L. M. Carl, T. Middel, M. Ross, D. L. G. Noakes, D. B. Hayes, and J. R. Baylis. 2001. Potentials and pitfalls of integrating data from diverse sources: lessons from a historical database for Great Lakes stream fisheries. Fisheries 26(7):14–23.

McLaughlin, R. L., M. L. Jones, N. E. Mandrak, and D. Stacey. 2010. FishMAP online: a web application supporting science-based decisions concerning fish movement and passage. Great Lakes Fishery Commission, Ann Arbor, Michigan.

Michener, W. K., S. Allard, A. Budden, R. B. Cook, K. Douglass, M. Frame, S. Kelling, R. Koskela, C. Tenopir, and D. A. Vieglai. 2012. Participatory design of DataONE—enabling cyberinfra-structure for the biological and environmental sciences. Ecologi-cal Informatics 11:5–15.

Michener, W. K., and M. B. Jones. 2011. Ecoinformatics: supporting ecology as a data intensive science. Trends in Ecology and Evolu-tion 27(2):85–93.

Muir, A. M., T. M. Sutton, M. T. Arts, R. M. Claramunt, M. P. Ebner, J. D. Fitzsimons, T. B. Johnson, R. E. Kinnunen, M. A Koops, and M. M. Sepúlveda. 2010. Does condition of Lake Whitefish spawners affect physiological condition of juveniles? Journal of Great Lakes Research 36(1):92–99.

National Institutes of Health. 2003. Data sharing policy and implemen-tation guidance. Available: http://grants.nih.gov/grants/policy/data_sharing/data_sharing_guidance.htm. (January, 2013).

National Science Foundation. 2010. Dissemination and Sharing of


D. Beard, G. Whelan, and A. Ostroff. 2011. A hierarchical spatial framework and database for the National River Fish Habitat Con-dition Assessment. Fisheries 36(9):436–449.

Watson, R., and R. Y. Kura. 2006. Fishing gear associated with global marine catches: database development. Fisheries Research 79:97–102.

Whiteed, D. C., J. S. Kimball, J. A. Lucotch, N. K. Manmenee, H. Wu, S. D. Chilcote, and J. A. Stanford. 2012. A riverscape analy-sis tool developed to assist wild salmon conservation across the North Pacific Rim. Fisheries 37(7):305–314.

Whitlock, M. C. 2011. Data archiving in ecology and evolution: best practices. Trends in Ecology and Evolution 26(2):61–65.

Witt, M. 2009. Institutional repositories and research data curation in a distributed environment. Library Trends 57:191–201.

Witt, M., J. Carlson, and S. Brandt. 2009. Constructing data curation profiles. International Journal of Digital Curation 3(4):93–103.

Research Results. Available: http://www.nsf.gov/bfa/dias/policy/dmp.jsp. (January, 2013).

Nelson, B. 2009. Data sharing: empty archives. Nature 461(7261):160–163.

Newman, H. B., M. H. Ellisman, and J. A. Orcutt. 2003. Data inten-sive e-science frontier research. Communications of the ACM 46(11):68–77.

(NOAA) National Oceanic and Atmospheric Administration Fisher-ies Service. 2010. Data and information management policy directive. Available: http://www.nmfs.noaa.gov/op/pds/docu-ments/04/04-111.pdf. (January, 2013).

———. 2011. Information and data administration policy. Available: http://ias.pifsc.noaa.gov/lds/docs/DMPolicyDraft1008.doc. (Jan-uary, 2013).

———. 2012. Data citation guidelines. Available: https://www.nosc.noaa.gov/EDMC/documents/edmcon/2012_breakout_sessions/Duerr-ESIP_Data_Citation_Guidelines.pdf. (January, 2013).

Parr, C. S. 2007. Open sourcing ecological data. Bioscience 57(4):309–310.

Porter, J. H., and K. W. Ramsey. 2002. Integrating ecological data: tools and techniques. Pages 396–401 in N. Callaos, J. Porter, and N. Rishe, editors. Proceedings of the 6th World Multi-Conference on Systematics, Cybernetics and Informatics, Orlando, Florida.

Postel, B. R., L. A. Shapiro, and J. C. Biesanz. 2002. On having one’s data shared. Journal of Cognitive Neuroscience 14(6):838–840.

Pratt, P. J., and J. J. Adamski. 2007. Concepts of database manage-ment, 4th edition. Boston, Massachusetts.

R Development Core Team. 2011. R: a language environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Available: http://www.R-project.org. (January, 2013).

(RIN) Research Information Network. 2008. To share or not to share: publication and quality assurance or research data outputs. Avail-able: http://eprints.soton.ac.uk/266742/1/Published_report_-_main_-_final.pdf. (January, 2013).

Sutton, T. M., M. A., Koops, M. T. Arts, A. M. Muir, A. E. Blukacz, R. M. Claramunt, M. P. Ebener, J. D. Fitzsimons, T. B. Johnson, and R. E. Kinnunen. 2007. Project completion report: does adult body condition affect recruitment potential in Lake Whitefish (Coregonus clupeaformis)? Great Lakes Fishery Trust, Lansing, Michigan.

Tenopir, C., S. Allard, K. Douglas, A. U. Aydinoglu, L. Wu, E. Read, M. Manoff, and M. Frame. 2011. Data sharing by scientists: prac-tices and perceptions. PLoS ONE 6(6):1–21.

Uhlir, P. 2010. Information gulags, intellectual straightjackets, and memory holes: three principles to guide the preservation of scien-tific data. Data Science Journal 7:ES1–ES5.

U.S. Fish and Wildlife Service. 2006. Department of Interior—depart-mental manual, part 378: data resource management. Available: http://www.fws.gov/policy/274fw1.html. (January, 2013).

Van den Eynden, V., L. Corti, M. Woollard, L. Bishop, and L. Horton. 2011. Managing and sharing data—best practice for researchers. UK Data Archive, University of Essex, Colchester, UK.

Vogeli, C., Y. Recai, E. Bendavid, L. M. Jones, M. S. Anderson, K. S. Louis, and E. G. Campbell. 2006. Data withholding and the next generation of scientists: results of a national survey. Academic Medicine 81(2):128–136.

Wake, M. H. 2008. Integrative biology: science for the 21st century. Bioscience 58(4):224–232.

Wallis, J. C., C. L. Borgman, M. Mayernik, and A. Pepe. 2008. Mov-ing archival practices upstream: an exploration of the life cycle of ecological sensing data in collaborative field research. Interna-tional Journal of Digital Curation 3(1):114–126.

Wang, L., D. Infante, P. Esselman, A. Cooper, D. Wu, W. W. Taylor,

From the Archives

Our great opponents in this have been the net-fishermen at the mouth of the river. Above that, every man wants a closed time; but, time says, “Everyone is going in, and I will go in too;” and they do, and catch all they can.

Mr. Eugene G. Blackford (1878): “Peculiar Features of the FishMarket”, Transactions of the American Fisheries Society, 7:1, 77-87.


FEATURE

SuperIDR: A Tool for Fish Identification and Information Retrieval

ABSTRACT: Students, fisheries professionals, and the general public may value computer-facilitated assistance for fish iden-tification and access to ecological and life history information. We developed SuperIDR, a software package supporting such applications, by utilizing the search and data retrieval capabili-ties of digital libraries, as well as key features of tablet PCs. We demonstrated SuperIDR utilizing a database with information on 207 freshwater fishes of Virginia. A user may annotate fish images and identify fishes by using a dichotomous key; search-ing for key words, similar images, subimages, or annotations on

Uma MurthyDepartment of Computer Science, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061-0106

Edward A. FoxDepartment of Computer Science, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061-0106

Yinlin ChenDepartment of Computer Science, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061-0106

Eric M. HallermanDepartment of Fish and Wildlife Conservation, Virginia Polytechnic Insti-tute and State University, Blacksburg, VA 24061-0321. Email: [email protected]

Donald J. OrthDepartment of Fish and Wildlife Conservation, Virginia Polytechnic Insti-tute and State University, Blacksburg, VA 24061-0321

Ricardo da S. TorresInstitute of Computing, University of Campinas, Campinas, SP 13083-970, Brazil

Lin Tzy LiTelecommunications Research and Development Center, CPqD Founda-tion, Campinas, SP, 13086-902, Brazil

Nádia P. KozievitchDepartment of Informatics, Federal University of Technology - UTFPR, Curitiba, PR 80230-901, Brazil

Felipe S.P. AndradeInstitute of Computing, University of Campinas, Campinas, SP 13083-970, Brazil

Tiago R.C. FalcaoInstitute of Computing, University of Campinas, Campinas, SP 13083-970, Brazil

Evandro RamosInstitute of Computing, University of Campinas, Campinas, SP 13083-970, Brazil

SuperIDR: herramienta para identifi-car peces y recuperar informaciónRESUMEN: estudiantes, profesionales de las pesquerías y público en general pudieran valorar la asistencia com-putarizada que facilite la identificación de peces y acceso a información sobre su ecología e historia de vida. Se desarrolla SuperIDR, un paquete computacional que con-tiene tales aplicaciones, basado en la búsqueda y capacid-ades de recuperación de datos de las bibliotecas digitales, así como también una plataforma para tabletas digitales. Se demuestra que SuperIDR utiliza una base de datos con información de 207 peces de agua dulce de Virginia. El usuario puede ingresar imágenes de peces e identificar-los mediante una clave dicotómica; además el programa busca palabras clave, imágenes similares, sub-imágenes o anotaciones hechas sobre las imágenes; así como también combinaciones entre estos enfoques. Los estudiantes que utilizan el paquete demostraron aumentar su habilidad para identificar correctamente los especímenes. Sus comen-tarios permitieron hacer mejorías, que incluyen la adición de nuevas características. El sistema computarizado para la identificación de peces de agua dulce presentes en Vir-ginia puede obtenerse en línea y modificarse. SuperIDR es un prototipo basado en computadoras personales para aplicaciones de identificación de especies —la arquitectura y programas abiertos que se desarrollaron, son aplicables a otras faunas de peces y a un rango más amplio de tareas de identificación de especies.

images; or combinations of these approaches. Students using the software demonstrated enhanced ability to correctly identify specimens. Their comments led to improvements, including the addition of new features. The PC-based system for identifying freshwater fishes of Virginia may be downloaded and modified. SuperIDR is a prototype for PC-based species identification ap-plications—the system architecture and the open-source soft-ware that we developed are applicable to other fish faunas and to a broader range of species identification tasks.

INTRODUCTION

Identification of freshwater fishes is critical to the study of fish ecology and management of fisheries and depends upon precise observation of external morphology, coloration, and internal characters. Learning to correctly identify freshwater fishes is challenging for fisheries students, who often find tra-ditional dichotomous keys intimidating. Such keys are rigid in their approach and do not accommodate the range of learning


styles actually utilized by students. In particular, dichotomous keys often do not focus upon diagnostic characters favored by many field ecologists and more user-friendly field guides (e.g., Page and Burr 2011). Many people in the general public would value ready access to knowledge about game and non-game fishes, including information on species identification, ecol-ogy, and life history. Such information generally is found in technical literature, which is not easily accessible to the general public.

A digital library (DL) is an information system with col-lections of documents/digital objects and their metadata and services to manage, organize, access, browse, index, and search through those collections to serve various user communities. Several DLs have been developed to present information on fish identification, ecology, and life history and aim to be widely ac-cessible and understood by diverse users, including practicing fisheries professionals, students, and the general public. For ex-ample, FishBase (Froese and Pauly 2011) provides information on over 32,000 species of fish, catering to professionals such as research scientists, fisheries managers, and zoologists, as well as providing links to a variety of useful Internet resources. Fish identification tools include classical dichotomous keys, a picture matching-based key, and identification using morphometrics. Targeting a more general audience, EFish (Helfrich et al. 2001) provides ecological and life history information on freshwater fishes of Virginia within a taxonomic framework. EKey (da S. Torres et al. 2007) builds upon EFish by offering resources for identification of freshwater fishes of Virginia, including browsing by taxon, text-based search, and image recognition. A photo-based computer system (Lyons et al. 2006) provides traditional dichotomous keys, a multi-feature query tool, and a slide show for identifying Wisconsin fishes. The system also provides data on characteristics distinguishing species, illustra-tions of diagnostic characters, and an illustrated glossary. One commercial product is Reef Fish ID (ReefNet, Inc. 2012), which uses a search engine that operates with a few key characteristics and then narrows down the possible species. IPez (Guisande et al. 2010), an automated system based on classification and regression trees and machine learning techniques, uses morpho-metric variation to identify fish taxa with input from either data files or interactive data entry by the user. Species identification using databases and automated systems has been developed for several biodiversity fields (reviewed by Gaston and O’Neill 2004). For example, EcoPod (Yu et al. 2006) is a PDA-based tool that uses a key-based approach, replacing traditional paper field guides with a mobile computing platform, in order to help semiskilled amateurs in identifying species (e.g., butterflies) in the field. EcoPod supports multiple ways of browsing through images and species information, including browsing of taxo-nomical classification and a key. In addition, it allows users to enter notes about an identified organism.

Recent developments in DL technology offer the oppor-tunity to build upon the approaches used in existing Internet-based fish identification and information retrieval systems. In particular, recent advances in software development offer the opportunity to annotate and search annotations on images. One

such notion is superimposed information (SI), or new informa-tion that is created to reference subdocuments (contextualized parts of documents) in existing information resources (e.g., annotation on part of an image). By integrating SI, a DL can provide enhanced support to tasks that involve working with subdocuments. As a result, annotations (and the subimages to which they refer) become separately accessible, searchable, and manageable within a DL. Because fish identification inherently involves distinguishing characters of fishes—that is, parts of a fish or its images—the task presents the opportunity to apply the concepts of SI and DLs. Against this background, the goals of this project were to

1. design and develop a DL, integrating capabilities to annotate, browse, and search for subimages and as-sociated information; and

2. assess the use and usefulness of this DL in learn-ing and identifying fish species and its support for working with subimages.

To address the first goal, we developed SuperIDR (Super-imposed Image Description and Retrieval tool) for freshwater fish identification and for access to ecological and life history information. SuperIDR builds upon the features of earlier sys-tems, enabling the user to add content (such as images, descrip-tions, subimages, and annotations) to the existing information base. It provides support for working with specific parts of im-ages and performing text- and content-based image description and retrieval. In addition, it has pen-input capabilities, mimick-ing freehand drawing and writing on paper. Species information in SuperIDR was extracted from the EKey system (da S. Torres et al. 2007). Though we utilized a database for the freshwater fish fauna of Virginia for demonstration purposes, our tool may be adapted for use in other biodiversity fields, and its system architecture may be applied to a wide range of species identi-fication tasks.

The second goal of assessing the use and usefulness of Su-perIDR was addressed by conducting multiple user studies in which we explored the use of subimages and evaluated the use of SuperIDR in fish species identification. In a controlled ex-periment, we found that students in an ichthyology class had a higher likelihood of correctly identifying an unknown fish spec-imen when they used SuperIDR than when they used traditional methods of fish identification, including a dichotomous key, fish identification web sites, and personal notes. In a follow-up qualitative user study, we investigated further how people worked with subimages in fish identification. Findings from the study suggest that working with subimages is an important, and sometimes necessary, part of fish species identification. The analysis of subimage/annotation content resulted in patterns of visual characteristics of parts of fishes (and their images) used in learning/identifying fishes, which can lead to improved ways of indexing and searching for subimages. The contexts and strategies of working with subimages in SuperIDR suggest new and enhanced support (i.e., DL) for scholarly tasks that involve working with subimages, including new ways of querying and searching for subimages and associated information.


METHODS

Software Development

We designed SuperIDR for archiving and interactively searching diverse interconnected data collections, including text, images, and annotations. Figure 1 outlines the software architecture of SuperIDR. Data collections consist of images, subimages (derived from images), textual descriptions, an-notations, taxonomical classification data that might be used for browsing only (such as family- or genera-level images and information), and complex objects (including superimposed objects/documents) that might comprise the digital objects men-tioned above.

Capabilities supported in SuperIDR include the following: • Annotation: The annotation service enables a user to

select subimages within images and to associate them with text annotations. In addition, a user can edit and delete annotations.

• Image management: A user can add images to and delete (user-added) images from a species database using the image management service.

• Text search: SuperIDR uses full-text or field-wise search to match keywords entered by the user against annotations, species descriptions, or both. Depending on the user’s selection, a separate ranked list of results is processed and displayed for species descriptions and annotations.

• Content-based image search: The content-based image search service inputs a query, which is either an image or a subimage. It then sends this query to the content-based image search component (CBISC;

da S. Torres et al. 2006; Kozievitch et al. 2008), which in turn matches the visual content of this query image or subimage against the visual content of all images and subimages in the CBISC index. A ranked list of results is produced, which could contain images and/or subimages. Search results are displayed separately for images and subimages.

• Combined search: The purpose of including a com-bined search service is to give the user an idea of how image and text content might be combined in a search, especially focusing on subimages combined with text. We used this implementation of combined search as a starting point to understand a user’s expectations. Text and image content have been combined in vari-ous ways in information retrieval (Hong et al. 1998). However, there has been limited work done in combin-ing text and images at the query level and in includ-ing subimages at the query level. We can consider this service to be a low-fidelity prototype for a combined search interface.

• Browsing: The browse service enables multiple

ways of browsing through species images, descrip-tions, subimages, and annotations, including: images and description of a species; annotations and, for each image, the subimage in the context of its base image; taxonomical classification, including families, genera, and species of fishes using column-wise organization or tree-based organization; or a digital version of the identification key guide of freshwater fishes of Vir-ginia (Jenkins and Burkhead 1994).

• Comparison: The comparison service is a special case of the browsing service. It enables a user to view two images side by side. A user is able to choose from im-ages of the same or different species. While viewing

the images side by side, a user can browse through annotations and view the associated subim-ages in the context of their base images. The goal of the compari-son service is to enable the user to manually analyze two images side by side.

• Logging: The logging function is used to log all user interac-tions with the tool, a service that would prove useful to teachers or researchers as opposed to gen-eral users.

SuperIDR was implemented using the C# programming language in the Visual Studio .NET platform. The metadata and logs were stored in a MySQL database (Oracle 2011).

Figure 1. Software architecture of SuperIDR shows repository with collections and services. The CBISC (da S. Torres et al. 2006; Kozievitch et al. 2008) and Lucene (Apache 2011) components are used to index and retrieve image and text data, respectively.


SuperIDR uses the CBISC (da S. Torres et al. 2006; Kozievitch et al. 2008) for indexing and retrieving the visual content of images and subimages. The Lucene (Apache 2011) component in SuperIDR is responsible for full-text and field-wise indexing and search of all text data in the SuperIDR collections, includ-ing textual description (of images) and annotations.

SuperIDR has been tested to run on Windows XP Profes-sional, Windows 7, and Windows XP tablet PC versions. The SuperIDR source code, technical details, and a user manual are available in Murthy (2011). We developed SuperIDR to work with tablet PCs in order to take advantage of pen-based input in order to emulate the use of a pen on paper. In addition, pen-based input would make SuperIDR relatively portable and convenient, especially in field situations. We evaluated three different versions of SuperIDR in user studies. Existing func-tionality was improved and new functionality was added with each new version, taking into account feedback gathered in the user studies.

Much of the species-specific database, including formal color pictures, text, key, and geographic distributions, was drawn from Jenkins and Burkhead’s (1994) Freshwater Fishes of Virginia. The pictures of preserved specimens were original.

Evaluation of SuperIDR

We performed three studies to evaluate SuperIDR in order to determine the following: (1) Is SuperIDR useful for fish iden-tification? (2) How does SuperIDR compare with traditional methods (e.g., dichotomous keys, personal notes, markings on images, fisheries websites, notecards, or combinations of these methods) for performing fish identification? (3) What are users’ perceptions about using SuperIDR for fish identification? and (4) How is SuperIDR employed to support the use of subimages in fish identification? We evaluated three different versions of SuperIDR in the respective user studies. Existing functionality was improved and new functionality was added with each new version, taking into account feedback gathered in the previous user studies. Procedures for data collection involving human subjects were approved by the Virginia Tech Office of Research Compliance Institutional Review Board under approval nos. 08-065, 08-066, and 10-692.

Longitudinal Formative Evaluation

We conducted a 2-month-long formative evaluation of Su-perIDR with three fisheries students, one faculty member, and one researcher (Murthy et al. 2009). At the beginning of the study, all scholars were given a tablet PC with the SuperIDR application installed and configured. We conducted a group training session for all participants and asked them to use Su-perIDR, as appropriate, in their daily activities, such as learning, research, analysis, and writing. Over the course of 2 months, we met biweekly with the participants as a group to receive feedback on the utility of SuperIDR. This formative evaluation helped identify significant improvements and led to the addition of new features to SuperIDR.

In-Class Study

To assess the effectiveness of SuperIDR, we conducted a study in spring 2008 with students in an undergraduate ichthyol-ogy course in the Department of Fish and Wildlife Conservation at Virginia Tech (Murthy et al. 2009). Students in this course learned about fishes and how to identify fish species; they prac-ticed species identification throughout the semester; and they developed familiarity with species identification methods. In the study, we collected the following data: • Species identification responses for 40 unknown

specimens. For a specimen to be regarded as correctly identified, its family, genus, and scientific name had to be correctly listed. We tracked the number of correctly identified specimens for each session. Considering the binary property of the response (correct–incorrect) and the existence of multiple factors impacting the task outcome, we used the generalized linear model with a binomial (logit) link function to analyze the species identification responses.

• Demographic data and data about prior experience with species identification and software tools.

• Qualitative feedback on use of SuperIDR in an exit questionnaire.

• Log data of user interaction with SuperIDR.

Qualitative Longitudinal Study

In a follow-up qualitative user study, we further investi-gated how people worked with subimages in fish identification and how SuperIDR supported that use (Murthy et al. 2009). The rationale for the qualitative study design was to maximize the use of SuperIDR in various settings and to obtain rich data from multiple sources on use of SuperIDR. Hence, we recruited people with interest and experience in fish identification and whose ongoing study or work involved fish identification. We deployed SuperIDR for 3 weeks in the study or work contexts of six individuals from October 21 to November 18, 2010, in the Department of Fish and Wildlife Conservation at Vir-ginia Tech. During this time, we recorded their use of the tool through interaction logs and diary entries. The diaries gave us a post-hoc means of observing situated behavior in contexts in which researchers were not present. We collected information on the participants’ background, fish identification materials and practices, and computer usage in a pre-study interview. In two separate sessions during the study, we had participants identify unknown specimens. In these sessions, we captured screen interactions with SuperIDR and conversations concern-ing SuperIDR. In a post-study interview, we asked participants about their use of subimages and about using SuperIDR in fish identification (using log and diary data to elicit information as required).


By analyzing logs, we were able to extract 940 subimages/annotations made by the participants. We analyzed them for recurring types or patterns of subimages/annotations. We used TagCrowd (Steinbock 2011), an online tag cloud generator, to obtain an overview of the words used in the participants’ annotations. TagCrowd considers a stop word list and other options, such as minimum frequency of keywords, maximum words to show, and whether to group similar words or not (such as memorize and memorizing) while generating a tag cloud. We preprocessed the annotation text to identify words that have the same meaning, such as “caudal” and “tail.” We entered this preprocessed text into TagCrowd, opting for a minimum oc-currence frequency of 1, the default English stop word list, and grouping similar words, in order to generate a tag cloud of the 50 most frequently used annotation keywords. Our analy-sis of the qualitative data, including interviews, diary entries, screen cap-tures, spoken thoughts, and images of materials used, was informed by the grounded theory approach. “Grounded theory” (Glaser and Strauss 1967; Glaser 1978; Straus 1987) refers to a qualitative research method with the aim of developing a theory from data collected in the course of research.

RESULTS

SuperIDR Features and Scenario

Use of SuperIDR might best be described by means of a scenario il-lustrating how a student might use the tool to identify a fish specimen. Matt (a pseudonym), a fisheries science student, has some experience with species identification and uses a com-bination of methods to identify and learn about fish species. In the past, he has supplemented the use of di-chotomous keys with personal notes, pictures from the Internet, a textbook, and other resources. The ichthyology instructor has decided to use a new software application in class, Super-IDR. Matt is comfortable with the use of technology and is enthusiastic about using SuperIDR, because it will likely help him with fish species iden-tification. Matt explores the features of SuperIDR, beginning with a tax-onomy browser, where fishes are or-ganized according to families, genera, and species. He browses to Percina peltata (Shield Darter), the species

being taught about in class (Figure 2). SuperIDR shows details of the species, including its physical description, habitat, and food habits (Figure 3). Matt can browse through the collection of images for a particular species. For each image, Matt may browse through annotations associated with parts of that image. Because the physical description does not have all of the distin-guishing characteristics of the Shield Darter, the student adds more images and annotations to the species data. To annotate an image, the student marks the fish picture and associates the marked regions with text explanations, plus notes in his own words, making it easier to remember and learn about this spe-cies (Figure 4). The student often reviews the fishes about which

Figure 2. Taxonomy browser in SuperIDR: (A) list of families; (B) genera in the selected family; and (C) species in the selected genus.

Figure 3. Species description interface in SuperIDR: (A) focus image, showing marked region associated with a selected annotation; (B) list of images in the species; (C) annotation menu and list of annotations on the focus image; and (D) ecological and life history information can be seen by scrolling down.


he is learning. He uses SuperIDR to browse through details of various species, using the browsing services of SuperIDR, in-cluding the taxonomical browser (Figure 5), the dichotomous key browser, and the species description interface (Figure 3). While reviewing Greenside Darter (Etheostomas blennioides) and Ashy Darter (E. cinereum.), two very similar-looking spe-cies, the student uses the comparison feature in SuperIDR to analyze images from these species and study their similarities and differences (Figure 6). After 2 weeks, the student’s ichthy-ology instructor holds a practice specimen identification test. Some of the unknown specimens are in jars and some are in the form of images posted on the course web site. The student ex-amines the fish specimens for specific characteristics and uses text search on annotations and species descriptions (Figures 7A and 7B). Browsing through the result list and occasionally click-ing on a result for details, he is able to narrow down the options

to the species of the unknown specimen. In some cases, the student uses the combined search to help identify the species of a specimen, provided as an image. He marks the mouth of the specimen’s image in the image part of the query and enters keywords describing characteristics of the mouth (Figures 7C and 7D). This narrows down the possibilities and he is able to identify the species.

In-Class Study

On entry questionnaires, 17 of the 18 students rated them-selves as moderately or very familiar with computers, although 13 indicated that they had very low expertise with tablet PCs. Sixteen of the 18 students had previously taken a course on species identification, not necessarily fish species, so most were familiar with the task. Most students used annotations, mark-

ing, note-taking, and organization of some form on paper for learning, but most did not use digital applications to perform the same activities. All students were familiar with online search engines, with Google search and Google image search being the most popular responses.

In each of two sessions, pairs of students were presented with 20 specimens and asked to identify them (Table 1). In one session, stu-dents used traditional dichotomous keys; in the other they used Super-IDR. Half of the groups used paper keys first and then SuperIDR, and half used SuperIDR first. Each team worked with one specimen at a time. Table 2 summarizes the scores for teams using the respective methods across the two sessions. The team (P = 0.015) and the method (P = 0.011) had a significant impact on the out-come of the species identification task, whereas the session (hence, the nature of specimen) did not im-pact the outcome significantly. Using SuperIDR yielded a higher likeli-hood of identifying a specimen cor-rectly than using traditional methods (Table 2) and with less than a week of use. However, the questionnaire responses indicated that students preferred using the identification key over SuperIDR. A reason for this could be insufficient use of SuperIDR and its subimage features (indicated by the log data). This, in turn, might have been due to the timing and short duration of the experiment. An indi-cator of student interest in SuperIDR

Figure 4. Annotation feature in SuperIDR, which accepts pen and keyboard input.

Figure 5. Dichotomous key browser in SuperIDR.


is that 6 of 14 teams chose to keep SuperIDR for three more weeks until the end of the semester. Students said that they wanted to explore the tool further.

Qualitative Longitudinal Study

To tease out characteristics of a subimage and associated informa-tion, we considered participants’ fish identification materials, interview responses, and the subimages and an-notations they made. It is clear from the tag cloud (Figure 8) that many of the keywords used in the annotations referred to parts of fishes, suggesting a reference to subimages (caudal, eye, mouth, etc.). We observed a variety of types of annotations and associ-ated subimages (Figure 9). Some an-notations directly described a visual characteristic of the subimage. Some annotations were about the morpho-logical location of a body part. Some brought together multiple parts of the fish but used a single subimage/anno-tation to describe those parts. Some considered the image or the fish spe-cies as a whole or might have referred to information outside the image or the fish species. Finally, there were subimages or annotations that com-bined these types. This analysis of subimage or annotation types matched with the participants’ responses on the characteristics of subimages or anno-tations.

Using SuperIDR to identify an unknown specimen, participants fol-lowed a top-down approach to iden-tifying fishes just as they did with traditional methods. In most cases, they were able to identify the fam-ily of the specimen within a minute of looking at the specimen (for both preserved specimens and images of specimens). Then, participants fol-lowed one of several paths to nar-row down their choices, including (1) using the taxonomy (Figure 2), tree, or dichotomous key browsers (Figure 5); (2) using the text, image, or com-bined search (Figure 7); or (3) visu-ally analyzing the specimen or image and identifying their choices. At the end of this step, they usually ended

Figure 7. Search in SuperIDR: (A) annotation search results for the text query—“red mark,” “small mouth,” “pointed snout,” “no spots”; (B) species description results for the same query; (C) combined search query interface; and (D) annotation/subimage combined search.

Figure 6. Comparison feature of SuperIDR, allowing observation of similarities and differences among species.


up with two to three species from which to choose. Then, they would open the descriptions (Figure 3) of these chosen species and browse through the text, the images, and the annotations (if any) of the species. They would switch back and forth among species, analyzing and comparing the species’ information. Sometimes, they used the comparison feature (Figure 6) to ana-lyze similarities and differences. Finally, they would select one species. The context of subimages, which might be described by the copresence of two or more subimages and associated infor-mation (including annotations and species description), was im-portant. Manual analysis of subimages proved an important part of working with subimages, too. Even when SuperIDR returned the putatively correct species as the top result, participants often exhibited the need to confirm the species by analyzing all of its images and its distinguishing characteristics as well as other information after they had successfully identified the species of an unknown specimen.

We received positive feedback about the participants’

use of SuperIDR and observed and recorded interesting, new ways of working with subimages in fish identification. Par-ticipants felt that the bringing together of capabilities to work with subimages—that is, annotation, text-based search, content-based search, browser capability—was helpful. Searching on subimages and other information related to subimages was considered important, whether using SuperIDR or recalling information from one’s memory. To improve the search effec-tiveness, participants suggested additional capabilities to work with subimages, including (1) the use of faceted search to filter/view results based on subimages/annotations and other fields, (2) the ability to filter results considering the family or genus, and (3) the use of an ontology to suggest similar meaning terms for the annotation or query. Five of the six participants repeat-edly made use of a combination of multiple subimages, associ-ated annotations, and other species information as a query and expected the tool to match against species with images with similar subimages and annotations. Such a combination query might be considered as a complex object query.

DISCUSSION

Considerable progress has been achieved in developing computer software (Lyons et al. 2006) and Internet resources (Helfrich et al. 2001; da S. Torres et al. 2007; Froese and Pauly 2011) to help identify fish and to access ecological and life history information. Recent developments in creation, mainte-nance, and active use of DLs provide opportunities to ease these tasks, enhancing learning and work. We developed a prototype DL system for fish identification and demonstrated its use in classroom and independent work contexts, demonstrating in particular the utility of working with parts of fish images. We developed the system for tablet PCs, which provide an intui-tive interface emulating, as closely as possible, annotation of a paper document with a pen for selecting parts of images in documents and annotating those selections. The image descrip-tion and retrieval tool will help users work with images and parts of images; associate them with multimedia information like text annotations derived from annotations made by an elec-tronic pen on a tablet PC; and later retrieve information based on text- and content-based retrieval techniques. Information may be retrieved through searching or by browsing, browsing species based on taxonomy or a dichotomous key, retrieving annotations or species information based on keyword match-ing, or retrieving images and marks (parts of images) based on matching of content of images/marks (marks, which were cre-ated earlier).

Applications

We foresee several ways in which SuperIDR might be uti-lized by teachers and students, practicing fisheries profession-als, and members of the general public.

A first set of uses and applications would be in fisher-ies education. We note that in this study, we developed and evaluated personal use of superimposed information in a DL; that is, the data in the DL were accessible and editable only

TABLE 1. Species used in classroom evaluation of SuperIDR in evaluative sessions 1 and 2.

Session 1 Session 2

Alosa sapidissima Etheostoma vitreum

Ambloplites rupestris Lepomis megalotis

Percina rex Cyprinella spiloptera

Cottus bairdi Hypentelium nigricans

Etheostoma nigrum Percina peltata

Percopsis omiscomaycus Lepomis auritus

Micropterus salmoides Campostoma anomalum

Cottus carolinae Dorosoma cepedianum

Percina notogramma Centrarchus macropterus

Etheostoma flabellare Rhinichthys atratulus

Lepomis gibbosus Dorosoma petenense

Lepomis macrochirus Etheostoma simoterum

Clinostomus funduloides Etheostoma blennioides

Cyprinella galactura Notropis rubellus

Micropterus dolomieu Alosa aestivalis

Notropis telescopus Lepomis gulosus

Lepomis cyanellus Percina roanoka

Percina aurantiaca Percina peltata

Etheostoma rufilineatum Rhinichthys cataractae

Fundulus heteroclitus Lepomis microlophus


by one person. However, DLs are well suited for sharing information accessible by multiple people. Thus, the development of a social or col-laborative superimposed information DL is a natural next step, where mul-tiple users—for example, instructors and students in a class—have access to and might create and edit content such as annotation of diagnostic char-acters on images of fish. The task of creating subimages or annotations on all of the fish in a regional fish fauna might be crowdsourced (Doan et al. 2011) by the class; that is, a single user would not be burdened with cre-ating and adding all of content, and multiple users would benefit from all new content created. In addition, crowdsourcing would enable users to collaborate over shared content.

A second set of users would be practicing fisheries professionals for whom fish identification and infor-mation on ecology and life history are of day-to-day interest. SuperIDR may be used in a laboratory setting or in the field, because access to the Internet is not needed.

A third set of users would be the many people in the general public with an interest in fishes. Upon en-countering a striking species, many people want access to information on the ecology and life his-tory of the species. EFish (Helfrich et al. 2001) and EKey (da S. Torres et al. 2007) have proven popular for the purpose but do not offer a user-centered design that a web-based version of SuperIDR would offer. One major differentiation between SuperIDR and prior web-based systems is that SuperIDR of-fers ways for a user to extend the information base (including new images and annotations). This feature enables users to keep the SuperIDR content up to date as well as to incorporate new information and personal notes. Several interactive web-based fish identification resources that have been launched over the past few years were built using different architectures and offer different features for nontechnical users. For example, the Wis-cFish identification database (Hanson et al., no date) offers a taxonomic key and a keyword-based query system. The Dis-cover Life website (Discover Life 2012), actually a compilation of resources, offers the user access to various approaches to identify different fish taxa. FishBase (Froese and Pauly 2011) offers dichotomous keys, picture matching, and morphometric identification. With the ability to annotate, search annotations, perform comparisons, and access other features of DL technol-ogy, we feel that SuperIDR offers features not available in other existing applications.

Figure 8. Tag cloud of top keywords in participants’ annotations, generated by TagCrowd (Steinbock 2011) from 940 annotations.

Figure 9. Examples of subimage/annotation types: (A) shape, pattern, or texture of a part—deeply notched dorsal fin; (B) locations of a part—eyes almost on top of head; (C) copresence of parts—terminal mouth; short, blunt nose; and (D) about the information object as a whole—Roanoke basin.

TABLE 2. Number of correctly identified fish species per student team (of a total of 20 specimens each session) using traditional methods or SuperIDR.

Traditional Methods SuperIDR

Team ID Session No. correct

Team ID Session No. correct

1 1 15 2 2 18

4 1 16 4 2 17

6 1 13 6 2 17

11 1 12 11 2 16

3 2 8 3 1 14

5 2 13 5 1 15

9 2 12 9 1 10

13 2 11 13 1 11

14 2 6 14 1 14

Mean = 11.7 Mean = 14.7


SuperIDR as a Prototype

Though SuperIDR offers innovations over existing soft-ware and websites for fish identification and information re-trieval, we recognize two ways in which it might be improved. First, knowing where a fish was collected, many users would find it simple and straightforward to use a key or resources concerning only those fishes occurring in that watershed. For example, an individual may want to search resources regarding cyprinids occurring in the James River Basin of Virginia. This feature may be offered wherever a reliable species list exists for given watersheds. Second, many users would value more com-plex Boolean searches than are currently offered. Such complex searches many involve a combination of a taxon, annotation, and location; for example, a search for a darter with dusky sad-dle marks living in the upper Tennessee River drainage.

We regard our system as a prototype demonstrating the po-tential of tablet PC-based data annotation, search, and retrieval by means of a case study involving freshwater fishes of Vir-ginia. The system architecture is general—it may be applied to other fish faunas by adding or removing species-specific in-formation to or from the database. This capability is supported by the existing software, which can support a change of the database using procedures described in the user’s manual. More fundamentally, the system architecture and coding may be ap-plied in any case where external morphology provides the basis for distinguishing species; for example, it is being applied for identifying flies and butterflies (R. da S. Torres, unpublished). Features of the tablet PC platform and application—especially the abilities to upload, annotate, and store images and then to search annotations—may prove useful to support fieldwork. We are extending our software to support identification of human parasites in Brazil (Kozievitch et al. 2010; Murthy et al. 2010), where field clinicians produce wet mounts of specimens but do not have access to a medical library or to the Internet and would value the ability to build and to access a searchable reference database.

ACKNOWLEDGMENTS

Development of the tablet PC software was supported by grants from Microsoft, the National Science Foundation (DUE-0435059), Coordination of Improvement of Personal Higher Education (CAPES), the Sao Paulo Research Foundation (FAPESP), and the National Council for Scientific and Tech-nological Development (CNPq). We thank faculty, researchers, and students in the departments of Computer Science and Fish and Wildlife Conservation at Virginia Tech and at University of Campinas, whose help and valuable inputs proved key to improving various versions of this tool. We are grateful to the Department of Computer Science and College of Engineering for the loan of tablet PCs for evaluation by students. We thank Robert Jenkins, Noel Burkhead, and the American Fisheries So-ciety for use of copyrighted materials on the freshwater fishes of Virginia. Tamar Hallerman photographed the preserved speci-mens.

REFERENCES

Apache. 2011. Lucene.net search engine library. Available: www.apache.org. (February, 2012).

da S. Torres, R., E. M. Hallerman, R. E. Jenkins, N. M. Burkhead, and B. Herrington. 2007. EKey—the electronic key for identifying freshwater fishes. Available: http://www.cmiweb.org:8080/ekey/. (February, 2012).

da S. Torres, R., C. B. Medeiros, M. A. Goncalves, and E. A. Fox. 2006. A digital library framework for biodiversity information systems. International Journal on Digital Libraries 6(1):3–17.

Discover Life. 2012. Discover Life—fish. Available: http://www.dis-coverlife.org/mp/20q?guide=Fish. (February, 2012).

Doan, A., R. Ramakrishnan, and A. Y. Halevy. 2011. Crowdsourcing systems on the world-wide web. Communications of the Associa-tion for Computing Machinery 54(4):86–96.

Froese, R., and D. Pauly, editors. 2011. FishBase. Available: http://www.fishbase.org. (February, 2012).

Gaston, K. J., and M. A. O’Neill. 2004. Automated species identifica-tion: why not? Philosophical Transactions of the Royal society of London B 359:655–667.

Glaser, B. G. 1978. Theoretical sensitivity: advances in the methodol-ogy of grounded theory. Sociology Press, Mill Valley, California.

Glaser, B. G., and A. Strauss. 1967. The discovery of grounded theory: strategies for qualitative research. Sociology Press, Mill Valley, California.

Guisande, C., A. Manjarres-Hernandez, P. Pelayo-Villamil, C. Gra-nado-Lorencio, I. Riveiro, A. Acuna, E. Prieto-Piraquive, E. Ja-neiro, J. M. Matias, C. Patti, B. Patti, S. Mazzola, S. Jimenez, V. Duque, and F. Salmeron. 2010. IPez: an expert system for the taxonomic identification of fishes based on machine learning techniques. Fisheries Research 102:240–247.

Hanson, P., J. Kitchell, J. Lyons, E. White, and P. Moy. No date. Wis-consin fish identification database. Available: http://wiscfish.org/fishid. (February, 2012).

Helfrich, L., T. Newcomb, E. Hallerman, and K. Stein. 2001. EFish: the virtual aquarium. Available: http://www.cnr.vt.edu/efish. (February, 2012).

Hong, Z., C. Syin, and K. F. Lai. 1998. Query expansion by text and image features in image retrieval. Journal of Visual Communica-tion and Image Representation 9(4):287–299.

Jenkins, R. E., and N. M. Burkhead. 1994. Freshwater fishes of Vir-ginia. American Fisheries Society, Bethesda, Maryland.

Kozievitch, N. P., R. da S. Torres, F. Andrade, U. Murthy, E. Fox, and E. Hallerman. 2010. A teaching tool for parasitology: enhancing learning with annotation and image retrieval. Pages 466–469 in M. Lalmas, J. Jose, A. Rauber, R. Sebastiani, and I. Frommholz, editors. Proceedings of the 14th European Conference on Re-search and Advanced Technology for Digital Libraries. Springer-Verlag, Heidelberg, Germany.

Kozievitch, N. P., T. R. C. Falcao, and R. da S. Torres. 2008. A .net implementation of a content-based image search component. in Proceedings of the 23rd Brazilian Symposium on Databases, Campinas, Brazil, 2008. Association for Computing Machinery, New York.

Lyons, J., P. Hanson, and E. White. 2006. A photo-based computer system for identifying Wisconsin fishes. Fisheries 31(6):269–275.

Murthy, U. 2011. Digital libraries with superimposed information: sup-porting scholarly tasks that involve fine grain information. Ph.D. dissertation. Virginia Polytechnic Institute and State University, Blacksburg, Virginia. Available: http://scholar.lib.vt.edu/theses/available/etd-04142011-175752. (February, 2012).

Murthy, U., E. A. Fox, Y. Chen, E. Hallerman, R. da S. Torres, E. Ramos, and T. Falcao. 2009. Superimposed image description and


retrieval for fish species identification. Pages 285–296 in J. L. Borbinha, S. Kapidakis, C. Papatheodorou, and G. Tsakonas, edi-tors. ECDL ’09: Proceedings of the 13th European Conference on Research and Advanced Technology for Digital Libraries.

Murthy, U., N. Kozievitch, E. Fox, R. da S. Torres, and E. Haller-man. 2010. SuperIDR: a tablet PC tool for image description and retrieval. Page 152 in R. H. Reed and D. A. Berque, editors. The Impact of Tablet PCs and Pen-Based Technology on Education. Purdue University Press, West Lafayette, Indiana.

Murthy, U., L. T. Li, E. Hallerman, E. A. Fox, M. A. Perez-Quinones, L. M. Delcambre, and R. da S. Torres. 2011. Use of subimages in fish species identification: a qualitative study. Pages 185–194 in G. Newton, M. Wright, and L. Cassel, editors. Proceedings of the 11th Annual International ACM/IEEE Joint Conference on Digi-tal Libraries (JCDL ’11). Association for Computing Machinery, New York.

Oracle. 2011. MySQL. Available: http://www.mysql.com. (February, 2012).

Page, L. M., and B. M. Burr. 2011. Peterson’s field guide to freshwater fishes. Houghton Mifflin Company, Boston.

Reef Net, Inc. 2012. Reef fish identification: Florida, Caribbean, Ba-hamas. Available: http://reefnet.ca/products/fishes/caribbean. (February, 2012).

Steinbock, D. 2011. TagCrowd. Available: http://tagcrowd.com (Feb-ruary, 2012).

Strauss, A. 1987. Qualitative analysis for social scientists. Cambridge University Press, Cambridge, UK.

Yu, Y., J. A. Stamberger, A. Manoharan, and A. Paepcke. 2006. Eco-Pod: a mobile tool for community-based biodiversity collection building. Pages 244–253 in JCDL ’06: Proceedings of the 6th ACM/IEEE-CS Joint Conference on Digital Libraries. ACM Press, New York.

blueleafenviro.com

The best telemetry tool?

Experience.

Blue Leaf has effectively used techniques ranging from presence/absence with PIT tags, to fine-scale three-dimensional tracking with acoustic tags, to fish movement and interactions with DIDSON sonar imaging.

Call us for a free consultation and learn how our technical expertise in fisheries telemetry can help make your project successful.

DIDSON (Dual-Frequency Identification Sonar)


The theme for the 142nd annual meeting of the American Fisheries Society (AFS) was “Fisheries Networks: Building Ecological, Social, and Professional Relationships,” and that provided me a hook to reflect on how studies of ecological networks have developed and how they can be used to inform fisheries research. Ecological networks, as we know them, date back to the pioneering studies of Raymond Lindeman around 1940 (Illustration 1). He studied Cedar Creek Bog in Minnesota and made a detailed model of nutrient cycling expressed as en-ergy flows (Lindeman 1942). For this, he used thermodynamic principles to evaluate and understand ecosystem functioning, and through this he established the field of trophic dynamics. The study of energy flows and concepts he introduced, such as food chains, food webs, ecological transfer efficiency, and energy pyramids, now provide core elements of community and ecosystem ecology.

Lindeman was born in Redwood Falls, Minnesota, and edu-cated at the University of Minnesota in the Twin Cities where

AFS 2012 was hosted. He received a fellowship to work with G. Evelyn Hutchinson at Yale Uni-versity, managed to pub-lish his Ph.D. studies on Cedar Creek Bog though ill, but unfortunately he died soon after, at only 27 years old. He was a bril-liant mind, and we can only guess how he would have shaped our research world had his days been more numerous.

Lindeman’s studies, however, inspired re-search in the decades to follow, most notably the International Biological Program (IBP), a major international initiative that during 1964–1974 conducted studies of bio-logical productivity in ecosystems throughout the world (Illustration 2).

PLENARY SESSIONS

Ecological Networks in Fisheries: Predicting the Future?Villy ChristensenNF-UBC Nereus Program, Fisheries Centre, University of British Columbia, 2202 Main Hall, Vancouver, BC V6T 1Z4, Canada. E-mail: [email protected].

Incidentally, this was also where I first participated in ecologi-cal research—as a first-year student joining the tail end of these studies, sampling fish in a lake in Denmark.

The IBP was mainly descriptive in its nature and had nu-merous modeling activities, including some dynamic ecosystem modeling—a topic to which we return later. A lasting legacy of the IBP was that it brought focus to ecosystem research. There were also numerous follow-up studies to the IBP. Methodolo-gies had been developed and coordinated through the IBP, and many researchers had been introduced to the field. The time had come for ecosystem research.

Among the follow-up studies was an extensive 5-year study (around 1980) of the French Frigate Shoals in the northwestern Hawaiian Islands. Researchers quantified energy flows and bio-masses ranging from plankton through to marine mammals and over the 5 years gathered an impressive amount of data. Realiz-ing the need to make sense of the mountain of data, the National Oceanic and Atmospheric Administration hired a newly gradu-ated oceanographer, Jeff Polovina, to construct an ecosystem model of the French Frigate Shoals.

At this time there were two major activities on ecosystem modeling with a fisheries perspective. Taivo Laevastu and colleagues at the National Marine Fisheries Service’s Alaska Fisheries Science Center worked on a complex multispecies model of the Bering Sea (Laevastu and Larkins 1981) and K. P. Andersen and Erik Ursin (1977), at the Charlottenlund Castle, Danish Institute for Fisheries and Marine Research, were con-structing an equally complex model of the North Sea. Polovina evaluated these modeling efforts and realized the impossibility of constructing species-based dynamic models for biologically diverse areas such as a tropical coral reef ecosystem. From the

Illustration 1. Raymond Lindeman (1915–1942). Illustration 2. Study sites of the International Biological Program (IBP).


Laevastu model, he adopted the principle of mass balance and used this to construct a simple ecological accounting system (Polovina 1984).

“Mass balance” here means that energy input has to bal-ance energy output (including storage) for each species (or functional group) that is being modeled. If we can mass balance one species we can balance the whole ecosystem. For this, we use information about how much food predators require and compare this to how much production is available from their prey. It has to match. This adds constraints to the modeling. Adding constraints is fundamental for all modeling and is one reason that mass balance modeling has been shown to be suc-cessful. In 2009, this led to the modeling approach that Polovina developed, Ecopath (Illustration 3), being recognized by the Na-tional Oceanic and Atmospheric Administration as one of the 10 biggest scientific breakthroughs in the organization’s 200-year history.

I have been developing the Ecopath approach and software for more than two decades, starting off with Daniel Pauly in the Philippines (Christensen and Pauly 1992). Daniel had the idea of merging Polovina’s Ecopath model with ecological network analysis such as that developed by Robert Ulanowicz and oth-ers (Ulanowicz 1986). Finding out how and seeing it through became my Ph.D. work, which was focused on network analysis of trophic interactions based on meta-analysis of aquatic eco-systems.

From this work, let me highlight ecosystem development. One of the greatest ecologists of all times, E. P. Odum (Illustra-tion 4), described a set of ecosystem attributes and how these would change as ecosystems develop (Odum 1969). I quantified most of Odum’s 24 attributes based on some 40 Ecopath mod-els and ranked the models after maturity (Christensen 1995). It worked really well, and since then a number of colleagues have repeated the analysis with the same result. We can rank ecosystems.

The ranking is typical indicator work. You set a number of criteria, extract the numbers, and out comes a ranking. But what attributes and indicators should we use and how do we obtain the overall ranking? I was really fascinated by this during my Ph.D., the idea that one could extract a few indicators from food webs and use those to characterize the state of ecosystems.

There are, however, many indicators and properties in eco-logical network analysis—in order to be noticed, any ecologist doing research in the field must develop his or her own way of capturing the essence of ecosystems. This, combined with very few attempts at evaluating methods and approaches across studies, seems to characterize the field: consensus building has not been an integral part of the development. The big challenge after half a century of ecological network analysis is still to explain what the seemingly endless suite of indicators tells us.

Yet I do not intend to compare network analysis to the “Emperor’s New Clothes” (Illustration 5) —though it is a chal-lenge to interpret the many concepts and indicators. I have worked enough with network analysis to see clear patterns, some of which are consistent and rather straightforward to ex-plain, whereas others are much more elusive. As an example of where I have yet unfulfilled expectations of network analysis, let me point to the identification of critical species in ecosys-tems—the canaries in the coal mine.

I come to think of the Hitchhiker’s Guide to the Galaxy, especially the third of five volumes in the trilogy (Adams 1982). If you don’t remember it: our planet was really a giant super-computer operated by mice. It tolled away for some millions years to answer the biggest, most fundamental question about life, the universe and everything. Eventually the answer came: 42; but by then no one remembered the question. I have often been in that situation with network analysis: It gives the answer, but what was the question? What do the indicators tell us? How do we interpret them? And, importantly, can we use this for making predictions?

Illustration 3. The basic Ecopath model creates a snapshot of an ecosystem at a given point in time: who eats who and how much? Mass balance links predator and prey: there has to be enough food for the predators. Illustration 4. Eugene P. Odum (1913–2002).


Making predictions and evaluating “what if” questions remains elusive, however, as ecological network analysis has demonstrated very little predictive capabilities, such as we are craving for fisheries management. Rather, network analysis tends to be static, almost without exception—it is the study and interpretation of snapshots such as mentioned earlier.

Dynamic considerations have, however, entered from a dif-ferent route. There was a productivity subgroup of IBP that fo-cused on modeling, including dynamic modeling of ecosystems. For this, they created a new field in ecology, systems analysis, and recruited a cohort of bright, quantitative young scientists who used the emerging computers to make models and perform analyses never imagined before.

In essence, what they did was turn the snapshot from the static food web studies into the movie version. And somehow a movie is less open to interpretation than a photo: it adds con-straints. But the modeling had problems. All predator–prey modeling is in essence built on Lotka-Volterra dynamics. This means that the consumption by predators is estimated from the number of predators times the number of prey times a search rate. More predators equals more consumption; more prey equals more consumption. Behind this is a thermodynamic prin-ciple called “mass action,” and this works absolutely fine when mixing reagents and wanting to predict the products. There are, however, problems when using it in ecology.

The systems analysts in the IBP found that their dynamic models were unstable and commonly experienced cycles and model self-simplification. Cycles are fine when modeling, for instance, snowshoe hare–lynx interactions in boreal systems (Krebs et al. 2001), but they are not a regular feature of more diverse ecosystems. What presented a bigger problem was self-simplification: Lotka-Volterra models are inherently unstable, and it is not possible to maintain ecologically similar groups in models with top-down, mass action control. The poorer compet-itors will die out. This was a problem that marred the modeling of ecosystems, and eventually most or all of the IBP modelers left the field to pursue other avenues.

One of the bright young fellows in the IBP was my col-league Carl Walters. He had struggled to make ecosystem mod-els behave and had given up (Hilborn and Walters 1992). Then one day in the early 1990s he was out fishing on a lake in British Columbia with his 9-year-old son, Will. When you fish with Carl you don’t often catch anything, so Will got bored, looked over the side, and saw a lot of nice big Daphnia in the water (Il-lustration 6). He asked: “Why don’t the fish eat them all, Dad?”

Carl went on to give the obvious explanation, one that any fish biologist could have given. “We are fishing for the big trout, they are out here in the open, deep part of the lake. The small trout are hiding along the shore where the big ones don’t come, and they are the ones eating the Daphnia. If the small trout come out here, they would immediately be eaten by the big ones.” A simple, straightforward explanation, and only afterwards did the profound implications of the reply dawn on him.

The fundamental aspect missing in predator–prey modeling was behavior. Organisms are not randomly moving particles as thermodynamics and mass action terms tell us. Think of a coral reef with its swarms of planktivores (Illustration 7). The small stay close to the safety of the reef; the larger stray a bit further away, but only a safe distance. The moment a roaming piscivore, such as a barracuda, comes patrolling by, they all take cover.

llustration 5. Food web representations can be beautiful, but what do they tell us?

Illustration 6. The birth of the foraging arena theory.

Illustration 7. Coral reef representation of the foraging arena—the fish are all planktivores and stay close to the protection of the reef, alert and ready to dive for cover.


The implication of this is that the prey concentration the pi-scivores sees is different from the total planktivore abundance; similarly, the plankton concentration we may measure with nets around the reef is different from what the planktivores actually experience when their foraging is restricted to the immediate safe surroundings of the reef. It takes three to tango: the plank-tivore (dancer one) restricts its activities in response to the pi-scivore (dancer two), and this in turn restricts its own access to plankton (dancer three; Walters and Martell 2004).

From a modeling perspective, Walters developed an elegant way of adding behavior to predator–prey modeling through the foraging arena theory (Ahrens et al. 2012). Organisms change between two behavioral states, being available or unavailable for predation, and including this only calls for adding one ad-ditional parameter to the Lotka-Volterra equation, a behavioral exchange coefficient (that relates to carrying capacity).

One small step of logic but a giant step for modeling—suddenly the ecosystem models started behaving. Where it had been almost impossible to get models to maintain diversity, in-corporation of the foraging arena considerations stabilized the models, which in turn made it possible to replicate population trends in ecosystems. This started in earnest a decade ago when fitting ecosystem modeling to time series data started prolifer-ating, and we have since witnessed a virtual explosion of case studies to the effect that there now are more than 50 of the kind (Illustration 8).

The case studies are based on the Ecosim module of the Ecopath with Ecosim (EwE) approach and software (Chris-tensen and Walters 2004a), and we have drawn a number of lessons from them (Christensen and Walters 2011). As a rule, to explain historic changes in ecosystems we have to consider (1) food web effects, (2) environmental change, and (3) human im-pacts (Illustration 9). An implication of this is that environmen-tal productivity patterns can be identified throughout the food web. There are variable time delays linked to turnover rates and food web constellations, but we can see environmental signals propagate through the food web. We have also seen evidence that environmental productivity can be amplified through the food web. The biological explanation for this may be that more food results in excess beyond maintenance, freeing resources to be allocated to growth and reproduction.

Fit-wise, the models tend to work well for species or groups with strong fisheries impacts; that is, we generally find good agreement with single-species assessment models. Where there are divergences, they can often be explained from model as-sumptions related to food web effects. It is also clear that though trends for some species can be explained, there can be others for which the models are unable to offer insight—often because we have no reliable information about what the important drivers of change may be for such species. There is, however, nothing to indicate that such model failures have implications for the over-all model fit—here one rotten apple does not spoil the bunch.

We see impacts of changes in predator abundance on for-age species (prey release) and, in some cases, the opposite ef-fect, where prey abundance impacts predators. In addition, there are cases where fisheries seemingly outcompete predators as increased fishing mortality on a forage species is accompanied by decline in predation mortality (Walters et al. 2008).

Where ecological networks currently have their greatest potential for contribution in fisheries is in evaluating trade-offs for management. We have reached the point where we can eval-uate tradeoffs between alternative uses of fisheries resources with some authority; see, for example, Christensen and Walters (2004b).

Summing up, ecosystem models can now replicate historic changes in ecosystems and be used to evaluate the relative im-pact of fisheries, food web dynamics, and environmental change (Illustration 10), and notably use this to evaluate trade-offs. With models that behave well enough to replicate the past, we can start thinking of using them to predict the future, to ask what if questions.

This is the foundation for an initiative of the Nippon Foun-dation and the University of British Columbia (NF-UBC) called “Nereus—Predicting the Future Ocean,” which, with partners at Princeton, Duke, Cambridge, and Stockholm universities and at the United Nations Environment Programme’s World Conserva-tion Monitoring Centre, has taken on the task of predicting how

Illustration 8. Case studies where Ecopath ecosystem models have been fitted to time series data.

Illustration 9. Replicating the history of ecosystems calls for inclusion of food web, environmental, and human impact.


the world’s supply of seafood may develop in a future that’s impacted by climate change. The key question we are ask-ing is “Will there be sea-food and healthy oceans for future generations to enjoy?” (Illustration 11). To answer the ques-tion we have to make predictions. There will be uncertainty and un-expected events, but we need to give guidelines for how to ensure that there will be seafood for future generations. What choices must we make for this?

Given that the sea-food market is inter-national, it is a global question, and we have to tackle the question through modeling scaled accordingly. There is, however, no tradition for global modeling in fisheries, and though the Intergovernmental Panel for Climate Change has done a tremendous job of predicting how our physical environment will be impacted by cli-mate change, it is only

now that the consequences of climate change on life on Earth are gaining attention.

The NF-UBC Nereus Program has taken on the task by developing a global ocean-modeling framework, which incor-porates modeling of the physical environment, lower and higher trophic levels, and human activities including governance. It is a framework that incorporates alternative modeling components in order to consider uncertainty through an ensemble approach, inspired by how the Intergovernmental Panel for Climate Change has tackled global environmental modeling.

Uncertainty indeed has to be a major factor in making pre-dictions. Though ecosystem models now offer some predictive capabilities for evaluating major human impacts and making predictions, we cannot make beautiful orchestrated symphonies or detailed predictions, and we will never be able to do that for complex ecosystems. There are two notable factors that prevent this. One is Walters’ “vampires in the basement” (Illustration 12); the other is incomplete knowledge of how systems will react to management interventions (Walters and Martell 2004).

We must expect the unexpected; there will be events that we cannot predict. Invasive species is a case in point and, more generally, behavioral responses in ecosystems are no more pre-dictive than they are for human systems. Let me illustrate with an example: seals have been increasing in the Strait of Geor-gia since culling ceased in the 1970s. Until a few years ago, mammal-eating transient killer whales were not observed in the strait. Then one summer a small pod came in and found plenty of prey—the next summer the whale watching boats counted 100 transient killer whales coming in, and transients have been regular visitors since then. From a modeling perspective, such behavioral events are unpredictable, and they have repercus-sions throughout the ecosystems.

There is also considerable uncertainty about how ecosys-tems will react to many management interventions, especially

where our knowledge about drivers and impact is very incomplete. Our best option wherever this is the case is represented by adaptive management with carefully planned monitoring, experimentation, and adaptation (Walters 1986). Modeling is an integral part of this and is needed to guide the entire pro-cess and limit the risk of making preventable mistakes.

Therefore, though we cannot make de-tailed predictions for how ecosystems will develop, we as a society need to carefully choose what direction to take and we need to avoid the preventable mistakes. For this, it is crucial that policy makers and managers set clear objectives for management and that sci-entists in turn define and evaluate alternative policy options (Illustration 14). We need to manage our ecosystems with a strong com-

Illustration 10. The butler did it: humans are the usual suspects when evaluating fish population trends, but ecosystem models can now be used to evaluate the relative contribution of food web, environ-mental, and human impact.

Illustration 11. Will there be seafood and healthy oceans for future generations to enjoy?

Illustration 12. When making predictions expect the unexpected, the vampire in the basement will bite you.


mitment to moving in a sustainable direction if there indeed is to be seafood and healthy oceans for future generations to enjoy.

ACKNOWLEDGMENTS

Special thanks to Dalai Felinto for the original artwork; to Carl Walters for discussions that helped shape this contribution and for the many years of work that went before it; to Buzz Hol-ling, Steve Carpenter, Eddie Carmack, Line Bang Williams, and Daniel Pauly for discussions and inspiration; and to Bill Fisher and the AFS for the opportunity to address the 142nd annual meeting. This work was supported through the NF-UBC Nereus Program, a collaborative initiative conducted by the Nippon Foundation, the University of British Columbia, and five addi-tional partners, aimed at contributing to the global establishment of sustainable fisheries.

REFERENCES

Adams, D. 1982. Life, the universe and everything. Harmony Books, New York.

Ahrens, R. N. M., C. J. Walters, and V. Christensen. 2012. Foraging arena theory. Fish and Fisheries 13:41–59.

Andersen, K. P., and E. Ursin. 1977. A multispecies extension to the Beverton and Holt theory of fishing, with accounts of phosphorus cir-culation and primary production. Meddelelser fra Danmarks Fiskeri og Havundersøgelser 7:319–435.

Christensen, V. 1995. Ecosystem maturity—towards quantification. Eco-logical Modelling 77:3–32.

Christensen, V., and D. Pauly. 1992. ECOPATH II—a software for bal-ancing steady-state ecosystem models and calculating network char-acteristics. Ecological Modelling 61:169–185.

Christensen, V., and C. J. Walters. 2004a. Ecopath with Ecosim: methods, capabilities and limitations. Ecological Modelling 172:109–139.

Christensen, V., and C. J. Walters. 2004b. Trade-offs in ecosystem-scale optimization of fisheries management policies. Bulletin of Marine Science 74:549–562.

Christensen, V., and C. J. Walters. 2011. Progress in the use of ecosystem modeling for fisheries management. Pages 189–205 in V. Chris-tensen and J. L. Maclean, editors. Ecosystem approaches to fisheries: a global perspective, Cambridge University Press, Cambridge, UK.

Hilborn, R., and C. J. Walters. 1992. Quantitative fisheries stock assess-ment: choice, dynamics, and uncertainty. Chapman and Hall, Lon-don, UK.

Illustration 14. Alice: “Would you tell me, please, which way I ought to go from here?” Cheshire Cat: “That depends a good deal on where you want to get to.” Policy makers need to set clear objectives for management, and scientists need to evaluate alternative options for managers.

Illustration 13. Monitor (left), experiment (center), and adapt (right). The fundamental aspects of adaptive management rely on modeling as the guid-ing factor.

Krebs, C. J., R. Boonstra, S. Boutin, and A. R. E. Sinclair. 2001. What drives the 10-year cycle of snowshoe hares? Bioscience 51:25–35.

Laevastu, T., and H. A. Larkins. 1981. Marine fisheries ecosystem: its quantitative evaluation and management. Fishing News Books, Farn-ham, England.

Lindeman, R. L. 1942. The trophic-dynamic aspect of ecology. Ecology 23:399–418.

Odum, E. P. 1969. The strategy of ecosystem development. Science (New York, N.Y.) 104:262–270.

Polovina, J. J. 1984. Model of a coral reef ecosystem. Coral Reefs 3:1–11.

Ulanowicz, R. E. 1986. Growth and development: ecosystem phenom-enology. Reprint 2000. Springer Verlag, New York.

Walters, C. J. 1986. Adaptive management of renewable resources. Reprint 2001. MacMillan, New York.

Walters, C. J., and S. J. D. Martell. 2004. Fisheries ecology and man-agement. Princeton University Press, Princeton, New Jersey.

Walters, C. J., S. J. D. Martell, V. Christensen, and B. Mahmoudi. 2008. An Ecosim model for exploring ecosystem management options for the Gulf of Mexico: implications of including multi-stanza life history models for policy predictions. Bulletin of Ma-rine Science 83:251–271.


LITTLE ROCK MEETING UPDATE

AFS 2013 Little Rock:Preparing for the Challenges Ahead

Begin making plans to attend the AFS 143rd Annual Meeting, themed “Preparing for the Challenges Ahead,” in Little Rock, Arkansas on September 8–12, 2013. Conference attendees will have the opportunity to network with fisheries professionals and students, stay current on the latest in fisheries science, and enjoy the sights and scenes of central Arkansas.

The AFS Technical Program will be held primarily at the State House Con-vention Center in Little Rock’s River Market district while social events will be held at various venues in the River Market district. The River Market dis-trict is located in downtown Little Rock along President Clinton Avenue on the banks of the Arkansas River. President Clinton Avenue recently was designated one of “10 Great Streets for 2009” by the American Planning Association’s Great Places in America Program. The Convention Center is located near the Clinton Presidential Library; Heifer

International, headquarters of an inter-national non-profit humanitarian organi-zation; and the Arkansas Game and Fish Commission Witt Stephens Jr. Central Arkansas Nature Center. All are ac-cessible by walking or trolley. There are numerous other opportunities in and adjacent to the River Market district for diners, shoppers, and history buffs. Little Rock is located near the intersec-tion of the Arkansas Valley, the Ouachita Mountains, the Ozarks, and the Arkansas Delta, and thus is a jumping-off point for visits to the diverse geography of Arkan-sas. Nearby national parks include Little Rock Central High School National His-toric Site; Hot Springs National Park; and the nation’s first national river, Buf-falo National River. Diverse fishing op-portunities are offered by the state for a variety of species—stripers, largemouth bass, smallmouth bass, walleye, rainbow and brown trout, and catfish.

The 2013 AFS Annual Meeting will feature a broad range of topics associated

with current and future challenges facing aquatic communities, biological scien-tists, managers, other professionals, and the general public. Topics will include assessments of status and trends of spe-cies and communities and their habitats; management challenges and methods; and the potential effects of changing social structures and ecological stresses such as increasing human population, increasing water demands, and a chang-ing climate. Planned symposia topics include large river floodplains, fisheries training for students, roles of fisheries in native American culture, and ancient fishes.

Contributed paper and poster ab-stracts are due by March 15, 2013.

Details will be found at afs2013.com/call-for-papers.

Meeting details and registration will be found at afs2013.com.


Larry DominguezFish/Aquatic Ecologist, EcoAssets Land and Water Resources Consultants, 6525 Arnesen Lane SW, Olympia, WA 98512. E-mail: [email protected]

Beyond Biology: Can We Reconcile the Land–Water–Humankind Interface?

HURRICANE SANDY RETROSPECTIVE

The definition book was thrown at Sandy. Coming and going with multiple characteristics, including hurricane, sub-, post-, extratropical cyclone/storm signs1, the eventual “Superstorm” title continued the vali-dation of changing climate character-istics.

As a Federal Emergency Man-agement Agency (FEMA) reservist, I assist public entities in navigating reg-ulatory compliance of infrastructure projects and conduct dam-age surveys during response and recovery. My first thoughts on providing some insight to my non-hurricane-zone American Fisheries Society colleagues came from my central New Jersey shore ground-zero observations. Returning a stranded cracked-shell horseshoe crab back to the sea, noting marine fish mor-talities in upland areas and freshwater species mortalities from inland lakes that were inundated by saltwater or killed by septic waters unleashed through communities region-wide from over-whelmed sewers, would be interesting to report. The aquatic observations turned to mortality of burrowing animals and altered shorebird habitats. These and some instances of oiled wildlife, stranded individuals, and temporarily affected migra-tions of southbound migratory birds suggest that impact levels varied among species. Effects on beach morphology varied as expected in dynamic and altered shorelines. Some beaches were built up, and others were eroded. Some rivers rose inland to unprecedented levels because of the wind-enhanced tidal surge. Southern area embayments and rivers were lowered because of the offshore winds. Perhaps the most damage related to fish-eries was the fishing industry damage in New York and New Jersey, prompting a federally-declared fishery resource disaster. Amidst the protected inhabited shoreline areas, the news was mixed coastwide: dunes work, seawalls and bulkheads work, sometimes neither work.

1 As a biologist I know that I will be clarified by the weather experts in a follow-up technical article sometime soon. For technical explanations you can visit the Na-tional Oceanic and Atmospheric Administration’s Weather Service and Hurricane Center or affiliated regional university program web pages dedicated to Sandy. One metric stands out: integrated kinetic energy. When adding up all of the total energy contained in the expansive area of tropical storm–force winds and the area of hurricane-force winds, Sandy contained the most integrated kinetic energy of any storm in history—none contained more overall sheer power in recorded history. For energy metric junkies, Sandy’s peak wind energy was 329 TJ—2.7 times higher than Katrina’s peak energy and the equivalent to five Hiroshima-sized atomic bombs.

The damage reminded me of either the poorest develop-ing nations I have visited or a postapocalyptic movie set. The waterlogged mattresses, stuffed animals, and board games piled curbside among the thousands of tons of debris quickly turned my thoughts to my children at home, and how I would feel with such loss. Amidst the volume of debris (which would eventu-ally surpass double that of the Great Pyramid of Giza’s volume) were lifelong possessions and memories. This was an event of a lifetime. People’s lives are changed—many by tragedy. One town recorded a 500-year flood event, another recorded record 30-minute rainfall; how does one even plan for that? In one town hall I noted a 1920 photo on the wall showing damaged beach houses and eroded dunes. The caption: “How the ocean is wrecking cottages at Longport near Atlantic City, N.J.” How timeless is that? These early 20th-century places of grandeur were at risk then. Today, weeks removed from the disaster, contractors in these damaged areas are finishing out partially constructed custom homes that were being built prior to Sandy. There is something wrong with this image as the neighbors empty their homes of damaged possessions, drywall, and in-sulation. Do we reconcile our persistence on these disturbance-prone landscapes with the fact that we are willing to endure these events? Does the pioneering or bold spirit of landholding compel us to stay no matter the risk or cost?

In the evenings of this deployment I had the opportunity to reflect on my lifelong experiences of ocean, river, estuarine, and lake walks that have spanned Asian, African, Middle East-ern, European, tropical, and North American beaches. I have likely walked hundreds of miles of the land–water interface, taking mental notes of this area that has drawn humankind since our origins. In fact, one of my favorite things to do, probably like many of you, is to watch waves run up on the beach, crash against rocks, or peel across a reef. A wave is the expression of

Image on the left shows even structures designed to pass water underneath during flood periods could not withstand wave height and strength created by Sandy. The pilings of many structures were skewed or shifted or lowered from liquefaction and although upper floors may be intact, the result could be demoli-tion. Image on right shows a multistory hotel skewed off its foundation and the deep end of a pool filled with sand a couple of hundred meters from the sea.


Clockwise from top left: Ancient yet practical design of a stilt home (right) stands next to utterly destroyed multi-million-dollar homes in Sandy’s wake on the New Jersey shore. Extreme designs in flood-prone areas of Bahamas (top right) and Lousiana (lower right) propagate the determination of man to live by water despite the cost. Storm-proofing/storm-resistant construc-tion typically means building higher and stronger. FEMA funds assisted in bolstering structures on the Dauphin Islands in Alabama (lower left) in recent years, but ensuing hurricanes created further damage, prompting buy-out strategies and avoidance of rebuilding in highly vulnerable areas.

In 2004, the Banda Aceh, Indonesia, shoreline and beyond was destroyed by a tsunami, and today many areas of the town are being rebuilt but disaster prepa-ration has focused more on comprehensive disaster programs and early warning systems than on structures or relocation that reduce or eliminate the risks from future events.

the vast ocean behind it. It can bring ideas and thoughts to mind. The classic B.C. cartoon honors this concept as the character writes a message on a tablet and sends it to sea and a profound quote returns. The tablet that returned to me read “Water is life—it will do what it must do.”

People have endured flood zone living for millen-nia. In modern history, the priority adjustments appear to be in reducing loss of life. FEMA was planning to revise the flood elevation rules in coastal towns next year, but these recommendations will likely be made for rebuild-ing actions even now. Changes are occurring so fast that today’s requirements are essentially obsolete. Although predictions can be made regarding where extreme flood-ing can occur, such as in embayments and rivers where water impounds during surges, recent storm events with their additional tide and wave factors necessitate broad-scale adjustments. Flood levels that would seriously impact New York were presented over two decades ago, but other priorities kept major actions from being implemented. Is Sandy the scale of action that causes wholesale change at this land–water–human interface?


It seems that we may tend to scale down the disaster and likelihood of whole-sale floodplain development changes because the death toll was astonishingly low and property damage is recoverable. Recall that the 2004 Indian Ocean tsunami death toll topped 280,000. How were you affected by that? I am not sure what it would take to prioritize our preparation and planning. Tsunamis may not give the same preparation and evacuation period that hurricanes do, but it is now near certain that as development in low-lying areas persists, a tsunami would cause an unimaginable death toll.

Finishing out the deployment, the circumstances reminded me of a fa-mous shark movie filmed a little bit north of the Jersey Coast. In the evenings I had the opportunity to skim through the newspapers to keep in touch with the pulse of coastline management. Frankly, with some of the posturing of civic leaders I was reminded of a moment of face wincing because the salty fisherman Quint in Jaws scraped his nails on the chalkboard to silence the crowd. In this case, my wincing was caused by the shrill of adjectives that I am not used to hearing together: “… to rebuild stronger, higher, and more sus-tainable, …” “… stronger, smarter, alternative, more resilient, …” “We need our beaches back open and fully functional by next summer.” In a short while, Chief Brody gets his first look at the size of the beast and, deathly afraid, says, “You’re going to need a bigger boat.” We have caught a glimpse of the beast and certainly our bigger boat is new thinking and management decisions that can reconcile our long-term relationship with the sea. This means that we need to innovate in ways that we have not yet pondered. Images associ-ated with this article show construction approaches, but natural processes and habitat functions of coastlands still remain threatened, and the human risks are only slightly reduced.

The Netherlands’s Flood Protection Act plans on a 1-m sea level rise by 2100 in addition to rigorous analysis and reporting requirements. Centuries-old practices of preparing designated areas as fail-safe flood zones for inun-dation are maintained. As needed, both physical and economical changes are incorporated in the Dutch safety policy. The United States is slowly turning to incorporate this type of thinking. With so many jurisdictional levels mixed with the general population’s coastline affinity, the changes are not sweep-ing and challenge any notion of reconciliation. Climate change remains at the forefront of discussion in the aftermath—but the most immediate climate change must occur with the policies and management actions that reconsider the wisdom of rebuilding human infrastructure on the barrier dunes.

Larry Dominguez is a consulting fish and aquatic ecologist and cofounder of Ecologists With-out Borders (ecowb.org). The opinions expressed in this article are those of the author and do not reflect positions of FEMA, Ecologists Without Borders, EcoAssets, or the American Fisheries Society.

Seaside Heights in typical years (left) and a few days after Sandy. Boardwalks were initi-ated as a pragmatic measure to keep sand from being tracked into beachside attrac-tions, eateries, and accommodations. The New Jersey coast boardwalk evolved into an economic artery in nearly every coastal town, and its function and attraction will certainly be formulated into recovery.

Longstanding boardwalks could not hold their ground during Sandy even with upgraded fasteners. Destroyed boardwalk will total miles. Sand that was washed into town streets and homes, as high as 4 ft. in some areas, was returned to the beach after inspection and some screening. Much of the material was placed in emergency fashion to prepare for the Nor’easter storm that followed within days. The volume of sand that moved into neighbor-hoods extensively damaging homes gives evidence of the historic and natural processes that would otherwise occur to replenish or build barrier islands.


UNIT NEWS

Ben GallowayMSUAFS Outreach Committee Chairperson. E-mail: [email protected]

The Montana State University Student Subunit of the American Fisheries Society (MSUAFS)

The Montana State University Student Subunit of the Amer-ican Fisheries Society (MSUAFS) continues to devote time to public outreach, educating members of the local community about the importance of fisheries conservation, and responsible resource use. Partnering with Montana Fish, Wildlife & Parks and the Children’s Museum of Bozeman, MSUAFS conducted its second annual Family Fishing Fun event in June 2012, where children and adults from the local community learned about the importance of fisheries conservation, and many were introduced to fishing for the first time. The event was a great success, and sixteen children ages 5–12 and their parents spent three weeks learning about fisheries and honing their newly acquired fishing skills on local ponds.

Participants took part in classroom activities that provided a firsthand look at fish ecology, life history, and conservation. The first week consisted of a hands-on encounter with macro-invertebrates collected from local streams and ponds. Children

were encouraged to handle a myriad of aquatic invertebrates, including large salmonfly and dragonfly larvae, mayfly larvae, leeches, and snails. Children were taught the importance of these invertebrates as food sources for fish and other organisms, but also as a key piece of a healthy ecosystem. The second week, children learned about general fish biology and the ad-vantages of different body forms in lentic and lotic environ-ments. The final week, children participated in fish dissections and were thrilled with the chance to investigate the internal anatomy of a trout. Each week, children were also provided equipment and instruction (including a safety talk) on how to fish, and given the opportunity to catch fish at a local pond. Many children experienced fishing for the first time, and many had even caught their first fish! By providing the community with opportunities like these, MSUAFS continues to help edu-cate and inform the public about fish ecology and conservation, while fostering the desire to protect the fisheries resource in children and adults alike.

Young participants investigate their catch, while MSUAFS members provide instruction in casting and invertebrate identifica-tion. Photo credits: Susan Denson-Guy, Christopher Guy, Callie Hamilton, and Eleanor Barker.


Revisiting the Process for Qualifying for the AFS/Sea Grant Best Student Presentation and Poster Awards

The American Fisheries Society (AFS)/Sea Grant Best Stu-dent Paper and Presentation awards are an important mecha-nism to highlight the students in our profession and the future leaders of AFS. In 2007, the process for selecting the AFS/Sea Grant Best Student Presentation and Poster awards was sub-stantially revised (Sutton et al. 2007). Prior to the changes, all students were randomly assigned to a small group of judges. Accommodating the ever-increasing numbers of students giving oral or poster presentations at the AFS annual meeting required a large team of volunteer judges. However, it was not always as-sured that each student would receive evaluations from the same number of judges. Some students received reviews from three or more judges, whereas others did not receive any evaluations (Sutton et al. 2007). Further, the high levels of variability in scoring among judges created difficulty in determining winners in either the poster or presentation medium. During the AFS an-nual meeting in San Francisco (2007), a new process for select-ing finalists and judging these prestigious awards was piloted with much success and positive feedback from both students and judges. This process has remained in place since that time with only a few minor changes.

Although the selection and evaluation process for this award has been in place for 6 years, the system still appears to be largely unknown to the society. For example, in 2012, only 22 applications for the Best Student Presentation award and six applications for the Best Student Poster Award were completed. This is far below expectations given the numbers of student members in the AFS and the number of students who attend the annual meeting. Here, we describe the application process for students to be considered for Best Student Presentation or Best Student Poster awards, the selection of finalists, and the judging that occurs during the AFS annual meeting. We also address questions that are frequently asked by students and advisors in an effort to enhance participation in the process.

COLUMNGuest Director’s Line

THE BEST STUDENT PRESENTATION/POSTER AWARD PROCESS

How do students apply for consideration for either award?

Application for consideration for these awards is very simi-lar to the standard abstract submission process for the AFS an-nual meeting but with two additional requirements: an extended abstract and letter of support from the student’s advisor (Figure 1). Students begin by uploading a standard abstract for one me-dium (i.e., poster or presentation) to the online abstract submis-sion system as described in the Call for Papers in Fisheries. Once the standard abstract is uploaded, the system will ask the student whether he or she wishes to be considered for the Best Student Presentation or Best Student Poster award. If the stu-dent selects “no,” then the process is complete and the student is assigned to an appropriate alternative symposium by the Annual Meeting Program Committee (Figure 1). If the student selects “yes,” then the abstract submission system prompts the student to upload an extended abstract, which is no more than three pages in length and includes background, methods, results (in-cluding up to five tables or figures), discussion, and references. Examples of extended abstracts can be found online at http://www.fisheriessociety.org/education/BSP.htm. The student will also be reminded to request an electronic letter of support from his or her advisor that must be e-mailed to the organizers of the Best Student Presentation/Poster symposium. The letter must indicate that the student’s research is at a stage appropriate for award consideration. In other words, the research project or specific objective must be complete or nearly complete. Both the extended abstract and the advisor’s support letter must be provided by the same deadline as outlined in the Call for Papers in Fisheries for standard abstract submissions. Failure to receive all of the required components will prohibit the student’s ap-plication from further consideration (Figure 1).

Melissa R. WuellnerDepartment of Natural Resource Management, South Dakota State University, Box 2140B, Brookings, SD 57007. E-mail: [email protected]

Shannon FisherWater Resources Center and Minnesota River Board, 184 Trafton Science Center South, Mankato, MN 56001


Figure 1. Flowchart describing the process for selection of the Best Student Paper/Poster at the AFS annual meeting. This process has been followed since the 2007 meeting in San Francisco.

The required elements were provided by the student and advisor. What happens next?

The next step in the process involves evaluation of the extended abstracts sub-mitted by those students who completed the application process. The full list of ap-plicants is reduced to a maximum of 20 finalists in each medium (though the total number of poster submissions has never exceeded 10 since 2007). From 2007 to 2011, the organizers of the Best Student Presentation/Poster symposium evaluated all of the extended abstracts for both mediums and selected the finalists. In 2012, a new system was piloted that assigned abstracts to division representatives from the

Education Section and the Student Subsec-tion who evaluated and scored each of the assigned abstracts based on a rubric (see the Extended Abstract Evaluation Form at http://www.fisheriessociety.org/education/BSP.htm). This system will likely continue into the future. The evaluators’ scores are tallied and the distributions of scores are examined by the symposium organizers to select the list of finalists. All student appli-cants are notified whether or not they have been selected as a finalist very soon after the abstract submission deadline (Figure 1). Those students not selected as finalists are placed into appropriate alternative sympo-sia by the AFS Annual Meeting Program Committee. Finalists in both mediums are assigned to the Best Student Presentation/Poster symposium and compete for the awards at the AFS Annual Meeting (Figure 1).

For those who are selected as finalists, what happens during the AFS annual meeting?

The final phase of the competition for these awards takes place during the AFS annual meeting; student finalists present their oral presentation or poster early in the week. The Best Student Presentation sym-posium typically takes place on Monday afternoon and Tuesday morning each year. The finalists for Best Student Poster stand next to their work during the Monday night poster and trade show social. Students must adhere to the guidelines for poster dimen-sions and presentation criteria as described on the AFS annual meeting website for all presenters in any symposium. The Best Student Presentation/Poster co-organizers assemble a team of volunteers to evalu-ate the finalists within each medium—five judges for presentations and three for post-ers. Each judge scores and provides com-ments to all students within each medium category based on established evaluation guidelines (see the Best Student Presenta-tion and Poster Judging Sheets at http://www.fisheriessociety.org/education/BSP.htm). At the conclusion of the Best Student Presentation symposium on Tuesday, all of the presentation and poster judges meet to examine the distributions of scores for all finalists within each medium, discuss any potential ties, and decide on the winners and honorable mentions in each category. The overall winners and honorable mentions for


the Best Student Presentation/Poster awards are then announced during the AFS business meeting on Tuesday afternoon. The overall winners in each category receive their plaques and prize money ($450 for each award) during the Education Section’s business meeting the following year. All student finalists re-ceive scores and comments from the judges by mail from the Best Student Presentation/Poster co-organizers within a few weeks of the AFS Annual Meeting (Figure 1).

FREQUENTLY ASKED QUESTIONS

Question #1: Why are the extended abstracts and advisor letter required as part of the application for the Best Stu-dent Presentation/Poster awards?

Both requirements provide support that the student’s over-all research project or at least one of the objectives is at or near completion. Results and tables and/or figures as required in the extended abstract indicate that data have already been collected and analyzed. The advisor’s letter lends further evidence that the student’s work is at the appropriate stage for consideration for these awards. Students will often present proposal or pre-liminary data at the AFS annual meeting to gain audience feed-back or practice in scientific presentations, both of which are laudable goals. However, such presentations are not competitive and do not fit the goals of the awards described here. Completed projects represent the best of the best of what AFS students can produce.

Question #2: I was a student at [insert name of univer-sity or college here] and recently completed my master’s or Ph.D. program. I am currently working as a fisheries professional. The presentation I am planning for the AFS annual meeting is from my recently completed thesis or dissertation. Am I still eligible to apply for this award?

In short, the answer to this question is yes. If the research that the recently graduated student will be presenting was com-pleted during the time spent in undergraduate or graduate study and all of the required application materials are received by the deadline listed in the Call for Papers in Fisheries, the student will be considered for either the Best Student Presentation or Poster award, depending on the medium he or she selected.

Question #3: Can I [a student] give the same presentation in both the Best Student Presentation symposium and in a second symposium that is related to my research topic?

The answer to this question is a simple no. Difficulties are often faced by the AFS Annual Meeting Program Committee to accommodate all of the presentation requests received. By submitting the same presentation in two separate time slots, students are potentially affecting someone else’s opportunity to present his or her work. A student may decide that presenting in a topic-specific symposium will be a better opportunity to share information with colleagues and network with potential collabo-rators and employers and that this will benefit him or her more than possibly winning the Best Student Presentation award. The

student and his or her advisor should collectively decide what is in the student’s best interest in the case of such conflicts. But we emphasize that students should only submit their presentations once for inclusion in the AFS annual meeting program because if the submission is not chosen for inclusion in the Best Student Paper/Poster symposium, it will be automatically submitted for inclusion in the general program.

Question #4: Did you mention that there was prize money associated with the Best Student Presentation/Poster awards?

The overall winners in both categories each receive a US$450 cash prize and a plaque during the Education Section business meeting the following year. Honorable mentions re-ceive a plaque but no cash prize.

CONCLUSION

The criteria for applying and evaluating students for the AFS/Sea Grant Best Student Presentation and Poster Awards since 2007 is more rigorous compared to the previous system, but final evaluations and feedback from student finalists and judges indicate that they system is greatly improved. The col-lective confidence in selecting deserving winners and recogniz-ing honorable mentions has increased since implementation of the new system. We hope that the number of students applying for consideration for these awards will continue to increase for 2013 and into the future. If there are any other questions that were not addressed in this article, we encourage the reader to contact the Best Student Presentation/Poster co-organizers or the Education Section officers (see http://www.fisheriessoci-ety.org/education/officers.htm for contact information). Best of luck to all who apply! We look forward to seeing the products of your hard work.

ACKNOWLEDGMENTS

We thank Trent Sutton, Donna Parish, and James Jackson for initiating the changes to the selection process for the AFS/Sea Grant Best Student Presentation and Poster awards. Previ-ous drafts of this article were improved with comments from Mike Quist and Craig Paukert. We appreciate the symposium organizers who served on the Best Student Presentation/Poster committee and all of the judges who have volunteered since 2007. Finally, we thank the AFS annual meeting program com-mittees and their chairs as well as AFS staff who have aided in the success of the awards symposia from 2007 through 2012.

REFERENCE

Sutton, T. M., D. L. Parrish, and J. R. Jackson. 2007. Time for a change: revising the process for judging student presentations at the annual meeting. Fisheries 32(1):42–43.


Small Impoundment management

In north amerIcaEdited byJ. Wesley Neal andDavid W. Willis

This book is an in-depth overview of biota, habitat, and human management in small water bodies up to approximately 40 ha in surface area. Authors were selected to cover the wide geographic diversity of ponds and pond management throughout North America.

The first section (Introduction and History) defines small impoundments, provides a concise history of pond management, overviews pond resources in the USA and world, and dis-cusses the importance of small impoundments. Section Two (Pond Environment) addresses proper construction considerations, explores the physical and chemical characteristics of these waters, discusses productivity, and examines methods to manipulate environmental conditions in small waters. Section Three (Fish Management) describes current stocking practices and species selection, addresses the importance of proper harvest and assess-ment, and explores mechanisms involved in population dynamics and the occurrence of crowded predator or prey populations. Section Four (Problem Troubleshooting) addresses problems that can arise in small impoundments and provides solutions. Section Five (Op-portunities) provides a platform for topics that previously had received limited treatment in the educational literature. Thorough discussions of fee fishing and community fishing opportunities for small impoundments are provided, as is an overview of careers in pri-vate sector pond management and extension/outreach. Finally, the technical aspects of managing small impoundments for wildlife are described in detail.

A primary use for this book will be university classes on pond or small impoundment management for advanced undergraduate or graduate students. Practicing fisheries pro-fessionals should also find substantial value in the depth of information provided by the book. Finally, private pond owners will find the book to be useful as they seek to learn more about ponds and pond management.

420 pages, index, hardcoverList price: $79.00AFS Member price: $55.30Item Number: 550.69CPublished November 2012

TO ORDER:Online: www.afsbooks.orgAmerican Fisheries Societyc/o Books InternationalP.O. Box 605Herndon, VA 20172Phone: 703-661-1570Fax: 703-996-1010


Evaluating the Effects of Electricity on Fish Embryos as a Potential Strategy for Controlling Invasive Cyprinids. Sam Nutile, Jon J. Amberg, and Reuben R. Goforth. 142: 1–9.

Influence of Sex and Reproductive Status on Seasonal Movement of Lake Sturgeon in Namakan Reservoir, Minnesota–Ontario . Stephanie L. Shaw, Steven

R. Chipps, Steve K. Windels, Molly A. H. Webb, and Darryl T. McLeod. 142: 10–20.

Persistent Organic Pollutants in Juvenile Chinook Salmon in the Columbia River Basin: Implications for Stock Recovery. Lyndal Johnson, Bernadita Anulacion, Mary Arkoosh, O. Paul Olson, Cath-erine Sloan, Sean Y. Sol, Julann Spromberg, David J. Teel, Gladys Yanagida, and Gina Ylitalo. 142: 21–40.

Development and Evaluation of a Bioenergetics Model for Bull Trout. Matthew G. Mesa, Lisa K. Weiland, Helena E. Christiansen, Sally T. Sauter, and David A. Beauchamp. 142: 41–49.

Connections between Campeche Bank and Red Snapper Popula-tions in the Gulf of Mexico via Modeled Larval Transport. Donald R. Johnson, Harriet M. Perry, and Joanne Lyczkowski-Shultz. 142: 50–58.

Estimating Effective Sample Size for Monitoring Length Distribu-tions: A Comparative Study of Georges Bank Groundfish. Yuying Zhang and Steve X. Cadrin. 142: 51–67.

Response of Wild Trout to Stream Restoration over Two Decades in the Blackfoot River Basin, Montana. Ron Pierce, Craig Podner, and Kellie Carim. 142: 68–81.

Validating Back-Calculation Models Using Population Data. K. O. Perez and S. B. Munch. 142: 82–94.

[Note] Swordfish Vertical Distribution and Habitat Use in Relation to Diel and Lunar Cycles in the Western North Atlantic. Justin D. Lerner, David W. Kerstetter, Eric D. Prince, Liana Talaue-McMa-nus, Eric S. Orbesen, Arthur Mariano, Derke Snodgrass, and Gary L. Thomas. 142: 95–104.

Trends and Synchrony in Black Bass and Crappie Recruitment in Missouri Reservoirs. Paul H. Michaletz and Michael J. Siepker. 142: 105–118.

Exploring the Influence of Stock–Recruitment Relationships and Environmental Variables on Black Bass and Crappie Recruitment Dynamics in Missouri Reservoirs. Michael J. Siepker and Paul H. Michaletz. 142: 119–129.

JOURNAL HIGHLIGHTSTransactions of the American Fisheries Society Volume 142, Number 1, January 2013

Assessing Juvenile Chinook Salmon Behavior and Entrainment Risk near Unscreened Water Diversions: Large Flume Simula-tions. Timothy D. Mussen, Dennis Cocherell, Zachary Hockett, Ali Ercan, Hossein Bandeh, M. Levent Kavvas, Joseph J. Cech Jr., and Nann A. Fangue. 142: 130–142.

Relationships between the Abundance of Pacific Lamprey in the Columbia River and Their Common Hosts in the Marine Environ-ment. Joshua G. Murauskas, Alexei M. Orlov, and Kevin A. Siwicke. 142: 143–155.

[Forum] Issues Regarding the Use of Sedatives in Fisheries and the Need for Immediate-Release Options. J. T. Trushenski, J. D. Bowker, S. J. Cooke, D. Erdahl, T. Bell, J. R. MacMillan, R. P. Yanong, J. E. Hill, M. C. Fabrizio, J. E. Garvey, and S. Sharon. 142: 156–170.

Estimating Spatial and Temporal Components of Variation for Fisheries Count Data Using Negative Binomial Mixed Models. Brian J. Irwin, Tyler Wagner, James R. Bence, Megan V. Kepler, Wei-hai Liu, and Daniel B. Hayes. 142: 171–183.

Longitudinal Length Back-Calculations from Otoliths and Scales Differ Systematically in Haddock. Hannes Baumann, Sandra J. Sutherland, and Richard S. McBride. 142: 184–192.

Quantifying the Effects of Aging Bias in Atlantic Striped Bass Stock Assessment. Hongsheng Liao, Alexei F. Sharov, Cynthia M. Jones, and Gary A. Nelson. 142: 193–207.

Potential Effects of Changes in Temperature and Food Resources on Life History Trajectories of Juvenile Oncorhynchus mykiss. Jo-seph R. Benjamin, Patrick J. Connolly, Jason G. Romine, and Russell W. Perry. 142: 208–220.

Spatial Segregation of Spawning Habitat Limits Hybridization be-tween Sympatric Native Steelhead and Coastal Cutthroat Trout. T. W. Buehrens, J. Glasgow, C. O. Ostberg, and T. P. Quinn. 142: 221–233.

Evaluation of Sample Design and Estimation Methods for Great Lakes Angler Surveys. Zhenming Su and David Clapp. 142: 234–246.

Evaluation of Electrofishing Catch per Unit Effort for Indexing Fish Abundance in Florida Lakes. Matt A. Hangsleben, Micheal S. Allen, and Daniel C. Gwinn. 142: 247–256.

Temporal Variation in the Physiological Responses in Largemouth Bass following Small Club Angling Tournaments. Matthew M. Van-Landeghem, David H. Wahl, and Cory D. Suski. 142: 257–267.

Modeling Prey Consumption by Native and Nonnative Piscivorous Fishes: Implications for Competition and Impacts on Shared Prey in an Ultraoligotrophic Lake in Patagonia. Romina Juncos, David A. Beauchamp, and Pablo H. Vigliano. 142: 268–281.

Reproductive Biology of the Tiger Grouper in the Southern Gulf of Mexico. Doralice Caballero-Arango, Thierry Brulé, Virginia Nóh-Quiñones, Teresa Colás-Marrufo, and Esperanza Pérez-Díaz. 142: 282–299.

Environmental Constraints on Piscivory: Insights from Linking Ultrasonic Telemetry to a Visual Foraging Model for Cutthroat Trout. Adam G. Hansen, David A. Beauchamp, and Casey M. Bald-win. 142: 300–316.


little additional burden on financial and human resources. Our society has taken a step in this direction through our contract with Taylor & Francis for publishing our journals and publiciz-ing them worldwide; our publications staff has seen an increase in manuscript submissions from other continents since the Tay-lor & Francis contract went into effect. I have also directed the AFS Publications Overview Committee to evaluate alternatives for increasing international visibility of our journals, including name changes that could make them more attractive to a global readership.

Thinking back to that governing board meeting in 1993, our society has made significant progress in recognizing that we are part of a much larger network addressing fisheries science, management, and conservation issues on a global scale. The ability to communicate at the speed of light has strengthened the network considerably, but there is still much more we can do. Perhaps Carlos should have handed out quarters to governing board members who referred to the society-wide level of the AFS organization as “the international”—something to ponder for future meetings.

Continued from page 51 AFS Seeks Chief Science Editor for Fisheries Magazine

The American Fisheries Society (AFS) seeks a scientist with a broad perspective on fisheries to oversee the science content of Fisheries. Editor must be committed to fast-paced dead-lines, and would be appointed for a five-year renewable term which begins in 2013. • Assign an appropriate science editor for each sci-

entific paper submitted to Fisheries.

• Make final publication decisions based on peer reviews orchestrated by science editors.

• Ensure the veracity of each issue’s total scientific content.

• Help recruit and retain science editors, as well as providing mentoring and guidance.

• Solicit cutting-edge submissions as well as ensur-ing broad coverage.

• Work with Fisheries Managing Editor and AFS Publications Director on the content, themes, and direction of the scientific aspects of Fisheries

To be considered, send current curriculum vitae along with a letter of interest explaining why you want to be the Chief Sci-ence Editor by e-mail to [email protected].

Note: The Chief Science Editor receives an honorarium, and support to attend the AFS Annual Meeting.

BIOLOGISTS | PROJECT MANAGERS | ENGINEERS208.275.0011 | [email protected] | ww.biomark.com

A Reputation You Can TrustBiomark specializes in designing, fabricating and installing customized detection arrays to meet your specific project needs. Our systems provide you the peace of mind and data collection reliability that you can expect with 22 years experience of PIT tag system development and implementation.

Leading the industry in product development, manufacturing, implementation and supply.

HPT Tags | Outstanding performance| FDX-B & HDX| Bulk & Pre-loaded| Competitive pricing

Hand Readers| Water resistant & durable| Time/Date stamp| Large memory| Easy memory

download

Specializing in PIT Tag Technology


DATE EVENT LOCATION WEBSITEFebruary 21–25, 2013 Fish Culture Section Mid-Year Business Meeting Nashville, TN was.org/WasMeetings/meetings/Default.

aspx?code=AQ2013

February 26–27, 2013 2013 AFS North Carolina Chapter Annual Meeting Burlington, NC sdafs.org/ncafsContact: Greg Cope at [email protected]

March 13–16, 2013 31st Annual Salmonid Restoration Conference Fortuna, CA calsalmon.org/salmonid-restoration-conference/31st-annual-salmonid- restoration-conference

April 8–12, 2013 7th International Fisheries Observer and Monitor-ing Conference (7th IFOMC)

Viña del Mar, Chile IFOMC.com

April 25–26, 2013 NPAFC 3rd International Workshop on Migration and Survival Mechanisms of Juvenile Salmon and Steelhead in Ocean Ecosystems

Honolulu, HI npafc.org/new/index.html

May 20–24, 2013 AFS Piscicide Class - Planning and Executing Successful Rotenone and Antimycin Projects

Logan, UT fisheriessociety.org/rotenone Contact: Brian Finlayson at [email protected]

June 25–27, 2013 2013 International Conference on Engineering & Ecohydrology for Fish Passage

Corvallis, OR fishpassage.umass.eduContact: Dr. Guillermo R. Giannico at [email protected]

August 3–7, 2014 International Congress on the Biology of Fish Edinburgh, United Kingdom

icbf2014.sls.hw.ac.uk

CALENDARFisheries Events

To submit upcoming events for inclusion on the AFS web site calendar, send event name, dates, city, state/province, web address, and contact information to [email protected].

(If space is available, events will also be printed in Fisheries magazine.)

More events listed at www.fisheries.org

From the Archives

The relations of the temperature of the water to tile movements of the menhaden schools having been studied, a new question is at once suggested. When the schools disappear from our coast, driven by falling temperatures, where do they go? The answer must be in the form of a theory, for no one has seen them during their win-ter absence – at least no one has been able to identify the New England and Middle States fishes after their departure in the autumn.

Professor J. Browne Goode (1878): Migration of Fishes, Transactions of the American Fisheries Society, 7:1, 27-67.


ANNOUNCEMENTSFebruary 2013 Jobs

Graduate Research Asst–Ph.D.WV Univ – Wildlife & Fisheries Resources ProgramStudent Salary: $19,848 annual stipend, plus full tuition waiver

Closing: Until filled

Responsibilities: Successful applicant will conduct genetic re-search on juvenile lake sturgeon in Lake Superior, with the pri-mary goal of assigning individuals to their spawning population of origin. Opportunity for additional research questions related to juvenile lake sturgeon ecology/genetics.

Qualifications: Highly motivated student with interest in fish ge-netic research. Requires a M.S. in fisheries science, ecology, genet-ics, or related field. Experience in genetics preferred. Submit letter of interest, CV, & unofficial transcript and GRE scores to [email protected].

Contact: Dr. Amy Welsh, at below email.

Email: [email protected]

Link: wildlife.wvu.edu

Employers: to list a job opening on the AFS online job center submit a position description, job title, agency/company, city, state, respon-sibilities, qualifications, salary, closing date, and contact information (maximum 150 words) to [email protected]. Online job announce-ments will be billed at $350 for 150 word increments. Please send bill-ing information. Listings are free (150 words or less) for organizations with associate, official, and sustaining memberships, and for individ-ual members, who are faculty members, hiring graduate assistants. if space is available, jobs may also be printed in Fisheries magazine, free of additional charge.

Aquatic Invasive Species (AIS) TechWY Game and Fish DeptTemporary Salary: $13.00 per hour

Closing: 3/1

Responsibilities: Aquatic Invasive Species (AIS) Technician (up to 41 positions) will inspect watercraft at key locations around the state in an effort to prevent the introduction and spread of aquatic invasive species. Participates in outreach and education efforts to inform the general public about the impacts of aquatic invasive species. Duties may include conducting physical watercraft inspec-tions and decontaminations. Boater interviews at watercraft check stations. Provide outreach information and assistance to the boating public. Water quality and invasive species sampling, data entry, and equipment maintenance.

Qualifications: Must be able to work well independently with minimal supervision and also able to work well with others on a crew. Experienced in the operation of watercraft, 4wd vehicles, and pulling trailers. Requires effective communication and public relations skills in daily interactions with the public. Weekend and holiday work will be required. Position may require extended days (4 days) in campers. Must have a valid driver’s license.

Contact: Preference will be given to those who submit a cover letter to Beth Bear, AIS Coordinator, 528 S. Adams, Laramie, WY 82070 or below email, in addition to submitting the state applica-tion. The cover letter should include career plans, suitability for the job, and preferences for work locations as well as dates of avail-ability (starting and ending dates are particularly important).


Link: statejobs.state.wy.us/JobSearch.aspx

Fisheries BiologistPrivate sector / Southeastern Pond Management, Birmingham, ALPermanent Salary: DOE/DOQ


Responsibilities: Performing pond evaluations using electrofish-ing equipment, using the acquired data to create pond management plans, customer follow up / customer relations

Qualifications: B. S. degree in Fisheries biology or related field is required, Masters degree in Fisheries biology is preferred. Prefer-ence may be given to applicants with boating and electrofishing experience.

Contact: Scott Cherones, at below email.


Link: sepond.com

Staff Fishery Economist /Senior Fishery EconomistNorth Pacific Fishery Management CouncilPermanent Salary: GS-11/12/13 equivalent — (Salary range 68,000–98,000 DOE). Exceptional applicants may be considered for placement at the Senior economist level, (salary range 90,000–120,000 DOE).


Responsibilities: The Department of Biology at Coastal Carolina University invites applications for a full-time tenure-track position at the Assistant Professor beginning Fall 2013.

Responsible for teaching Ichthyology, Principles of Ecology, and will have the opportunity to teach other advanced courses in their area of specialization. Successful candidates will be expected to emerge as exemplars of teaching, contribute to both introductory and graduate courses and develop potentially fundable research programs in fish ecology involving undergraduate and graduate students. Candidates whose research interests include freshwater systems are encouraged to apply.

Qualifications: PhD


Link: jobs.coastal.edu

Our transmitters aren’t as interesting as what researchers put them on.

But, they are more reliable.ATS offers the smallest, longest lasting fish transmitters in the world; VHF, acoustic and archival. We provide complete tracking systems, including receiver/dataloggers, antenna systems and more. Plus, our coded system virtually eliminates false positives from your data set, providing you with 99.5% accuracy, a level not available from any other manufacturer.

World’s Most Reliable WildlifeTransmitters and Tracking Systems

Contact ATS for details.

ATStrack.com • 763.444.9267

fisheries · fisheries • vol 38 no 2 • february 2013• 49 contents fisheries vol 38 no 2 feb...

Documents