Approaches for studying fish production: Do river and lake researchers have different perspectives?

Biased perspectives of fisheries researchers may hinder scientific progress and effective management if limiting factors controlling productivity go unrecognized. We investigated whether river and lake researchers used different approaches when studying salmonid production and whether any differences were ecologically supported. We assessed 564 peer- reviewed papers published between 1966 and 2012 that studied salmonid production or surrogate variables (e.g., abundance, growth, biomass, population) and classified them into five major predictor variable categories: physical habitat, fertility (i.e., nutrients, bottom-up), biotic, temperature, and pollution. The review demonstrated that river researchers primarily analyzed physical habitat (65% of studies) and lake researchers primarily analyzed fertility (45%) and biotic (51%) variables. Nevertheless, understudied variables were often statistically significant predictors of production for lake and river systems and, combined with other evidence, suggests that unjustified a priori assumptions may dictate the choice of independent variables studied. Broader consideration of potential limiting factors on fish production, greater research effort on understudied genera, and increased publication in broadly scoped journals would likely promote integration between lentic and lotic perspectives and improve fisheries management. Les perspectives biaisees de chercheurs du domaine des peches pourraient faire entrave a des avancees scientifiques et a une gestion efficace, si cela devait se traduire par la non-reconnaissance de facteurs qui limitent la productivite. Nous avons verifie si les chercheurs travaillant en riviere, d'une part, et en lac, d'autre part, utilisent des approches differentes pour etudier la production de salmonides et si certaines differences sont appuyees par des considerations ecologiques. Nous avons examine 564 articles evalues par des pairs publies de 1966 a 2012 et portant sur la production de saumons ou des variables substitutives (p. ex. abondance, croissance, biomasse, population) et les avons classes selon cinq grandes categories de variables explicatives, soit celles reliees a l'habitat physique, a la fertilite (c.-a-d. nutriments, effet ascendant), au biote, a la temperature et a la pollution. Cet examen demontre que les chercheurs travaillant en riviere analysent principalement l'habitat physique (65% des etudes), alors que les chercheurs travaillant en lac analysent principalement des variables associees a la fertilite (45%) et au biote (51%). Cela etant, des variables sous-etudiees constituent souvent des variables explicatives statistiquement significatives de la production de systemes lacustres et fluviaux ce qui, combine a d'autres observations, semble indiquer que des hypotheses a priori non justifiees pourraient dicter le choix des variables independantes etudiees. Une plus grande prise en consideration de facteurs qui pourraient limiter la production de poissons, plus de recherche axee sur des genres sous-etudies et un nombre accru de publications dans des revues a grande portee favoriseraient probablement l'integration des perspectives lentique et lotique et amelioreraient la gestion des peches. [Traduit par la Redaction]
Introduction One of the primary objectives in the study of freshwater ecosystems is to better understand controls on primary and secondary production, particularly the production of fish species of economic, [...]