Comparing Otoliths and Scales as Structures Used to Estimate Ages of Largemouth Bass: Consequences of Biased Age Estimates

Abstract

Management agencies often estimate the ages of Largemouth Bass Micropterus salmoides based on the examination of scales—a structure that is known to produce biased estimates—without knowing how the associated bias affects management decisions. We sought to understand the effects of this bias by comparing population metrics that were predicted using scale‐derived and otolith‐derived age data. We collected scales and otoliths from Largemouth Bass that were sampled during standard electrofishing surveys. The age of each fish was estimated independently by three separate readers using both scales and otoliths. We assessed the average coefficient of variation for scale‐derived and otolith‐derived age estimates, examined the bias of scale‐derived age estimates, and estimated von Bertalanffy growth model parameters by using ages estimated from scales and otoliths. These parameter estimates were used in yield‐per‐recruit simulations that predicted yield and the percentage of individuals in the cohort surviving to 380 mm (proportional size distribution [PSD] 380) or to 470 mm (PSD 470) at several levels of natural mortality and fishing mortality. Otolith‐derived age estimates were more precise; scale‐derived age estimates showed significant positive bias for fish younger than age 6 and significant negative bias for fish older than age 6. Von Bertalanffy parameter estimates were significantly different when using ages estimated from scales and those estimated from otoliths. Modeling indicated that estimates of yield and PSD 380 resulting from the two structures were similar. However, the use of scale‐derived ages resulted in underestimating the impact of fishing mortality on PSD 470 by as much as five times at low levels of natural mortality and fishing mortality. Our estimates of precision and bias were similar to other comparisons of scales and otoliths, and the results of our yield‐per‐recruit simulations are likely generally applicable for Largemouth Bass management. Trophy fishing is a common management objective, and managers relying on scale‐based age data could be less likely to adopt the restrictive harvest regulations that are critical for producing trophy Largemouth Bass. Received November 17, 2016; accepted June 22, 2017 Published online August 22, 2017

Modeling Growth: Consequences from Selecting Samples by Size

“…When added to real‐world size‐selectivity of gears, the additional effect of size‐stratified sampling would make the estimation of biologically meaningful growth coefficients a matter of conjecture. These features need not compromise within‐study objectives to compare growth rates in time or space that can be addressed with appropriate statistical control or experimental design (e.g., Nate and Bremigan ; Tyszko and Pritt ). However, they may result in misleading analyses in subsequent studies that adopt the VB estimates for other purposes.…”

Standard Assessments Reveal Context‐Dependent Responses of Crappie Populations to a Length‐Based Regulation in Ohio Reservoirs

“…) using assigned‐age data from age–length keys (as above). We used nonlinear mixed‐effects models to fit VBGFs, using the MLL (preregulation versus postregulation) as a fixed effect on each von Bertalanffy parameter ( L ∞ , K , and t 0 ); we included crappie species and reservoir as random effects on each von Bertalanffy parameter (see also Tyszko and Pritt ). The significance of the MLL as a fixed effect on von Bertalanffy parameters for each reservoir group model was assessed using P ‐values (α = 0.05).…”

Growth of adult river herring that spawn in tributaries of the Potomac River in northern Virginia

“…When age-2 individuals were removed from the analysis in this study, K increased from 0.179 to 0.243 mm/year for alewife (both sexes combined) and from 0.525 to 0.697 mm/year for blueback herring (both sexes combined). One hypothesis for why age-2 individuals were available in this study and not in previous studies is that the populations could have spawned at earlier ages in this study because of years of overfishing, which has been documented in Atlantic cod (Gadus morhua) (Trippel, 1995), several Pacific salmon species (Ricker, 1981), and numerous other fish species (Darimont et al, 2009). Fishermen target large individuals within a population.…”

“…Using data sets with biased ages can result in poor population modeling and conflicting strategies for population management (Beamish and McFarlane, 1987;Bertignac and de Pontual, 2007;Katsanevakis and Maravelias, 2008;Tyszko and Pritt, 2017;Porta et al, 2018). Age biases can influence stock assessment by overestimating or underestimating growth or mortality, affecting policy decisions about a population (Beamish and McFarlane, 1987;Katsanevakis and Maravelias, 2008).…”