Chapter 3: Validation of Annual and Daily Increments in Calcified Structures and Verification of Age Estimates
David L. Buckmeier, Peter C. Sakaris, and Daniel J. Schill
Fisheries scientists need to accurately and precisely estimate fish age, at both annual and daily scales, to understand life history characteristics, quantify dynamic rates, and assess age-specific habitat needs. Estimating fish age is often a critical component of the management process, yet fisheries personnel working with freshwater species rarely receive formal training or use quality assurance–quality control (QA–QC) protocols to ensure that age data are reliable. Fortunately, validated age estimation methods are available for many freshwater fish species in North America (Maceina et al. 2007; Spurgeon et al. 2015). Because the negative consequences of imprecise and biased age estimates can be substantial, we hope that this chapter will (1) encourage further validation of age estimation methods, particularly for nongame species, and (2) facilitate development of QA–QC programs that verify reader accuracy and precision for routine age estimation.
The need for age validation studies was first emphasized by Beamish and McFarlane (1983) and Casselman (1983), although some earlier works attempted to validate methods for specific species (e.g., Holden and Vince 1973). Following these initial efforts, Geffen (1992) and Campana (2001) reviewed common validation methods, noting the benefits and limitations of each technique. Francis (1995) emphasized that validation studies should quantify accuracy, rather than simply declaring whether or not a method is accurate. The cumulative information in these articles and from past age estimation texts and symposia (e.g., Bagenal 1974; Summerfelt and Hall 1987; Secor et al. 1995; Panfili et al. 2002; Green et al. 2009; Quist et al. 2012) provides a basis for accurate age estimation. However, use of a validated age estimation method will never guarantee accuracy because individuals can prepare calcified structures improperly or have different interpretations of observed increments (Campana 2001; Buckmeier 2002). Verification (through QA–QC processes) is also needed to ensure that age estimates meet acceptable levels of accuracy and precision.