Otolith ageing quality control
Learn how to avoid otolith ageing errors and incorrect conclusions.
On this page
Age bias plots showing no bias (top panel) and severe underageing after Age 4 (bottom panel).
Beyond age validation
Age validation validates ageing methodology rather than the ageing accuracy of an individual age reader. We most often apply it in situations such as demonstrating that otolith sections along an axis parallel to the sulcus produce accurate ages.
Quality control is normally equated with age validation, which is often difficult and expensive to undertake. However, validation is only 1 of the 3 components of quality control, and in some cases, is the smallest source of error.
All ageing studies which involve more than 1 set of ages, whether at the daily or yearly level, should incorporate a complete quality control program. Components of the program should include:
- age validation to demonstrate that the age based on counts of periodic growth increments is, on average, equal to the true age of the fish
- tests for bias and long-term drift to demonstrate that the age reader interprets the growth increments in the same way (on average) as other age readers and at other times
- measures of precision to show repeatability among age readers or within the same age reader on different occasions
Tests for bias and long-term drift
Validated or not, different age readers can easily interpret a given otolith in different ways. There’s a bias if the difference is consistent. This is seen when a reader is higher or lower than the other for 1 or more age groups, at least on average.
A bias may also occur within a reader over a period of time. This occurs if a given age reader interprets an otolith differently now than was the case a few years ago. Long-term drift such as this isn’t unusual, and can be both dangerous and difficult to detect.
You can’t detect such a bias, particularly if it occurs only in old fish, with standard measures of precision such as:
- percent agreement
- coefficient of variation
- average percentage error
Nor can you detect long-term drift by replicating readings of a sample taken from the current year.
An age bias plot is the best way to detect bias, and should be a standard component of any ageing program.
Measures of precision
Measures of precision are meaningless if bias is present. However, if bias is absent, the coefficient of variation and average percent error are both useful measures.
In the past, we widely used percent agreement. However, it’s no longer used by most laboratories as it’s very sensitive to the age range in the sample. Different age readers will always have higher percent agreement on a sample of young fish than on a sample of older fish.
By contrast, both coefficient of variation and average percent error are relatively insensitive to the age range. Many laboratories now require their age readers to age a subsample of a reference collection of otoliths for each stock or species on a periodic basis.
Age bias plots and coefficient of variations are based on this comparison to ensure that long-term quality is being maintained. Ideally, you’ll first carry out age validation on the reference collection, although this isn’t always possible.
Related links
- Date modified: