Fish size structure analysis via ordination: A visualization aid

Abstract

AbstractObjectiveVisual aids like length‐frequency histograms are widely used to examine fish population status and trends; however, comparing multiple histograms simultaneously becomes cumbersome and inefficient. Complicating matters further, overlaying covariates on histograms to highlight connections with length frequencies can be challenging. An alternative, and the subject of this Perspective, is to display length distributions as an ordination using similarity indexes; in many cases, this allows for improved visual organization and representation of relationships with covariates.MethodsI review the application of ordination methods for analysis of size structures using alternative visualizations that may facilitate the identification of connections that are concealed when analyzing a series of histograms. After a brief introduction to similarity indexes, types of ordinations, and sample sizes, I examine four case studies to illustrate size structure analysis via similarity indices: (1) unconstrained ordination to identify “bass‐crowded” populations in a set of 34 small fishing lakes, (2) unconstrained ordination to evaluate the effect of three consecutive length limits on a Largemouth Bass Micropterus nigricans population over a span of 28 years, (3) constrained ordination to assess the relationships between fish community size structure and in‐lake and off‐lake environmental descriptors in 30 oxbow lakes, and (4) constrained ordination to identify what aspects of Largemouth Bass size structure were related to six types of reservoir habitats.ResultSize structure analysis via similarity indexes enabled the exploration of extensive length‐frequency data. It is important to acknowledge that ordinations serve solely as a visual aid for assessing size structure—no statistical testing is involved.ConclusionOrdination techniques and software are advancing at a quick pace, holding great promise for the future of size structure analysis via similarity indices.