Skip to main content
Dryad

Ontogenetic shifts in growth and habitat use of freshwater Coastal Cutthroat Trout revealed through otolith analysis of multiple elements

Atkinson, Jamieson1

Published Apr 13, 2026 on Dryad. https://doi.org/10.5061/dryad.w9ghx3g42

Data files

Apr 13, 2026 version files 6.09 MB

Click names to download individual files

Abstract

This dataset contains otolith microchemistry transects and derived products for adfluvial Coastal Cutthroat Trout (Oncorhynchus clarkii clarkii) sampled in the Comox Lake–watershed system (Vancouver Island, BC). Raw transects for stream-caught juveniles and lake-caught adults were acquired as elemental concentrations (ppm) and converted in this workflow to molar ratios (element, mmol/mol) for Mg, Mn, Zn, Sr, Ba (Li included where available). Primordia and check positions (e.g., first winter) were compiled from microscope-based measurements for both adults and juveniles, enabling extraction of life-stage-specific windows (e.g., natal/rearing sections) along each transect.

We use multiple analytical tracks.

(1) Natal origin discrimination: After cleaning and unit conversion, we compute OTF-level means over a standardized rearing window and log-transform elements for MANOVA/ANOVA with Tukey HSD across capture sources; letters are overlaid on boxplots to summarize pairwise differences. We then fit linear discriminant functions (LDFA) on juvenile rearing means (Sr, Ba, Mn, Mg; Zn available in the data and plots) and apply the trained model to adult lake fish to estimate natal contributions, reporting jackknife confusion matrices, posterior probabilities, and assignment summaries.

(2) Movement/habitat signal detection: We assess Mn spikes as putative indicators of habitat shifts (e.g., stream→lake entrance) using a hybrid rule (rolling mean + individual mean fallback), summarizing spike count, magnitude, start/end locations, and before/after contrasts within fish.

(3) Zn oscillations as annulus proxies by linking Zn inflection distances to visually measured annulus starts (1–5AL) and fitting per-annulus simple linear models, reporting coefficients, r², model p-values, and slope-≈1 tests.

Together, the dataset and scripts support reproducible conversion, QC, visualization of transects, hypothesis testing across sources, discriminant assignment of adults, and mechanistic exploration of Mn and Zn features as indicators of habitat transitions and age structure.