Data from: Geomorphic, hydroclimatic, and biotic factors influence juvenile emigration timing in a southern coho salmon population
Data files
May 08, 2026 version files 1.45 GB
-
adult_returns.csv
8.33 KB
-
antenna_status.csv
349.11 KB
-
covariate_model_betas_var_cov_matrix.csv
600 B
-
covariates.csv
1.79 KB
-
daily_flow.csv
134.34 KB
-
early_emigration_ms.Rmd
41.11 KB
-
emigration_date.csv
1.59 MB
-
fl_emigrants.csv
838.76 KB
-
individual_FL.csv
3.27 MB
-
MSE_NO_COVARIATES.DBF
1.82 KB
-
MSE_NO_COVARIATES.FPT
11.11 MB
-
MSE_Z_COVARIATES.DBF
4.12 KB
-
MSE_Z_COVARIATES.FPT
1.43 GB
-
psi_estimates.csv
2.10 KB
-
psi_model_results.csv
681 B
-
README.md
8.60 KB
-
release_dates.csv
1.53 KB
Abstract
Environmental variation gives rise to species’ life history diversity, which is essential to population resilience. Here, we characterize diversity in juvenile emigration timing in an endangered coho salmon population complex and examine physical and biotic factors contributing to phenological variation among streams and cohorts. Over 11 years, PIT-tagged coho salmon from common family groups were released from a conservation hatchery into four tributaries of the Russian River, California, USA, and we tracked their emigration timing using PIT detection systems. We estimated the probability that juveniles would emigrate early (< March 3) using a multistate emigration model and found that early emigration probability was higher in streams with less floodplain area and increased with streamflow, salmonid density, stream temperature, and individual size. Both life history strategies—i.e., early emigration and emigration during the typical spring smolt migration window—contributed to adult returns. The life history plasticity observed suggests that the remnant population responds to environmental heterogeneity and that efforts to diversify and reconnect a variety of habitats could stimulate multiple emigration strategies and increase population resilience.
Dataset DOI: 10.5061/dryad.x3ffbg823
Description of the data and file structure
We use passive integrated transponder (PIT) tags and detection systems to track the movement timing of hatchery-released juvenile coho salmon. Data files include PIT detection dates, daily status of PIT antennas (i.e., whether they were operational), size data collected on juveniles before they were released from the hatchery, and environmental covariates used in our analysis.
Files and variables
File: early_emigration_ms.Rmd
Description: R markdown file that includes code for analysis conducted and figures generated for this manuscript.
File: release_dates.csv
Description: Dates and numbers of juvenile hatchery coho salmon released into the study streams.
- subwatershed = subwatershed where the fish were released
- cohort = year class of coho salmon representing the year in which the fish hatched
- release_date = date fish were released from hatchery
- std_week_start = standardized start week date
File: emigration_date.csv
Description: List of PIT-tagged juvenile coho salmon detected emigrating from four tributaries of the Russian River watershed. date_rel = date the fish were released from the hatchery, date = maximum detection date on PIT antennas.
- date = date
- cohort = year class of coho salmon representing the year in which the fish hatched
- date_rel = date the fish was released from the hatchery
- std_date = date with a standardized year across the dataset
- std_week_start = standardized start week date
- wy_day = water year day
- subwatershed = subwatershed where the fish was detected
File: adult_returns.csv
Description: List of PIT tagged coho that returned as adults to the Russian River watershed.
- return_age = age that the fish returned from the ocean
- cohort = year class of coho salmon representing the year in which the fish hatched
- subwatershed = subwatershed the fish was released into
- date = date the fish was detected returning as an adult
- std_date = date with a standardized year across the dataset
- emigrant_group = early or late emigrating fish (as defined in the manuscript)
File: individual_FL.csv
Description: Predicted fork lengths of all PIT-tagged coho salmon released into the study streams, whether they were detected emigrating or not.
- subwatershed = subwatershed the fish was released into
- cohort = year class of coho salmon representing the year in which the fish hatched
- date_rel = date fish was released from hatchery
- pred_FL_mm_rel = predicted fork length of the fish at the time it was released into the stream (mm)
- z_pred_FL_mm_rel = z-scored value of pred_FL_mm_rel
File: fl_emigrants.csv
Description: Predicted fork lengths of individual PIT-tagged coho salmon detected emigrating from tributaries and included in the analysis for this study.
- cohort = year class of coho salmon representing the year in which the fish hatched
- std_date = date with a standardized year across the dataset
- pred_FL_mm_rel = predicted fork length of the fish at the time it was released into the stream (mm)
- psi = probability of emigrating before March 3
- vba = 1 is higher proportion of valley bottom area, and 0 is lower proportion of valley bottom area (see manuscript for definitions)
File: antenna_status.csv
Description: Data showing the status of PIT antennas on each study stream each day. For antenna_status, 1 = fully operational, 0.5 = partially operational, 0 = not operational).
- subwatershed = subwatershed where the antenna was operated
- std_date = date with a standardized year across the dataset
- antenna_status = status of antenna- see definitions in description
- cohort = year class of coho salmon representing the year in which the fish hatched
File: daily_flow.csv
Description: Daily streamflow data collected by USGS on Austin Creek downloaded using dataRetrieval.
- cohort = year class of coho salmon representing the year in which the fish hatched
- date = date
- std_date = date with a standardized year across the dataset
- mean_daily_flow = mean daily flow (ft3/s)
File: covariates.csv
Description: Covariates used in the design matrix of the mark-recapture model. See methods in the manuscript for detailed descriptions of each covariate.
- group_label = an ID representing the subwatershed (first 3 digits of the subwatershed) and the cohort
- vba = categorical variable: high valley bottom area = 1 and low valley bottom area = 0
- z_flow = z-scored value of flow_days_above_med (see description below)
- z_density = z-scored value of fall_rel_density_m (see description below)
- z_temperature = z-scored value of mean_temp (see description below)
- flow_days_above_med = number of days between Oct 1 and Mar 2 that mean daily flow was above the dataset median daily flow
- fall_rel_density_m = number of fish released per meter of stream
- mean_temp = mean daily temperature between Oct 1 and Mar 2 in degrees Celsius
File: MSE_NO_COVARIATES.DBF
Description: Program MARK file that includes five multistate emigration models used to evaluate the influence of cohort (2012 - 2022) and stream (Willow, Dutch Bill, Green Valley, Mill) on early emigration probability of juvenile coho salmon. See the "Characterizing spatiotemporal patterns in juvenile emigration timing" section of the Methods in the manuscript for additional detail.
Program MARK is needed to open this file and can be downloaded for free at https://www.phidot.org/software/mark/downloads/.
File: MSE_NO_COVARIATES.FPT
Description: This file is needed to open MSE_NO_COVARIATES.DBF and must be in the same folder.
File: psi_estimates.csv
Description: Estimates of juvenile coho early emigration probability (psi) of 40 stream/cohort groups. Estimates were generated from the MSE_NO_COVARIATES Program MARK models included in this repository.
- year = year that fish hatched
- group = an ID representing the subwatershed (first 3 digits of the subwatershed) and the cohort
- estimate = estimated probability of early emigration
- lci = 95% lower confidence interval
- uci = 95% upper confidence interval
- subwatershed = subwatershed name
File: MSE_Z_COVARIATES.DBF
Description: Program MARK file containing 13 multistate emigration models used to evaluate the influence of valley bottom area (proportion of watershed area characterized as valley bottom), flow (days above median daily flow), temperature (°C), density (fish/m), and fish size (fork length (mm)) on early emigration probability of juvenile coho salmon. Detailed descriptions of the covariates and model construction and evaluation can be found in the “Evaluating drivers of early emigration probability” section of the Methods in the manuscript.
Program MARK is needed to open this file and can be downloaded for free at https://www.phidot.org/software/mark/downloads/.
File: MSE_Z_COVARIATES.FPT
Description: This file is needed to open MSE_Z_COVARIATES.DBF and must be in the same folder.
File: psi_model_results.csv
Description: Output of best model in MSE_Z_COVARIATES file in this repository.
- model = name of model for cross-referencing in MARK file
- parameter_index = parameter number in mark recapture model
- beta = specifies parameter in the model (intercept, covariate or interaction term)
- coefficient = coefficient value of the beta parameter
- lci_95 = 95% lower confidence interval
- uci_95 = 95% upper confidence interval
File: covariate_model_betas_var_cov_matrix.csv
Description: Variance-covariance matrix exported from MSE_Z_COVARIATES multistate emigration model in Program MARK.
Code/software
Please see the early_emigration_ms.Rmd file for the R packages that are used for the analysis and figures in this manuscript. Program MARK is required to run the multistate emigration models included in this repository. This can be downloaded at https://www.phidot.org/software/mark/downloads/.
Access information
Streamflow data was downloaded from USGS: https://waterdata.usgs.gov. Site number 11467200 (Austin Creek).
