Modelling capture probability of Atlantic salmon (Salmo salar) from a diverse national electrofishing dataset: Implications for the estimation of abundance

Millar, Colin P.; Fryer, R. J.; Millidine, K. J.; Malcolm, I. A.

doi:10.1016/j.fishres.2016.01.001

Cited by 26 publications

(30 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It was a continuous variable, calculated separately for fry and parr for the first pass and the combined second and third passes (given smaller sample sizes). Models were fitted by maximum likelihood using the ef package (Millar et al ., ) in R (http://www.r-project.org).…”

Section: Methodsmentioning

confidence: 98%

“…The model selection criterion was the BIC adjusted to prevent overfitting due to overdispersion (Millar et al ., ), given by:

{BIC}_{adj} = ((), - 2 L / 1.23) + K \log N

where L is the log‐likelihood, 1.23 is the estimate of overdispersion in the Girnock electrofishing data, N is the number of site‐visits and K is the number of variables in the model. The effects of the covariates on capture probability were illustrated using partial effects plots where the covariate of interest was allowed to vary whilst the others were held constant at their reference (categorical) or median (continuous) level.…”

Section: Methodsmentioning

confidence: 98%

“…Capture probability ( p ) of fry and parr was modelled using the multi‐pass electrofishing data for the five index sites following the methods described by Millar et al . () where full details can be found. In brief, p was modelled as a logistic function of covariates (Table ) with continuous variables included as linear terms or smooth terms with fixed degrees of freedom.…”

Section: Methodsmentioning

confidence: 99%

“…Electrofishing is one of the most commonly used methods for sampling fish in rivers (Dunham et al ., ). Multi‐pass electrofishing data can provide estimates of absolute abundance that have negligible bias and are relatively precise (Millar et al ., ; Rosenberger & Dunham, ), but these data are time‐consuming and costly to obtain which can result in poor spatial coverage (Foley et al ., ; Simonson & Lyons, ). There is, therefore, a desire from some fisheries managers to replace multi‐pass electrofishing with cheaper and more rapid single‐pass or timed (catch per unit effort; i.e ., the number of fish captured per unit time independent of area fished) electrofishing (Bateman et al ., ; Bertrand et al ., ; Hanks et al ., ; Matson et al ., ; Millar et al ., ).…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

These are not the trends you are looking for: poorly calibrated single‐pass electrofishing data can bias estimates of trends in fish abundance

Glover

Fryer

Soulsby

et al. 2019

Journal of Fish Biology

Self Cite

View full text Add to dashboard Cite

There is increasing interest in the potential of single‐pass and timed electrofishing to assess status and trends in fish populations. However, where capture probability varies over time, there is a risk that uncalibrated electrofishing data could fail to detect, or provide biased estimates of trends. This study analysed a long‐term electrofishing dataset collected over 50 years in an intensively studied catchment where egg deposition and emigrant production declined by c. 82% and 35% over the same time. The electrofishing data were used to illustrate the effects of changing capture probability on estimated trends in juvenile Atlantic salmon Salmo salar abundance. Temporal variability in capture probability was modelled. Trends in abundance were then estimated from uncalibrated single‐pass electrofishing count data and compared with estimates from data calibrated for capture probability. The calibrated data revealed significant declines in S. salar fry (age 0) and parr (age ≥ 1) abundance. However, the trend estimates from the uncalibrated data were positively biased and not significant. Exploration of alternative (realistic) scenarios with different trends in true abundance and capture probability suggests that uncalibrated electrofishing data can provide very misleading estimates of trends. The problem is exacerbated in data where capture probability is low. It is recommended that single‐pass and timed electrofishing methods should not be used to assess trends in fish populations without regular (annual) calibration.

show abstract

Section: Methodsmentioning

confidence: 98%

“…The model selection criterion was the BIC adjusted to prevent overfitting due to overdispersion (Millar et al ., ), given by:

{BIC}_{adj} = ((), - 2 L / 1.23) + K \log N

Section: Methodsmentioning

confidence: 98%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

These are not the trends you are looking for: poorly calibrated single‐pass electrofishing data can bias estimates of trends in fish abundance

Glover

Fryer

Soulsby

et al. 2019

Journal of Fish Biology

Self Cite

View full text Add to dashboard Cite

show abstract

“…One approach could be to allow the intercept to vary smoothly between catchments using a Gaussian Markov random field (Cressie, 1993), so the intercept in unmonitored catchments could be estimated from nearby monitored catchments. This approach has been developed in other contexts (Millar et al, 2015(Millar et al, , 2016 and offers promise in the context of large-scale T w modelling.…”

Section: Extending Predictionsmentioning

confidence: 99%

Can spatial statistical river temperature models be transferred between catchments?

Jackson

Fryer

Hannah

et al. 2017

Hydrol. Earth Syst. Sci.

Self Cite

View full text Add to dashboard Cite

Abstract. There has been increasing use of spatial statistical models to understand and predict river temperature (T w ) from landscape covariates. However, it is not financially or logistically feasible to monitor all rivers and the transferability of such models has not been explored. This paper uses T w data from four river catchments collected in August 2015 to assess how well spatial regression models predict the maximum 7-day rolling mean of daily maximum T w (T w max ) within and between catchments. Models were fitted for each catchment separately using (1) landscape covariates only (LS models) and (2) landscape covariates and an air temperature (T a ) metric (LS_T a models). All the LS models included upstream catchment area and three included a river network smoother (RNS) that accounted for unexplained spatial structure. The LS models transferred reasonably to other catchments, at least when predicting relative levels of T w max . However, the predictions were biased when mean T w max differed between catchments. The RNS was needed to characterise and predict finer-scale spatially correlated variation. Because the RNS was unique to each catchment and thus non-transferable, predictions were better within catchments than between catchments. A single model fitted to all catchments found no interactions between the landscape covariates and catchment, suggesting that the landscape relationships were transferable. The LS_T a models transferred less well, with particularly poor performance when the relationship with the T a metric was physically implausible or required extrapolation outside the range of the data. A single model fitted to all catchments found catchment-specific relationships between T w max and the T a metric, indicating that the T a metric was not transferable. These findings improve our understanding of the transferability of spatial statistical river temperature models and provide a foundation for developing new approaches for predicting T w at unmonitored locations across multiple catchments and larger spatial scales.Copyright statement. The works published in this journal are distributed under the Creative Commons Attribution 3.0 License. This license does not affect the Crown copyright work, which is re-usable under the Open Government Licence (OGL). The Creative Commons Attribution 3.0 License and the OGL are interoperable and do not conflict with, reduce or limit each other.

show abstract

The essential value of long‐term experimental data for hydrology and water management

Tetzlaff

Carey

McNamara

et al. 2017

Water Resources Research

124

127

View full text Add to dashboard Cite

Observations and data from long‐term experimental watersheds are the foundation of hydrology as a geoscience. They allow us to benchmark process understanding, observe trends and natural cycles, and are prerequisites for testing predictive models. Long‐term experimental watersheds also are places where new measurement technologies are developed. These studies offer a crucial evidence base for understanding and managing the provision of clean water supplies, predicting and mitigating the effects of floods, and protecting ecosystem services provided by rivers and wetlands. They also show how to manage land and water in an integrated, sustainable way that reduces environmental and economic costs.

show abstract

Modelling capture probability of Atlantic salmon (Salmo salar) from a diverse national electrofishing dataset: Implications for the estimation of abundance

Cited by 26 publications

References 37 publications

These are not the trends you are looking for: poorly calibrated single‐pass electrofishing data can bias estimates of trends in fish abundance

These are not the trends you are looking for: poorly calibrated single‐pass electrofishing data can bias estimates of trends in fish abundance

Can spatial statistical river temperature models be transferred between catchments?

The essential value of long‐term experimental data for hydrology and water management

Contact Info

Product

Resources

About