Imperfect Estimation of Lepeophtheirus salmonis Abundance and Its Impact on Salmon Lice Treatment on Atlantic Salmon Farms
Accurate monitoring of sea lice levels on salmon farms is critical to the efficient management of louse infestation, as decisions around whether and when to apply treatment depend on an estimation of abundance. However, as with all sampling, the estimated abundance of salmon lice through sampling sa...
| Published in: | Frontiers in Marine Science |
|---|---|
| Main Authors: | , , , |
| Format: | Article in Journal/Newspaper |
| Language: | English |
| Published: |
Frontiers Media S.A.
2021
|
| Subjects: | |
| Online Access: | https://doi.org/10.3389/fmars.2021.763206 https://doaj.org/article/ca59edc37e0648d2aa4c85b7090f60a6 |
| Summary: | Accurate monitoring of sea lice levels on salmon farms is critical to the efficient management of louse infestation, as decisions around whether and when to apply treatment depend on an estimation of abundance. However, as with all sampling, the estimated abundance of salmon lice through sampling salmon cannot perfectly represent the abundance on a given farm. While suggestions to improve the accuracy of lice abundance estimates have previously been made, the significance of the accuracy of such estimation has been poorly understood. Understanding the extent of error or bias in sample estimates can facilitate an assessment as to how influential this “imperfect” information will likely be on management decisions, and support methods to mitigate negative outcomes associated with such imperfect estimates. Here, we built a model of a hypothetical Atlantic salmon farm using ordinary differential equations and simulated salmon lice (Lepeophtheirus salmonis) abundance over an entire production cycle, during which salmon were periodically sampled using Monte Carlo approaches that adopted a variety of sample sizes, treatment thresholds, and sampling intervals. The model could thus track two instances of salmon lice abundance: true abundance (based on the underlying model) and monitored abundance (based on the values that could be estimated under different simulated sampling protocols). Treatments, which depend on monitored abundance, could be characterized as early, timely, or late, as a result of over-estimation, appropriate estimation, and under-estimation, respectively. To achieve timely treatment, it is important to delay treatments until true abundance equals some treatment threshold and to execute treatment as soon as this threshold is reached. Adopting larger sample sizes increased the frequency of timely treatments, largely by reducing the incidence of early treatments due to less variance in the monitored abundance. Changes in sampling interval and treatment threshold also influenced the accuracy of abundance ... |
|---|