A FINITE-STATE CONTINUOUS-TIME APPROACH TO INFERRING REGIONAL MOVEMENT RATES OF ATLANTIC BLUEFIN TUNA USING CONVENTIONAL TAGGING STUDIES
Atlantic bluefin tuna (Thunnus thynnus) are known to move throughout the North Atlantic Ocean and are targeted by various commercial fishing operations in different regions of their range. Due to the large distances individual fish can travel and potential interaction with different fishing fleets a...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Text |
| Language: | English |
| Subjects: | |
| Online Access: | http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.582.4451 http://www.iccat.int/Documents/CVSP/CV060_2007/no_4/CV060041109.pdf |
| Summary: | Atlantic bluefin tuna (Thunnus thynnus) are known to move throughout the North Atlantic Ocean and are targeted by various commercial fishing operations in different regions of their range. Due to the large distances individual fish can travel and potential interaction with different fishing fleets around the North Atlantic, quantitative analysis of the migratory behavior of Atlantic bluefin tuna is important for sound assessment of population size and productivity. The movement of individual fish among regions can be viewed as a finite-state continuous-time stochastic process. Under certain assumptions, instantaneous migration rates between regions and time-specific probabilities of being alive or captured in each region can be inferred from regional locations and times at release and capture available from conventional tagging studies. The estimated probabilities and migration rates can inform stock assessment models on the relative proportions of the population in each region at a given time. This document presents a likelihood approach to estimating these migration and mortality rates and results for maximum likelihood estimates provided by data simulated under ranges of hypothetical migration and mortality rates for adult Atlantic bluefin tuna among two regions and various tagging study designs. Results indicate that bias and precision of maximum likelihood estimates are a function of the migration and mortality rates, number of tags deployed, regions where the tags are deployed and duration of study. RÉSUMÉ |
|---|