Limits...
Atlantic Bluefin Tuna (Thunnus thynnus) Biometrics and Condition.

Rodriguez-Marin E, Ortiz M, Ortiz de Urbina JM, Quelle P, Walter J, Abid N, Addis P, Alot E, Andrushchenko I, Deguara S, Di Natale A, Gatt M, Golet W, Karakulak S, Kimoto A, Macias D, Saber S, Santos MN, Zarrad R - PLoS ONE (2015)

Bottom Line: The predicted weight by month was estimated as an additional component in the exponent of the weight-length function.We update and improve upon the biometric relationships for bluefin currently used by the International Commission for the Conservation of Atlantic Tunas, by incorporating substantially larger datasets than ever previously compiled, providing complete documentation of sources and employing robust statistical fitting.WLRs and other conversion factors estimated in this study differ from the ones used in previous bluefin stock assessments.

View Article: PubMed Central - PubMed

Affiliation: Instituto Español de Oceanografía, C.O. Santander, Santander, Spain.

ABSTRACT
The compiled data for this study represents the first Atlantic and Mediterranean-wide effort to pool all available biometric data for Atlantic bluefin tuna (Thunnus thynnus) with the collaboration of many countries and scientific groups. Biometric relationships were based on an extensive sampling (over 140,000 fish sampled), covering most of the fishing areas for this species in the North Atlantic Ocean and Mediterranean Sea. Sensitivity analyses were carried out to evaluate the representativeness of sampling and explore the most adequate procedure to fit the weight-length relationship (WLR). The selected model for the WLRs by stock included standardized data series (common measurement types) weighted by the inverse variability. There was little difference between annual stock-specific round weight-straight fork length relationships, with an overall difference of 6% in weight. The predicted weight by month was estimated as an additional component in the exponent of the weight-length function. The analyses of monthly variations of fish condition by stock, maturity state and geographic area reflect annual cycles of spawning and feeding behavior. We update and improve upon the biometric relationships for bluefin currently used by the International Commission for the Conservation of Atlantic Tunas, by incorporating substantially larger datasets than ever previously compiled, providing complete documentation of sources and employing robust statistical fitting. WLRs and other conversion factors estimated in this study differ from the ones used in previous bluefin stock assessments.

No MeSH data available.


Related in: MedlinePlus

Coefficient of variation (CV) of weight at size for 5 cm SFL bins.Eastern (left panel) and Western (right panel) Atlantic bluefin tuna (BFT).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4623980&req=5

pone.0141478.g003: Coefficient of variation (CV) of weight at size for 5 cm SFL bins.Eastern (left panel) and Western (right panel) Atlantic bluefin tuna (BFT).

Mentions: Additional analyses were performed to identify robust procedures to fit the RWT-SFL relationship. Traditional allometric scaling laws for relationships between biological parameters like volume and length (e.g., weight and length) are well-known examples of power-law functions. However, the mean is usually well defined for these types of functions when the exponent is greater than 2, but is only true in the case of variance when the exponent exceeds 3 [16]. This poses problems when applying traditional statistics based on variance or standard deviation, such as regression analysis, or when estimating confidence intervals for predicted values [16]. In general it is expected that with allometric scaling, the variance of the dependent variable (e.g., weight) increases continuously along with the independent variable. Analyses of standardized input data showed a trend of variance for weight at size (Fig 3). For small Eastern bluefin (< 60 SFL) CV of weight at size ranged from 15% to 25%; between 60 to 150 cm SFL the CV decreased to between 6 to 12.5%; while larger than 150 cm SFL the overall CV remained about 12.5%. For Western bluefin smaller than 120 cm SFL, the CV of weight at size ranged between 8 to 18%; then the CV increased between 120 and 200 cm SFL, ranging from 20% to 50%; while larger than 200 cm SFL the CV average decreased to about 12.5%. However for fish closer to 300 cm SFL, an extremely large variance of weight at size was observed. Because of these trends, a weighted nonlinear regression was employed to reduce the influence of observations with larger than expected variance of weight at size, by using the inverse of the estimated CV of weight at 5 cm length bins as the weighting factor.


Atlantic Bluefin Tuna (Thunnus thynnus) Biometrics and Condition.

Rodriguez-Marin E, Ortiz M, Ortiz de Urbina JM, Quelle P, Walter J, Abid N, Addis P, Alot E, Andrushchenko I, Deguara S, Di Natale A, Gatt M, Golet W, Karakulak S, Kimoto A, Macias D, Saber S, Santos MN, Zarrad R - PLoS ONE (2015)

Coefficient of variation (CV) of weight at size for 5 cm SFL bins.Eastern (left panel) and Western (right panel) Atlantic bluefin tuna (BFT).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4623980&req=5

pone.0141478.g003: Coefficient of variation (CV) of weight at size for 5 cm SFL bins.Eastern (left panel) and Western (right panel) Atlantic bluefin tuna (BFT).
Mentions: Additional analyses were performed to identify robust procedures to fit the RWT-SFL relationship. Traditional allometric scaling laws for relationships between biological parameters like volume and length (e.g., weight and length) are well-known examples of power-law functions. However, the mean is usually well defined for these types of functions when the exponent is greater than 2, but is only true in the case of variance when the exponent exceeds 3 [16]. This poses problems when applying traditional statistics based on variance or standard deviation, such as regression analysis, or when estimating confidence intervals for predicted values [16]. In general it is expected that with allometric scaling, the variance of the dependent variable (e.g., weight) increases continuously along with the independent variable. Analyses of standardized input data showed a trend of variance for weight at size (Fig 3). For small Eastern bluefin (< 60 SFL) CV of weight at size ranged from 15% to 25%; between 60 to 150 cm SFL the CV decreased to between 6 to 12.5%; while larger than 150 cm SFL the overall CV remained about 12.5%. For Western bluefin smaller than 120 cm SFL, the CV of weight at size ranged between 8 to 18%; then the CV increased between 120 and 200 cm SFL, ranging from 20% to 50%; while larger than 200 cm SFL the CV average decreased to about 12.5%. However for fish closer to 300 cm SFL, an extremely large variance of weight at size was observed. Because of these trends, a weighted nonlinear regression was employed to reduce the influence of observations with larger than expected variance of weight at size, by using the inverse of the estimated CV of weight at 5 cm length bins as the weighting factor.

Bottom Line: The predicted weight by month was estimated as an additional component in the exponent of the weight-length function.We update and improve upon the biometric relationships for bluefin currently used by the International Commission for the Conservation of Atlantic Tunas, by incorporating substantially larger datasets than ever previously compiled, providing complete documentation of sources and employing robust statistical fitting.WLRs and other conversion factors estimated in this study differ from the ones used in previous bluefin stock assessments.

View Article: PubMed Central - PubMed

Affiliation: Instituto Español de Oceanografía, C.O. Santander, Santander, Spain.

ABSTRACT
The compiled data for this study represents the first Atlantic and Mediterranean-wide effort to pool all available biometric data for Atlantic bluefin tuna (Thunnus thynnus) with the collaboration of many countries and scientific groups. Biometric relationships were based on an extensive sampling (over 140,000 fish sampled), covering most of the fishing areas for this species in the North Atlantic Ocean and Mediterranean Sea. Sensitivity analyses were carried out to evaluate the representativeness of sampling and explore the most adequate procedure to fit the weight-length relationship (WLR). The selected model for the WLRs by stock included standardized data series (common measurement types) weighted by the inverse variability. There was little difference between annual stock-specific round weight-straight fork length relationships, with an overall difference of 6% in weight. The predicted weight by month was estimated as an additional component in the exponent of the weight-length function. The analyses of monthly variations of fish condition by stock, maturity state and geographic area reflect annual cycles of spawning and feeding behavior. We update and improve upon the biometric relationships for bluefin currently used by the International Commission for the Conservation of Atlantic Tunas, by incorporating substantially larger datasets than ever previously compiled, providing complete documentation of sources and employing robust statistical fitting. WLRs and other conversion factors estimated in this study differ from the ones used in previous bluefin stock assessments.

No MeSH data available.


Related in: MedlinePlus