TY - JOUR
T1 - Historical analysis of blue marlin (Makaira nigricans Lacepède, 1802) catches by the pelagic longline fleet in the eastern Pacific Ocean
AU - Marín-Enríquez, Emigdio
AU - Abitia-Cárdenas, Leonardo A.
AU - Moreno-Sánchez, Xchel G.
AU - Ramírez-Pérez, Jorge S.
N1 - Publisher Copyright:
© 2020 CSIRO.
PY - 2020/3
Y1 - 2020/3
N2 - We analysed a historical (1959-2017) database of blue marlin catches reported by the industrial pelagic longline fleet operating in the eastern Pacific Ocean. The time series of catch-per-unit-effort (CPUE) for the time period 1959-2015 was modelled as a function of temporal and environmental variables by using generalised additive mixed models (GAMM). CPUE predictions were made on an independent dataset (2016-2017). Results suggested that a higher nominal CPUE occurred near the equator during boreal winter, spring and autumn, with a shift towards the north during the boreal summer. Two high CPUE events were observed, one in the early 1960s and one in the early 1990s. The final GAMM explained 61% of the total variance of the CPUE time series. Only a small percentage of total deviance was explained by the environmental variables, so we suggest that changes in fleet-targeting practices are the main cause of the large variability in the time series. The final GAMM can be used to predict blue marlin CPUE with a prediction accuracy of 3-8 fish per 1000 hooks. We observed an important decline in CPUE from the early 2000s; the factors responsible for this decline should be further investigated.
AB - We analysed a historical (1959-2017) database of blue marlin catches reported by the industrial pelagic longline fleet operating in the eastern Pacific Ocean. The time series of catch-per-unit-effort (CPUE) for the time period 1959-2015 was modelled as a function of temporal and environmental variables by using generalised additive mixed models (GAMM). CPUE predictions were made on an independent dataset (2016-2017). Results suggested that a higher nominal CPUE occurred near the equator during boreal winter, spring and autumn, with a shift towards the north during the boreal summer. Two high CPUE events were observed, one in the early 1960s and one in the early 1990s. The final GAMM explained 61% of the total variance of the CPUE time series. Only a small percentage of total deviance was explained by the environmental variables, so we suggest that changes in fleet-targeting practices are the main cause of the large variability in the time series. The final GAMM can be used to predict blue marlin CPUE with a prediction accuracy of 3-8 fish per 1000 hooks. We observed an important decline in CPUE from the early 2000s; the factors responsible for this decline should be further investigated.
KW - environmental variability
KW - fisheries oceanography
KW - generalised additive mixed models
UR - http://www.scopus.com/inward/record.url?scp=85072122592&partnerID=8YFLogxK
U2 - 10.1071/MF19088
DO - 10.1071/MF19088
M3 - Artículo
SN - 1323-1650
VL - 71
SP - 532
EP - 541
JO - Marine and Freshwater Research
JF - Marine and Freshwater Research
IS - 4
ER -