共查询到1条相似文献,搜索用时 0 毫秒
1.
Generalized linear models (GLM) and generalized additive models (GAM) were used to standardize catch per unit fishing effort
(CPUE) of Ommastrephes bartramii for Chinese squid-jigging fishery in Northwest Pacific Ocean. Three groups of variables were considered in the standardization:
spatial variables (longitude and latitude), temporal variables (year and month) and environmental variables, including sea
surface temperature (SST), sea surface salinity (SSS) and sea level height (SLH). CPUE was treated as the dependent variable
and its error distribution was assumed to be log-normal in each model. The model selections of GLM and GAM were based on the
finite sample-corrected Akaike information criterion (AICC) and pseudo-coefficient (Pcf) combined P-value, respectively. Both GAM and GLM analysis showed that the month was the most important variable affecting CPUE and could
explain 21.3% of variability in CPUE while other variables only explained 8.66%. The interaction of spatial and temporal variables
weakly influenced the CPUE. Moreover, spatio-temporal factors may be more important in influencing the CPUE of this squid
than environmental variables. The standardized and nominal CPUEs were similar and had the same trends in spatio-temporal distribution,
but the standardized CPUE values tended to be smaller than the nominal CPUE. The CPUE tended to have much higher monthly variation
than annual variations and their values increased with month. The CPUE became higher with increasing latitude-high CPUE usually
occurred in 145°E–148°E and 149°E–162°E. The CPUE was higher when SST was 14–21°C and the SLH from −22 cm to −18 cm. In this
study, GAM tended to be more suitable than GLM in analysis of CPUE. 相似文献