MEPS 147:1-9 (1997)  -  doi:10.3354/meps147001

Identifying the effects of oceanographic features and zooplankton on prespawning herring abundance using generalized additive models

Maravelias CD, Reid DG

Spatial distribution patterns of prespawning herring were analyzed in relation to zooplankton biomass, sea-surface temperature and salinity, temperature and depth of the thermocline (fronts) and the temperature difference between surface and bottom water. Data were collected in mid-summer 1995 during the ICES coordinated herring acoustic survey of the ICES Division IVa. We used generalized additive models (GAMs), nonparametric generalization of multiple linear regression, to test the hypothesis that prespawning herring distribution is related to zooplankton availability and the oceanography around the Shetland Islands (UK), with particular reference to inflows from the Slope Current. The results of this study supported this hypothesis. We found that zooplankton biomass and the location of ocean fronts influence the distribution of prespawning herring. Mean herring abundance was consistently highest in areas having a surface salinity of 35.1 ppt and where the zooplankton abundances were higher. Results indicated that herring appeared to prefer the well-mixed waters and transition zones and avoided the stratified and frontal areas. The present results also suggested that prespawning aggregations of herring followed the movements of zooplankton to deeper and cooler waters beneath the thermocline during summer. Waters with specific salinity and temperature properties are attractive to herring due to the process of frontal mixing which enhances primary and secondary production. These waters are ultimately linked with the Slope Current which is responsible for the advection of warm, nutrient-rich, saline water into the North Sea ecosystem.


Herring · Oceanography · Zooplankton · North Sea · Generalized additive models


Full text in pdf format