Aggregative patterns of pre-spawning Atlantic herring on Georges Bank from 1999-2010

¹ NOAA Fisheries, Northeast Fisheries Science Center, 166 Water St., Woods Hole, 02543 MA, USA
² Integrated Statistics, 166 Water St., Woods Hole, 02543 MA, USA
³ Current Address: 2 South Ascot Ct., Newtown 18940 PA USA

^a Corresponding author: Michael.Jech@noaa.gov

Received: 17 September 2011
Accepted: 31 January 2012

Abstract

Atlantic herring (Clupea harengus) in the offshore regions of the Gulf of Maine migrate each fall from their feeding grounds to the northern portion of Georges Bank to spawn. The Northeast Fisheries Science Center’s (NEFSC) herring acoustic survey has taken advantage of this behavior by conducting systematic surveys of the pre-spawning fish each year since 1999. Multi-frequency acoustic and midwater trawl data were collected along transects oriented perpendicular to bathymetric contours. Acoustic backscatter was analyzed to describe the aggregative patterns (e.g., size, location in the water column, and spatial and temporal distribution) of Atlantic herring during these surveys and regression trees were used to examine the aggregation characteristics. The positional variables of distance to spawning grounds and vertical location in the water column were the primary characteristics for describing pre-spawning aggregations. Secondary to these were the temporal variables of diel and survey timing, and the morphological characteristic of aggregation area. Lower numbers of aggregations were observed close to the herring spawning grounds but with higher acoustic energy than larger numbers of aggregations observed further from the spawning grounds but smaller in size and lower in acoustic energy. Most aggregations were in the lower portion of the water column, but those that were in the upper portion of the water column had higher acoustic energy. Consistently throughout the decade, 90% or more of herring aggregations were located within 40 nautical miles of their spawning grounds. The regression tree method provided valuable insight to the data series where it highlighted spatial and temporal patterns and was an effective way to quantitatively summarize relationships.