ISPPA was an international project that involved France, Peru and Chile and would last from 2001 to 2008; it was hoped that it would be of interest to CoML. The objective was to explain how the very large fluctuations in fish catches off the coast of South America were linked to climatic variation and to understand how the multi-species ecology of the Humboldt Current responded to the huge environmental variations associated with ENSO. The approach was to study the behaviour of both individuals and schools and to relate these to stock behaviour and stock characteristics. There were significant technical challenges in making direct observations and also in surveying whole populations, which spanned a vast region subject to rapid environmental fluctuations, and also extended into inaccessible areas.

The solutions to these problems were to use:

1. EUREKA to achieve effectively instantaneous survey coverage over a very large area
2. LIDAR to survey the inaccessible areas
3. multi-beam sonar to evaluate biases by studying fish behaviour, quantifying fish avoidance and reconstructing school dimensions in 3-D

EUREKA entailed a two-day survey by a large fleet of commercial fishing boats, which was equivalent to a 1.5-month survey by a research vessel. The data were not quantitative, however, and there were major acoustic problems associated with calibration and noise. The technical challenge was to find a low-cost, standard scientific echo sounder with simple, automated calibration and also automated data processing, including GPS, SST and other quantities.

Airborne LIDAR allowed large areas to be surveyed quickly at low cost. The instrument was non-intrusive and produced data very similar to those from acoustic instruments. The technical challenges were to match the LIDAR survey to the EUREKA survey and to cross calibrate the two sets of survey results.

Multi-beam sonar (e.g. Reson Seabat 6012), which could quantify and correct for fish avoidance, would allow the dimensions, density and internal structure of fish schools to be reconstructed in three-dimensions. Its shortcomings were short range (~100 m), high frequency and restricted (90°) coverage, plus background noise (side lobes), a high volume of data, prototype software and a restricted survey speed. Technical improvements were needed to overcome these limitations.

In discussion François Gerlotto and David Farmer agreed that calibration was absolutely vital to the success of the project. Otherwise, it was agreed that the project was an exciting one, given that – as Olav Rune Godø pointed out – the inability to make ‘snap-shot’ surveys was a key constraint in attempting to understand ecosystem dynamics. The help of the international community was needed to stimulate the necessary technical developments and encourage the participation of other regional laboratories.

top of page