Quality and status of fish stocks in lakes: gillnetting, seining, trawling and hydroacoustics as sampling methods

All sampling methods give selective or biased estimates of fish species abundance, distribution and size structure. This creates problems, e.g. in regard to the Water Framework Directive of the European Union, which demands evaluation of the quality and status of fish stocks in lakes. We compared fish sampling by means of Nordic multimesh gillnets, seining, trawling and hydroacoustics in two Finnish lakes in summer 2007 and 2008. Sampling methods were used 'as such', i.e. no special design was implemented for method comparison. In the shallow eutrophicated lake the species' composition of gillnet sampling and seining were very different. The biomass-% of percids dropped from gillnet (61%) to seining (9%) and that of cyprinids grew from 39 to 90%, respectively. In the deep pelagic area of the oligotrophic lake, vendace and smelt predominated in trawl catches. The number of fish caught by gillnetting in that area was too small to make any conclusions about the species composition. In the eutrophicated lake, the combined length distribution for all fish species differed significantly between gillnetting and seining. In the oligotrophic lake, the gillnet catches were too small for any comparison of fish size. The difference in the length distribution of fish between trawl and echosounding was significant in most analysed depth layers. In upper depth layers acoustics sampled larger fish than trawling, and in deeper layers smaller fish. Using a combined acoustic-trawl method, the pelagic fish biomass was estimated to be approx. 17 kg ha(-1) in the deep and oligotrophic lake. We conclude that in large and deep-water areas, the use of active gear is enough in fish sampling to evaluate the quality and status of fish stocks. Gillnetting together with seining is an appropriate method to work out the quality and status of fish stocks in shallow and littoral areas of large lakes. Variation in the catch selectivity of fish sampling gear requires a discrete ecological classification for each type of gear.