Gunnar Thoresson¹, Mart Kangur², Rimantas Repecka³, Toomas Saat⁴ and Maris Vitinsh^5
1National Board of Fisheries, Institute of Coastal Research, Gamla Slipvägen 19, S-740 71 Öregrund, Sweden.
²Estonian Marine Institute, 32 Lai street, EE-0001 Tallinn, Estonia.
³Institute of Ecology, Akademijos 2, LT-2600 Vilnius, Lithuania.
⁴University of Tartu, Institute of Zoology and Hydrobiology, Vanemuise 46, EE-2400 Tartu, Estonia.
⁵Latvian Fisheries Research Institute, 6 Daugagrivas street, LV-1007 Riga, Latvia.

Key words: perch, Baltic, gill-net fishing, monitoring.

Abstract.
Traditional methods used for stock assessment are expensive and make high demands on the quality of input data. Furthermore, the large number of separate stocks in most Baltic  coastal areas and the difficulty in achieving complete catch statistics require alternative methods. Long term standardized fishing with gill-nets can provide catch data for demersal fishes of good quality to be used for comparing relative abundances between different areas and years. If such data could be related to reliable total catch statistics in one area, this ratio could be used to estimate potential yields in other areas by test fishing. Standardized test fishing with gill-nets has been performad on demersal fishes in Sweden, Finland, Estonia, Latvia and Lithuania. In some areas extensive tests have been performed on perch, indicating that mean CPUE can serve as an index for �true� stock size variations, thus enabling spatial and temporal comparisons. The possibility of using this index in resource assessment is discussed