A tank experimental analysis on the precision of size and abundance estimation using echo integration method

• Main Authors: Huang, Hsin-Hung, 黃信泓
• Other Authors: Lu, Hsueh-Jung
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/06812382021116067050

碩士 === 國立臺灣海洋大學 === 環境生物與漁業科學學系 === 102 === Echo integration has been widely used to estimate standing crops of fish resources, however, its reliability is highly constraint by the certainty of target discrimination and scaling factors. Commercial or experimental gear sampling is necessary to support acoustic data interpretation. Even under single species circumstance, the accuracy of acoustic survey may be affected by target strength availability, number of total pings, behavior of fish and etc. In order to understand the reliability of echo integration method in terms of fish length and abundance estimation, we designed this tank experiment. Sixty black seabreams (Acanthopagrus schlegelii) of different body size (8-28 cm) were raised to create 16 combinations of size and schooling density. Then, we tested the error of acoustic estimation under different conditions of total ping number and integration parameters. Major results are summarized below: (1) The conversion equation from TS to TL for black seabreams was established as: TS = 21.84 #westeur024# log (TL) – 68.61 (r = 0.84). Total lengths of fish arranged from 9.5 to 27.3 cm, and TS values ranged from -46.11 to -32.71 dB. (2) The estimated average value of total length of fish were agree with that of actual length of fish. However their frequency distributions were not well consistence with each other, especially in the cases of mixing fish size. The reason of the discrepancy were mainly due to schooling behavior of fish. (3) In spite of total ping number increased substantially, the variation of average value of volume back scattering strength (Sv) were within 1-2 dB. The estimated density of fish also remained stable. The major factor affecting accuracy of estimated density was due to the parameters of TS. (4) The estimated abundance of 16 cases of TL and density combinations indicated that the accuracy of 80 % of them were higher than 75 %. As most of the condition setting in the experiment were tougher than normal, therefore in general cases the echo integration method can maintain its reliability in a certain level.