|Other Abstract||Hydroacoustic method has been applied in marine fisheries and fisheries research for decades and are well established, Which has been used to investigate fish abundances, spatial –temporal distribution and behaviors in the oceans. In the recent years, this technique has been applied in the freshwater and shallow water.
In the present thesis, Hydroacoustic surveys in conjunction with GIS modeling have been applied in estimation of Chinese sturgeon (Acipenser sinensis) and other fish species’ abundance, temporal-spatial distribution, in the Gezhouba Dam Reach of Yangtze River, during the Chinese sturgeon spawning seasons. And acoustic surveys had been conduct by the Simrad EY60 split-beam scientific Echo sounder of 200KHz, in two specified periods of each year: pre-spawning and post-spawning of Chinese sturgeon.
As for the estimation of Chinese sturgeon spawning stock, it is crucial to distinguishing the Chinese sturgeon echo-signal from other signals, such as bubble, debris, sediments, and other fish species in the echogram of the sonar data. Because of limited technique, in the past we differentiated the echo-signal of Chinese sturgeon from that of other targets, based solely on the echo-shape and echo length, which is too subjective.
So the subjective identification methods of species had been improved in this thesis, based on echo-shape, echo length, Target Strength (TS) and tracking of the target in the sonar beam, based on related study and field work experiences. Stepwise discriminant function analysis (Stepwise DFA) was then applied to rectify any inaccurate classification.
Stepwise DFA was applied to training a set of 481 tracked single echo detections, which belonging to four classification groups, possible signal of Chinese sturgeon, other fish species, sediments and noise (back reverberation, bubble, debris, etc). The results show that, in all 25 Chinese sturgeon echo-signals were detected in the spawning ground of Gezhouba Dam, in the last 3 years. And the accuracy of identification of Chinese sturgeon is 90.9%. In Steps DFA analysis, 24 variables, out of 67 variables were applied in discrimination and identification. PCA (principle component analysis) combined with DFA analysis was then used to ensure the significance of the 24 variables and detailed the identification pattern. The results indicated that we can discriminate Chinese sturgeon from other fish species and noise using certain descriptors such as the behaviour variable, echo characteristics and acoustic cross-section characteristics. However identification of Chinese sturgeon from sediments is more difficult and needs a total of 24 variables. This is due to the limited knowledge about the acoustic-scattering properties of the substrate regions.
Based on identified individuals, 18 of Chinese sturgeon individuals were distributed in Gezhouba Dam and the Miaozui region, with a surface area of about 3.4km2. 7 of the individuals were distributed in Miaozui and the Yanshouba Dam region, with a surface area of about 13km2. As for the vertical distribution, Chinese sturgeon tend to living in the depth of 10~20m, with an average depth of 17.5m, and the sturgeon preferred to inhabit near the bottom.
From 2005 to 2007, the spawning stock of Chinese sturgeon is 257ind., 237ind. and 203ind. respectively, No significant difference in annual amounts were observed in the those years. And in the recent three years, the eatimated laying is 3560,000 grains、1196,000 grains and 238,6000 grains respectively. By Comparing with foregone data, the reproduction size decrease sharply in the recent ten years, The drop tendency is extremely remarkable (p<0.01).
While studying the Chinese sturgeon, We also studied the abundance and temporal-spatial distribution of other fish species, in pre-spawning and post-spawning periods of Chinese sturgeon. As for their abundance, in pre-spawning period, the mean numeric density of fish (X±S.E.) is 209.48 97.12ind., 227.34 69.95 ind., 128.56 37.74ind. respectively. And in post-spawning periods, the mean numeric density (X±S.E.) is 372.09 152.41ind., 321.41 124.57ind.and 178.43 37.52ind. respectively. Therefore, in the post-spawning periods, the fish numeric density is higher than that of Pre-spawning period. However, the difference is not significant (P>0.05), according to student’s T-test.
For the fish size, we had detected more than 4,000 individuals in recent three years. More than 60% individuals’ TS are an interval of -60dB and -50dB, less than 40% individuals’ target strength are more than -50dB. So in this area, most of fish individuals are small fish, medium fish is less, large fish and extra-large fish is rare, based on the empirical measures of target strength. And nonparamtric tests result show that the fish size has no significant difference in recent 3 years (P=0.87).
In conjunction with the GIS modeling (Geographic Information System), the fish distribution pattern were also revealed in the thesis. As for the fish vertical distribution, the detected fish are distributed in a depth of 5~20m, and most of them aggregating near the bottom. As for the fish horizontal distribution and variation, the result show that the fish tend to located in some specifical region, such as just downstream the strobe, in the front of the Yichang dockyard site and in the front of the medical factory site, near the Yilin bridge and about the Yanshouba Dam Reach. But the distribution varied in the different period, which might be in relation to behavior of Chinese sturgeon spawning.|