Proceedings of the Third World Fisheries Congress: Feeding the World with Fish in the Next Millenium—The Balance between Production and Environment

Analysis for Spatial Heterogeneity of East China Sea Fishery Resources Using a Geographical Information System Based on a Semivarigram

Fenzhen Su, Chenghu Zhou, Quanqin Shao, Yunyan Du, Changqin Yao

doi: https://doi.org/10.47886/9781888569551.ch53

Spatial heterogeneity is an important theory of ecology (Kareiva 1994) and the most interesting problem in the research of the functions and processes of ecological systems with different scales. Describing heterogeneity quantitatively is the key problem in landscape ecology (Wiens 1992) but is necessary before the ecological spatial model can be modeled with landscape and physical processes (Risser et al. 1984; Turner 1987; O’Neil et al. 1988; Turner and Gardner 1991).

Life forms always distribute in spatio-temporal space by individual, population, and community with spatio-temporal heterogeneity; ecologists discovered this phenomenon in the 1940s (Watt 1947). In the 1980s, the main methods for analyzing spatial heterogeneity were classical statistics and multianalysis. However, classical statistics has many limitations for its basic hypothesis when applied to spatio-temporal heterogeneity, and multianalysis ignores heterogeneity (Philips 1985; Li and Wu 1992; Rossi et al. 1992; Datilleal and Leglandre 1993). Geostatistics is an efficient method for analyzing and interpreting spatial heterogeneity quantitatively (Rossi et al. 1992). With a geographical information system (GIS), geostatistics can be used to extract important information from large datasets to provide knowledge or as an efficient method of knowledge discovery. In this paper, we apply geostatistics with GIS to the field of fish ecology. Extracted information can be used to determine fish distribution and in fieldwork to investigate fishery resources.

To show how geostatistics can be used to analyze spatial variation, we selected the fish ecological system in the East China Sea, a marginal sea between the Asian continent and the Pacific Ocean. Its current system is composed of littoral currents and the Kuroshio Current system, so its water is a combination of diluted littoral water and water from the far sea. The diluted littoral water has low salinity, low transparency, and great inter-annual variability in temperature. The temperature and salinity of water from the far sea are high, so temperature and salinity gradients in the East China Sea are steep. Biomass in this region is low compared with other regions in the world, only 3.92 metric tons/km² (Zhao 1988). Approximately 11% of production in this region is of pelagic fishes and 40% is of demersal fishes (Zhao 1988). This region is the most important fishing area for China, generating 40% of China’s total fishery production (Liu and Zhan 1999).

Sample data for 1987–1997 were obtained from four of the largest fishing companies in China. The catch of these four companies equals 12% of the total catch by all Chinese companies in the region. Each sea catch is recorded by date, catch production, position, company name, and gear type. We divided the East China Sea into a regular grid (10 × 10 ft for each cell) and calculated the average production per catch in each grid cell as the fish density. Trawl catches were used as a measure of demersal fish density; purse seine catches were used as a measure of the density of pelagic fishes. We mapped the fishing locations and found that the catch locations occupy all of the area, so although the locations were not randomly sampled, the catch data are reliable when used to describe the distribution of fish density.

There is a 54% probability that two companies were fishing in the same cell on the same day, with an overlap of more than 80% if the low-production records are deleted. Furthermore, the results change only slightly if the records of any one company are deleted, which demonstrates the robustness of the sample data set.