우리나라 연근해 고등어 어획량의 공간의존성과 영향요인 분석 : 격자 기반 공간패널모형을 활용하여

다국어 초록 (Multilingual Abstract)

This study aims to identify the spatial distribution patterns and influencing factors of chub mackerel(Scomber japonicus) catch across the coastal and offshore waters of South Korea. The analysis incorporates the entire Korean fishing grounds, using grid-level panel data on mackerel catch volumes, environmental variables (chlorophyll-a, dissolved oxygen, salinity, and sea surface temperature), and fishing effort variables (gross tonnage, number of vessels, and fishing days). First, a spatial autocorrelation analysis was conducted to examine spatial clustering and migratory patterns of mackerel catch. Eight spatial weight matrices were constructed based on adjacency and inverse distance, and Global Moran’s I tests confirmed significant positive spatial autocorrelation across all quarters. The Moran’s I results showed that spatial dependence was weakest in the second quarter and strongest in the first and fourth quarters, consistent with Korea’s seasonal mackerel fishing patterns. In addition, the Local Moran’s I analysis revealed dynamic seasonal clustering patterns on the map, which were consistent with the migratory routes of mackerel for spawning and overwintering. Second, the spatial panel analysis revealed that both environmental and fishing effort factors had significant effects on mackerel catch volumes. Among the fishing effort variables, gross tonnage, number of vessels, and fishing days showed high multicollinearity; therefore, a principal component(PC1) summarizing these variables was derived and used in the model. An increase in dissolved oxygen and sea surface temperature was associated with a decrease in catches, whereas salinity exhibited a strong positive effect. Chlorophyll-a was not statistically significant. These findings are consistent with previous studies linking mackerel abundance to marine environmental variability, providing a quantitative basis for predicting catch fluctuations and identifying productive fishing grounds. Rising sea surface temperatures, in particular, may reduce nearshore productivity, underscoring the need for adaptive resource management under climate change. Fishing effort had a strong positive effect, and its indirect impact exceeded its direct effect, reflecting the large-scale operational characteristics of purse-seine fleets. Such pronounced spillover effects suggest that fishing pressure in one area can propagate to adjacent zones. Third, this study applies a two-stage instrumental variable dynamic spatial panel analysis, representing the first application in the fisheries field. To address endogeneity, the model incorporates both the spatial lag of the dependent variable and its one-period autoregressive term, while the instrument set consists of lagged explanatory variables, their spatial lags, and differenced terms as instruments. The temporal autoregressive coefficient(–0.17) indicated that catch levels did not persist in the same grid over time, reflecting the migratory ecology of mackerel. The spatial lag coefficient(0.96) revealed strong interdependence of catches across neighboring grids. Among explanatory variables, sea surface temperature(–0.28) and PC1(0.48) showed consistent effects with those from the spatial panel models, both exhibiting stronger short-run than long-run marginal effects. The findings provide a scientific foundation for Korea’s ongoing fishery resource assessment and management initiatives. Furthermore, as national marine spatial databases continue to advance, this research framework offers a scalable analytical basis for expanding spatial analyses of fishery resources in the future.

목차 (Table of Contents)

• Ⅰ. 서론 1
• 제1절 연구의 배경 및 목적 1
• 1. 연구 배경 1
• 2. 연구 목적 4
• 제2절 연구의 범위 및 방법 6
• Ⅱ. 선행연구 검토 8
• 제1절 선행연구 검토 8
• 1. 고등어의 서식 환경 8
• 2. 공간자기상관 14
• 3. 공간패널분석 18
• 4. 동적공간패널분석 23
• 제2절 선행연구와의 차별성 24
• Ⅲ. 고등어 어업 현황 및 자원 관리 현황 분석 26
• 제1절 고등어의 생태적 특성과 어업 현황 26
• 1. 고등어의 생태적 특성 26
• 2. 고등어의 생산 현황 27
• 제2절 고등어의 자원 관리 현황 32
• 1. TAC 제도 운용 32
• 2. 금어기 설정 및 금지체장 적용 35
• 제3절 수산자원관리 정책과 공간계량학적 연계 36
• 1. 수산자원관리계획 36
• 2. 해양공간기본계획 39
• Ⅳ. 분석 방법 및 추정 모형 41
• 제1절 공간자기상관분석 41
• 1. 공간가중행렬 41
• 2. Moran’s I 분석 43
• 제2절 공간패널분석 47
• 1. 공간패널분석 47
• 2. 동적공간패널분석 52
• Ⅴ. 실증 분석 58
• 제1절 공간자기상관 분석 58
• 1. 자료 분석 58
• 2. 공간가중행렬 64
• 3. 공간자기상관 분석 66
• 제2절 공간패널분석 81
• 1. 다중공선성 분석 81
• 2. 주성분 분석 82
• 3. 하우스만 검정 85
• 4. 공간패널분석 결과 86
• 5. 분석 모형 결정 93
• 6. 한계효과 추정 결과 97
• 제3절 동적공간패널분석 116
• 1. 동적공간패널분석 결과 116
• 2. 단기 및 장기 한계효과 추정 결과 120
• 3. 일반패널 및 공간패널모형의 추정 결과 비교 122
• Ⅵ. 결론 125
• 제1절 연구 및 분석 결과 요약 125
• 1. 연구 및 분석 결과 요약 125
• 2. 정책적 시사점 127
• 제2절 연구의 의의 및 한계점 131
• 1. 연구의 의의 131
• 2. 연구의 한계 및 향후 연구 132