Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 1229-3431(Print)
ISSN : 2287-3341(Online)
Journal of the Korean Society of Marine Environment and Safety Vol.29 No.7 pp.802-811
DOI : https://doi.org/10.7837/kosomes.2023.29.7.802

Prediction of Species Distribution Changes for Key Fish Species in Fishing Activity Protected Areas in Korea

Hyeong Ju Seok*, Chang Hun Lee**, Choul-Hee Hwang***, Young Ryun Kim****, Daesun Kim*****, Moon Suk Lee******
*Senior researcher, Risk Assessment Team, Marine Eco-technology Institute (MEI), Busan 48520, Korea
**Principal researcher, Marine Spatial Research Institute, Marine Eco-technology Institute (MEI), Busan 48520, Korea
***Director, Business Headquarters, Marine Eco-technology Institute (MEI), Busan 48520, Korea
****Director, Research and Development Headquarters, Marine Eco-technology Institute (MEI), Busan 48520, Korea
*****Senior Research Scientist, Ocean Law and Policy Institute, Korea Institute of Ocean Science and Technology(KIOST), Busan 49111, Korea
*****Principal Research Scientist, Ocean Policy Research Center, Korea Institute of Ocean Science and Technology(KIOST), Busan 49111, Korea

* First Author : hjseok@marine-eco.co.kr, 051-611-4275


Corresponding Author : chlee@marine-eco.co.kr, 051-611-5424
November 27, 2023 December 14, 2023 December 29, 2023

Abstract


Marine spatial planning (MSP) is a crucial element for rational allocation and sustainable use of marine areas. Particularly, Fishing Activity Protected Areas constitute essential zones accounting for 45.6% designated for sustainable fishing activities. However, the current assessment of these zones does not adequately consider future demands and potential values, necessitating appropriate evaluation methods and predictive tools for long-term planning. In this study, we selected key fish species (Scomber japonicus, Trichiurus lepturus, Engraulis japonicus, and Larimichthys polyactis) within the Fishing Activity Protected Area to predict their distribution and compare it with the current designated zones for evaluating the ability of the prediction tool. Employing the Intergovernmental Panel on Climate Change (IPCC) 6th Assessment Report scenarios (SSP1-2.6 and SSP5-8.5), we used species distribution models (such as MaxEnt) to assess the movement and distribution changes of these species owing to future variations. The results indicated a 30-50% increase in the distribution area of S. japonicus, T. lepturus, and L. polyactis, whereas the distribution area of E. japonicus decreased by approximately 6-11%. Based on these results, a species richness map for the four key species was created. Within the marine spatial planning boundaries, the overlap between areas rated “high” in species richness and the Fishing Activity Protected Area was approximately 15%, increasing to 21% under the RCP 2.6 scenario and 34% under the RCP 8.5 scenario. These findings can serve as scientific evidence for future evaluations of use zones or changes in reserve areas. The current and predicted distributions of species owing to climate change can address the limitations of current use zone evaluations and contribute to the development of plans for sustainable and beneficial use of marine resources.



국내 어업활동보호구역 주요 어종의 종분포 변화 예측

석형주*, 이창헌**, 황철희***, 김영윤****, 김대선*****, 이문숙******
*해양생태기술연구소 위해성평가실 선임연구원
**해양생태기술연구소 해양공간연구소 책임연구원
***해양생태기술연구소 사업 총괄 본부 본부장
****해양생태기술연구소 R&D 총괄 본부 본부장
*****한국해양과학기술원 해양법·정책연구소 선임연구원
******한국해양과학기술원 해양법·정책연구소 책임연구원

초록


해양공간계획(Marine spatial planning)은 해양의 합리적인 이용과 지속 가능한 해양 공간 활용을 위한 중요한 요소이다. 특히 어업활동 보호구역은 지속 가능한 어업을 위한 핵심 용도구역으로, 해양공간계획 경계 내에서 약 45.6%를 차지한다. 그러나, 현재 어업활동보호구역의 지정 과 평가는 미래 수요와 잠재적 가치를 충분히 반영하지 못하고 있으며, 중장기 계획 수립을 위한 보다 합리적인 평가 방법과 예측 도구가 필요한 상황이다. 이러한 문제를 해결하기 위해, 본 연구는 어업활동보호구역 내 주요 어종인 고등어, 갈치, 멸치, 참조기를 대상으로 어종 분포 예측을 시도하고, 현재 용도구역과의 비교를 통해 예측 도구의 가능성을 평가하였다. 한편, IPCC 6차 기후변화 시나리오(SSP1-2.6 및 SSP5-8.5)를 적용한 종분포 모델(MaxEnt)을 사용하여 미래 기후변화에 따른 어종의 이동 및 분포 변화를 분석한 결과, 고등어, 갈치, 참조기의 분포 면적은 현재보다 약 28~86% 증가했으나, 멸치의 분포 면적은 약 6~11% 감소했다. 이 결과를 바탕으로 주요 4종의 종풍부도 지도를 작성하였으며, 해양공간계획 경 계 내에서 ‘높음’으로 평가된 종풍부도 해역과 어업활동보호구역이 중복되는 비율은 약 15%, SSP1-2.6 시나리오에서 21%, SSP5-8.5 시나리오에서 34%로 증가하였다. 연구 결과는 향후 용도구역 평가나 유보구역 변경 시 과학적 근거로 활용될 수 있으며, 어종의 현재 종분포와 기후변화에 따 른 분포 예측을 통해 현재 용도구역 평가의 한계를 보완하고, 지속 가능한 유용 해양 자원의 이용을 위한 계획 수립에 기여할 것으로 판단된다.



    1. 서 론

    국내 해양공간계획(Marine Spatial Planning)은 해양의 합리 적인 용도 배분과 지속 가능한 해양 공간 이용을 위해 9개 의 용도구역으로 나누어 관리하고 있다. 이 중 어업활동보 호구역은 어업 활동의 자유를 보장하고 지속 가능한 어업을 위해 지정되었으며, 전체 용도구역 중 약 45.6%를 차지해 그 중요성이 높다(Han et al., 2021). 어업활동보호구역의 평가에 는 주요 어종의 어획량, 조업 밀집도, 면허어장 및 보호구역 의 유무 등이 주요 지표로 사용된다. 하지만 현재의 평가 방 식은 미래 수요와 잠재적 가치를 충분히 반영하지 못하고 있어, 중장기 계획을 위한 새로운 평가 방법과 예측 도구가 필요한 상태이다.

    최신 IPCC 6차 보고서에 따르면, 기후변화 시나리오에 따 라 2100년까지 세계 평균 해양 표층 수온(Surface Seawater Temperature, SST)이 SSP1-2.6 시나리오에서는 0.86℃, SSP5-8.5 시나리오에서는 최대 2.89℃까지 상승할 것으로 예상된다 (IPCC, 2022). 해양 표층 수온의 상승은 어류의 성장, 생존 및 공간 분포에 직·간접적으로 영향을 미칠 수 있으며, 이에 적 응하지 못한 어종은 사멸할 위험이 있거나 회피하여 이동할 것으로 보인다. 어류 분포의 변화는 기후변화에 대한 초기 반응으로 간주되며(Walther et al., 2002), 이미 북대서양에서 어종의 절반 이상이 기후변화로 인해 서식지를 북쪽으로 이 동한 사례가 보고되었다(Rose et al., 2000). 따라서 해양 표층 수온 상승과 같은 환경 변화는 해양공간계획과 어업활동보 호구역의 관리에 중요한 요소로 고려되어야 할 것이다.

    어업활동보호구역의 주요 대상 어종인 고등어, 멸치, 갈 치, 참조기는 환경 변화에 민감하며 국내에서 상업적으로 중요한 어종이다(Cury and Roy, 1989;Jung et al., 2013;Lee et al., 2009). 국내 어류 서식지에 영향을 미치는 주요 요소 중 하나로 수온이 알려져 있으며(Kim et al., 2010;Park et al., 2004;Hwang et al., 2007), Chlorophyll-a 농도와 같은 먹이생물 의 요인도 어류 분포 및 서식지와 상관성이 있는 것으로 나 타난 연구들이 있다(Ko et al., 2010). 기후 변화 시나리오를 바탕으로 한 종분포 예측 연구는 국내·외에서 활발히 진행 되고 있지만(Elith et al., 2006;Kwon et al., 2012), 특히 국내 주요 해양 어종의 서식지와 분포 예측에 관한 연구는 아직 부족한 실정이다(Bang et al., 2022;Park et al., 2020).

    본 연구는 통계적 모형을 활용하여 종의 서식 특성과 공 간적 잠재 서식 분포를 예측하는 방법에 초점을 맞추고 있 다. 이러한 모형은 다양한 환경 요인들의 영향을 고려하여 종의 서식 가능성을 통계적으로 분석하는 데 유리하다는 점 에서 중요하다(Bae and Jung, 2015). 기존 연구에서 주로 다루 는 모형은 로지스틱 회귀모형, 일반화 가법모형, 그리고 MaxEnt 모형(Maximum Entropy Model, Maxent) 등이 있다 (Guisan and Thuiller, 2005). 특히, MaxEnt 모델은 환경변수와 출현 데이터에 기반하여 종의 출현 확률 분포를 예측하는 기계학습 모델로, 주로 출현 데이터를 활용하여 변수 간의 복잡한 관계를 효과적으로 분석할 수 있다는 점에서 주목할 만하다(Phillips, 2017;Bang et al., 2022).

    따라서, 본 연구에서는 어업활동보호구역 내 주요 어종을 대상으로 하여 환경요인 간의 관계를 분석하고 종분포 모델 (MaxEnt)을 적용하고자 한다. 이를 통해 현재의 어종 분포와 기후변화가 주요 어종의 분포에 미치는 영향을 예측하며, 이러한 예측 결과는 어업활동보호구역의 중장기적 계획 수 립 및 용도구역 평가에 중요한 기초 자료로 활용될 수 있을 것으로 판단된다.

    2. 재료 및 방법

    2.1 주요종 출현 자료

    본 연구에서는 어업활동보호구역 내 주요 어종을 분석 대상으로 선정하였다. 연구에 사용된 어종은 표층 자료를 기반으로 선정하여, 저어류는 이번 연구에서 제외하였다. 산 업적 가치가 높고, 기후변화로 인한 서식지 변화가 예상되 는 어종을 우선적으로 고려하여, 최종적으로 고등어(Scomber japonicus), 갈치(Trichiurus lepturus), 멸치(Engraulis japonicus), 참조기(Larimichthys polyactis) 등 총 4종을 연구 대상으로 선 정하였다.

    4종의 출현 데이터는 2020년부터 2022년까지 수협중앙회 가 제공하는 연근해 어업의 월별 어획량 자료를 활용하여 수집하였다. 수집된 자료는 계절별로 분류하였으며, 동계는 12월부터 2월까지, 춘계는 3월부터 5월까지, 하계는 6월부터 8월까지, 그리고 추계는 9월부터 11월까지의 기간을 각각 포 함한다. 또한, 매년 실제로 어획된 지점들만을 출현 격자로 활용하여 데이터의 정확성을 높였다.

    2.2 환경변수

    주요종의 서식지 분포에 영향을 미칠 가능성이 있는 환경 변수로서 Sea Surface Temperature(이하 SST), Chlorophyll-a(이하 Chl-a)를 선정했으며, 연구의 공간적 범위가 한반도 전 해역 을 포함할 정도로 넓기 때문에 위성영상을 활용했다. 위성영 상은 자료의 이용과 취득의 편의성을 고려해 NOAA(National Oceanic and Atmospheric Administration)에서 제공하는 Aqua MODIS의 자료 중 대기보정이 된 Monthly Composite 영상을 활 용했다. 위성영상의 구축 기간은 주요종 출연 기간에 맞춰 2020년부터 2022년까지이며, 3년간 데이터를 계절별로 구분해 최대(max), 최소(min), 및 평균(mean) 자료로 구축했다(Table 1).

    미래의 기후변화에 따른 표층 해수온 변화 자료는 기상청 에서 제공한 자료를 사용하였으며, 2041년부터 2060년까지의 기간에 대한 IPCC(Intergovernmental Panel on Climate Change)의 SSP1-2.6(저탄소 시나리오) 및 SSP5-8.5(고탄소 시나리오) 시 나리오를 포함한다. 이 자료는 제6차 결합모델 상호비교 프 로젝트(Coupled Model Intercomparison Project Phase 6, CMIP6) 의 앙상블 예측 결과에 통계적 상세화 기법을 적용하여 구 축되었다(KMA, 2022)(Table 2).

    2.3 종분포 모형 최적화

    어업활동보호구역 주요종의 종분포를 예측하기 위해 MaxEnt 3.4.1k(https://biodiversityinformatics.amnh.org/open_source/maxent/) 모형을 사용하였다. 주요종의 분포 확률 지도를 구축하고, 10분위 threshold 값(10th percentile training presence) 을 적용하 여 출현 및 비출현 값을 갖는 지도를 구축하였다.

    또한, 주요종의 종풍부도(Species Richness, SR) 지도를 구축 하였으며, 3개의 등급(낮음(0~5), 중간(6~11), 높음(12~16))으 로 구분하고 어업활동보호구역과 비교·분석 하였다(Bae and Jung, 2015).

    SR = Σ Si (Si = 개별 어종이 출현하는 격자 )

    MaxEnt는 서식지 분포 확률을 추정하기 위해 출현 자료와 환경변수 공변량 자료를 활용하여 최대 엔트로피의 확률 분 포로 나타낸다. 모형의 예측 정확도는 ROC(Receiver Operating Characteristic) 곡선의 AUC(Area Under Cover) 값을 통해 평가 하였다. ROC 곡선은 Y축(참양성률)과 X축(위양성률)의 면적 에 의해 생성된다. AUC는 모델 성능의 척도로 활용되며, 5개 의 등급(0.5-0.6(failing), 0.6-0.7(bad), 0.7-0.8(reasonable), 0.8-0.9 (good), 0.9-1.0(great))으로 평가하였다. 변수가 모델에 미치는 중요도와 기여도는 Percent Contribution와 Permutation Importance 를 통해 확인하였다(Swets, 1988).

    주요종의 출현 확률을 training set(75%)와 test set(25%)로 나 누어 무작위적 추출 과정을 10회 반복하여 평균값을 제시하 였으며(Merow et al., 2013), ‘Replicated run type’은 ‘Crossvalidate’ 로 선택하였고, 본 연구에서 적용한 최종 배경 지점이 약 10,000개로 모형의 기본 및 권장 값(Max number of background points)을 10,000개로 설정하였다(Barbet-Massin et al., 2012).

    MaxEnt에는 특성 조합(FC; Feature Class)과 정규화 승수 (RM; Regularization Multiplier)라는 두 가지 조절 가능한 매 개변수가 존재한다. FC는 L(Linear), Q(Quadratic), P(Product), T(Threshold), H(Hinge), C(Categorical)의 6가지의 조합으로 모 델 내에서 복잡한 관계를 모델링하기 위한 수학적 변환을 나 타낸다(Morales-Castilla et al., 2017). RM은 복잡성과 과적합을 방지하기 위해 사용되며, 기본값은 1이다(Phillips and Dudı́k, 2008). 정규화 승수가 낮을수록 모델은 존재 지점에 더 가깝 게 맞추려는 경향을 나타내며, 높을수록 광범위한 예측을 생성한다(O'Banion and Olsen, 2014). 따라서, 본 연구는 최적 의 모델을 선택하기 위해 6개의 FC 조합(L, LQ, H, LQH, LQHP, 및 LQHPT)과 10개의 RM(0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5 및 5.0)을 적용하여 60개의 모형 중 AICc값이 가장 낮 은 모형을 선택하였다. 매개변수의 최적화는 ENMeval package(Muscarella et al., 2014)를 활용하였다.

    3. 결과 및 고찰

    3.1 모형 예측 성능

    모델의 예측 성능을 결정하는 FC와 RM은 ‘H-0.5, 2.5, 4.0’, ‘LQH-0.5’, ‘LQHP-0.5, 1.0’, ‘LQHPT-0.5, 1.0, 1.5, 2.0’이 최적의 조합으로 평가되었다. 최적 모델 조합을 적용하여 어류 4종 의 계절별 분포를 예측한 결과 train AUC 값은 고등어 (0.859~0.929), 갈치(0.770~0.880), 멸치(0.929~0.974), 참조기 (0.788~0.899)이며, test AUC 값은 고등어(0.790~0.888), 갈치 (0.756~0.891), 멸치(0.918~0.948), 참조기(0.752~0.893)으로 나타 나 예측 정확도가 높고, 모형의 설명력이 적절함을 보여준 다(Swets, 1988)(Table 3).

    3.2 종분포 예측

    MaxEnt 모형에서 사용된 환경변수인 SST (max, mean, min) 와 Chl-a (max, mean, min)의 기여도를 평가한 결과, 고등어와 멸치는 Chl-a에서 24.8~76.7% 범위로 가장 큰 기여를 보였다. Chl-a는 멸치 서식지와 밀접한 관련이 있으며, 특히 춘계와 추계에 발생하는 식물성 플랑크톤 번성에 큰 영향을 받는 것으로 알려져 있다(Yamada et al., 2004;Kim et al., 2006;Oh and Suh, 2006). 이전 연구에서도 국내 멸치의 분포 예측 결 과 Chl-a가 주요 결정 요인으로 평가되었다(Bang et al., 2022). 갈치와 참조기의 경우, 하계와 춘계에는 Chl-a가 45.1%에서 51.3% 범위로 높은 기여도를 보였고, 동계와 춘계에는 SST 가 51.7%에서 76.6% 범위로 더 큰 영향을 미쳤다(Fig. 1). 참 조기는 수온이 16℃ 이상으로 높아지면 깊은 수심으로 이동 하고, 갈치는 수온 하강 시기인 11월경에 남하하여 월동을 시작하는 것으로 알려져 있다. 이러한 이동 특성이 계절별 기여도 차이에 영향을 미치는 것으로 보인다(Lin et al., 2008;Kim et al., 2020).

    2020년부터 2022년까지의 계절별 출현 자료를 바탕으로 현재의 종분포 예측을 수행하였다(Fig. 2). 고등어는 춘계에 동해 연안 및 남해·제주 연안이 가장 적합한 해역으로 나타 났으며, 하계부터 동계에 이르기까지 서해와 동중국해역으 로 분포가 확장되었다. 갈치는 남해·제주 해역과 동중국해역 에 넓게 분포하는 경향을 보였으나, 동계에는 남해·제주 해 역으로 분포가 축소되었다. 멸치는 주로 국내 연안에 분포 하는 것으로 나타났다. 참조기는 춘계와 하계에는 국내 연 안 및 제주해역에 분포하고, 추계에는 동중국해로 분포가 확장되며, 동계에는 다시 제주와 남해 해역으로 축소되는 경향을 보였다. 현재 예측된 분포는 출현 분포와 유사한 양 상을 보이나, 동해 해역의 예측 분포가 상대적으로 좁게 나 타난 것은 동해 외해역의 출현 자료 부족으로 인한 영향으 로 추정된다.

    3.3 기후변화에 따른 종분포 변화

    현재 예측된 분포 지도에 미래의 기후변화 시나리오인 SSP1-2.6과 SSP5-8.5에 따른 변화된 해수온 자료를 투영한 결 과(Fig. 3), 고등어의 분포는 추계부터 춘계에 걸쳐 전 해역으 로 확대되었으나, 하계에는 현저히 감소하는 경향을 보였다. 추계부터 춘계에는 현재 예측 분포 대비 미래 예측 분포 면 적이 SSP1-2.6에서는 65%, SSP5-8.5에서는 85% 증가했지만, 하계에는 각각 42%와 30%로 감소했다. 모델에서 평균 SST 가 약 19~25°C 범위에서 적합한 분포를 보였으며, 25°C를 초 과하는 온도에서는 분포 확률이 급격히 감소했다(Fig. 4). 이 는 기존 연구에서 보고된 고등어의 최대 임계 수온이 25°C 라는 점과 일치한다(Kim et al., 2020). 따라서, 기후변화로 인 한 SST 상승이 특히 하계에 서해와 동중국해의 고등어 예측 분포에 영향을 미친 것으로 판단된다.

    갈치의 경우, 기후변화 시나리오(SSP1-2.6, SSP5-8.5)에 따 라 현재 예측 분포 대비 모든 계절에서 예측 분포가 확대 되는 경향을 보였다. 특히 춘계에는 동중국해와 동해 외해 역으로 77%(SSP1-2.6), 112%(SSP5-8.5) 증가하여 확장되었고, 하계에는 서해 외해역으로 129%(SSP1-2.6), 119%(SSP5-8.5) 증가하여 확장되었다. 반면, 춘계와 동계에는 평균적으로 39%(SSP1-2.6), 55%(SSP5-8.5) 증가하여, 춘계와 하계에 비해 상대적으로 미미한 분포 변화를 나타냈다.

    멸치는 주로 연안에서 예측 분포를 보였다. 춘계에는 SSP1-2.6 시나리오에서 136%, SSP5-8.5 시나리오에서 181%로 크게 증가했으나, 하계부터 동계에 걸쳐 2%에서 80%까지 감 소하는 경향을 보였다. 외해역의 출현 자료 부족과 환경변 수의 기여도가 Chl-a에서 58.7%에서 90.6%에 이르기 때문에, 특히 클로로필이 번성하는 춘계에 더 큰 영향을 받았고, 수 온 변화의 예측 영향은 상대적으로 제한적인 것으로 분석된 다. 또한, Bang et al.(2022)의 국내 멸치 예측 분포에 관한 선 형연구 결과와 비교해 볼 때, 본 연구는 멸치의 분포가 연안 지역에 더 집중되어 있는 것으로 예측되었다. 이러한 차이 는 출현 자료, 표층 수온, Chl-a 변수의 해상도, 데이터 전처 리 방법, 연도별 변화 등 다양한 요인들로 인해 발생된 것으 로 추정되며, 예측 범위가 실제보다 과대 또는 과소 추정될 수 있음을 나타내었다. 따라서, 보다 정확한 예측을 위해서 는 신뢰도 높은 자료의 구축과 검증단계가 필요할 것으로 판단된다.

    참조기의 분포는 춘계에 기후변화 시나리오 SSP1-2.6과 SSP5-8.5에 따라 각각 82%, 113% 증가하여 전 해역으로 확 대되었다. 하계부터 동계까지는 SSP1-2.6에서 평균 59%, SSP5-8.5에서는 74% 증가하여, 기후변화에 따른 분포 확장이 점진적으로 감소하는 경향을 보였다. 춘계에 참조기의 분포 면적이 높게 나타난 주요 원인으로 SST의 변수 기여도가 66.3%로 높게 나타났으며, 모델 분석 결과 참조기에 적합한 수온 범위가 약 14~17℃로 예측되었다(Fig. 5). 이러한 수온 범위는 미래의 춘계 동해 해역의 수온과 일치하는 경향을 보이며, 기후변화에 따른 동해의 수온 상승이 참조기의 분 포 확대에 직접적인 영향을 미쳤을 것으로 판단된다. 참조 기는 다른 표영성 어류와 달리 수심 40~200m에서 분포하며, 동계에는 깊은 수심에서 월동하는 특성이 있다(Shojima and Otaki, 1982). 이러한 점을 고려하여, 향후 연구에서는 중층 및 저층의 환경 자료를 추가적으로 고려할 필요가 있을 것 으로 판단된다.

    3.4 종풍부도 및 어업활동보호구역

    주요 4종을 대상으로 현재와 미래 기후변화에 따른 종풍 부도 지도를 작성하고, 이를 어업활동보호구역과 비교·분석 하였다. 해양공간계획 경계 내에서 '높음'으로 평가된 종풍부 도 구역과 어업활동보호구역의 중복 비율은 약 15%였으며, SSP1-2.6 시나리오에서 21%, SSP5-8.5 시나리오에서 34%로 증가하였다. 타 용도구역과 유보해역 내에서는 약 12%로 나 타났으며, SSP1-2.6에서 22%, SSP5-8.5에서 34%로 증가하여 종풍부도가 높은 해역이 늘어나는 경향을 보였다(Fig. 6). 대 상 어종은 온대 또는 아열대 해역에 서식하는 어류로 분류 되며(Kang et al., 2014a;Kang et al., 2014b;Kim et al., 2011;Park and Oh, 2020), 기후 변화에 따른 해수온 상승이 종풍부 도 면적의 확장과 연관이 있는 것으로 추정된다. 이러한 결 과는 기후변화에 따라 주요 어종의 연안 서식지 확대가 어 업활동보호구역의 지정 및 평가에 영향을 미칠 수 있음을 나타내며, 삼치, 전갱이, 붕장어 등 본 연구에 미포함된 다양 한 어종을 추가하여 종풍부도를 적용함으로써 보다 정확한 예측 결과를 도출할 필요성이 있다.

    4. 결 론

    본 연구는 MaxEnt 모델을 활용하여 기후변화가 어업활동 보호구역 내 주요 어종의 분포 변화에 미치는 영향을 예측 하였다. 고등어와 멸치는 Chl-a이 가장 큰 기여를 보였으며, 갈치와 참조기는 계절에 따라 Chl-a와 SST의 기여도가 달랐 다. 향후, 염분, 수심, 보호구역, 양식장과 같은 범주형 자료 를 포함하여 환경 변수의 다양성을 확장해 예측력을 높일 필요가 있을 것으로 판단된다.

    미래 기후변화 시나리오(SSP1-2.6, SSP5-8.5)에 따라 고등 어, 갈치, 참조기의 출현 확률은 현재 확률 분포에 비해 28~86% 증가하였으나, 멸치는 6~11% 감소하였다. 종풍부도 지도와 어업활동보호구역을 비교 분석한 결과, '높음' 등급의 종풍부도 구역과 어업활동보호구역의 중복 비율은 약 15% 였고, SSP1-2.6에서는 21%, SSP5-8.5에서는 34%로 증가하였 다. 이러한 결과는 기후변화를 대비하여 향후 어업활동보호 구역을 평가하고 적절한 보존 대책과 정책을 수립할 때 과 학적 근거로 활용될 수 있을 것으로 기대된다. 또한, 다양한 어종 자료의 보완을 통해 더욱 정확한 종풍부도를 작성하 고, 이를 어업활동보호구역의 중장기적 계획 수립 및 용도 구역 평가에 활용할 수 있을 것으로 판단된다.

    사 사

    이 논문은 2023년 해양수산부 재원으로 해양수산과학기 술진흥원의 지원을 받아 수행된 연구이다(20220431, 해양공 간 정책시뮬레이터 기술 개발).

    Figure

    KOSOMES-29-7-802_F1.gif

    The relative contribution (%) of four environmental variables in the Korean waters from 2020 to 2022, derived using the MaxEnt model. SST represents Sea Surface Temperature, and Chl-a represents Chlorophyll-a Concentration.

    KOSOMES-29-7-802_F2.gif

    The distribution of four Key species in Korean waters (2020-2022) was derived using MaxEnt. The values range from 1 to 0, with values closer to 1 indicating more suitable areas or higher occurrence probabilities.

    KOSOMES-29-7-802_F3.gif

    RCP 2.6 and RCP 8.5 depict spatial distribution and seasonal changes for the present (2020-2022) and future (2041~2060) based on the 10th percentile of the training presence as thresholds. The sky-blue color represents the current predicted distribution, orange represents the future predicted distribution, and purple indicates overlapping distributions.

    KOSOMES-29-7-802_F4.gif

    The species response curves for summer temperature as environmental variables affecting Scomber japonicus predicted by MaxEnt.

    KOSOMES-29-7-802_F5.gif

    The species response curves for spring temperature as environmental variables affecting Larimichthys polyactis predicted by MaxEnt.

    KOSOMES-29-7-802_F6.gif

    Korea's Activity Protected Areas and Predicted Fish Species Richness Using MaxEnt Modeling with RCP 2.6 and 8.5 Scenarios.

    Table

    Environmental parameters and presence point used to construct MaxEnt Model for Key fish species in Korea sea

    <i><sup>1</sup>National Federation of Fisheries Cooperatives in Korea</i>
    <i><sup>2</sup>National Oceanic and Atmospheric Administration</i>

    Ensemble forecast data for each CMIP6-based scenario applying statistical downscaling techniques (Korea Meteorological Administration, 2022)

    Overall Description and Performance of MaxEnt Modeling with Averaging Points, Features class, Regularization Multiplier, and AUC

    <i><sup>1</sup>AUC: Area Under the Curve</i>

    Reference

    1. Bang, M. , D. Sohn, J. J. Kim, W. Choi, C. J. Jang, and C. Kim (2022), Future changes in the seasonal habitat suitability for anchovy (Engraulis japonicus) in Korean waters projected by a maximum entropy model, Frontiers in Marine Science, 9(922020), p. 16.
    2. Bae, E. and J. Jung (2015), Prediction of shift in fish distributions in the Geum River Watershed under climate change. Ecology and Resilient Infrastructure, 2(3), pp. 198-205.
    3. Barbet-Massin, M. , F. Jiguet, C. H. Albert, and W. Thuiller (2012), Selecting pseudo‐absences for species distribution models: How, where and how many?, Methods in ecology and evolution, 3(2), pp. 327-338.
    4. Cury, P. and C. Roy (1989), Optimal environmental window and pelagic fish recruitment success in upwelling areas, Can. J. Fish. Aquat. Sci., 46(4), pp. 670-680.
    5. Elith, J. , C. H. Graham, R. P. Anderson, M. Dudı́k, S. Ferrier, A. Guisan, … and N. E. Zimmermann (2006), Novel methods improve prediction of species’ distributions from occurrence data, Ecography, 29(2), pp. 129-151.
    6. Guisan, A. and W. Thuiller (2005), Predicting species distribution: offering more than simple habitat models. Ecology letters, 8(9), pp. 993-1009.
    7. Han, J. R. , T. H. Kim, E. Y. Choi, and H. W. Choi (2021), A Study on the Mapping of Fishing Activity using V-Pass Data-Focusing on the Southeast Sea of Korea, Journal of the Korean Association of Geographic Information Studies, 24(1), pp. 112-125.
    8. Hwang, S. D. , G. A. McFarlane, O. I. Choi, J. S. Kim, and H. J. Hwang (2007), Spatiotemporal distribution of pacific anchovy (Engraulis japonicus) eggs in the West Sea of Korea, Fish. Aquat. Sci., 10(2), pp. 74-85.
    9. IPCC (2022), Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Sixth Assessment Report: Technicla Summary, Intergovernmental Panel on Climate Change, p. 118.
    10. Jung, K. M. , S. Kang, H. K. Cha, K. H. Choi, and M. S. Myksvoll (2013), Buoyancy and vertical distribution of mackerel Scomber japonicus eggs in Korean waters, Korean Journal of Fisheries and Aquatic Sciences, 46(6), pp. 957-965.
    11. Kang, J. H. , Y. G. Kim, J. Y. Park, J. K. Kim, J. H. Ryu, C. B. Kang, and J. H. Park (2014a), Comparison of fish species composition collected by set net at Hupo in Gyeong-Sang- Buk-Do, and Jangho in Gang-Won-Do, Korea, Korean Journal of Fisheries and Aquatic Sciences, 47(4), pp. 424-430.
    12. Kang, M. , S. G. Choi, and B. K. Hwang (2014b), Acoustic characteristics of Anchovy schools, and visualization of their connection with water temperature and salinity in the Southwestern Sea and the Westsouthern Sea of South Korea, Journal of the Korean Society of Fisheries and Ocean Technology, 50(1), pp. 39-49.
    13. Kim, D. , J. Shim, and S. Yoo (2006), Seasonal variations in nutrients and chlorophyll-a concentrations in the northern East China Sea, Ocean. Sci. J., 41(3), pp. 125-137.
    14. Kim, Y. H. , J. T. Yoo, E. H. Lee, T. Y. Oh, and D. W. Lee (2011), Age and growth of largehead hairtail Trichiuruslepturus in the East China Sea, Korean Journal of Fisheries and Aquatic Sciences, 44(6), pp. 695-700.
    15. Kim, J. Y. , J. B. Lee, and Y. S. Suh (2020), Oceanographic indicators for the occurrence of anchovy eggs inferred from generalized additive models, Fisheries and Aquatic Sciences, 23(1), pp. 1-14.
    16. Kim, S. R. , J. J. Kim, H. W. Park, S. Kang, H. K. Cha, and H. J. Baek (2010), Maturity and spawning of the chub mackerel Scomber japonicus in the Korean waters, Korean Journal of Fisheries and Aquatic Sciences, 53(1), pp. 9-18.
    17. KMA (2022), Climate change around the Korean Peninsula, Korea Meteorological Administration, p. 2.
    18. Ko, J. C. , Y. I. Seo, H. Y. Kim, S. K. Lee, H. K. Cha, and J. I. Kim (2010), Distribution characteristics of eggs and larvae of the anchovy Engraulis japonica in the yeosu and tongyeong coastal waters of Korea, Korean. J. Ichthyol., 22(4), pp. 256-266.
    19. Kwon, H. , J. E. Ryu, C. Seo, J. Kim, D. O. Lim, and M. H. Suh (2012), A study on distribution characteristics of corylopsis coreana using SDM. Journal of environmental impact assessment, 21(5), pp. 735-743.
    20. Lee, C. I. , L. H. Jang, and S. E. Park (2009), Influence of water temperature during the main spawning period on anchovy catch, Korean Journal of Fisheries and Aquatic Sciences, 42(3), pp. 297-301.
    21. Lin, L. S. , J. H. Cheng, Y. Z. Jiang, X. W. Yuan, J. S. Li, and T. X. Gao (2008), Spatial distribution and environmental characteristics of the spawning grounds of small yellow croaker in the southern Yellow Sea and the East China Sea, Acta Ecologica Sinica, 28(8), pp. 3485-3494.
    22. Merow, C. , M. J. Smith, and J. A. Silander Jr (2013), A practical guide to MaxEnt for modeling species' distributions: what it does, and why inputs and settings matter, Ecography, 36(10), pp. 1058-1069.
    23. Morales-Castilla, I. , T. J. Davies, W. D. Pearse, and P. Peres-Neto (2017), Combining phylogeny and co-occurrence to improve single species distribution models, Glob. Ecol. Biogeogr, 26(6), pp. 740-752.
    24. Muscarella, R. , P. J. Galante, M. Soley-Guardia, R. A. Boria, J. M. Kass, and M. Uriarte (2014), ENMeval: An R package for conducting spatially independent evaluations and estimating optimal model complexity for Maxent ecological niche models, Methods Ecol. Evol., 5(11), pp. 1198-1205.
    25. O’Banion, M. S. and M. J. Olsen (2014), Predictive seismicallyinduced landslide hazard mapping in oregon using a maximum entropy model (MaxEnt), In Proceedings of the 10th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, Anchorage, AK: Earthquake Engineering Research Institute.
    26. Oh, H. J. and Y. S. Suh (2006), Temporal and spatial characteristics of chlorophyll a distributions related to the oceanographic conditions in the Korean waters, J. Korean. Assoc. Geogr. Inf. Stud., 9(3), pp. 36-45.
    27. Park, I. S. and J. S. Oh (2020), Comparison of morphometric traits between small yellow croaker (Larimichthys polyactis) and yellow croaker (L. crocea), Korean Journal of Environmental Biology, 38(4), pp. 507-517.
    28. Park, J. H. , Y. J. Lim, H. K. Cha, and Y. S. Suh (2004), The relationship between oceanographic and fishing conditions for anchovy, Engraulis japonica, in the southern Sea of Korea. J. Korean. Soc. Fish. Res., 6(2), pp. 46-53.
    29. Park, J. U. , T. Lee, D. G. Kim, and S. Shin (2020), Prediction of potential habitats and distribution of the marine invasive sea squirt, Herdmania momus, Korean Journal of Environmental Biology, 38(1), pp. 179-188.
    30. Phillips, S. J. (2017), A brief tutorial on maxent, p. 39.
    31. Phillips, S. J. and M. Dudík (2008), Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation, Ecography, 31(2), pp. 161-175.
    32. Rose, G. A. , B. DeYoung, D. W. Kulka, S. V. Goddard, and G. L. Fletcher (2000), Distribution shifts and overfishing the northern cod (Gadus morhua): a view from the ocean, Canadian Journal of Fisheries and Aquatic Sciences, 57(3), pp. 644-663.
    33. Shojima, E. and H. Otaki (1982), Lunar rhythmic catch fluctuation in the yellow croaker, Pseudosciaena polyactis, Bull. Fish. Res. Dev. Agency, 28, pp. 147-166.
    34. Swets, J. A. (1988), Measuring the accuracy of diagnostic systems, Science, 240 (4857), pp. 1285-1293.
    35. Walther, G. R. , E. Post, P. Convey, A. Menzel, T. J. C. Beebee, J. M. Fromentin, O. Hoegh-Guldberg, and F. Bairlein (2002), Ecological responses to recent climate change, Nature, 416 (6879), pp. 389-395.
    36. Yamada, K. , J. Ishizaka, S. Yoo, H. C. Kim, and S. Chiba (2004), Seasonal and interannual variability of sea surface chlorophyll a concentration in the Japan/East Sea (JES), Prog. Oceanogr., 61(2-4), pp. 193-211.