高屏溪流域生態系服務在氣候與土地利用變遷下的熱點變化與權衡分析

Abstract

生態系服務（ecosystem services）是環境自然產生，且對人類有益的各種功能。土地利用變遷與氣候變遷這兩者之間的交互作用會讓生態系統發生變化，導致供水、糧食生產、生物多樣性等關鍵功能改變，若未妥善規畫可能會對人類社會產生威脅。本研究透過蒐集歷史資料及模擬未來氣候變遷與土地利用變化，探討高屏溪流域在 2007 年、2014 年、2021 年三個不同歷史年份以及2050年共享社會經濟路徑(Shared Socioeconomic Pathways, SSP)SSP585時，不同大氣環流模式(General Circulation Models, GCM)：MIROC6、MRI-ESM2-0、BCC-CSM2-MR在SSP585情境下的生態系服務熱點及彼此之間的權衡關係，並透過這些結果提供對高屏溪流域內未劃設保護區之生態系服務熱點進行劃設及治理之建議。 本研究結合空間統計分析與生態系服務建模，評估流域在歷史與未來氣候情境下的土地利用變遷及其對生態系服務的影響。研究運用InVEST（Integrated Valuation of Ecosystem Services and Tradeoffs）模式，針對流域內過去特定年份進行生態系服務量化，涵蓋六項生態系服務：碳儲存、沉積物留存、氮營養鹽留存、磷營養鹽留存、棲地品質、產水量。為辨別具高生態系服務價值的區域，本研究採用區域空間自相關分析（Local Indicators of Spatial Association, LISA），找出每項服務中高值聚集的「High-High」區域，並與該服務前 25% 的高值分布進行交集分析，以此劃定每項生態系服務的空間熱點，再與政府所劃設之保安林、自然保留區、特定水土保持區聯集，用以界定未來發展中應予限制或優先保護的空間區域，作為保護區初步規劃依據。本研究使用 CLUE-s（Conversion of Land Use and its Effects at Small regional extent）模式，模擬2050年在不同氣候情境下流域土地利用的分布變化。該模擬透過輸入各種土地利用轉變的驅動因子，評估不同氣候驅動下土地利用格局的差異與趨勢，作為推估未來情境下生態系服務變化的基礎。之後再次以 InVEST 模型計算 2050 年各項生態系服務的空間分布，並與歷史年份相比較。最後，為分析不同生態系服務間是否存在協同關係（synergy）或權衡關係（trade-off），計算各服務指標間的 Pearson 線性相關係數，並進一步透過雙變量空間自相關分析（Bivariate LISA），探討不同生態系服務之間在空間分布上的聚集與交互作用情形。 研究結果表示高屏溪流域內約有11%的面積包含4種以上綜合熱點。既有的政策已能有效保護60%的生態系服務綜合熱點，其中林地佔限制區域的83.5%，草地佔8.1%，這兩種土地利用類型較能提供生態系服務。在模擬不同氣候模式（GCM）下未來土地利用變遷的情境中，雖然各種生態系服務的空間分布略有差異，但整體變化不大。研究發現，產水量較容易受到氣候變遷的影響，其空間分佈與產水量變化明顯。結果指出，山區林地長期都是碳儲存、營養鹽留存、棲地品質等指標的高值熱點（High-High），而下游都市化區域則常是生態服務低值熱點（Low-Low），特別是在人口密集與開發密集的地區，顯示人為活動會造成生態系功能退化。最後，研究建議未來應將這些綜合熱點納入國土空間規劃與保護政策的依據，特別是將具有高度生態價值的地區優先列為限制開發的保護區，以確保生態系服務的長期穩定與持續提供。

Ecosystem services refer to the various benefits and functions naturally generated by ecosystems due to their inherent structure and processes, without direct human manipulation. The interaction between land use change and climate change significantly undermines ecosystem resilience, leading to the collapse of critical services such as water provision, food production, and biodiversity maintenance, thereby posing substantial threats to human well-being. This study integrates historical data and future climate and land use projections to investigate the ecosystem service hotspots and trade-offs in the Gaoping River Basin across three historical years (2007, 2014, and 2021), as well as in 2050 under the Representative Concentration Pathway (RCP) 8.5 scenario. Future simulations incorporate three General Circulation Models (GCMs): MIROC6, MRI-ESM2-0, and BCC-CSM2-MR. The findings aim to inform the delineation and governance of critical ecosystem service areas currently lacking official protection within the watershed. This research combines spatial statistical analysis with ecosystem service modeling to evaluate the impacts of historical and projected climate scenarios on land use change and ecosystem service provision in the watershed. The InVEST (Integrated Valuation of Ecosystem Services and Tradeoffs) model was employed to quantify five key ecosystem services—carbon storage, sediment retention, nitrogen and phosphorus retention, habitat quality, and water yield—for each selected year. To identify areas of high ecosystem service value, the study applied Local Indicators of Spatial Association (LISA) to detect “High-High” clusters of each service and intersected these with the top 25% of service values to delineate spatial hotspots. These hotspots were further overlaid with officially designated areas, including protection forests, nature reserves, and designated soil and water conservation zones, to identify priority areas for future development constraints or conservation, forming a preliminary planning basis for protected area designation. The CLUE-s (Conversion of Land Use and its Effects at Small regional extent) model was used to simulate land use distribution in 2050 under different climate scenarios, incorporating multiple land use drivers to assess projected changes in land use patterns. Based on the simulated land use for 2050, the InVEST model was again applied to quantify future ecosystem service distributions, which were then compared with historical baselines. To explore potential synergies or trade-offs among ecosystem services, the study calculated Pearson correlation coefficients among service indicators and conducted Bivariate LISA to examine the spatial co-occurrence and interactions among services. Results indicate that approximately 11% of the Gaoping River Basin contains areas with four or more overlapping ecosystem service hotspots. Existing policies effectively protect around 60% of these composite hotspots, with forest lands accounting for 83.5% and grasslands 8.1% of the restricted zones—land types typically associated with higher ecosystem service provision. Under simulated future land use scenarios based on different GCMs, the spatial distribution of ecosystem services shows only minor variation. Among the services, water yield is the most sensitive to climate change, exhibiting significant spatial and interannual variability. Furthermore, mountainous forested regions consistently serve as High-High hotspots for carbon storage, nutrient retention, and habitat quality, whereas downstream urbanized areas are frequently Low-Low zones with degraded service provision, especially in densely populated and developed regions. These findings underscore the detrimental impact of anthropogenic pressure on ecosystem functions. Finally, the study recommends incorporating these composite hotspots into national spatial planning and conservation policy frameworks, prioritizing areas of high ecological value for development restrictions to ensure the long-term stability and continuous provision of ecosystem services.