農場尺度生物多樣性保育系統性優化規劃方法－以梅峰農場為例

Abstract

生物多樣性為人類相當重要之自然資源，但近年來因棲地破壞、外來種引進等問題使得生物多樣性快速喪失，且生物多樣性之保育常與人類之經濟發展產生衝突，能運用於保育之資源常常有限，因此如何提高有限資源之效率、識別保育管理之區域且發展一支援決策之工具為生物多樣性保育之重要任務。 本研究以南投縣之台大梅峰農場為研究區域，以系統性保育規劃為架構進行農場尺度之鳥類生物多樣性之規劃。考量保育目的為考慮鳥類對於棲地空間配置與組成適宜性，並選取適當區域使得資源效益最大化，以羅吉斯迴歸(Logistic Regression)計算每種鳥類棲地適宜性指數，同時也選取出迴歸具代表性之鳥種作為規劃目標物種，並利用少數鳥種進行候選規劃各單元之成本結合方式分析，再使用系統性保育規劃模式Marxan進行不同目標物種比例與空間限制成本之保護區選取，最後以景觀生態指數評估各目標物種比例與空間限制成本下之連結度。 研究結果顯示，梅峰農場鳥類棲地喜好區位大多與森林面積相關，且此區域鳥類大多喜好不同類別之邊緣區域，如耕地與天然林、人工針葉林之邊緣區域；此外，若將規劃單元成本定為棲地適宜性指數經極值正規化後相加，比將棲地適宜性指數直接相加更能公平客觀地考量各種鳥類之適宜棲地；Marxan選取保護區結果方面，若在資源有限時，梅峰農場北天然林與東方天然林與果園交界區為保育效益最大之區域，而西方區域之效益最差，若須進行開發可考慮以此區域進行。另一方面，從將無保育等級與非特有(亞)種之鳥種排除後之結果可得知，保護區之區位並無明顯差別但範圍大小需求有差異。最後，以景觀生態指數評估連結度可發現，不同之目標物種比例之需求應有不同之空間限制成本，才能使得連結度達到同樣之水準。本研究提出一新方法概念，並確認能結合鳥類對於棲地空間組成與配置之適宜程度進入保護區規劃當中，以提供規劃者更好之決策方向。

Biodiversity is one of the most important natural resources for human. However, biodiversity has been rapidly lost in recent years due to destruction of habitat, invasion of alien species and some other issues. Moreover, biodiversity conservation and economic growth often conflict, and conservation resources that can be used are often limited. Therefore, how to improve the efficiency of limited resources, identify where conservation management are needed and develop a decision support tool is an important task for biodiversity conservation. In this study, a conservation planning was conducted for maintaining biodiversity of birds at the Meifeng Highlands Experiment Farm of National Taiwan University in Nantou, and the planning of bird biodiversity was followed by the framework of systematic conservation planning. The goal of the conservation planning is to efficiently select the protected areas with higher habitat suitability. We calculated the Habitat Suitability Index (HSI) for each bird by logistic regression, and selected the species which are representative in regression analysis as target species. Furthermore, a systematic conservation planning model, Marxan, was applied to identify the protected areas that cover different proportion of target species. In Marxan, the distribution of HSI was evaluated as the cost of planning units. In order to zone a protected area practically, a sensitivity analysis of Boundary Length Modifier (BLM) which represents the weighting of the cost of spatial constraint was carried out. Landscape metrics were further applied to assess the connectivity of the protected area. The results revealed that birds at Meifeng prefer forest and edge areas between cropland and natural forest and between cropland and needleleaf plantation. Compared with the selected protected areas, the summation of normalized HSI as the cost of planning units is more objective than using the summation of HSI of all species. When the resources are limited, northern natural forest and the border of eastern natural forest and orchard are the areas with the largest conservation benefit, and the western area has the least benefit. Moreover, when the species that are neither classified as conservation nor endemic (sub) species are excluded, there is no significant difference in the location of protected areas but the required size. Based on the connectivity results, it is found that different spatial constraint cost should be used to meet different proportion needs of target species. This study proposed a novel concept for conservation planning with an integration of habitat spatial composition and configuration for planners a better decision-making strategy.