Abstract Facing the loss of biodiversity caused by landscape fragmentation, implementation of ecological networks to connect habitats is an important biodiversity conservation issue. It is necessary to develop easily reproducible methods to identify and prioritize actions to maintain or restore ecological corridors. To date, several competing methods are used with recurrent debate on which is best and if expert-based approaches can replace data-driven models. We compared three methods: knowledge-driven (expert based), data-driven (based on species distribution model), and a mixed approach. We quantified their differences in habitat and corridor mapping, and prioritizations of landscape elements in terms of importance for connectivity. Key parameters generating these differences were identified. To put this into practice, the case study of the wildcat (Felis silvestris Schreber, 1777) was chosen. The results highlighted differences and similarities between approaches used. The data-driven approach was more successful in identifying the suitable habitat with regard to wildcat ecology, while the knowledge-driven approach was better able to account for obstacles to wildcat movements in the landscape matrix. However, these two methods converged for the identification of patterns of habitat patches and corridors that are important for global landscape connectivity. For both methods, we identified adjustments that can improve the outcome. The mixed approach largely differed in that it required more inputs to be performed. In the end, conservation actions were identified and could guide nature conservation practitioners in their efforts to restore landscape connectivity.

中文翻译：

---

自然保护决策中物种栖息地网络建模方法的比较：以比利时南部野猫为例

摘要 面对景观破碎化造成的生物多样性丧失，实施生态网络连接生境是一个重要的生物多样性保护问题。有必要开发易于重现的方法来确定维护或恢复生态廊道的行动并确定其优先顺序。迄今为止，使用了几种相互竞争的方法，并反复辩论哪种方法最好，以及基于专家的方法是否可以取代数据驱动模型。我们比较了三种方法：知识驱动（基于专家）、数据驱动（基于物种分布模型）和混合方法。我们量化了它们在栖息地和走廊测绘方面的差异，以及景观元素在连通性重要性方面的优先级。确定了产生这些差异的关键参数。为了把这付诸实践，选择了野猫（Felis silvestris Schreber，1777）的案例研究。结果突出了所用方法之间的差异和相似之处。数据驱动的方法在确定适合野猫生态的栖息地方面更成功，而知识驱动的方法能够更好地解释景观矩阵中野猫运动的障碍。然而，这两种方法融合在一起，用于识别对全球景观连通性很重要的栖息地斑块和走廊的模式。对于这两种方法，我们确定了可以改善结果的调整。混合方法的主要区别在于它需要执行更多的输入。最后，确定了保护行动，并可以指导自然保护从业者努力恢复景观连通性。