Delivering Ecoacoustic Net Gain in the UK

生态声学是利用声音来更多地了解生态系统的性质及其功能。科学领域和实际应用正在迅速发展，因为新的硬件和分析技术，特别是人工智能方法，使方法更容易获得，结果更具可操作性。广泛的分类群产生的声音可以用来推断存在、活动、行为，在某些情况下还可以推断其丰富性。监督小组有使用生态声学方法监测鸟类、蝙蝠、直翅目动物、小型哺乳动物、两栖动物和一系列栖息地类型的经验。生物多样性净收益(BNG)是英国政府采用的一种发展方法，并在即将出台的环境法和国家规划政策框架中得到实施。该政策旨在使自然环境处于比开发前更好的状态--这一成果是通过一个30年的监测计划来评估的。作为这一政策的结果，人们对可用于证明净收益和可大规模部署的生物多样性监测方法产生了极大的兴趣。生态声学有潜力通过在大范围内以低成本提供生物多样性措施来满足生物多样性政策的需求，但关键问题仍然存在。1)生态声学方法是否能够以足够的能力通过生境质量梯度检测生物多样性的变化，以用于测量生物多样性净收益；2)同样的结果是否适用于为支持连通性而固有价值的线性生境；3)生态声学方法固有的误差和偏差是什么，以及这些对其作为衡量生物多样性净收益的工具的有效性有何影响？如何设计/模仿这些问题？4)我们能不能定义基准来衡量“好的”和“差的”质量的栖息地听起来像什么？

Ecoacoustics is the use of sound to understand more about the nature of ecosystems and how they function. The scientific field and practical applications are rapidly developing as new hardware and analytical techniques, in particular AI methods, make the methodologies more accessible and the results more actionable. A broad spectrum of taxa produces sounds which can be used to infer presence, activity, behaviour, and in some cases abundance. The supervisory team has experience using ecoacoustics methods to monitor birds, bats, orthoptera, small mammals, amphibia and a range of habitat types.Biodiversity Net Gain (BNG) is an approach to development adopted by the UK government and implemented in the forthcoming Environment Act and National Planning Policy Framework. The policy aims to leave the natural environment in a measurably better state than before development - and this gain is assessed through a 30 year monitoring programme. As a consequence of this policy there is a significant interest in methods for monitoring biodiversity that can be used to evidence net gain and which can be deployed at scale.Ecoacoustics has the potential to address the needs of BNG policy by providing measures of biodiversity at low cost over large scales, but key questions still remain. 1) Are ecoacoustics methods capable of detecting changes in biodiversity across habitat quality gradients with sufficient power to be of use for measuring biodiversity net gain, 2) do the same results hold for linear habitats, inherently valued to support connectivity, 3) what errors and biases are inherent to an ecoacoustics approach, and how do these impact on its effectiveness as a tool for measuring biodiversity net gain? How can these issues be designed/modelled away? and 4) can we define reference baselines to benchmark what 'good' and 'poor' quality habitats sound like?