Ensuring NHP welfare through in-cage cognitive testing and remote sensing of physiological parameters

NHPs的神经科学实验通常涉及一套在其他研究领域很少发现的实验程序。在本项目中，我们致力于系统量化它们对NHPs中动物福利的潜在影响，重点关注流体控制、运动限制以及由于长期参与实验而可能累积的严重程度。我们已经朝着提高神经科学研究中NHPs福利的长期目标取得了重要进展。首先，通过测量圈养恒河猴的每日液体摄入量，我们显示了大量的个体内部和个体之间的差异，并得出结论，简单的归一化模型（每动物体重的液体体积）不太适合描述恒河猴的液体消耗行为。其次，我们使用并改进了基于笼子的自动化测试系统，以证明在动物行为训练的某些阶段可能减少运动限制。第三，我们开发了一个数据库，用于记录来自不同实验室的动物在长时间内的实验、行为和生理参数，并使用跨实验室标准化数据收集工具。第四，作为概念验证，我们量化了不同实验条件下短期唾液皮质醇反应，并制定了长期毛发皮质醇测量取样的SOP。对于更新项目，我们建议通过添加新的技术特征和先进的自动行为训练算法，使基于笼子的测试系统在界面和行为控制方面与传统的实验设置最相似。对于后者，我们将描述恒河猴在笼子里接受认知任务训练时学习行为的可变性和共性。此外，我们将系统地从数据库中的几个实验室收集生理参数，并对数据进行荟萃分析，以确定与福利相关的生物标志物。

Neuroscientific experiments in NHPs typically involve a set of experimental procedures rarely found in other areas of research. In this project we work towards a systematic quantification of their potential effects on animal welfare in NHPs, with a focus on fluid control, movement restraint, and putative cumulative severity due to long-term participation in experiments. We already made important progress towards our long-term goal of improving the welfare of NHPs in neuroscience research. First, by measuring the daily fluid uptake of captive rhesus monkeys, we show substantial intra- and inter-individual variability and conclude that simple normalization models (fluid volume per animal weight) are not well suited for describing fluid consumption behaviour in rhesus monkeys. Second, we used and advanced our automated cage-based testing system to demonstrate the possible reduction of movement restraints in certain phases of the behavioural training of animals. Third, we developed a database for registering experimental, behavioural and physiological parameters of animals from different labs over long periods of time with tools for standardized data collection across labs. Forth, as a proof-of-concept, we quantified short-term salivary cortisol responses in different exper-imental conditions and developed an SOP for sampling longer-term hair cortisol measures. For the renewal project, we propose to make our cage-based testing system most similar to conventional experimental settings regarding interfaces and behavioural control by adding new technical features and advanced algorithms for automated behavioural training. With the latter we will characterize the variability and commonalities in learning behaviour of rhesus monkeys trained on cognitive tasks in their cages. Further, we will systematically collect physiological parameters from several labs in our database and run meta-analyses on the data to identify welfare-relevant biomarkers.