This work aims to test and introduce a newly developed methodological approach for palaeoenvironmental reconstructions using advanced intelligent controls, namely fuzzy cognitive maps (FCMs). The new methodology is presented to the scientific community through the palaeoenvironmental reconstruction of a Lower Pleistocene sedimentary sequence in the Sousaki Basin (north-eastern Corinth Gulf, Greece). The palaeoenvironmental analysis using FCMs revealed the presence of an unstable lagoonal environment with minor zonation in the study area. Five palaeoenvironmental units were identified, while only four were found using conventional statistical methods. The newly developed approach removes subjectivity from biostratigraphical data and provides automated and more accurate results, as fuzzy theory examines the given data from a more general perspective. It also highlights natural variations existing among evolving environments in species abundance gradients. This work suggests the great potential for creating a new method that could serve as a useful complement to the established statistical methods to exploit more detailed results on the palaeoenvironmental features of past fossil assemblages.

这项工作旨在测试和引入一种新开发的方法论方法，用于使用先进的智能控制进行古环境重建，即模糊认知图 (FCM)。通过对 Sousaki 盆地（希腊科林斯湾东北部）下更新世沉积层序的古环境重建，向科学界展示了这种新方法。使用 FCM 进行的古环境分析揭示了研究区域存在具有轻微分带的不稳定泻湖环境。确定了五个古环境单元，而使用传统统计方法只发现了四个。新开发的方法消除了生物地层数据的主观性，并提供自动化和更准确的结果，因为模糊理论从更一般的角度检查给定的数据。它还强调了物种丰度梯度不断变化的环境中存在的自然变化。这项工作表明创建一种新方法的巨大潜力，可以作为对既定统计方法的有用补充，以利用过去化石组合的古环境特征的更详细结果。