Species are considered the basic unit of biological classification and evolution. Hence, they are used as a benchmark in several fields, although the ontological status of such a category has always been a matter of debate. This paper aims to discuss the problem of the definition of species within the new mechanistic approach. Nevertheless, the boundary between entities, activities, and mechanisms remains difficult to establish and always requires an analysis of what is meant by explanation. As a case study, the paper describes the debate concerning the species category by considering different kingdoms: Animals, Bacteria, and Fungi. The inherently biological differences between those groups prevent the use of a single, universally applicable concept of species that could fit the mechanisms responsible for the variability present in these kingdoms. The same issue is encountered within each group, as highlighted through a focus on mammals and microbes. This controversy has given rise to opposite approaches, namely: monism, which looks for a single definition that might account for all species, and pluralism, which admits that different groups of organisms require somewhat different definitions. In order to develop an adequate definition of species, we propose to apply a new mechanistic framework, which considers the ontic-epistemic dimensions of scientific explanation in close parallel. The apt correlation between epistemic and the ontic aspects highlights the way in which the concept of species and the reference to data are strictly co-determined. This suggests that the concept of species is better understandable within a dual ontic-epistemic approach.

物种被认为是生物分类和进化的基本单位。因此，它们在多个领域被用作基准，尽管这一类别的本体论地位一直存在争议。本文旨在讨论新机械方法中物种的定义问题。然而，实体、活动和机制之间的界限仍然难以确定，总是需要对解释的含义进行分析。作为一个案例研究，本文通过考虑不同的界：动物、细菌和真菌，描述了有关物种类别的争论。这些群体之间固有的生物学差异阻碍了使用单一的、普遍适用的物种概念来适应这些王国中存在的变异性的机制。正如对哺乳动物和微生物的关注所强调的那样，每个群体都遇到了同样的问题。这一争议引发了相反的方法，即：一元论，它寻找一个可以解释所有物种的单一定义；以及多元论，它承认不同的生物体群体需要有所不同的定义。为了对物种做出适当的定义，我们建议应用一个新的机械框架，该框架同时考虑科学解释的本体认知维度。认识论和本体论之间的恰当关联凸显了物种概念和数据参考严格共同确定的方式。这表明物种的概念在双重本体认知方法中更容易理解。