Pakulnicka Joanna, Kruk Marek
Department of Zoology, University of Warmia and Mazury in Olsztyn, Lodzki sq. 3, 10-727, Olsztyn, Poland.
Department of Applied Informatics and Mathematical Modelling, University of Warmia and Mazury in Olsztyn, Sloneczna 54, 10-719, Olsztyn, Poland.
Sci Rep. 2025 May 7;15(1):15898. doi: 10.1038/s41598-025-00525-1.
Water deficits are a serious problem around the world, which also affects young landscapes, where lakes are most abundant. This poses a threat to many habitats and biological diversity found here. The relationships between species in the ecological networks of lakes at different stages of development and in nearby post-exploitation water bodies remain poorly understood. To better understand the functioning of beetle communities in different ecosystems, we created five network models that we subjected to graph analysis. By analysing the general attributes of the network (number of neighbours, shortest path, characteristic path length, clustering coefficient, network centralisation, network density and network heterogeneity) and those related to the nodes (NCC-Node Closeness Centrality, NBC-Node Betweenness Centrality, NDC-Node Degree Centrality) and to the edges (EBC-Edge Betweenness Centrality and correlations between the biomass of species as nodes), we were able to determine the role of each species in the networks and the relationships between the species. We then used the machine learning ensemble modelling XGBoost-SHAP to identify species that are particularly important in migrations between water bodies and to assess the direction and strength of migrations using Shapley values. Our analyses are based on faunal material from 25 lakes (mesotrophic, eutrophic, dystrophic) and 31-post-exploitation water bodies (clay pits and gravel pits) in northern Poland, in the Masurian Lake District. We found a total of 169 species representing different ecological and functional components. We have shown that the structures of the network between the biomass of species in the analysed five water types differ significantly. The highest value for network density was recorded in eutrophic lakes and clay ponds, the lowest in dystrophic lakes. In eutrophic lakes these are mainly eurybionts, in clay pits-rheophiles and in gravel pits-argilophiles and tyrphophiles. The relationship between the species with the highest NBC and EBC values is particularly important in order to maintain the stability of the network. The periphery of the network usually consists of larger predators that do not compete with each other. By analysing the migration directions of beetles between different ecosystems, we were able to demonstrate a greater affinity of the beetle fauna, especially the argilophiles (e.g. Scarodytes halensis and Laccobius minutus) inhabiting gravel pits, to dystrophic lakes. The beetles in clay pits originate mainly from mesotrophic lakes. These are mainly rheophiles, mostly weakly flying species, such as: Haliplus fluviatilis, Haliplus fulvus, Ilybius fenestratus, Hygrotus vericolor and Haliplus flavicollis. These species are important for the stability of ecological networks in the studied lake types. Their movements between the ecosystems studied in turn contribute to the functional connectivity between the individual lakes, which ensures the stabilisation of biotic relationships at the landscape level. At the same time, they generally also indicate the optimisation of environmental conditions in post-exploitation water bodies, which makes them potential substitute habitats for natural lakes.
水资源短缺是全球面临的一个严重问题,这也影响到年轻的景观区域,这些区域湖泊最为丰富。这对这里的许多栖息地和生物多样性构成了威胁。人们对处于不同发育阶段的湖泊以及附近开采后的水体生态网络中的物种之间的关系仍然知之甚少。为了更好地理解不同生态系统中甲虫群落的功能,我们创建了五个网络模型并对其进行图分析。通过分析网络的一般属性(邻居数量、最短路径、特征路径长度、聚类系数、网络中心性、网络密度和网络异质性)以及与节点相关的属性(NCC-节点接近中心性、NBC-节点中介中心性、NDC-节点度中心性)和与边相关的属性(EBC-边中介中心性以及作为节点的物种生物量之间的相关性),我们能够确定每个物种在网络中的作用以及物种之间的关系。然后,我们使用机器学习集成模型XGBoost-SHAP来识别在水体之间迁移中特别重要的物种,并使用Shapley值评估迁移的方向和强度。我们的分析基于波兰北部马祖里湖区25个湖泊(中营养、富营养、贫营养)和31个开采后的水体(粘土坑和砾石坑)的动物材料。我们总共发现了169个代表不同生态和功能成分的物种。我们已经表明,在分析的五种水类型中,物种生物量之间的网络结构存在显著差异。网络密度最高值出现在富营养湖泊和粘土池塘中,最低值出现在贫营养湖泊中。在富营养湖泊中,这些主要是广适性生物,在粘土坑中是喜流生物,在砾石坑中是喜泥生物和喜沼泽生物。具有最高NBC和EBC值的物种之间的关系对于维持网络的稳定性尤为重要。网络的外围通常由彼此不竞争的大型捕食者组成。通过分析甲虫在不同生态系统之间的迁移方向,我们能够证明甲虫动物群,特别是栖息在砾石坑中的喜泥生物(如Scarodytes halensis和Laccobius minutus)对贫营养湖泊有更大的亲和力。粘土坑中的甲虫主要来自中营养湖泊。这些主要是喜流生物,大多是飞行能力较弱的物种,如:Haliplus fluviatilis、Haliplus fulvus、Ilybius fenestratus、Hygrotus vericolor和Haliplus flavicollis。这些物种对于所研究的湖泊类型中的生态网络稳定性很重要。它们在研究的生态系统之间的移动反过来有助于各个湖泊之间的功能连通性,这确保了景观层面生物关系的稳定。同时,它们通常也表明开采后水体中环境条件的优化,这使它们成为天然湖泊的潜在替代栖息地。