Abstract:

Aim: The inherent growth characteristics of plants necessitate reliance on external agents for pollen transfer, forming close mutualistic relationships with flower-visiting insects. However, differences in flower-visiting insect communities between invasive and non-invasive plants in urban ecosystems, as well as the impacts of invasive plants on flower-visiting insect diversity and plant–pollinator interaction networks, remain poorly understood. This study, conducted in six urban parks in Fuzhou, compares the composition, diversity, and network-level parameters of flower-visiting insects between invasive and non-invasive plants to explore these differences. 

Methods: Based on field surveys of flowering plants and their flower-visiting insects in the six parks, we employed linear mixed models (GLMM and LMM), β-diversity partitioning, non-metric multidimensional scaling, and network analysis. 

Results: (1) A total of 26 invasive plant species were recorded, predominantly Asteraceae, while 98 non-invasive species were recorded, mainly Rosaceae. Among flower-visiting insects, 144 species (belonging to 6 orders and 36 families) were associated with invasive plants, and 156 species (belonging to 7 orders and 46 families) were associated with non-invasive plants. Hymenoptera dominated both groups, while Lepidoptera and Hemiptera exhibited significant visitation preferences for invasive plants. (2) The number of invasive plant species was significantly positively correlated with flower-visiting insect diversity and significantly negatively correlated with the species-level specialization index (d′), although the explanatory power was low (R² = 0.135). Invasive and non-invasive plants shared some flower-visiting insects, while invasive plants additionally attracted other insects. Functional group differences between the two insect communities were not significant, indicating functional redundancy. (3) Invasive plants led to decreased connectivity of plant-flower visitor networks, a less negative nestedness Z-score, and declines in both modularity and network specialization (H₂'). 

Conclusions: In Fuzhou parks, the flower-visiting insect community structure on invasive plants is more uneven than that on non-invasive plants, dominated by a few dominant species. Invasive plants attract new insects and form new interaction modules, rather than simply replacing or simplifying the original pollination network. We speculate that invasive plants may ultimately occupy a dominant position in the network.

Key words: biodiversity, interaction network, species-level specialization index, urban green spaces, Fuzhou