Researchers compiled floristic records for the islands, spanning an area range of 0.004 to 515.376 square kilometres. The assembled dataset contained 1,158 native seed-plant species and 394 non-native species across 154 families and 729 genera. Most records had been collected during the preceding 20 years.

The observational study separated plants by native or non-native status and by woody or herbaceous growth form. Investigators compared linear, intercept, continuous one-threshold and left-horizontal one-threshold models for relationships between island area and species richness, Faith’s phylogenetic diversity and standardised phylogenetic diversity. A small-island effect was inferred when a threshold model was supported and the linear model had a corrected Akaike information criterion difference greater than two.

For combined seed plants, the left-horizontal model was the most strongly supported for species richness and Faith’s phylogenetic diversity. Native plants had estimated area thresholds of 2.12–2.49 square kilometres for these measures, falling between previously reported thresholds for bryophytes and butterflies in the same archipelago.

Non-native seed plants also exhibited the small-island effect for species richness and Faith’s phylogenetic diversity. Their thresholds were lower than those estimated for native plants and for the combined flora. Excluding cultivated plants made this difference more pronounced: richness ranged from zero on 30 islands to 67 species on Zhoushan, the largest island in the sample.

Growth form produced contrasting patterns. Across both native and combined seed plants, herbs had larger area thresholds than woody plants, contrary to the expectation that the generally stronger dispersal of herbs would produce lower thresholds. Among non-native plants, however, herbs had lower thresholds than woody plants. The study associated this contrast with the composition of the records: 98% of non-native woody plants were cultivated, while about 40% of non-native herbs were naturalised or invasive.

Human influence was measured as the combined area of built-up land and cropland divided by island area. Its relationship with island area was non-linear, with a threshold of 0.13 square kilometres. Large islands had a higher mean human-influence value than small islands, 0.24 compared with 0.05, although the researchers stated that human impact was not determined solely by area.

Phylogenetic patterns added detail not apparent from richness alone. Faith’s phylogenetic diversity was strongly correlated with species richness, while standardised phylogenetic diversity showed more variable relationships with area. Native woody plants and non-native herbs supported small-island effects for standardised phylogenetic diversity, whereas several other growth-form combinations did not.

The authors cautioned that information on species distributions, community dynamics and human influence remained incomplete despite the assembled database. They said detailed long-term field surveys and remote-sensing monitoring of human activities would be needed to clarify the ecological processes associated with the observed patterns and to support conservation strategies.

Subject of Research: Small-island effects in native and non-native seed-plant diversity across the Zhoushan Archipelago

Article Title: Small-island effect is detected for both native and non-native seed plants in the largest archipelago of China

Article References: Shang, X., Yan, E., & Zhang, J. (2026). Small-island effect is detected for both native and non-native seed plants in the largest archipelago of China. Frontiers of Biogeography, 19, Article e168932. https://doi.org/10.21425/fob.19.168932

Image Credits: AI Generated

DOI: 10.21425/fob.19.168932

Keywords: small-island effect, species-area relationship, seed plants, phylogenetic diversity, Zhoushan Archipelago, island biogeography, plant invasions, native plants, non-native plants, ecological modelling, Small-island, effect