Ecological strategies and extinction risk in butterflies and macro-moths of Great Britain and Ireland: A trait similarity network approach.

Trait-based analyses of extinction risk often examine individual traits independently, which can obscure how combinations of traits define broader ecological strategies. Here, we apply a network-based representation of trait similarity to analyze how extinction risk is distributed across combinations of ecological traits, extending an approach previously applied to seabirds. Using a curated species-level trait dataset for butterflies and macro-moths (order Lepidoptera), we constructed a species similarity network for 968 species and identified nine ecological communities using unsupervised community detection. The resulting communities were structured primarily by phenology and life-cycle characteristics. Although conservation status was not used in clustering, threatened species were unevenly distributed across communities, with the multivoltine, externally developing strategy group containing nearly six times the expected number of threatened species relative to background prevalence in both Great Britain and Ireland. These results show that extinction risk in Lepidoptera is structured by integrated ecological strategies, with patterns of vulnerability emerging from multivariate trait configurations. Trait similarity networks therefore provide a scalable approach for identifying strategy-level patterns of conservation vulnerability beyond single-species assessments, particularly in data-limited taxa such as insects where long-term population monitoring is often incomplete.