Illuminated Roads, Hidden Impacts: An Integrated Assessment of Invertebrate Biodiversity in the Atlantic Forest

Artificial light at night associated with road infrastructure in Atlantic Forest landscapes may act as an important ecological disturbance, affecting the composition, behavior, and connectivity of invertebrate communities. In this study, we evaluated biodiversity associated with a road section in the Tapiraí region, São Paulo, Brazil, with a focus on the potential effects of artificial night lighting on local entomofauna. We applied an integrative approach combining environmental DNA, DNA metabarcoding, and traditional entomological sampling in order to improve taxon detection across multiple environmental compartments. A total of 37 samples were analyzed, including water, soil, bulk insect samples, and spider webs. Sequencing generated 6,189,756 reads, of which 1,372,723 showed at least 95 percent similarity to reference sequences in the BOLD database. Molecular analysis detected 1,393 OTUs distributed across 12 phyla, 24 classes, 53 orders, 193 families, 122 genera, and 24 species, with a clear dominance of Arthropoda, which accounted for 1,342,733 reads. Sampling points 2 and 3 showed the highest numbers of identified reads, with 514,864 and 504,931, respectively. Spider web and bulk samples contributed most strongly to read number and OTU richness, while water and soil samples complemented detection by revealing exclusive taxa. In parallel, the entomological survey recorded 176 unique taxa distributed across 12 orders and 95 families, with Diptera, Coleoptera, and Lepidoptera together accounting for more than 70 percent of the recorded diversity. The expressive richness of nocturnal Lepidoptera, scarab beetles, fireflies, aquatic insects, and native bees indicates high ecological sensitivity of the area to artificial lighting effects, including phototactic attraction, disorientation, reproductive disruption, increased predation risk, barrier effects, and impairment of ecosystem services such as pollination and nutrient cycling. Our results demonstrate that integrating molecular ecology with classical entomology provides a robust strategy for rapid biodiversity assessment and for evaluating the potential ecological impacts of illuminated road infrastructure in highly diverse tropical landscapes.