High-Throughput DNA Barcoding Reveals Multi-Trophic Networks of Plants, Leafminers and Parasitoids in a Tropical Dry Forest.

Tropical dry forests host highly diverse insect communities and complex trophic interactions, yet these networks remain difficult to resolve using conventional molecular or rearing-based approaches. Here, we integrate long-read DNA barcoding, non-target sequence co-amplification, and host plant taxonomy to reconstruct tri-trophic interactions among leaf-mining insects, their host plants, and associated parasitoids in a Mexican tropical dry forest. Using Single Molecule, Real-Time (SMRT) sequencing on the PacBio Sequel IIe platform, we generated high-fidelity barcodes from individual leaf-miner larvae and from parasitoid larvae and pupae recovered from leaf mines. From 253 specimens, we obtained 214 sequences, and identified 83 operational taxonomic units including leaf-mining Lepidoptera, Diptera, Coleoptera, and their hymenopteran parasitoids, plus 31 non-target sequences. Of the 214 sequences, 69 were parasitoid wasps (32.2%). Of these, 40 were obtained from parasitoids dissected directly from leaf mines, while 29 were co-amplified from leaf-miner samples, demonstrating that SMRT-based co-amplification allows detection of parasitism without rearing. Our analysis revealed 156 trophic interaction types and 247 interaction events. Network analyses showed a highly modular tripartite network, with module structure primarily driven by host plant identity and a reduced set of influential leaf-miner and parasitoid taxa. Bipartite networks supported the hypothesis that interactions are more strongly compartmentalized between leafminers and host plants than at higher trophic levels. These results demonstrate that long-read DNA barcoding with co-amplification detection provides a scalable framework for reconstructing multi-trophic interactions from individual specimens, overcoming key limitations of rearing-based methods and enabling robust biodiversity and ecological network assessments in species-rich ecosystems.