Host-plant interactions have long been a topic that intrigues scientists. However, for smaller species, like insects and their larvae, it has not always been possible to conclusively identify interactions. Just because you find a larva in a certain tree, doesn’t necessarily mean that tree is itself the larva's meal of choice. It could very well be eating lichens, mosses, fungi or other much smaller species associated with the larger plants. 

But now, thanks to advances in DNA sequencing and the vast DNA reference library datasets, scientists now can turn to genetics to confirm associations and make new insights into the world of insect-plant interactions.

That is what three research groups in Germany (Bavarian State Collection of Zoology, University of Würzburg, and Advanced Identification Methods Gmbh [disclosure: I work for AIMG]) are piloting on a large sample of caterpillars from Peru. They sequenced a specific gene in each of 119 individual caterpillars that matched to 92 species in the database. DNA from the caterpillars' guts revealed that these caterpillars feed on many plant species, including lianas (a type of climbing vine) and mosses. Around 80% of them did not have the DNA of the trees where they were found in their gut content. 

Scientists are still trying to optimize the whole protocol, but this methodology shows a new way of looking into host-plant and food web interactions, and informing species diversity for conservation efforts. Now that they know this system works, their future plans include identifying much bigger datasets of arthropod groups and their hosts.