DNA from rainwater provides a window into tropical canopy biodiversity

Tropical rainforests still represent a treasure trove of undiscovered species. While many species of plants, animals, and insects have been identified over the years, some parts of these tropical ecosystems are difficult to access and pose challenges for researchers. In particular, the high canopies of tropical forests can be hard—and often expensive—to study without causing local damage.

However, past studies have indicated that rainwater may hold the solution to this problem. Rainwater has the ability to carry DNA from plants and animals in the surrounding environment (referred to as eDNA), which can then be analyzed to provide a window into the local biodiversity. Previously, scientists were able to detect various arthropods in the canopies of temperate forests by collecting eDNA from rainwater trickling down into samplers. They used a method called metabarcoding to identify the organisms represented in the eDNA samples.

In a new study, published in Science Advances, researchers tested out these methods in an old-growth Amazonian rainforest in French Guiana, then compared the results with eDNA collected from a nearby tree plantation. They also analyzed the effectiveness of the canopy rainwash method by spraying carrot juice into the surrounding environment and periodically testing the rainwash over several weeks to determine how long DNA was detectable after contact, as well as by conducting spatial comparisons.

With the help of eDNA collected over 10 days, the researchers were able to identify a wide range of species, including plants, mammals, birds, amphibians, and insects. Fish DNA was also present in the samples—likely as residue or droppings from mammals and birds after consumption. The researchers noted a marked difference in diversity between the natural forest and plantation canopies.

They write, “The taxonomic assignments were 1.3 to 1.9 times more diverse in the old-growth forest plot relative to the plantation, with assignments corresponding to 155 plant taxa (versus 111 in the plantation), 61 vertebrate taxa (versus 32), and 276 (versus 153) insect taxa.”

The eDNA rainwash method proved to be a robust method for species identification. The study authors found that DNA signals are detectable between an average of 8–20 days, depending on the particular DNA marker, and also that the signals are largely localized, at a range of tens of meters. In other words, the method appears to be reliable for detecting species present over some time and capable of detecting species specific to a small area due to the vertical trajectory of rain.

This method provides a contrast to other methods used for similar purposes, as eDNA from streams and air tends to be less localized, and while eDNA collected from soil is localized, it typically only describes the organisms living within the ground. Rainwater, on the other hand, is proving to be the better method for analyzing canopy diversity, and its collection in this study was done in a noninvasive and inexpensive manner—further proving its utility.

The study authors write, “We have established a robust methodological framework for an optimal rainwash eDNA sampling, opening doors for further refinements in community-level assessments across environments or disturbance gradients and for the detection of species crucial for conservation or environmental health.

“By enabling largescale and noninvasive monitoring, rainwash eDNA can substantially improve our description of the elusive diversity hidden in tropical forest canopies and inform and assess tropical forest conservation strategies, contributing vital data for the protection of these irreplaceable ecosystems.”

© 2025 Science X Network