 The traits of plant foliage have a large influence on many ecosystem processes, including herbivory, decomposition, productivity, and more. The canopy also provides tremendous habitat for biodiversity, and it also cycles vast quantities of carbon and water. And yet, especially in tall canopies, these ecosystems are very challenging to access and are therefore difficult to study. And we really don't know as much about them as we would like. Here at UKFS, we have a mix of prairie and woodland species. We collect sunlit canopy foliage because we are looking to demonstrate the physical and chemical properties of a variety of species all across the domain. There are many plant foliar traits that researchers are interested in. Here at NEON, we measure leaf mass per area, leaf water content, the concentrations of carbon and nitrogen, as well as their stable isotopes, content of lignin and cellulose, as well as micro and macronutrients, such as phosphorus, calcium, potassium, iron, and others. We collect leafs from the sunlit canopy because we are trying to create a linkage between our ground data and the aerial flyover data. With the advent of new remote sensing technologies, we're able to map and monitor certain plant foliar traits at higher spatial and temporal scales than ever before. But this approach relies on developing models to link remote sensing data to traits, and this requires sampling from diverse plant canopies. Access to sunlit leaves in tall canopies is a real challenge. Until recently, the primary technique NEON used to collect these tissues were line launchers. This involves shooting a rope up into the canopy from the ground. The line can then be used to raise a saw or simply pull to bring down a small branch containing sunlit leaves. It is possible to procure foliage with this manner, but again it is very challenging. It can take a long time to sample a single tree. It's hard to be precise about the exact branch you want to target, and there can be some safety concerns because sometimes larger branch or larger amounts of the tree come down than you were anticipating. Our drone collections are performed by flying a drone to our target tree and collecting a branch from the top of the tree and bringing it back to our sampling team. So more specifically, our drone has GPS coordinates associated with each individual target that we are going for, and the drone itself will actually just automatically hover over the tree that we are looking for. The botanist will be on the ground looking up at the tree to verify that the drone is over the correct individual, and when everything is looking good, the drone operator will fly it down very slowly. There's a little claw mechanism attached to the very bottom of the drone that can saw the tree and grab the specific branch we are looking for. This tool has many advantages, the number one being that we can very accurately and precisely get those top of canopy sunlit samples that are otherwise very hard to reach. The tool enables much faster sampling, much more precise sampling, and is in fact safer because we can make sure we're targeting only small manageable size branches that are efficiently delivered to the sampling team at the base of the canopy. So the upsides to drone collection is way more fast and way more efficient than any of our land-based methods that we use. For example, we could probably get three to four trees done in the same amount of time that it would take to get one tree done for using the line launcher. It requires quite a bit of training to become competent with the sampling tool, and operators must have FAA remote pilot certification. Given the increasing availability of hyperspectral remote sensing data sets, such as new satellite missions from NASA and the European Space Agency, the need for high quality, full-year trait measurements continues to grow. The new changes to the protocol make me very excited. It definitely has helped streamline the entire process and from a ground-based perspective we can collect a lot more trees at a wider spatial area, so it definitely improves our overall data quality.