The Effects of Various Commercial Herbicides Applied by Drone

The Effects of Various Commercial Herbicides Applied by Drone

The Effects of Various Commercial Herbicides on a Variety of Nuisance Species When Applied Via a DJI T10 Agricultural Drone by Michael Migliorino




A common pest that requires constant human management is the vast array of weeds and unwanted tree and bush growth that can block views, cover roadways, damage electrical lines, ruin landscaping and more.  In order to combat this issue, civil engineers, farmers, and private individuals alike have tried a variety of methods:hand-picking, using machinery for cutting, weed-barriers, hand-spraying herbicide, and more.  Recent developments in drone technology however have made an additional option possible: drone-spraying herbicide.  


This offers the opportunity to reach hard-to-access areas such as the interiors of very densely weeded fields to the branches over roadways that may be elevated a dozen stories or even more.  In order to test though the effectiveness of using drone spraying on a variety of weeds, we have sprayed a local field using a T10 agricultural drone using a variety of herbicides to both determine which is most effective when sprayed from a drone and whether or not using a drone for spraying herbicides could be an effective pest management strategy for handling the pest of excess vegetation. Method In order to test whether or not the drone could be effective at managing weeds and unwanted vegetation, and also to determine which herbicides are effective at managing these plants when sprayed from a drone, we first marked off equally-sized sections in a parcel infested with a variety of unwanted plants.  Then, we marked each parcel using a flag labeled with different herbicides that were being tested; the herbicides tested were Millennium Ultra, Confront, Sterling Blue, and Triclopyr.  Before spraying these herbicides, we took a visual note of the entire parcel in addition to a 3d map of plant health developed using Pix4DFields software, an industry leader plant health monitoring software, and a Phantom 4 Multispectral drone, a drone capable of utilizing a variety of multispectral sensors to get highly detailedimages of chlorophyll content, photosynthesis efficiency, and much, much more.  Finally, 2 after this a T10 spray drone sprayed each of the herbicides off of a predetermined quarter of the parcel.  The Sterling Blue was sprayed in a solution that was 25% product and 75% water by volume, the Millennium Ultra was sprayed in a solution that was 37.5% product and 62.5% water by volume, the Confront was sprayed in a solution that was 25% product and 75% water by volume, and the Triclopyr was sprayed in a solution that was 37.5% product and 62.5% water by volume.  


All solutions were sprayed in their respective areas at an application rate of 1 gallon of solution per acre and at an altitude of 10 feet above the canopy. One week after the spraying, another drone survey with the multispectral drone was conducted to determine any noticeable changes in plant health. Results Overall, the majority of plants within the parcel, regardless of herbicide applied, appeared to be less healthy based on various indicators; the plants were less green, had fewer leaves3 present, leaves that were present were considerably more shriveled, and woody structures as well as foliage were more brittle- all indicators that less photosynthesis is occurring and thus that the plants are declining in health. This is in contradiction with the plants prior to the spraying, which were, according the multispectral drone surveying(as seen in Appendix A), generally completing photosynthesis and were thus in good health.  Additionally, it is worth noting that in most cases there appeared to be more damage in the upper portion of the plants (the upper crown in the case of the trees). Species specific visual observations are included in Figure 1 below Figure 1: Plant appearances SpeciesAppearance before spray Appearance After Spray Control (from non-sprayed area) SumacGreen, healthy, and flowering (Appendix B 1.1) Almost all leaves are completely red, somewhat drooping. Trees tend to wilt too (Appendix B 1.2I)  With the Millenium Ultra specifically leaved were completely brown and crisp (Appendix B1.2II) Still green, flowering, and healthy (Appendix B 1.3) Eastern RedbudGreen and healthy (Appendix B 2.1) Most leaves on plant have brown spots in addition to a yellow coloration (Appendix B 2.2) iuyiy AshGreen and healthy (more so than most of its species … no Leaves are still green, except for in a smaller number of iuyiy4 signs of ash bore) (Appendix 3.1) black spots.  There does appear to be a form of film over the leaves (Appendix B 3.2) Wintersweet*Green, healthy, full of foliage.  A handful of yellow leaves. (Appendix B 4.1) Decimated.  Almost all foliage is shriveled up and is tan/gray in color(perhaps fewer than 10 leaves on the entire tree were merely discolored and were somewhat recognizable). Much of it has fallen out of the tree.  Very little in the means of foliage.  Remaining shriveled leaves are significantly brittle (Appendix B 4.2) iuyiy BrambleGreen and healthy (Appendix B 5.1) Had some yellow spots and patches on some, but not all, leaves on the plant. Was still producing fruit(Appendix B 5.2) Green and healthy (Appendix B 5.3) Poison IvyGreen, healthy, and abundant (Appendix B 6.1) Red, shriveling, wilting, and low hanging(Appendix B 6.2) Green, with a hint of yellow. No signs of wilting, leaves still held upright. (Appendix B 6.3) GrassGreen, healthy, growing abundantly. Somewhat tall (Appendix B 7.1) Tannish green or grayish green in color.  More signs of brittleness and wilting (Appendix B 7.2) Growing green and clustered.  Is somewhat shorter due to more frequent mowing. (Appendix B 7.3) PineTrimmed back, very thin and tall.  Not much foliage. (Appendix B 8.1) Trimmed back, very thin and tall.  Not much foliage (Appendix B 8.2) Trimmed back, very thin and tall.  Not much foliage (Appendix B 8.3)5 BlackhawGreen and healthy (Appendix B 9.1) Most leaves are 75% red or greater.  Plant is wilting slightly (Appendix B 9.2) iuyiy MapleGreen and healthy (Appendix B 10.1) Leaves are brown and are developing holes.  Health of foliage is considerably worse near the top of the plant than the bottom. (Appendix B 10.2) OakGreen and healthy (Appendix B 11.1) Green and healthy. A few minor brown spots (Appendix B 11.2) MugwortGreen and healthy abundant in cracks of concrete (Appendix B 12.1) 50% of specimens were gray, brittle, and had shriveled away.  Remaining individuals had few signs of damage.  It is not clear whether or not these specimens were present at the time of spraying or if they had only grown in later. (Appendix B 12.2) iuyiy Wild CarrotGreen, healthy, flowering (Appendix 13.1) Crisped and withered. (Appendix 13.2) iuyiy *The proper identification of these specimens was difficult due to the poor quality of remaining specimens after the spraying Figure 2: Comparison of Herbicides Sprayed6 HerbicideResult Sterling BlueCaused wilting and discoloration in the majority of plants that it came in contact with. Did not do any very major damage (like the Millennium Ultra) to any specific specie, and did not damage all species (like the Triclopyr did) Millenium UltraA few individual species, such as maple, were completely spared of all of but the most minor damage (a handful of tiny brown spots for example on the maple), while other species, such as sumac, were decimated far worse than any other individuals of their species were under other herbicides (for example in the case of the sumac the Triclopyr made the Sumac trees in contact with it have foliage turn red and wilt while the Millenium Ultra had all foliage completely shrivel up, turn brown, and become brittle on sumacs that it came in contact with) TriclopyrHad a very wide range of effect.  Almost every type of plant had some damage whether it was brown spots, foliage loss/discoloration, withering, or more. However, there were also some individuals that were barely affected, even though other individuals of their species may have suffered extreme damage.  It is worth noting that many of the times when this happened the individuals that did suffer the most damage were more exposed, while the individuals that were spared from damage tended to be located in relatively thick brush. ConfrontMainly only damaged a few select plants by giving brown spots. Analysis7 After reviewing the results, it can be concluded that for various species the effects of the herbicides are as displayed in Figure 3 below. Figure 3: Efficacy On A Variety of Species SpeciesOverall Noted EffectHow Effective (1-10)? SumacTriclopyr turns the foliage completely red and caused wilting.  Millenium Ultra resulted in complete shriveling, browning, and brittleness of foliage. 9 (Triclopyr) 11+(Millenium Ultra) Eastern RedbudResults in a large number of brown spots in addition to a significant yellow coloration. 7.5 AshProduces a couple of black spots within the leaves. 4 WintersweetCapable of defoliating these plants, and causing for the little remaining foliage to shrivel and become brittle. 10 BrambleMakes the foliage get yellow spots and patches. 4.5 Poison IvyCauses for the leaves to turn distinctly red and to shrivel back. 8 GrassTurns grass grayish-green and tannish-green.  Not capable of fully killing it, but can reduce its growth. 5 PineThese herbicides are ineffective against the pine trees present on the plot. 1 BlackhawWill cause for the leaves to lose almost all greenness in favor of a red complection. 7 MapleResults in brownish spots and even some holes to form on the leaves, especially near the top of the crown. 4 OakCauses for a few minor brown spots to form on leaves in the extremities of the crown. 2Mugwort 6.5 8 Will cause most individuals to shrivel up and become brittle, especially near the base.  Will sometimes allow for the top part of the plant to continue to grow, and in rare cases some individuals will be unaffected. Wild Carrot Herbicide will cause for complete shriveling, becoming brittle, and a termination of growth 9 While more research would likely have to be done to explain why the damage done by the herbicides was more severe in the upper-crown than elsewhere on the plant, this can likely be explained by two major causes.  First, the upper-crown likely had more exposure to the herbicides being sprayed than the rest of the plant, since by being up and at the top the foliage of the upper crown helped to block some of the lower foliage from the chemicals, and as a result likely absorbed more product than the other leaves on the plant.  Second, the extremities of a plant are typically the first to die off when a plant is experiencing stress, and therefore the plant likely would have stopped supporting the foliage of the upper crown, which is perhaps the farthest out extremity on the plant, first when the herbicide starting to take root in its system and cause for it to receive damage.  In either case, supplying more product and in higher concentrations would likely help ensure that the damage is more uniform by ensuring that the lower foliage also gets a fatal dosage just as the top does. Additionally, the result that, especially in areas sprayed with Triclopyr, that some individuals of certain species were affected while others (typically in areas of denser foliage) weren’t likely has a root cause similar to the first reason why the crowns of other larger specimens were the most affected. In areas that are densely populated by weeds, it is more likely that some may shield others from the majority of the herbicide.  Additionally, since these areas have more biomass per unit of land, and the amount of herbicide sprayed from the drone9 remained the same per unit of acre regardless of the biomass, it is possible that in these dense areas the larger amount of plant material was able to absorb the herbicide without dying since each individual unit of biomass had less product that it had to absorb.  In either case, spraying more herbicide and in higher concentrations will likely reduce these effects and result in a more complete kill-off of the weeds. Lastly, the observation that Millennium Ultra tends to be very effective against some species but not have much of an effect on other species is useful in that it suggests that the Millennium Ultra may be useful for selectively killing certain unwanted species.  For example, in gardening there are often unwanted species growing in-between the more aesthetically pleasing species that were planted there on purpose.  Therefore, if it is found that the plants being grown on purpose are immune to the Millennium Ultra while the weeds are not, it can very effectively allow for the control of the garden by killing the weeds while leaving the wanted plants with a single spray. Implications The results of this study are useful for determining the best courses of action for limiting the growth of unwanted plant species in a variety of situations. Specifically, this study has exposed which species are most susceptible to herbicides being sprayed from a drone.  Therefore, when agronomists in the future are trying to manage the growth of certain species they can know ahead of time whether it would be worth it for them to use herbicides sprayed from drones for pest control or rather another method.  This information is especially important if agronomists were looking to only target one or two species in a location while sparring others.  Economic sectors that would be impacted by this10 include forest management, landscape designing, agricultural works, civil planning, energy, construction, and more. By using this data and being able to potentially utilize a drone, which can target very specific and oftentimes hard to access areas remotely, for pest species management significant increases in efficiency can be achieved, thus saving resources for the entity managing the plants to put towards other productive means.  This is also good news for the environment, since such targeted measures means that less harmful product will be released into the environment overall. Limitations The major limitation of this study is the lack of uniformity in the plot of land used.  Each section had different species of plants, different plant densities, varying amounts of concrete cover, and varying canopy sizes.  This means that it is difficult to truly determine which herbicide performed the greatest since it is possible that some of these extraneous variables may have influenced the plant growth, thus making it more difficult to assign exactly what the exact effect of each herbicide was in comparison with each other. Additionally, the size of the plots sprayed were quite small, all being under 5% of an acre. This means that the sampling size was quite small, and so even small outliers that otherwise would be unnoticed in a larger study may seriously affect the results of this study.  For example say a single tree of a certain species in our plot died for some unknown other reason it would significantly skew our data towards showing that the herbicide sprayed on it is more effective than it really is.Lastly, our study was constrained by the fact that in some cases there was no control group. 11 While some species, such sumacs, did have other individuals on the property outside of the spray zone that have been used as a control group (thus allowing us to prove that our spraying did effectively restrict their growth since the sumacs that we sprayed turned red or died while other non-sprayed specimens on the property were still healthy afterwords), other species did not, and so we can not conclusively prove that it was our spraying that affected them and not the ambient environment. Future Studies Future studies that can be done to improve our knowledge on the impacts of these herbicides on nuisance species when sprayed from a T10 drone include those that are more standardized. One such study could have a large field that is uniformly covered with a specific nuisance species, and equal parts of it can be sprayed with each herbicide (with one equal part left unsprayed to use as a control).  In such a scenario, uniform soil conditions as well as uniform watering could be done to ensure that there are fewer confounding variables to interfere with results. Additionally, studies similar to the one detailed above can be conducted using various concentrations of herbicide as the independent variables rather than various herbicides.  This can help determine what the most effective and efficient concentration is of selected herbicides, thus saving time, money, and product. Conclusion To conclude, the species most susceptible to herbicides sprayed from drones, and thus the most likely candidates for being controlled by drone spraying, are sumac, eastern redbud,wintersweet, poison ivy, blackhaw, and mugwort.  Meanwhile, species such as grass, 12 bramble, pine, maple, oak and ash may have to be controlled using other methods if they are found to be unwanted in a particular location. Additionally, this study has found that the most effective herbicide to spray from a drone for the sake of controlling weeds in general is Triclopyr. Meanwhile, Millenium Ultra would be more effective if the goal is to control certain species especially susceptible to it such as sumac, or if the setting requires that only certain species, notably species that are affected by it, are to be killed in an area with wanted species that are immune to it. On the contrary, this study found Confront was not effective at controlling the unwanted species, and that if one did need to control pest species they should select a different herbicide.Appendix A

Back to blog