Urban Tree Mortality: The Purposes and Methods for (Secretly) Killing Trees Suggested in Online How-To Videos and Their Diagnoses

Introduction

It is commonly asserted that everyone loves trees (Braverman 2008). However, despite their many benefits, including reducing urban heat (Ziter et al. 2019), stormwater runoff (Berland et al. 2017), and air pollution (Nowak et al. 2006), trees also provide ecosystem disservices. Ecosystem disservices are the problematic environmental features and processes that run counter to human values and preferences (Lyytimäki and Sipilä 2009). In urban forestry, disservices can include smelly flowers, messy fruit, leaf litter, vegetation that interrupts sightlines, and infrastructure damage caused by growing roots and branches and failing trees and limbs (Klein et al. 2019; Roman et al. 2021). Several disservices arise from sudden, one-time events such as hurricanes or ice storms (Conway and Jalali 2017), including tree failure, which can damage infrastructure or result in injury or death (Klein et al. 2019). Other disservices occur seasonally or over the course of years. The latter often involves predictable, natural functions of trees (Roman et al. 2021). For example, the production of seeds and fruit can cause litter accumulation, autumn leaf fall can be a disservice as leaves require raking, and tall-growing trees can block views (Martin and Doucet 2022). Because of these natural processes and resulting disservices, some develop negative perceptions towards urban trees (Schroeder et al. 2006; Camacho-Cervantes et al. 2014; Fernandes et al. 2019).

On private properties, ecosystem disservices are a driver in tree removals (Conway 2016; Guo et al. 2019). In communities without private tree protection bylaws or ordinances, people can remove their trees at their own discretion. To prevent rampant tree loss, many communities have adopted regulatory mechanisms that prevent people from removing trees on their property (Ordóñez-Barona et al. 2021). However, there are often caveats in these regulations that treat dead, dying, or hazardous trees differently, reducing the permitting process or not requiring a tree be replanted (Conway and Lue 2018). A homeowner can therefore secretly kill their own tree to reduce time and cost associated with tree removal.

Disservices are also produced by publicly owned trees, resulting in some city residents expressing discontent about municipally owned trees (Delshammar et al. 2015). Similarly, disservices can arise from a tree on a neighbouring property, resulting in conflict between neighbours (Olivero-Lora et al. 2020). There have been several cases of tree poisoning or removals where people have trespassed to kill an unwanted tree (Dunster 2018; Stoffel et al. 2020). For example, in British Columbia, Canada, the defendants drilled holes in the claimants’ tree and applied an unknown toxic substance, causing the tree to die (Bowbrick v. Jakob 2017). The tree had previously shaded the defendants’ pool area. Also in British Columbia, a defendant admitted to poisoning 3 street trees to reclaim her partial ocean view (R v. Matheson 2006). In Texas, USA, the defendant killed a cypress tree by dropping a brown liquid into holes that were drilled into the roots (Withrow v. Armstrong 2006). In Oregon, USA, a sycamore was killed by a growth regulator herbicide applied to the root system (Brown v. Johnston 1971). In many such cases, the parties and the courts look to arborists as expert witnesses to identify the cause of death and, in the case of poisoning, the chemical agent used. Such investigations are important to the Courts, as intentionally poisoning trees or removing trees during willful or intentional trespass are subject to heavier liability and punitive damages (Dunster 2018).

People interested in killing nuisance trees are likely to use online resources to identify suitable methods. Online how-to videos are a common educational resource for procedural learning (Utz and Wolfers 2022). Of the online video-sharing services, YouTube is the largest and most visited (Snelson 2011), which provides a public forum that supports the sharing of unverified content (Osman et al. 2022). Unlike many websites, YouTube also records the date of publishing and the number of views.

Several research methodologies have been developed to utilize YouTube videos in review studies, predominantly for healthcare and medical studies (Sampson et al. 2013; Madathil et al. 2015; Osman et al. 2022) as well as research on environmental topics, including climate change (Shapiro and Park 2018; Duran-Becerra et al. 2020), climate engineering (Allgaier 2019), bird watching (Kikuchi et al. 2022), wildlife management (McLean et al. 2022), and wildlife conservation (Freund et al. 2021; Vins et al. 2022). By reviewing YouTube videos, we can address the use of social media and videos to discuss issues related to trees, a previously underexamined source in arboriculture and urban forestry (Martin and Doucet 2022). It is therefore an important area of preliminary examination for identifying tree-killing techniques that the public may discover.

To support practitioners in their diagnostics of these most nefarious abiotic disorders and disturbances, we conducted a review of online videos that explain how to (secretly) kill a tree. Our study examines 3 research questions: (1) What are the stated reasons the treekilling methods could be used? (2) What are the most frequently recommended methods for killing a tree? (3) How can the proposed methods be detected? We explore these research questions through inductive coding of videos on YouTube (Alphabet Inc., Mountain View, CA, USA). This study is helpful for arborists trying to diagnose tree-killing techniques as part of forensic investigations. This study also provides the foundation for further studies on diagnostic methods for these tree-killing techniques.

Methods

The video search was performed on 2023 December 15. Similar to literature reviews, video review methods use search strings. We conducted 5 searches with different strings of terms to identify videos associated with killing trees. The search terms were reviewed and approved by a team of 12 arborists and urban foresters. The 5 search strings were: ‘secret kill tree’; ‘secret poison tree’; ‘kill neighbor’s tree’; ‘poison neighbor’s tree’; and ‘how to kill a tree’. YouTube shows the results that include the words in the search string, regardless of word ordering in the phrase (Granwehr 2021).

The first 100 results for each search were selected, resulting in a total list of 500 videos, representing the videos that YouTube’s search engine deemed the most relevant to our search. Because YouTube sorts videos according to relevance as the default search optimization (Granwehr 2021), the selected videos represent the videos that a viewer would likely find when conducting the same search, although the ordering of results may differ between users based on their search history. Including the first 100 results helps ensure that the reviewed videos best reflect the treekilling techniques viewed by YouTube users. Non-English and duplicate videos were removed.

The results of the search were saved to a spreadsheet (Microsoft Excel; Redmond, WA, USA) that included the video title, URL, account, posting date, views, likes, comments, and channel subscribers. The videos were then screened to ensure that they were related to methods for killing trees. Videos about how to remove stumps without a stump grinder were removed, but videos about how to kill living stumps, especially those that are sprouting, were retained.

After all videos were assessed, we retained 79 videos for coding and analysis and created transcripts of the videos so that we could code textual data. After the transcripts were verified, we then used a mix of deductive and inductive coding. Deductive coding is a content analysis method that uses predefined codes (Calvo and de la Cova 2023). By contrast, codes used in inductive coding are developed after repeat readings of the data (Corbin and Strauss 1990). We used deductive coding to identify themes based on our research questions. For each video, we identified text that addressed the 4 themes: (1) Purpose—the suggested reason for employing the method; (2) Secrecy— whether the purpose is intended to be secretive; (3) Legality—whether the video mentions or warns against illegal usage; and (4) Method—the type of method used to kill the tree.

Following the deductive coding, we read the transcripts several times to become familiar with the purposes and methods discussed, the first step in inductive coding per Thomas (2006) and Chandra and Shang (2019). After becoming familiarized with the transcripts, we developed potential codes to identify the many purposes and methods discussed in the videos. We then went through the transcripts and labeled the codes. For the Purpose theme, we identified both upper- and lower-level codes. Upper-level codes were the broad purposes, either to kill nuisance trees or ecological reasons. Videos about killing nuisance trees were ones that concerned how to kill a tree that is causing harm or inconvenience to people or property. Videos with an ecological purpose included removing trees to benefit or restore ecosystems, which may support either flora or fauna. We then assigned lower-level or subcodes to provide greater specificity about these 2 purposes. For the Method themes, the codes identify the methods that were used and, in the case of herbicide, the active ingredient that is recommended. We finished the inductive coding by removing any overlapping or redundant codes.

Lastly, we employed qualitative content analysis per Hsieh and Shannon (2005) to identify the number of occurrences and public engagement (e.g., likes, comments) associated with different purposes and methods. In the Discussion, we provide diagnostic details for each method to provide arborists with the tools to identify these disturbances in urban trees.

Results

Methods for Tree Killing

There were 17 different methods proposed across the 79 videos (Figure 1). On average, the use of rubbing alcohol, copper nails or other metals, and flush cutting had the highest views. While most videos only suggested one method, 22 videos suggested multiple methods for killing a tree. The majority of videos used herbicide (n = 50), which was also the most viewed method based on total views (Figure 2). The second most commonly discussed method was applying salt around the tree (n = 12), yet it had fewer average views, likes, and comments than less commonly described methods. The most liked method (fewest number of views per like) was girdling, followed by copper nails or other metals. The most commented-on method (fewest number of views per comment) was copper nails or other metals, followed by girdling.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Average number of views, likes, and comments on reviewed YouTube videos about tree-killing methods. Note that the count of methods exceeds the number of videos because 22 videos suggested multiple methods.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Total number of views, likes, and comments on reviewed YouTube videos about tree-killing methods. The numbers above the likes and comments indicate the number of views per like or comment. Note that the count of methods exceeds the number of videos because 22 videos suggested multiple methods.

Herbicide was recommended in 50 of the videos, with 5 videos recommending multiple (alternative) herbicides. The most commonly recommended herbicide was glyphosate (n = 17) followed by triclopyr (n = 11) and picloram, sold as Tordon (n = 8) (Corteva Agriscience, Indianapolis, IN, USA). Also recommended, although to a lesser extent, were 2,4-Dichlorophenoxyacetic acid (n = 3), Imazapyr (n = 3), copper sulfate pentahydrate (n = 2), and copper sulfate (n = 1). Ten videos did not report the type of herbicide used. While 5 videos recommended multiple (alternative) herbicides, only one video recommended mixing herbicides, recommending 2,4-D and triclopyr in one tank.

Discussion

Our review of tree-killing methods reveals that many of the available video results discuss broad purposes related to nuisance tree killing and ecosystem management, even when searching for secret killing or poisoning methods. Consequently, several of the methods are obvious and nondiscreet ways of killing trees. This also explains why no video warned of any legal ramifications from killing trees.

Nonetheless, the most popular method—herbicide— is a discreet option and one that has appeared in court cases (Dunster 2018; Stoffel et al. 2020). Herbicides such as glyphosate and triclopyr are commonly available and have proven efficacy. This contrasts with other methods including copper nails, boiling water, and rubbing alcohol that have little to no support in the literature. Of the various discreet methods, herbicide, salt, fertilizer, and fuel and oil are likely to be the most reasonable and most effective based on their known impacts to trees (Costello et al. 2003). There are, however, methods of detection for these 4 issues.

Salt as a tree-killing method involves applying large amounts of salt to the soil. This can be detected using soil electrical conductivity (EC). Following the application of salt, the saturated soil paste will have a high EC value, indicating a high salt concentration (Brady and Weil 2008). There are also visual symptoms. Root-based absorption of salt results in growth stunting and yellowing foliage, becoming necrotic before defoliation (Costello et al. 2003).

Fertilizer burn is also a common abiotic disorder of urban trees, although it is often presented as a consequence of misapplication of fertilizer rather than malapplication of fertilizer (Lilly et al. 2022). The application of fertilizer can be seen as patches of rich green or dead grass where there is high—even toxic—amounts of nutrients. The fertilizer can be detected using nutrient soil tests (Brady and Weil 2008).

Fuel and oil as tree-killing methods involves soil drenching with, commonly, gasoline or diesel. There are several test kits available to determine the presence of petroleum hydrocarbons (Environmental Protection Agency 2015). There is little published literature on fuel and oil, however. The expansion of pipelines renewed some interest in the effects of oil spills on trees, finding that most vegetation died with deciduous species experiencing more rapid injury than evergreen species (Jenkins et al. 1978). A study of a jet fuel spill found that mangrove vegetation experienced widespread and rapid damage, despite little residual contamination to the water and sediments (Ballou and Lewis 1989). Further research is needed, however, to provide remediation options for urban trees.

Herbicides are the most complex issue. Herbicide symptoms vary by active ingredient and species and are influenced by precipitation (Costello et al. 2003). There are also several lookalike symptoms that prevent conclusive diagnostics. Symptoms can therefore be described as being consistent with herbicide damage, but laboratory testing is required to definitively confirm the presence of herbicide. However, some application methods result in long-term signs. The “hack and squirt” method of using a small hatchet to make an incision in the trunk and spraying herbicide into the incision will result in an obvious cut into the bark. Similarly, drilling holes into the tree to inject herbicide will result in holes in the tree. Even though the chemical agent may not be apparent, the method of application would be.

The detection of chemical herbicide agents is ubiquitous in its approach, using some form of chemical residue testing. There are many established methods for the common agents. Testing for glyphosate, which was the most commonly recommended herbicide, uses high-performance liquid chromatography (HPLC) to test for glyphosate parent ions and aminomethylphosphonic acid (AMPA) ions (Botten et al. 2021). Similarly, triclopyr is an auxin type systemic herbicide that regulates plant growth sold under the trade names Garlon, Remedy, and Turflon (all produced by Corteva Agriscience, Indianapolis, IN, USA). Tests for triclopyr use the acid form, although tests can also be run based on the specific product used. Tests for Garlon 3A^® test for triclopyr amine, tests for Garlon 4^® and Pathfinder II^® (Corteva Agriscience, Indianapolis, IN, USA) test for triclopyr butoxyethyl ester (Durkin 2011). A version of the QuEChERS (“quick, easy, cheap, effective, rugged, and safe”) solid phase extraction method can be used to detect triclopyr using HPLC-MS/MS (European Food Safety Authority et al. 2022), although further research is needed on detection for trees.

The remaining methods are more self-evident. Girdling, which is the removal of bark, phloem, and cambium from the trunk of the tree—either in whole or in part—is readily visible as a ring of missing bark. The detection of flush cutting, which is cutting the tree or shoot at or below grade, is self-evident. Similarly, the causal agent is obvious with copper nails hammered into the tree, fire set to the trunk, or black plastic covering the foliage or soil. Fungal inoculation also involved drilling a hole into the tree, which leaves a visible hole in the tree.

Conclusion

While trees provide many benefits in urban environments, they also provide disservices. There have been many cases of people trying to kill trees that are legislatively protected or privately owned. To assist arborists in detecting these methods, we used YouTube videos to examine what methods are suggested to the public. We highlight the most common, inconspicuous, and likely most effective options: herbicide, salt, fertilizer, and fuel and oil. Forensic arboriculture and expert witness testimony relies on objective fact, and this review provides a starting point for investigations into suspect tree mortalities.

Conflicts of Interest

The authors reported no conflicts of interest.