I posted the following tweet recently, which sparked some discussion. The image on the left shows a glyphosate-treated tree pit, where the plants have been sprayed and have subsequently died back, leaving a distinctive orange-brown carpet of dead growth. The tree pit on the right, seen in South London hosts a range of small plants such as chickweed (Stellaria media) and meadow grass (Poa annua).
Sprayed tree pits are not an uncommon sight in the UK. This European survey carried out between 1999 and 2001 showed that contrary to many countries which prefer mechanical removal, the UK controls weeds mainly via chemical methods, and usually with the chemical glyphosate (famous for being a component of Roundup). Local authorities and highway management companies spend large amounts of money and get contractors to clear tree pits of plants. In fact, it has become a habit, and is usually part of the maintenance regime for streets and other public spaces. Weeding is also often requested by residents, as here in Bromley, South London: https://fix.bromley.gov.uk/report/523245
On this online reporting tool, a resident is complaining about the fact that two tree pits in their street have had “significant weed growth” “for weeks“. The council replies that “there are three applications of weed spray, in spring, summer and early autumn”, and that “Once sprayed ‘dieback’ should occur within seven to ten days“, then “the residue” is removed by street cleaning. They inform him that the next weed spraying programme “will take about 4/5 weeks to cover the whole borough“. They also mention that “Tree pits will not be sprayed if the trees are 100mm or less thick”, in which case the weeds have to be removed manually.
This is where it becomes interesting. This guideline suggests that glyphosate would be more harmful to young trees than to more established ones. In other countries, where glyphosate has been banned in public spaces (in France for example), tree pits, including those with young trees, are filling up with “weeds”, as manual removal is also limited. Here in Germany:
In Switzerland, in several French cities too, projects that allow local residents to “adopt” tree pits are flourishing. The city of Basel for example offers a 3-year permit for residents to plant in a tree pit, with a list of suggested plants that are adapted to this environment and that provide biodiversity benefits.
But what does science say in that respect? Do tree pits have to be kept free of plants? Does glyphosate harm young trees more than old ones?
One of the early (1987) publication on the topic, by the Forestry Commission, entitled “Trees & Weeds” points out several effects of weeds on trees:
- competition for water: weeds dry the soil by extracting moisture, which decreases its availability for trees and reduces their growth rate
- competition for nutrients: weeds take up nutrients from the soil, which reduces the availability for trees (especially younger ones). Legumes such as clover (which fix nitrogen from the air) are thought to be particularly detrimental to young trees.
- allelopathy: weed roots release compounds into the soil that could be toxic for trees
- soil and air temperature: weeds insulate the soil, making it cooler in summer and warmer in winter.
This guide does contain some useful evidence, but is also full of assumptions because there was very little research on the topic at that time. For example:
“Foresters sometimes assert that ‘weeds’ are beneficial in dry sunny weather by providing a humid micro-climate for the tree and protecting it from desiccation. This assertion is perhaps true if the only choice is whether or not to mow the weeds, but the benefits of shelter are unlikely to outweigh root interference.“
“over a whole year it is unclear whether the soil temperatures under bare soil or weeds are most favourable for root growth. Bare soil is probably best”
“all broadleaved species tested on landscape-tree planting sites have grown much better with effective weeding. The evidence for conifers is less complete, but there is no reason to believe they would behave very differently under the same conditions”
The effects mentioned above were studied in a forestry context, on open soils, rather than in an urban environment. Many factors can also influence the impact of weeds:
- plant species: woody plants and small annuals are less harmful than grasses or nitrogen-fixing plants such as clover
- soil type: sparse weed cover on a sandy soil is more harmful than lush weed growth on loam (because loam retains more moisture and nutrients)
- quality of planting: a poorly planted tree with superficial roots will be more impacted
- tree species: fast-growing tree species appear to be harmed more by the presence of weeds when they are young than slow-growing species such as oak which are impacted at low level throughout their lifespan. Tree species with deeper roots may also be less affected (this is because water and nutrient competition from weeds often happens in the superficial layer of soil).
Interestingly, the guide asks this question: Is it possible to find ground covers, wild flowers or non-competitive grasses perhaps, that will not harm young trees? It points out the risks that ground cover plants may become invaded with more competitive harmful weeds, and concludes that underplanted trees are “unlikely to fare better than if they were planted into bare soil.” Herbicide-resistant mosses are quoted as a potential option for underplanting, but they are “not generally considered attractive enough”.
Looking at this guide, it appears that tree bases should be kept weed-free at all costs. So are European cities mad to let plants grow in their tree pits?
A study undertaken in Liverpool showed that competition from weeds is indeed responsible for the death of young trees, but as a secondary cause rather a direct effect (drought stress is responsible for most tree deaths – if appropriate irrigation was used, or if the tree had been planted in a moisture-retaining soil and mulched, the negative effect of weeds may not have been as dramatic). In addition, a more recent study in New York showed that tree stewardship has a significant impact on tree survival: trees that had been underplanted with flowers by residents had a higher survival rate, in spite of potential negative effects caused by ornamental planting.
The Urban Tree Manual (2018) suggests effective weed control “during the first three to five years of the tree’s life” to guarantee tree establishment, using “chemical control, plastic mulch mats or wood/bark chippings“. But have the impacts of glyphosate on trees, particularly when correlated with other factors such as drought stress been measured?
There is in fact very little scientific literature regarding the impact of glyphosate on street trees. Most papers look at the effects of glyphosate on commercial tree production (fruit trees, timber, paper wood, Christmas trees etc). Here are some of the effects listed:
- Bark cracking: additives in glyphosate formulas can cause cracking in the tree bark, which reduces winter hardiness
- Leaf damage: chlorosis – yellowing due to a lack of chlorophyll -, mottling, leaf curling and necrosis (leaf death) if glyphosate is applied to the leaves or exposed roots through wind drift or spraying mistakes
- Growth problems in young trees: when applied to young trees, glyphosate penetrates through the bark and accumulates in the phloem (plant tissue that transport sugars in the tree). It can take years (8-10 years) to break down. In autumn, it is carried to the roots. When the sap rises again in spring, glyphosate is transported to young buds and causes abnormal development such as witches brooms (a plant deformity where dense mass of shoots grows from a single point, with the resulting structure resembling a broom or a bird’s nest) or stunted growth.
- Fruit drop: glyphosate increases the production of ethylene, the plant hormone that promotes fruit ripening. This can lead to fruit bruising and fruit drop.
- Increase in pathogenic soil fungi: soil fungi such as Fusarium or Phytophtora ramorum (which causes Sudden Oak Disease) are stimulated by glyphosate application. Glyphosate is also thought to cause basal trunk cankers (for example in apple trees).
- Decrease in nutrient uptake: glyphosate decreases the the availability and the uptake of micronutrients (manganese, zinc, copper, iron) which are essential for plant growth and resistance to diseases. This effect is particularly strong in acid soils.
- Decrease in disease resistance: a decrease in the availability of micronutrients means that plants struggle to produce anti-microbial and other anti-pathogen compounds. They are also less able to produce the hormones involved in wound cicatrisation (to wall off pathogens).
The action of glyphosate is based on a seven-step metabolic process called the Shikimate pathway. The pathway leads to the formation of amino acids (phenylalanine, tyrosine, and tryptophan), which are the building blocks of many plant compounds: from the growth hormone auxin, to leaf and flower pigments, nitrogen fixing compounds, or flavonoids that help plants defend themselves against diseases. Glyphosate blocks the enzyme at step 6 of the pathway, which prevents the amino acids from being formed. This kills the plant.
What is more interesting is that the Shikimate pathway is found in other kingdoms, such as Bacteria, Archaea, Protozoa and Fungi. It is therefore reasonable to assume that glyphosate has an impact on soil microorganisms.
More recent studies, such as this one by Hagner et al. (2019) acknowledge the negative effects of glyphosate on soil mycorrhizae, bacteria and earthworms that have been reported in scientific literature. But they compare it to the effect of hoeing, a common manual alternative. In their experiment, they found that hoeing has drastic effects on soil fauna, for example a decrease in the abundance of nematode worms, and a lower litter mass loss (rate of organic matter decomposition). The difference in the impacts of hoeing compared to glyphosate application + hoeing was however small. They suggest that glyphosate may not have major effects on the structure and functioning of soil decomposer communities.
Glyphosate is generally regarded as safe as it is degraded rapidly by soil microorganisms. Recent studies have however shown that the rate of degradation can vary enormously with the soil type, climatic conditions, phosphorus content.
What is not disputed is that killing live plants, whatever the method, does have effects on:
- the abundance of soil fauna (weeding decreases live plant and increases dead plant resources)
- litter mass loss (weeding means the litter is drier, and therefore less favourable to microbial activity)
- soil mineral nitrogen availability (less intake by live plants)
Going back to our problematic of tree pits, what do these results teach us? I have not been able to find any literature on the impact of glyphosate on street trees (but please contact me if you have information on this). As we have seen previously, the impact of weeding vs non-weeding, manual vs chemical removal can vary with local conditions. In addition, tree pits present unique challenges, from isolation, to drought stress, lack of soil fauna, disturbance (trampling, urine) etc. which could further influence the impacts of weeding.
Glyphosate has strong effects on plant growth, and it appears therefore sensible to avoid spraying young trees with thin bark, to minimise these risks (particularly as glyphosate can persist for years in plant tissues). But given its significant effects on soil and surrounding plants, one can wonder about the necessity to use glyphosate at all on urban trees. Alternative methods such as mulching can provide additional benefits for young trees, while underplanting may have negligible negative impacts on established trees. More importantly, scientific evidence and informal observations tend to show the importance of proper tree stewardship, whether by local residents or authorities.
In urban areas with parks deficit, tree pits can provide unique green spaces in front of people’s homes. Whether they are filled with interesting spontaneous flora, or used to grow flowers, I believe that they should be celebrated, not seen as a part of the pavement that needs to be cleaned.