Japanese knotweed (Fallopia japonica): an analysis of capacity to cause structural damage (compared to other plants) and typical rhizome extension

Study species: Fallopia japonica

Fallopia japonica is a tall, vigorous, clump-forming, herbaceous perennial, which grows up to two to three m in height (Fig. 2A) and often forms dense thickets. The stems are robust, bamboo-like, slightly fleshy and hollow, with a diameter of up to four cm. Tall-brown to bronze canes remain over winter and persist for approximately 3 years. Leaves are 10–15 cm long, lush, light green and shield-shaped with a flattened base (Fig. 2B). Growth over successive years builds up a sturdy dense crown at the base of canes (Fig. 2C). New growth primarily emerges from crowns at the start of the growth season, but also directly from rhizomes. Rhizomes are initially white, extremely fleshy and fragile while extending (Fig. 2D), but mature into yellow/orange sturdier woody structures (Fig. 2D). The majority of rhizome is found in the upper 50 cm of soil, but it can penetrate down to three m and, depending on soil type and site features, spread up to 10 m from parent plants is possible under very rare circumstances (Booy, Wade & Roy, 2015). Only female (male sterile) plants are known to be present in the UK, which form drooping grape-like clusters of flowers with distinct stigmas. Seeds are shiny, triangular, dark brown, three to four-mm long, two-mm wide and sterile in the UK. See Booy, Wade & Roy (2015) for additional information on the biology of the species. F. japonica can regenerate from rhizome fragments weighing as little as 0.7 g (Brock & Wade, 1992), providing a node is present, and from stem sections, where suitable conditions are present (very moist, well-lit soils with high nutrient availability). The species is dispersed effectively in transported soil and by water (Environment Agency, 2013; Booy, Wade & Roy, 2015). F. japonica is tolerant of a wide range of habitat and soil types, but is most frequently found in disturbed urban habitats, particularly brownfield sites, railway verges and the banks of waterways, where it thrives in damp soils.

Fallopia japonica is closely related to two other members of the Fallopia genus, F. sachalinensis and Fallopia x bohemica, which have similar invasive ranges and have similar impacts. Of note, in some parts of its invasive range, Fallopia x bohemica spreads via the production of large numbers of wind-dispersed viable seeds that germinate at rates approaching 100% in some populations (Gillies, Clements & Grenz, 2016). However, spread by this means does not currently occur in the UK.

Literature assessment

In order to contextualise impacts associated with F. japonica within the larger subject of the capacity of plants that cause structural damage, this study assessed various guidance documents and papers published on the topic of plants causing damage and the relationship between various plant traits and capacity to cause damage. The primary points of interest from these documents are highlighted in ‘Plants and structural damage’. Additionally, a focused literature search on Web of Science was conducted on 27th June 2017 to identify academic papers that provide reference to or evidence of F. japonica-mediated damage to structures. The search terms used for the Web of Science search were ‘F. japonica’ and ‘Polygonum cuspidatum’, an old name for the same species, and within the returned publications ‘damage’. The abstracts were reviewed to determine what type of damage was referred to within the paper.

Fallopia japonica impact survey

A survey of F. japonica management contractors (PCA) and property surveyors (RICS) was conducted to collect evidence either for or against the assertion that F. japonica is a major cause of structural damage to properties. Survey forms were sent out to contractors and surveyors to determine, based on their last field observation of F. japonica, the presence, if any, of damage linked to the presence of the plant across a range of built structure types (see Table 1 for included questions; see Supplemental Information S1 for individual responses). In total, 51 PCA members and 71 RICS surveyors provided records relating to 122 properties (Table 1). Each respondent was also asked how far the closest evident aboveground F. japonica plant was from the residential building on the site that they had visited. This was cross-referenced against reports of damage (Table 2). Yes/No responses are presented as raw numbers and converted to percentage values and differences between PCA and RICS respondents were considered. Statistical analyses were undertaken in PAST version 3.15 (Hammer, Harper & Ryan, 2001).

Fallopia japonica rhizome extent survey

The survey of PCA contractors also asked respondents to provide details, based on the last five F. japonica excavation-based remediation works that they had conducted, on the above ground area of F. japonica and to provide the horizontal (i.e. distance from visible above ground plants) and vertical (i.e. distance from soil surface) extent of rhizomes encountered. In total, 26 contractors provided records of 81 excavations with sufficient detail (e.g. clear rhizome extent linked to an identified individual stand) to be included in the assessment. Eight records were removed due to reporting multiple stands, partial excavation or disturbed sites where it was not possible to accurately determine the rhizome extent from an individual stand (see Supplemental Information S1). Subsequently, stands were sub-classified into either ‘small’ or ‘large’ categories. The small category included any plants that covered a soil area of four m² or less, aimed at encompassing the typical size of stands found in small residential gardens. Stands covering an area greater than this were placed into the large category. This allowed for an examination of the relationship between above-ground area and rhizome extension, as well as an analysis of typical rhizome extension. Data were tested for normality (Anderson–Darling test) and differences between stand categories (large or small) were tested using the Mann–Whitney U test for non-normally distributed data. Data analyses were conducted using PAST version 3.15 (Hammer, Harper & Ryan, 2001).

Case study

A survey was conducted on 68 residential properties located on three streets in northern England. The houses on all three streets were built prior to 1900 (Consumer Data Research Centre (CDRC), 2018). All properties have been abandoned for at least 10 years and were in a state of disrepair, with most having cracked patios and crumbling brickwork (particularly on boundary walls). F. japonica was previously known to be present on properties located on all three streets. An assessment was carried out in September 2017 to determine any constraints that the species might pose to restoration and re-development (see Supplementary Information S2 for details). These sites represented a close to ‘worst case’ scenario in terms of susceptibility to damage from unchecked plant growth. With this in mind, a survey was conducted to determine presence and associated damage for F. japonica, trees, woody shrubs and woody climbers. All damage was compared against a baseline of existing damage that was present due to neglect, weathering and wear and tear over the lifetime of the properties, regardless of plant presence. Where plants were associated with damage to a structure, the damage was quantified based on the scale presented in Table 3 (see also Supplemental Information S2). Figure 3 presents examples of the rating scale that was applied.

By chance, a large number of Buddleja davidii (buddleia) plants were present at the case study sites. As such, this species was included in the assessment separately from other woody plants. B. davidii is a non-native woody shrub that is known to be invasive in the UK and elsewhere (Centre for Agriculture and Bioscience International (CABI), 2018b). Damage associated with the following species or plant groups are discussed in this case study: F. japonica, B. davidii, ‘trees’ (other woody, independently standing mature plants) and ‘woody climbers’ (woody plants that are not independently standing, e.g. attached to walls). In addition to presence, for F. japonica, mature (with crowns) and immature (without crowns) plants were assessed. Similarly, for B. davidii, mature (woody) and immature (not woody) plants were considered.

Literature assessment

Survey results

Survey results (reported rhizome extension)

There was a statistically significant difference (Mann–Whitney U; p < 0.05) in the horizontal extent of F. japonica rhizomes between small and large stands, with larger stands found to have further reaching rhizomes (Fig. 4). None of the small stands included in the assessment had rhizomes extending further than four m, and the majority (75%) had rhizomes extending two m or less. The average rhizome extension reported for small stands was 1.4 m. Only one plant in the large category had rhizome extension greater than five m (identified as a statistical outlier); all other records were below four m and the majority (75%) had rhizome extensions of 2.5 m or less.

There was also a statistically significant difference (Mann–Whitney U; p < 0.001) between the large and small stands for vertical rhizome extent, with larger stands found to have deeper reaching rhizomes (Fig. 5). No records with vertical rhizome extent in excess of 3.5 m were recorded. The small stands had rhizomes with a mean 1.02 m depth and a maximum of two m, whereas the maximum vertical extent recorded for the large stands was 3.2 m and the mean was 1.64 m.

Case study

In all but the most severe examples, the level of damage caused by plants did not exceed damage that was observed elsewhere within the study area in locations where plants were not growing. It would appear, in the context of dilapidation, that plants are generally not the cause but rather an accelerator to natural weathering and dilapidation.

Fallopia japonica was identified within the boundary of six properties (five mature stands and one immature stand) and the plant was identified within seven m of the main building of a further 12 properties, leading to a total of 18 properties where F. japonica was within the area identified by the ‘seven-m rule’ as being at risk. B. davidii was identified on 62 properties (31 mature and 31 immature). Trees were observed on six properties and woody climbers were observed on four.

In general, F. japonica was linked to less damage than the other species/species groups assessed (Table 4). Where F. japonica was linked to damage, mature plants were more likely to exacerbate the damage than to have been the original cause. There were no reported incidences of immature F. japonica causing or exacerbating damage.

Fallopia japonica was not linked to any damage to the main buildings. The three other groups were linked to damage, at varying degrees, typically in the form of simple co-occurrence (e.g. as in appearing together without a clear causal link) or interference with brickwork through exacerbation of existing weakness. Mature woody B. davidii was more likely to exacerbate damage than immature B. davidii, with immature B. davidii rarely exceeding co-occurrence or minor exacerbation. There was only one example of a plant being linked to causing direct damage to a building, rather that exacerbating it. This was a tree falling against a house.

With respect to damage to walls, F. japonica was correlated with two occurrences of damage; in both cases it was emerging from a crack and causing no detectable variation away from baseline damage elsewhere in the wall. The three other plant groups were linked to more damage than F. japonica, to varying degrees, typically in the form of simple co-occurrence or interference with brickwork through exacerbation of existing weakness. In all groups, the average damage score was higher than that of F. japonica (Table 4). Mature woody B. davidii was more likely to exacerbate damage than immature B. davidii, with immature B. davidii rarely exceeding co-occurrence or minor exacerbation. There were only two examples of a plant being linked to causing damage to walls, rather than exacerbating it, a tree pushing over a boundary wall and B. davidii pushing over a small retaining wall.

With respect to damage to paving, F. japonica was correlated with six occurrences of damage. In three cases it was emerging from a crack and causing no detectable variation away from baseline damage elsewhere in the paving, and in three other cases it was exacerbating existing damage (one minor, two moderate examples). B. davidii was linked to more damage to paving than F. japonica, typically in the form of simple co-occurrence or interference with paving through exacerbation of existing weakness. The average damage score was considerably higher for B. davidii than F. japonica. Mature woody B. davidii was more likely to exacerbate damage than immature B. davidii, with immature B. davidii rarely exceeding correlation or minor exacerbation. There was only one example of a plant being linked to causing damage to paving, rather that exacerbating it, which was a tree where the roots had lifted a large area of concrete paving with significant associated cracking.

Indirect damage: in the context of F. japonica

Plants are considered to cause structural damage to buildings primarily through indirect damage, for example through subsidence caused by modification to soil water content. High water-use tall trees are the main plant type implicated. Subsidence, with respect to plants, is only an issue on shrinkable clay soils, which are reasonably restricted in extent (Fig. 1). Importantly, to properly assess risk, individual site investigation is required to determine the exact type of clay present in a clay–soil area. The rate that water is removed from soil by plants varies depending on the characteristics of the plant and also by the total biomass of the plant. There is a strong linear relationship between water use and plant biomass (i.e. larger plants remove more water from the soil), as noted by Nielsen et al. (2015). Plants with higher water use and larger biomass are therefore the most likely to cause subsidence through the action of their roots removing water from soil. Some unpublished work suggests that F. japonica may be a high water use plant (Vanderklein et al., 2013); however, even if this is the case, it is not a high biomass plant by comparison to mature woody trees such as oak. The plants that are most likely to influence subsidence in the UK are listed in the NHBC (2017) guidance for building near trees. These species range in height between 10 and 28 m. In comparison, F. japonica typically only grows to between two and three m. The potential for plants to influence subsidence is calculated based on a zone of influence of between 0.5, 0.75 and 1.25 times the height of the plant (NHBC, 2017), depending on the water demand at maturity of the species in question (low, moderate or high, respectively). For F. japonica, this would suggest a maximum zone of influence of 3.75 m (the typical maximum height of the plant is three m, hence 3 × 1.25). However, when compared to mature trees, given the comparatively diminutive size of F. japonica, both in terms of above ground and below ground biomass, it is more likely to be at the lower end of the scale. As such, a calculation of 0.5 × 3 = 1.5 or 0.75 × 3 = 2.25 m is more likely to reflect the potential zone of influence of F. japonica at maturity. Furthermore, the mean rhizome length of small F. japonica stands, such as those more likely to be found in residential properties, is 1.4 m (‘Direct damage: in the context of F. japonica’ and Fig. 4), which falls comfortably within the lower zone. Such areas of influence are unlikely to be able to create a large enough area of soil shrinkage to impact all but the flimsiest of structure and, even then, only on properties shown to have shrinkable clay soil. As such, the risk associated with F. japonica causing subsidence based damage falls well below many other species commonly found in properties in the UK.

Direct damage: in the context of F. japonica

In some situations, trees and vegetation can adversely affect structures by direct action, for example structural failure of trees (collapse and impact), impact of branches with superstructures, displacement/lift/distortion and disruption of underground services and pipelines (British Standard, 2012).

The leading causes of damage due to direct physical contact by plants, that is collapsing vegetation striking buildings and power lines and branch impact, are not relevant in any meaningful way to F. japonica as the species is not tall enough and does not possess heavy enough aboveground structures. This is due to the fact that F. japonica aboveground material dies back at the end of each growth season; as such, the plant cannot accumulate sufficient above ground size and weight from successive years of growth.

Plants can also cause damage by exerting accumulating physical pressure on structures as they grow over time; however, as stated above, this is comparatively rare in terms of meaningful damage. Damage of this type is typically characterised by superficial or cosmetic damage to paving. However, more significant damage can occur where plants become trapped between two structures, for example two walls in close proximity to each other, and are allowed to exert pressure for an extended period of time without intervention (i.e. woody plants are allowed to mature in areas where management would be advisable) or where roots find their way into drains and pipes, as described above. The mechanisms by which plants grow and cause such damage are well understood (Biddle, 1998, 2001), as are the planting distances required to limit or avoid such damage (British Standard, 2012). While F. japonica can cause such damage due to direct action over time, it does not exceed that caused by woody species. The case study described in this paper demonstrates that F. japonica is less capable of causing this type of damage than trees and woody shrubs. Where F. japonica is implicated in such damage, this is likely to typically be a result of the plant exploiting a weakness or defect that was already present, rather than the plant initiating the damage, or it is simply a case of F. japonica emerging from an existing crack without influence. Regardless, even if it is assumed that F. japonica can equal trees in causing such damage (which is not the case), based on well understood principles (British Standard, 2012), a safe distance for mature F. japonica (crowns between 30 and 60 cm) would be 0.5 m for buildings and heavily loaded structures, and 1.5 m for paths and drives with flexible surfaces or paving slabs.

Additionally, the frequently stated ability of F. japonica to ‘grow through concrete’ is simply not supported by any evidence, as it is not possible due to the laws and principles of physics and biology. The extending tip of the F. japonica rhizome is remarkably soft and fleshy (Fig. 1) and it would be impossible for it to grow through intact concrete; however, these same characteristics make the extending rhizome adept at finding cracks and F. japonica has been shown to have significant ability to alter the direction of rhizome growth (Smith et al., 2007), highlighting the plant’s biological preference to go around obstructions, rather than through them. Where F. japonica is implicated in such damage, existing cracks or weaknesses are always present.

Typical rhizome extension

When the above is considered, the typical maximum rhizome extension of F. japonica is not all that relevant with respect to structural damage. Regardless, the results of the survey detailed above demonstrate that even large stands of F. japonica do not usually produce rhizomes that extend further than four m, showing that the ‘seven-m rule’ is not a statistically robust tool for estimating likely rhizome extension from above ground plants. The mean rhizome extent for small stands was 1.4 m and for large stands (above four m²) was 2.02 m. Similarly, the mean vertical extent recorded averaged between 1.02 m for the small stands and 1.64 for the large stands, with a maximum of 3.2 m.