Heracleum mantegazzianum (giant hogweed)

Prevention

SPS Measures

Heracleum mantegazzianum is prohibited as a noxious weed in the USA (USDA-NRCS, 2002). In the UK, it is included in legislation requiring land owners to take steps to control it and prevent further spread (Willoughby, 1996). In the EU, the species is listed among invasive alien species of Union concern, and is subject to restrictions and measures on keeping, importing, selling, breeding, growing and releasing into the environment (Regulation (EU) 1143/2014) . In New Zealand, it is listed among National Pest Plant Accord Species and it is prohibited to breed, distribute, release or sell. In Canada, giant hogweed seed trade is regulated under the Seed Act, R.S.C., 1985, c. S-8.

Public Awareness

Nielsen et al. (2005) provided a useful section on preventive measures, early detection and eradication, providing a checklist of actions, from the establishment of policies and guidelines, identification of routes of possible entry, identification of habitats most at risk, awareness campaigns, surveys of incidence and spread, eradication campaigns where necessary, followed up by monitoring. They note that mapping incidence and spread is made easy by the size and conspicuousness of the weed, ensuring that the public can provide reliable help, and even allowing clear monitoring from aerial photographs taken when the weed is in flower (Müllerová et al., 2005; 2013). Fonji et al. (2014) presented successful a project of public participation GIS involving school students monitoring H. mantegazzianum in Latvia. Shackleton et al. (2020) mention that management programs should concentrate on private land owners to encourage efficient control measures.

Heracleum mantegazzianum can only reproduce generatively, thus timing of the control measures to prevent flowering is crucial, and must be repeated until the seed bank is depleted (up to 7 years, Moravcová et al., 2018). It shows high regeneration ability after the damage (Dodd et al., 1994), and systematic control measures and effective monitoring are therefore necessary.

Cultural Control and Sanitary Measures

Heracleum mantegazzianum is not effectively controlled by light grazing, but intensive grazing, especially by sheep has provided good control and the rooting of pigs can also be highly effective (Tiley et al., 1996). Anderson and Calov (1996) reported on a five-year study in Denmark in which the population of the weed was much reduced after 2 years of sheep grazing and eliminated after 5 years, when no viable seeds were found to remain in the soil. The weed may be slightly less palatable to cattle, but grazing by cattle as well as pigs is recommended in Ireland (Lucey, 1994).

Buttenschøn and Nielsen (2007) confirm that sheep and goats seek out young plants of H. mantegazzianum and recommend grazing to begin in mid-spring. Herds should preferably include individuals already familiar with the weed to reduce the risk of over-eating and poisoning. Dark-skinned, thick-pelted animals are less likely to suffer dermatitis. Nielsen et al. (2007) give more detail on the management of grazing for control of the weed.

Physical/Mechanical Control

Hand-pulling (with gloves) can be effective on young seedlings but is impractical on larger plants. Mechanical cutting is frequently used to clear riverbanks, but provides no long-term control, as there is rapid re-growth from below ground, and it may encourage the perennation of flowering shoots which would otherwise die after flowering. Cutting in May and June is somewhat more effective in reducing seed production and/or re-growth than cutting in March, but no single cutting can be relied on to prevent eventual seeding. Even four mowings per year for 2 years failed to kill the plant (Nielsen et al., 2007). There is a trade-off between early treatment with high regeneration and later umbel removal with more efficient reduction in fecundity, but with a necessity to handle more developed fruits that are able to ripen even on cut-off umbels (Pyšek et al., 2007c). Digging or ploughing to destroy the crown (below 10 cm soil depth), or otherwise severing the tap-root from the crown (‘root-cutting’) can completely kill the plant (Pyšek et al., 2007e).

Biological Control

Apart from the use of grazing animals, no other biological control method has yet been developed, though the possibilities have been discussed by Fowler et al. (1991); Fowler and Holden (1994); Burki and Nentwig (1998); and most recently by Cock and Seier (2007). The latter comment that there had been cause for optimism when the intensive ‘Giant Alien’ project was initiated in 2002, but after detailed appraisal of all the potential biocontrol agents listed by Hansen et al. (2007) and Seier and Evans (2007), they can only state that “there are still some areas that need clarification before concluding that there is no prospect of finding host-specific natural enemies for biological control of giant hogweed.” All the most damaging insects and fungi have insufficiently narrow host range and can damage other Apiaceae such as parsnip (Pastinaca sativa).

In the UK, herbicides recommended for control of H. mantegazzianum include glyphosate, triclopyr and imazapyr but all must be applied early in the season (March to May) for best effect. Imazapyr has a residual effect in the soil that will prevent further germination but may also have a later effect on non-target species. Glyphosate is the most widely used compound, usually applied in April or May, when plants are 20-50 cm high, but owing to risks of toxicity to fish and algae, a buffer zone of 2 m should be left unsprayed adjacent to any river or other water body (Marcher, 2001). Nielsen et al. (2005; 2007) comment that the use of herbicides is increasingly controversial in parts of Europe, and there is a general (voluntary) ban on their use in public areas in Denmark. However, this ban was relaxed specially for the use of glyphosate on H. mantegazzianum. Instead of spraying, the herbicide can be stem injected to prevent water contamination (Grguric, 2018), and herbicide should generally be regarded only as one of the first steps in an integrated control programme. As a biochemical alternative to synthetic pesticides, mulch with pyrolysis liquid from birch wood can be used for controlling H. mantegazzianum seedlings (Hagner et al., 2020).

IPM Programmes

Tiley and Philp (1992) described an integrated two-year programme of spraying with glyphosate in April/May, combined with cutting below ground when or where spraying is not feasible, and dealing with large plants threatening to flower before dealing with vegetative plants. Nielsen et al. (2007) discussed the various factors to be considered in arriving at an optimal integrated control system, such as the relative suitability of cutting, umbel removal, root cutting, chemical control and cultivation, according to the size and density of the infestation, and whether eradication or containment is the aim. They also presented a table showing the seasonal schedule for the different operations. Nielsen et al. (2005) produced a table showing the different range of options appropriate to small, medium and large populations of the weed and detailed estimates of the costs of each control method. They also emphasized the importance of re-vegetation as a component of any integrated management strategy. Ravn et al. (2007) discussed this in more detail, though their work mainly relates to H. sosnowskyi. Williamson and Forbes (1982) emphasized that due to the likely spread of seed down a river, it is important to ensure that upstream infestations are dealt with before attempting eradication further down. Müllerová et al. (2005) suggested targeted eradication towards new satellite dispersal foci and linear structures serving as invasion drivers. Deciding on an optimum control programme may be helped by the development of a ‘model-assisted evaluation of control strategies’ by Nehrbass and Winkler (2007), and risk maps predicting the species spread in the future (Cook et al., 2007; Nielsen et al., 2008; Thiele et al., 2008; Wallentin, 2013). Thiele et al. (2008) stated that although the models might not be precise enough for targeted eradication measures in heavily invaded landscapes, they may serve well to identify habitats prone to H. mantegazzianum invasion. Shackleton et al. (2020) propose multiscale modelling and integrated global and regional species distribution models to monitor the spread and assess effectiveness on management measures in Switzerland. Nielsen et al. (2005) finished with a section on ‘Planning a management programme’ with valuable suggestions for defining priorities.

Remote sensing means were successfully applied for H. mantegazzianum monitoring and control to ensure early detection, targeted eradication and regular monitoring (Müllerová et al., 2005; 2013). For the monitoring, timing is crucial, because the plant can be well detected when flowering using drone, aerial as well as satellite imagery (Müllerová et al., 2017). However, the invader can be detected in leaf stage with high resolution drone imagery (Lawrence, 2019). New powerful algorithms of deep learning can assist to real-time detection (Menshchikov et al., 2021). To increase efficiency, the tool can be coupled with predictive modelling (Rocchini et al., 2015).

Ecosystem Restoration

The value of re-vegetation following control programmes is discussed under IPM programmes.