Skip to main content

Giant hogweed (heracleum mantegazzianum) a pernicious invasive weed: developing a sustainable strategy for alien invasive plant management in europe

Exploitable results

As a result of revegetations studies integrated management system was worked out providing control of Giant Hogweed and re-establishment of biodiversity in arable areas. Glyphosate (herbicide) treatment after substantial spring growth ensures control of Giant Hogweed rosettes, seedlings. Deep ploughing (up to 25cm) 2-3 weeks after herbicide application is effective soil cultivation method, because of cutting Giant Hogweed rosette plant roots and turning upside-down upper soil layer where majority of seeds is concentrated. Sown grass mixtures (componenets - competitive, produce dense swards, with high re-growth potential under frequent cutting) provide uniform covering of the soil after elimination of previously prevailing component of phytocoenosis -Giant Hogweed. Dicot plant species appear in treated areas later increasing biodiversity of phytoceonosis. Dicot herbicide application ensures hogweed seedling and other dicot weed control in early grass sward development stage. Cutting treatment beside hogweed control increases grass sward density, because of tillering stimulation. For control of Giant Hogweed in riverside's areas, where chemical treatment is prohibited, a method for creation of a strong competitive plant community and depression of Heracleum spp has been selected. Method combines sowing of grass mixtures (high seeding rate) and frequent cutting by hand mower.
Studies of Heracleum mantegazzianum in the Czech Republic and Germany revealed high seedling density, low mortality of established plants and fast population development. Flowering occurred in the 3rd year but was postponed up to 12 years under stress. Overlap between male and female phases allows for self-pollination and production of viable seed. This together with high fecundity allows single plants to start invasion after long-distance dispersal. A short-term persistent seed bank is formed; 90% of the seed germinate next spring but some survive longer. Morphophysiological dormancy is broken by cold stratification. These features, together with efficient dispersal by humans, water and wind, result in enormous invasion potential. At a local scale, populations increased by 1,260m2/year. Early in succession, life strategy accords with that of ruderal monocarpic plants, later with competitive perennials. Population growth rates of open stands depend on transition into next life-stages; generation time is ±3 years. In dense stands, stasis is more important and generation time exceeds 5 years. Productive human-disturbed habitats and inappropriate landscape management favoured invasion in Europe. Native diversity is adversely affected locally but not regionally. In its native region in the Caucasus, Heracleum is confined to disturbed sites and rare in natural subalpine meadows.
The best practice guidelines and an integrated control strategy have been developed on the basis of outputs from all the work-packages in the project. The project team has studied as many relevant aspects as possible of the biology and ecology of Heracleum mantegazzianum in Europe, the invaded area, and in the Caucasus, its native area: Taxonomy and genetics, development and phenology (seasonal changes and growth cycle), population dynamics, pathology, herbivorous insects and their impact on the plants, as well as interactions with soil, nutrients, vegetation cover and land use changes. Especially the effects of herbicides, grazing, pathogens and herbivores and vegetation management schemes as potential control strategies for tall invasive hogweeds have been investigated. The overall objective of the 'best practice' manual is to provide all European authorities and private with scientifically based but simple and practical management methods. The manual is available in eight different European languages on the project homepage where printed copies also can be ordered.
We examined the impact of two guilds of herbivores: external feeders and endophagous insects. A general result is that mature H. mantegazzianum plants are quite tolerant to herbivory. This is achieved by several defence mechanisms: (1) chemical defense by furanocoumarins, (2) mechanical defense by trichomes, (3) defense by ants, which are attracted on a mutualistic basis. These mechanisms are inducible, this means, they can be increased if necessary. Since it is rather difficult to overcome this defensive shield, the selection of a suitable biological control agent will be very difficult.
In order to develop guidelines for the future management of H. mantegazzianum, a number of presently applied chemical and mechanical methods as well as new potential control measures were tested and expenses and time efforts were estimated. In addition, experiences gained from various control campaigns, research studies and management in practice have been evaluated with respect to best practice control methods and ecological and economic consequences. The evaluation of this data/information has resulted in a description of the most efficient, sustainable and cheapest control method applicable against H. mantegazzianum in Europe.
Distribution data for Heracleum mantegazzianum (Giant Hogweed) were collected at three different levels: - Europe-wide based on the compilation of country reports including such data as year of first record, current distribution map and data on any other closely related species; - a country-wide study of three European countries: Czech Republic, Latvia and the United Kingdom, using relatively detailed data, e.g. 10km by 10km square records in the United Kingdom, and - a regional study focussing on the detailed current and historic distribution in the county of Hertfordshire, United Kingdom. This process combined with genetic studies of Heracleum material from across Europe identified three main taxa of large leaved hogweeds in Europe: H. mantegazzianum, H. persicum and H. sosnowskyi. A map was compiled showing the distribution of these three species across Europe with H. mantegazzianum being widespread and present ion most countries, H. persicum largely confined to Norway, and H. sosnowskyi from the Baltic states and other ex-soviet countries. The map is used to show those parts of Europe that are free of these species, distinguishing between those in which the environment is unsuitable for large leaved hogweeds, e.g. Mediterranean areas, and those that might be invaded due to suitable environmental conditions, e.g. the Pyrenees. The primary cause of spread of H. mantegazzianum and H. persicum at the international scale was the exchange of seeds between horticulturalists, mainly in botanic gardens and others such large gardens. Heracleum sosnowskyi was spread primarily by agriculturalists due to its use in the Soviet Union as a fodder crop. Within a country, the sending of seeds from one garden to another remains a key transfer agent though other media are relevant, e.g. passage down and up rivers and other linear habitats (e.g. the Czech Republic and the United Kingdom). In Latvia, however, the spread was mainly due to the distribution of the plant as a crop though, again, other factors were significant. At the regional scale, the vectors for spread were similar to those at the country level.
187 samples were analysed of 12 Heracleum species from 72 populations collected in 15 countries in Europe and in 3 countries in the Caucasus region including: (i) invasive species: H. mantegazzianum, H. sosnowskyi and H. persicum; (ii) native species: H. sphondylium and H. sibiricum; (iii) other species: H. antasiaticum, H. leskovii, H. pastinacifolium, H. transcaucasicum, H. ponticum, H. trachyloma (all from Caucasus) and H. stevenii. To assess genetic similarity and for rooting of the dendrogram out-group species (Ferulago and Daucus) were added. The nine primer combinations produced a total of 630 AFLP fragments, 95% of them were polymorphic. To assess the genetic relationships between species/populations an unweighted pair group mean analysis (UPGMA) and principal Principal coordinates analysis (PCO) were performed. Results revealed great genetic similarity between the three species of Heracleum that are invasive in Europe and also some other tall Heracleum species from the Caucasus. The out-group species (Daucus and Ferulago) were only about 17-20% similar to Heracleum species. The target invasive species were also clearly separated from (i) H. sphondylium, H. sibiricum and H. ponticum, and (ii) H. antasiaticum and H. leskovii. Heracleum mantegazzianum samples from both the Caucasus and Europe were genetically very close (similarity about 80%). Samples of Heracleum sosnowskyi formed two clusters - one including samples from the invaded range (Europe) and the other grouping samples from the native range (Armenia, Transcaucasia). Such division indicates that this species in Europe either comes from other parts of the greater Caucasus, or that it was subject to some modifications prior to its introduction to Europe. Russian records suggest that seeds of H. sosnowskyi that were used in plant breeding programs in NW Russia (Murmansk and Leningrad) originated from Republic of Kabardino-Balkaria (Russian Caucasus). Samples of Heracleum persicum from Nordic countries clustered with some tall Heracleum species from the Caucasus (H. trachyloma H. pastinacifolium, H. transcaucasicum). As reference samples from native range (Iran and Turkey) of H. persicum were not obtained, we could not confirm the conspecificity of the Nordic plants with the Iranian H. persicum. Some interesting results were revealed by this analysis. Very close genetic relationship between Heracleum sp. samples from Buckingham and Kensington Palace Gardens (London) and H. persicum from Norway confirms hypothesis that the H. persicum was brought to north Norway in the early 19th century by English gardeners. Samples from two populations of H. persicum collected in Denmark formed a separate cluster apart from samples from Norway and Finland, suggesting that multiple introductions are responsible for invasion of H. persicum to Nordic countries. This analysis also pointed to a possible identification problem as samples collected in two parks in Latvia and labelled as H. mantegazzianum showed great genetic similarity to H. persicum and the above mentioned related species. It is unlikely that these plants could be H. mantegazzianum and are either of hybrid origin or are in fact H. persicum or other close species. Due to unresolved taxonomy in H. persicum group more work has to be undertaken to determine the status of these plants.
The studies carried out eliminated most of the potential biological control agents from further consideration, and none studied could be selected for use as biological control agnets. Elimination of these potential biological control agents means that any further efforts towards classical biological control will need to focus on agents not studied in sufficient detail, or new agents from other parts of the native range of Heracleum not accessible to the project (i.e. NW Georgia). The lack of success over the three field seasons of the project mean that success with these tactics is less likely, and so funding to complete the investigation will be difficult to come by. There is a risk that if further funding is not secured to follow this up, unanswered questions will remain. The conclusion that nothing suitable for exploitation as a biological control agent is a startling one, and most unusual for a weed biological control project. We are still considering possible explanations for this. It also became apparent during the project that classical biological control using pathogens is inadvertently blocked by the wording of EU Directive 91/414. In light of this, project participants were able to feed this concern into plans to respond to an EU Concerted Action call to consider revision of the directive. Since at this point, there is no immediate scope for classical biological control, research to assess the potential for an augmentative approach using a mycoherbocide should be accorded a higher priority.
Population-biological simulation models (Matrix models and Spatially Explicit Individual-Based models) were developed and parameterised along field data from partners P6 and P7. They served as a tool for identifying decisive mechanisms for invasive spread of Heracleum mantegazzianum on a local scale: the ability to stay in a vegetative state in order to bridge years with unfavourable conditions for reproduction and long-distance dispersal for bridging areas which are not suitable for maintenance of local populations. Due to the high variability of individual behaviour modelling must go beyond standard matrix modelling in order to reproduce the invasive potential of the species in a sufficient manner. The ability of our models reached as far as being able to reproduce and evaluate data for a time series of regional spread obtained by aerial photographs. The model was also used to evaluate different options for and check for their ability to reduce population size in a sustainable manner. Irrespective of any details the simulation results showed that any measures for controlling H. mantegazzianum populations must last for approximately. 5 years or longer and must be applied with a high intensity. Otherwise populations will increase and invade again. Results will be disseminated by scientific publications.
To examine the herbivore insect communities on the invasive Heracleum mantegazzianum we performed sampling tours in the native (Russia, Georgia and Azerbaijan) and the invaded areas (Europe). Investigations in Europe included Switzerland, Germany, Denmark, the Netherlands, Belgium, the Czech Republic and Latvia. In total we have gained information about 358 species found on 21 Heracleum species. The herbivore communities between invaded and native area were rather similar. So far, only a few oligophagic and no monophagous species were found. In several cases we could not decide on the feeding selectivity of the species (rare species, no published information available). This concerns mainly root feeding beetles, which demand follow-up studies. So far, two new species were detected (one depressariid moth, one agromyzid fly), these species descriptions are in press. It is assumed that these species are highly specific to Heracleum mantegazzianum. According to the actual stand of knowledge no promising candidate for a biological control program could be detected.