Impatiens glandulifera (Himalayan balsam)

I. glandulifera is a highly invasive annual species which has spread rapidly in many parts of Europe and North America after its introduction as an ornamental. The spread is likely to continue to more northerly or high montane areas as a result of global climatic change. Due to its ability to form dense stands and its conspicuous appearance it has been blamed for negative biodiversity effects. Even though these effects are less severe than often thought, further spread is undesirable and should not be facilitated by further use, in particular in natural areas. Control is advisable in certain situations, e.g. nature reserves and conservation sensitive areas, but eradication from larger parts of its invasive range is not feasible due to the need to control the plant on a catchment scale, which is often impossible due to the sheer scale of occurrence and division of land ownership.

Before the recent advances in molecular phylogenetics, Impatiens (Balsaminaceae) was treated as a distinctly separate order, Balsaminales (Dahlgren, 1989), and more traditionally as a member of the order Geraniales under Rosidae (Cronquist, 1988; Thorne, 2000). Anderberg et al. (2002) and Geuten et al. (2004) disputed such classifications which were based mainly on morphological characteristics. As a result of their molecular phylogenetic studies, Balsaminaceae was reclassified as a family in the Ericales (an order of 26 families), sitting as a sister group to all other Ericales in the Balsaminoid Ericales. The Balsaminoid Ericales consist of the families Balsaminaceae, Marcgraviaceae, Pellicieraceae and Tetrameristaceae. Together this group comprises of approximately 1130 species. Three ‘forms’ of the species have been noted, forma albida (Hegi) B. Boivin, forma pallidaflora Weath., and forma glandulifera Vahl (Missouri Botanical Garden, 2008).

Attention should be paid to the taxonomic authority, as the true species is I. glandulifera Royle, whereas I. glandulifera Arn. is a synonym of I. taprobanica Hiern, a native of Sri Lanka (USDA-ARS, 2008).

I. glandulifera is a tall glabrous annual reaching 50 to 250 cm in height. It is now Europe’s tallest annual species. Its stems can be 0.5 to 5 cm in diameter and are sometimes branched in the upper part. Roots are up to 15 cm deep, the plants often forming numerous adventitious roots from the lower nodes. The leaves are opposite, the upper ones sometimes in whorls of three, up to 25 cm long and 7 cm wide, lanceolate to obovate, petiolate and sharply serrated at the edges. The inflorescences are racemes of 2-14 flowers that are 25-40 mm long. Flowers are strongly zygomorphic, their posterior sepal forming a sac that ends in a straight spur. Their colours vary from white to pink and purple. The capsule is 3-5 cm long and up to 1.5 cm wide. It contains up to 6 (Grime et al., 1988) or 4 to 16 seeds (Beerling and Perrins, 1993), that are 4-7 mm long and 2-4 mm wide with a mean weight of 7.32 mg.

I. glandulifera is native to the foothills of the Himalayas from north-west Pakistan to northern India. The native range in the western Himalayas is relatively small compared to its invasive range. According to Beerling and Perrins (1993), I. glandulifera is native from Kashmir to Garhwal between 2000 and 2500 masl, and Polunin and Stainton (1984) report the plant can grow up to 4000 masl in its native range. The plant is also recorded as native in Nepal (USDA-ARS, 2008), and possibly in Bhutan.

I. glandulifera is introduced and invasive in much of Europe, and parts of Canada, the USA and New Zealand.

I. glandulifera was first introduced to the UK in 1839 as a garden ornamental. Since its introduction to the UK the plant has spread at a rate of 645 km² per year (Perrins et al., 1993 in Weber, 2003). The spread was enhanced by beekeepers and the general public who released the plant into the wild on many occasions (Rotherham, 2000). The occurrence of I. glandulifera was noticed as early as 1855 in the UK, with the plant being recorded as naturalised in the county of Middlesex. The plant is now recorded throughout the UK including offshore islands such as the Isles of Scilly, the Shetland and Orkney islands.

From the UK I. glandulifera was taken to gardens in many European countries where it is still popular. The plants range expansion in mainland Europe began some 50-100 years later than in the UK. In Finland, the plant was recorded as naturalised in 1947 and individual populations were seen to expand during the 1980s (Kurtto, 1996). In the Czech Republic the earliest record of the plant occurring in the wild was in 1896, in Northern Bohemia (Pysek and Prach, 1995). Garden escapees were found in Switzerland in 1904, from where the species migrated along the Rhine to Germany. Although the explosive mechanism disperses seeds from the plant for only up to 7 m, the invasive spread was fast. Seeds can be transported with rivers over large distances and the spread was helped by humans. The spread is likely to continue with global warming to more northerly or high montane areas (Beerling, 1993).

Further spread of I. glandulifera is likely. As a result of seed transport with flowing water, it will predominantly lead to an increase in abundance within countries or regions. International transport may be motivated by the ongoing use and promotion of the species as a garden plant. Transport of seeds as a contaminant of soil, building material, etc. is possible, but less likely to cause new introductions (Beerling and Perrins, 1993; Hartmann et al., 1995).

I. glandulifera is regarded as an important invader in several European countries and is on the EPPO list of invasive alien plants, “as posing an important threat to plant health, the environment and biodiversity in the EPPO region. It is on the Swiss 'black list' of harmful invasives (Anon., 2002), listed as invasive in Austria (Essl and Rabitsch, 2002), and is among those invasives in Germany against which specific control measures are directed (Kowarik, 2003). I. glandulifera is also a declared noxious weed in the USA, in Connecticut, Oregon and Washington (USAD-ARS, 2008).

Natural Dispersal

Seeds are expelled from the plant by explosive dehiscence of the capsule and can lead to dispersal distances of 7 m. Long-distance dispersal of seeds occurs with the aid of flowing water such as along rivers. Fresh seeds are transported with sediment on the beds of rivers, and dry seeds are buoyant and can float over large distances.

Vector Transmission 

Isolated observations of seeds dispersed up to 10 m from the mother plant may indicate the possibility of seed transport by small rodents (Beerling and Perrins, 1993).

Accidental Introduction

Transport with topsoil is probable (Beerling and Perrins, 1993) but it is not clear, however, to what extent this has occurred in the introduction or spread to new areas. The transport of seed with river gravel in trains was reported in Germany (Hartmann et al., 1995), as well as contamination of building rubbish transported to waste disposal sites.

Intentional Introduction

I. glandulifera was imported as an ornamental species for its showy and scented flowers. It is being used as a garden plant in many European countries and is still sold by seed companies (Beerling and Perrins, 1993). The general public have aided the transport of this species throughout the UK, deliberately planting seeds in hedgerows and grassland (Rotherham, 2000). The flowers are very rich in pollen, and bee-keepers have dispersed seeds in order to enhance forage for honey bees (Hegi, 1912; Hartmann et al., 1995).

I. glandulifera is not a weed of agricultural fields. However, native herbaceous plants and tree regeneration can be out-competed by the dense growth of the species (Larson and Martinson, 1998; Maule et al., 2000).

Other Impatiens species are somewhat similar but differ in conspicuous features: The Asian I. parviflora is much smaller and has small pale-yellow flowers. Also, the yellow flowered European I. noli-tangere and the orange flowered American I. capensis are both smaller morphologically. The garden ornamental I. balsamina that occasionally escapes to waste ground in North America and in Europe has pubescent stems and capsules and usually single flowers.

In Europe, I. glandulifera is predominantly a weed of riparian systems where it can form dense monocultures along river banks (Pysek, 1995; Kowarik, 2003). I. glandulifera is also found in damp natural woodland, where it can attain is maximum known height (up to 3m), in addition the plant is found in forest plantations, forest clearings, railway embankments, waste ground, urban areas, roadside ditches and wet meadows.

In the native range, I. glandulifera is predominately a plant of high altitudes, moist, fertile valleys where it grows in clusters of 30-60 plants mixed in with surrounding native vegetation (Tanner et al., 2008). This is in stark contrast to the dense monocultures found in the invasive, introduced range.

Genetics

The degree of genetic heterogeneity is uncertain, though there is obvious variation in flower colour (Grime et al., 1988). The chromosome number is 2n=18 or 2n=20 (Grime et al., 1988; Beerling and Perrins, 1993).

Physiology and Phenology

Seeds germinate in early spring with a high germination rate of approximately 80% (Sebald et al., 1998). The cotyledon phase lasts until April in the UK when rapid shoot growth begins (Beerling and Perrins, 1993). The root system is augmented by adventitious roots from the lower nodes (Beerling and Perrins, 1993). The time from germination to the onset of flowering is 13 weeks in Germany, with flowering continuing for a further 12 weeks (Sebald et al., 1998). The mortality of seedlings and young plants can be high due to slug predation and physical damage from rainfall and late frosts (Prowse, 1998). Adult plants are killed by the first frost in autumn/winter months.

Reproductive Biology

The self-compatible flowers of I. glandulifera have the highest sugar nectar production per flower than any native European plant species (Chittka and Schürkens, 2001). This enables the plant to attract numerous insect pollinators, especially bees (Apis mellifera), bumble-bees (Bombus spp.) and syrphids. The species is exclusively propagated by seeds. The number of seeds produced is given as 700-800 seeds per plant and 5.7 seeds per pod by Beerling and Perrins (1993), and up to 4000 seeds per plant and 6.4 seeds per pod according to Sebald et al. (1998). A maximum of 32,000 seeds were produced per square metre in a pure stand in Germany (Koenies and Glavac, 1979). Explosive capsules expel seeds from the plant, and dissemination is also aided by flowing water, seeds transported with sediment and the fact that dry seeds are buoyant. In addition, seeds can be spread by human actions, as in the transportation with soil. Seeds require chilling to become viable. Although the species is reported as not having a persistent seed bank (Grime et al., 1988), there are indications that at least some seed can persist for 18 months (Beerling and Perrins, 1993).

Environmental Requirements

I. glandulifera has a preference for high atmospheric humidity. It grows in half-shade but also in full sunlight. In the native range the plant occurs at high altitudes between 1600 and 4300 m, but in Europe it is found at lower elevations. In the UK it has not been found above 210 m and in the eastern Alps in Austria it occurs at up to 1200 m (Drescher and Prots, 2000). In Europe, plants of all ages are frost intolerant with adults killed by the first frost in autumn and seedlings by late frosts in spring (Sebald et al., 1998). The species is also drought-intolerant and quickly wilts, and plants can survive only if the drought period is short (Beerling and Perrins, 1993). As an annual, the species is dependent on open sites for germination each spring; it is consequently favoured by disturbance. It occurs on a wide spectrum of soils from nutrient-poor to nutrient-rich and grows on mineral soils as well as on peat (Kowarik, 2003).

Associations

I. glandulifera is found in five main community types, riparian habitats, fens, mesotrophic grasslands, waste ground and woodlands. Within each habitat few associated plant species have been recorded growing within monocultures of the plant. In the UK, the main associated species are similar across the community types, and include Rubus fruticosus, Urtica dioica,Galium aparine, Cirsium arvense, and in wooded habitats the plant grows under a variety of tree species.

In southern Germany, it is most often accompanied by Urtica dioica, Aegopodium podagraria, Lamium maculatum and Galium aparine (Oberdorfer, 1983). Dense riverbank vegetation with I. glandulifera was described as Impatienti-Calystegietum, e.g., along the Odra in Poland (Dajdok et al., 1998).

I. glandulifera supports an impoverished diversity of phytophagous insects in the UK, but the extent to which these affect the ecology of the plant is not sufficiently studied (Beerling and Perrins, 1993). In the UK, only 3 arthropod species are known to feed on I. glandulifera, including two aphid species, Aphis fabae and Impatientinum balsamines, and the elephant hawk moth Deilephila elpenor (Beerling and Perrins, 1993).

The UK Environment Agency have estimated it would cost between £150-300 million to eradicate I. glandulifera from the UK should such a control programme be initiated. In Switzerland, Gelpke and Weber (2005) estimated it would cost between CHF 2,183,500 and CHF 13,812,696 (£923,133 to £5,839,691) to eradicate 95% of the current population of I. glandulifera in the Canton of Zürich alone. Such high costs coupled with the difficulty of implementing catchment scale control programmes due to the division of land makes controlling I. glandulifera on a national or regional level virtually impossible. Current control methods are labour intensive and difficult to implement also due to the often inaccessible habitats in which I. glandulifera grows. Control costs range from £0.50/m² for a single chemical application, or manual control by strimming up to £10/m² when habitat restoration is included (Tanner et al., 2008).

Although social impacts are difficult to quantify, the fact that I. glandulifera can restrict access to rivers for recreation and other amenities should be noted.

I. glandulifera is used as a garden ornamental and as a honey plant. It produces flowers with high nectar content and acts as a late source of nectar for bees and butterflies. Cattle are known to feed on the whole plant (Beerling and Perrins, 1993) but the browse value is not known.