Tradescantia fluminensis (wandering Jew)

T. fluminensis is a persistent invasive weed of natural areas where it carpets the ground and prevents native regeneration. It alters litter decomposition and nutrient cycling affecting ecological succession. It is likely to be further introduced and spread as a popular ornamental houseplant, and it propagates and spreads easily from stem fragments. It is a serious environmental weed especially in New Zealand but has also naturalized in a number of other countries.

Tradescantia L. is one of 41 genera in the Commelinaceae R. Br. (Faden, 1988), a family of monocotyledonous herbs that are largely tropical and subtropical, though several are temperate (Evans et al., 2000). The genus comprises about 70 species (Faden and Hunt, 1991) of erect or trailing habit, many of which are cultivated for ornamental purposes (Mabberley, 1997). The genus is divided into three alliances; the T. fluminensis alliance is exclusively South American in origin and defined by the possession of C-glycosides (del Pero Martínez and Martínez, 1993). The genus name derives from John Tradescant (1608-1662), gardener to King Charles I of England (Chittenden, 1965) and the species name 'fluminensis' is a Latin translation of 'from Rio de Janeiro'. The common name 'wandering Jew' refers perhaps to its trailing habit and/or its common occurrence, after a mythical man of the Middle Ages. Supposedly he offended Jesus on his way to the crucifixion and was therefore cursed to walk the earth alone until the world ends (Young, 2001).

The Commelinaceae is characterized by several features including a distinct closed leaf sheath, a succulent leaf blade, and three-merous flowers with distinct petals and sepals (Cronquist, 1981; Faden, 1985; Faden and Hunt, 1991). T. fluminensis has broadly ovate to oblong-lanceolate leaves arranged alternately on thin (2-3 mm diameter) weakly ascending (or pendant) leafy shoots up to 60 cm tall that grade into leafless stems with roots at the nodes. Glabrous leaves are 1.5-12 x 1-3.5 cm, variable, with acute to acuminate tips, dark green or flushed purple beneath and/or variegated off-white or cream. Individual plants (leafy shoot + leafless stem) vary in length from 0.3 to 1.5 m and branches arise along the leafy shoot. Apical growth is balanced by basal decay (Maule et al., 1995; Standish et al., 2004). High biomass mats comprise interlaced vertical leafy shoots on horizontal leafless stems held to the substrate by abundant fine roots that also form at aerial nodes within the mat. The true T. fluminensis from Rio de Janeiro, Brazil is rhizomatous, with leaves tending towards the larger end of the range quoted; and different clones have leaves that are green, or variegated white, or yellow (Huxley et al., 1999). Flowers are actinomorphic or nearly so, bisexual, about 2 cm in diameter with three white petals, three green sepals and six fertile stamens. Flowers are produced in clusters at the end of the leafy shoot. Fruits are capsular and locules (three) are one to two seeded. Seeds are black, pitted and arranged in a row (Faden and Hunt, 1991; Langeland and Burks, 2003).

T. fluminensis is endemic to the tropical rainforests of south-east Brazil (Barreto, 1997) and neighbouring areas of Uruguay and Argentina (USDA-ARS, 2003). It has now also naturalized in New Zealand (Kelly and Skipworth, 1984), south-eastern Australia (Dunphy, 1991), Portugal (Aguiar et al., 2001), Italy (Orlando and Grisafi, 1978), Russia (Tolkach et al., 1990), Japan (Enomoto, 2000) and the south-eastern USA (Wunderlin, 1998; USDA-NRCS, 2002; Wunderlin and Hansen, 2003). It is however, very likely to be present in many if not most countries of the world as an ornamental pot plant (not recorded).

T. fluminensis was first introduced to New Zealand in 1910 by a Manawatu farmer to stabilize a steep bank (Roy et al., 1998) and was recorded as being naturalized in Northland, New Zealand soon afterwards (Carse, 1916). There were subsequent intentional introductions of ornamental material. T. fluminensis has probably naturalized in other countries following intentional introduction as an ornamental garden plant. Material from Australia and the Russian Far East is referred to as T. albiflora (Tolkach et al., 1990; Dunphy, 1991) indicating that it is a different cultivar (T. fluminensis 'albovitta') to that introduced elsewhere. Variegated cultivars can revert to green in the shade, so one cannot be certain of the origin of plant material.

T. fluminensis plants are widely available via the internet and mail order companies, and its popularity as an ornamental houseplant means that the likelihood of further introductions is very high. Individual plants can re-sprout following severe dehydration that might occur during transport. However, the risk of escape into the wild and successful naturalization will vary widely from country to country.

Natural Dispersal (Non-Biotic)Propagation in New Zealand and Australia, and possibly other areas where introduced, is wholly vegetative. Fragments as small as 1 cm can successfully establish new plants and dispersal of fragments along streams is common in New Zealand and Australia, and probably elsewhere. Vector Transmission (Biotic)No examples of dispersal by animals have been recorded.Agricultural PracticesRoad machinery can disperse fragments (Department of Conservation, 1998). Humans are the primary vectors for the dispersal of T. fluminensis fragments through gardening practices and dumping of garden refuse. Livestock can spread fragments in their hooves (Ogle and Lovelock, 1989).Accidental IntroductionThere are no reports of T. fluminensis being distributed accidentally and it is not a contaminant of trade.Intentional IntroductionAs one of the 'commonest of all houseplants ' (Mabberley, 1997), T. fluminensis has spread around the world through international trade. Sale of the plant is now prohibited in Australia and New Zealand, but remains popular for hanging baskets in Ireland (Seager and Campion, 1984), Germany (Kadner, 1992), the UK (Digby and Firn, 1995) and the USA (Langeland and Burks, 2003).

T. fluminensis does not appear to be a significant weed of crops (CEPLAC, 2003) despite earlier suggestions to the contrary (Landcare Research, 1998). It is primarily an environmental weed and a weed of forests and forestry, preventing forest regeneration by shading woody seedlings and it can smother ground covers at any growth stage (Esler, 1978; Polly and West, 1996).

Tradescantia zebrina hort. ex Bosse. also has a trailing habit and has naturalized in many warm countries, but has a solitary inflorescence, red-purple flowers and leaves which are often striped and silver with a hairy sheath (Mabberley, 1997; Huxley et al., 1999). The trailing T. crassula Link & Otto. has channelled leaves with ciliolate margins (Huxley et al., 1999). In its native range, non-flowering T. fluminensis may be difficult to distinguish from plants of other genera in the family Commelinaceae, e.g., Commelina (though its roots are often tuberous), Callisia (succulent leaves) and Dichorisandra (spirally arranged leaves are broad and thin). In Australia and New Zealand, T. fluminensis is distinctive.

In its native range, T. fluminensis occurs in rainforest and other damp, humid and shaded places including roadsides and gardens (Barreto, 1997). Outside its native range, it also occurs in damp, humid and shaded places such as gardens, parks, banks, stream-sides and forest remnants (but not large tracts of forest). Damp fertile soils support dense swards of T. fluminensis (Ogle and Lovelock, 1989) whereas growth is sparse on rocky substrates (Barreto, 1997; Smale and Gardner, 1999). T. fluminensis tolerates waterlogging (Barreto, 1997) and though shade tolerant, available light is an important factor limiting its spread and biomass accumulation (Maule et al., 1995). An important limit to the distribution of T. fluminensis is its intolerance of frost (Bannister, 1986).Outside its native range, the distribution of T. fluminensis is limited by its reliance on streams, people or animals for its dispersal. T. fluminensis invades the edges of forest remnants or areas under canopy gaps where light levels promote its biomass accumulation (Standish et al., 2004), although, if the forest canopy is restored and light levels decrease it can persist in deep shade with as little as 1.4% of full sunlight (Adamson et al., 1991). High biomass mats appear to remain indefinitely under canopy gaps and at forest edges. It invades lowland podocarp-broadleaf forest remnants in New Zealand (Standish et al., 2004), lowland temperate rainforest remnants in Australia (Dunphy, 1991) and mesic mixed hardwood forests in Florida, USA.

GeneticsThe chromosome number is 2n=72 (Scaramuzzi et al., 2000) and chromosomes are mostly large (Faden and Hunt, 1991). It is feasible that one well-adapted genotype of T. fluminensis exists in areas of naturalization and molecular studies are required to determine genetic origin and relationships of T. fluminensis in New Zealand and elsewhere.Physiology and PhenologyT. fluminensis generally does not invade large tracts of undisturbed forest, rather its success as an invader is limited to disturbed sites where it responds to an increase in resource availability post-disturbance more rapidly than the resident species and/or is a better competitor. The physiology of T. fluminensis enables a rapid response to the availability of two key resources - light and nitrogen. Light and nitrogen availability increase following a disturbance, such as tree-fall, allowing it to rapidly invade the disturbed site. It grows rapidly in the increased light environment while accumulating a store of nitrogen in its tissues. Then, if the canopy gap is closed, T. fluminensis can acclimatize to the reduced light availability and utilize its store of nitrogen (Maule et al., 1995). In favourable conditions, aboveground production is 0.64-1.28 t/ha/year (Standish et al., 2004) or up to ~3 t/ha/year although biomass is reduced at drier locations. Ancymidol retards the growth and elongation of T. fluminensis resulting in decreased internodal length, more compact growth and more intense colour than non-treated plants (Blessington and Link, 1980). Variegation in T. fluminensis is caused by a genetic mutation inhibiting the production of chlorophyll (Land and Norton, 1973). Variegated cultivars can revert to green in the shade. T. fluminensis has been a study organism for research into gravitropism (Digby and Firn, 1995, 2002; Funke and Edelmann, 2000) and photosynthesis (Kutyurin et al., 1974; Rozonova et al., 1978; Adamson et al., 1991; Bil' and Fomina, 1992; Demmig-Adams and Adams III, 1992; Fomina et al., 1993; Tsionsky et al., 1997).Reproductive BiologyTradescantia spp. are usually self-sterile, so that individual plants and clones derived from them produce no seed (Huxley et al., 1999). T. fluminensis does not set seed in New Zealand or Australia and reproduction is wholly vegetative with fragments as small as 1 cm successfully establishing new plants. In its native range, T. fluminensis attracts non-nectar seeking insect pollinators (Faden, 1985; Evans et al., 2003). T. fluminensis can be propagated in vitro from bud cultures and repeated subcultures (Scaramuzzi et al., 2000).Environmental RequirementsT. fluminensis prefers higher rainfall zones, generally over 1000 mm per annum and up to 1800 mm has been recorded though the absolute limit is probably in excess of this. Mean annual temperatures are in the range 15-21°C. It is frost sensitive; the temperature which causes 50% damage is estimated at -4.0 to -4.4°C (Bannister, 1986). It can persist in the deep shade as low as 1.4% of full light (Adamson et al., 1991). It can tolerate shallow soils of any type though slightly acid soils are preferred. It has been recorded at altitudes as high as 3300 m in Bolivia (Missouri Botanical Garden, 2003).AssociationsYeates and Williams (2001), Toft et al. (2001) and Standish (2004) have identified various soil microfauna and invertebrate fauna associated with T. fluminensis in New Zealand. Juveniles of a large carnivorous land snail endemic to New Zealand species appear to use the exotic T. fluminensis as a refuge (Standish et al., 2002a).

An absence of natural enemies where introduced is often cited as an important reason for its success as an invader, and visible signs of disease and herbivore attack are more prevalent in Brazil than in New Zealand or Australia. There are a number of pests and pathogens that have been identified on T. fluminensis (e.g. Lockhart et al., 1981; Gjaerum, 1984; Baker and Zettler, 1988; Farr et al., 1989, Fowler et al., 2013; Macedo et al., 2016) and some promising potential biological control agents have been identified (Fowler et al., 2013; Macedo et al., 2016).

The economic impacts of T. fluminensis have not been quantified.

The impact of T. fluminensis on the natural environment has been studied most extensively in New Zealand, where it has affected the soil and vegetation of the forest remnants it invades. At high biomass, T. fluminensis increases litter decomposition and alters nutrient cycling (Standish et al., 2004). Leaf litter placed in bags beneath mats of T. fluminensis decomposed at almost twice the rate of litter placed outside the mat. The impact of T. fluminensis on decomposition was evident through the reduced litter layers in T. fluminensis-affected areas relative to non-affected areas, despite similar quantities of leaf litter fall into each habitat. Moreover, there was increased plant-available nitrate in T. fluminensis-affected relative to non-affected plots. Also, the annual uptake of nutrients by T. fluminensis was a significant proportion of the total nutrient inputs into the system via litter fall, which, with the exception of calcium, exceeded the amounts of these nutrients held within the forest litter layer, but was only a small amount of these nutrients held within the topsoil. The litter quality and microclimate of T. fluminensis are the likely mechanisms by which it alters these ecosystem properties (Standish et al., 2004). T. fluminensis prevents native forest regeneration and, given its potential to persist, T. fluminensis-affected forests may eventually become shorter in stature as a result of the tall-canopy species being unable to replace themselves (Standish, 2002b).

Dogs and some humans have an allergic contact reaction to T. fluminensis (Marsella et al., 1997), but it has become part of the cultural landscape in New Zealand - also appearing in the forest scenes of the internationally acclaimed 'Lord of the Rings: Fellowship of the Ring' film.

The main use of T. fluminensis is as a very popular and easy-care houseplant, available with leaves of various shades. However, it was first introduced to New Zealand by a farmer to stabilize a steep bank and as such may also have been used for erosion control in other areas.

Cultural Control

Cattle and poultry browse T. fluminensis and could be used to control invasions but they also damage other forest plants and disturb the soil in the process. An alternative approach is tree planting to enhance canopy cover and so reduce light availability to T. fluminensis (Standish, 2002a). There is experimental evidence to show that shading (artificial) causes a reduction in T. fluminensis biomass (Standish, 2002a). In addition, 'armouring' the edge of forest remnants at risk of invasion, by planting a buffer zone or shelterbelt around the forest remnant that can protect against windthrow.

Mechanical Control

Hand weeding and rolling the weed up like a carpet are considered suitable for removal of small infestations (Porteous, 1993). Care must be taken to remove every last piece because even the smallest fragment may regrow. In heavily infested forest remnants, gaps left by removal of T. fluminensis are likely to be filled by other invasive species (Standish, 2002a).

Chemical Control

Chemical control by herbicides is considered a practical means of controlling large infestations of T. fluminensis (McCluggage, 1998). However, re-spraying is often necessary (Standish, 2002a) and one of the most widely used herbicides, triclopyr, could have detrimental effects on wildlife (Standish et al., 2002b).

Biological Control

T. fluminensis has been identified as a good candidate for biological control in New Zealand because it is widespread and the risk of non-target effects are minimal to non-existent (Standish, 2001). Reducing both the total weed biomass and re-invasion of other weeds are the biggest challenges for a biological control programme (Standish, 2001). The gradual reduction of T. fluminensis that is likely to occur with biological control may reduce the chance of invasion by other weeds. Several surveys for potential biological control agents have been conducted in Brazil with the result that the chrysomelid beetle Neolema ogoblini was released in New Zealand in 2011 (Fowler et al., 2013; Jackman et al., 2015), further field testing of the beetles Neolema abbreviata and Lema basicostata are underway (Fowler, 2013), and the fungi Kordyana brasiliensis and Uromyces commelinae are showing potential for biological control (Macedo et al., 2016).

Integrated Control

A combination of chemical and manual removal methods has been used with success in New Zealand, but has required repeated efforts to ensure continued control (Anon., 1995). The key to successful control of T. fluminensis is to reduce light availability by improving canopy cover that also reduces invasion by other weeds (Standish, 2002a). This might be achieved by integrating biological control and tree planting to improve canopy cover.