Source: California Invasive Plant Council

URL of this page:

Invasive Plants of California's Wildland

Lythrum salicaria
Scientific name   Lythrum salicaria
Additional name information: L.
Common name purple loosestrife
Synonymous scientific names none known
Closely related California natives 15
Closely related California non-natives: 0
Listed CalEPPC List B
By: Carri Benefield

Listed as noxious by 16 states


Distinctive features:

Purple loosestrife (Lythrum salicaria) is a perennial wetland herb, typically less than five feet tall with showy spikes of reddish purple flowers. Each flower has five to seven pink-purple petals surrounding a small yellow center. Plants enlarge at their bases each year with more stems, sometimes becoming a rounded bush-like clump of thirty to fifty stems arising from a single root stock. Leaves are long and narrow, opposite to whorled, and closely attached to the four-sided stem.

Purple loosestrife may be distinguished from similar looking species by its generally larger size, opposite to whorled leaves, and less widely spaced flowers. It may be confused with other native species such as fireweeds, vervain, or the closely related Lythrum californicum, so accurate identification should be made before control measures are attempted. Fireweed (Epilobium angustifolium) has spikes of magenta flowers, but these spikes are narrower (four to five inches wide at the base) and more conical. Fireweed stems are round, and the leaves are alternate. Blue vervain (Verbena hastata) has smaller purple flower spikes, and the edges of the leaves are toothed. L. californicum has smaller, narrower leaves and half (six versus twelve) the number of stamens.


Lythraceae. Perennial herb. Stems: 1-5 ft (0.3-1.5 m), occasionally to 10 ft (3 m) tall, erect stems; four-sided, smooth to lightly hairy, woody with age, persisting through winter. Leaves: opposite or whorled, 2-6 in (5-14 cm) long, generally two-ranked, lanceolate to narrowly oblong, edges smooth, sessile (Hickman 1993). Inflorescence: terminal, spike-like, composed of numerous flowers closely arranged in leaf axils of ascending flower-bearing branches (Rawinski et al. 1982). Flower: trimorphic in regard to the relative lengths of stamens and style; hypanthium cylindric; sepals <0.5 in (4-5 mm), deltate; corolla bright red-purple with 5-7 petals <0.5 in (8-14 mm); stamens generally 12, included or exerted. Fruit: dark brown seed capsules containing numerous small seeds <0.05 in (1 mm) (Hickman 1993).


Purple loosestrife can be found in scattered freshwater wetlands in northern and central California. Counties with infestations include Humboldt, Mendocino, Modoc, Shasta, Siskiyou, and counties in the Sacramento Valley and San Francisco Bay Area. Other infestations occur along rivers in Fresno and Kern counties, such as along the Kern River in the southern Sierra. It is also widespread in the eastern United States. It is common in disturbed wetland habitats, such as stream and river banks, edges of ponds, lakes, and reservoirs, flooded areas, ditches and roadsides, but it can colonize fairly pristine wetland areas, including marshes, wet prairies, meadows, pastures, and bogs. While loosestrife generally grows best in full sun in moist, organic soils, it can tolerate a wide range of soil moisture regimes (intermittent or continuous flooding up to 1.25 feet or 0.4 m), soil types (clay, sand, silt, and muck; pH 4-9), nutrient levels, temperature, and light (up to 50 percent shade) (Skinner et al. 1994). If water levels drop, loosestrife can easily become established, and populations can expand on exposed soils.



Purple loosestrife is native to Eurasia and was introduced into the northeastern United States in the early 1800s in ships’ ballast, as an herbal, as an ornamental plant, and by beekeepers. It spreads primarily by seed, but can grow from broken-off stem and root segments and can resprout from the root crown (Bender et al. 1987). Seed dispersal occurs primarily by wind and water (most rapidly by flowing water, but also by permanent standing water), and in mud attached to wildlife, boats, vehicle tires, and footwear (Heidorn 1990, Skinner et al. 1994).



Purple loosestrife is a hardy perennial that can rapidly degrade wetlands, diminishing their value for wildlife habitat. Once established, it forms extensive monotypic stands that displace native vegetation relied on by wetland species for food and habitat (Nelson and Getsinger 1994, Bender 1987). Infestations jeopardize the federally listed endangered bog turtle (Clemmys muhlenbergi) among other species, though not in California. Purple loosestrife also clogs waterways and wetlands used for boating and other recreational activities. Impacts on agriculture include changes in hydrology and soil conditions of wetland pastures and meadows and clogged irrigation systems. Purple loosestrife has the potential to infest rice fields. An estimated 190,000 hectares of wetlands, marshes, pastures, and riparian meadows are affected in North America each year, with an economic impact of millions of dollars.



Purple loosestrife flowers from late June through September. Seed set begins in mid-July and continues into late summer. New shoots arise the following spring from buds at the top of the rootstocks. Long-lived seeds are copiously produced (more than two million seeds per mature plant) and shed gradually from capsules through the winter, contributing to an immense seedbank. It has not been determined how long purple loosestrife seeds remain viable, but it is thought to be at least several years.


Most seedlings establish in late spring to early summer and produce a floral shoot up to one foot (30 cm) high the first year. Moisture is the most critical habitat requirement for seed germination. Critical temperatures for germination range from 60 to 72 degrees F (15-20 degrees C) (Skinner et al. 1994).

(click on photos to view larger image)



The best method for removal of purple loosestrife depends on the site, the age and size of the infestation, the importance of impacts to non-targeted species, and the type and quantity of resources available to gain control. Ideally, an integrated pest management (IPM) approach should be developed to take advantage of the best available control methods. All methods require appropriate timing and follow-up monitoring to be successful.

In natural communities with no known invasion, prevention and early detection are the best approaches. Potential habitat should be searched annually for newly established populations in late July through August, when blooming plants are easily spotted from a distance. In addition, public education efforts and avoidance of unnecessary disturbances of wetland habitat are desirable (Heidorn 1990, Mountain 1994).


Physical control:

Manual methods: In areas with individual plants and small localized stands (up to 100 plants), younger plants (one to two years old) can be pulled by hand. Plants should not be pulled after flowering, because this will scatter seed. Older plants, particularly those in loose soils, can be dug out by teasing roots loose with a cultivator, but great care should be taken to bag and remove all plant parts from the site because broken-off pieces can re-root (Heidorn 1990, Skinner et al. 1994). Dispose of plants by burning (preferable) or by drying.

Several methods have been shown to be ineffective and should be avoided: Cutting at water level and below water level has resulted in resprouting and reestablishment. Mowing, if timed correctly, has been shown to prevent flowering and seed set and reduce carbohydrate reserves, but it can result in dissemination of seeds and re-rooting of stem fragments (Heidorn 1990). Flooding is ineffective unless plants can be inundated throughout the summer, but this is detrimental to most other emergent species and promotes spread of loosestrife to shallow areas (Skinner et al. 1994). Disking and flooding has shown some success, but most areas cannot be drained and dried enough to support heavy equipment (Skinner et al. 1994). Burning does not kill buried rootstocks, which resprout later. Herbicide plus burning was attempted without success (Skinner et al. 1994).


Biological control:

Insects and fungi: Since the 1980s the U.S. Fish and Wildlife Service and the International Institute of Biocontrol have been screening, testing, and importing natural purple loosestrife enemies. In the early 1990s the USDA and Washington and Oregon state departments of agriculture gained approval for the release of five agents, three weevils and two leaf eating beetles (Galerucella spp.). California required extra testing (for potential impacts on crape myrtle, a widely used ornamental) before obtaining a permit in 1998 to conduct test releases of leaf-eating beetles in California. Considerable damage to purple loosestrife from the leaf-eating beetles has been reported in other states and in California test release sites. However, further research is necessary before approval of large-scale release of Galerucella spp. in California. Other states have reported that it can take four to six years for Galerucella spp. to become fully established.

Plant competition: Research on reseeding with competitive species is still in its early stages. It has been difficult to find species that compete successfully with purple loosestrife in a wide variety of conditions and that are not themselves invasive. Japanese millet plantings in Minnesota have been successful in the short term, but this non-native species regenerates poorly and must be replanted every year (Bender et al. 1987).


Chemical control:

In areas up to four acres with clusters in excess of 100 plants, the most effective herbicides available are glyphosate (as Roundup® and Rodeo®) applied in spot applications. Glyphosate is non-selective, so care should be taken to avoid contact with non-targeted species that are critical in recolonizing the site and inhibiting loosestrife reestablishment (Bender 1987, Skinner et al. 1994). Herbicides should be applied to the foliage during the blooming season (late June to early September). Effective applications have been reported at both flower initiation (usually late June) (Heidorn 1990, Nelson and Getsinger 1994) and after peak bloom (usually late August) (Balogh 1986, Rawinski 1982). Timing of application is essential for maximizing chemical effectiveness and preventing seed production. Glyphosate should be applied by hand sprayer as a 1.5 percent solution (2 oz/gal water). Rodeo®, registered for aquatic use, should also be applied as a 1.5 percent solution, with the addition of a surfactant approved for use over water, as specified on the label (Heidorn 1990).

2,4-D approved for aquatic use is the second most commonly used herbicide, despite its inconsistent ability to control purple loosestrife. This herbicide’s selectivity for broadleaf plants and low cost make it appealing. Applied before flowering (late May to early June), it is most effective in controlling first-year seedlings and preventing seed production in mature plants, but it does not kill the plants and many will likely resprout (Bender 1987; Skinner et al. 1994).

Triclopyr (as Garlon 3A®or Renovate®) is not not yet approved for aquatic use, but it has great potential because it is selective for broadleaf plants and does not harm grasses and most other monocots, which are important in wetland habitats. While some reports point to the inconsistency of Garlon 3A® in the field (Katovich et al. 1996), others indicate that it is effective and promotes regrowth by grasses and other wetland monocots (Nelson and Getsinger 1994). Garlon 3A® continues to undergo evaluation for its use in aquatic environments.