In: Van Driesche, R., et al., 2002, Biological Control of Invasive Plants in the Eastern United States, USDA Forest Service Publication FHTET-2002-04, 413 p. 11 PURPLE LOOSESTRIFE B. Blossey Department of Natural Resources, Cornell University, Ithaca, New York PEST STATUS OF WEED Purple loosestrife, Lythrum salicaria L., (Fig. 1) is a weed of natural areas and its spread across North America has degraded many prime wetlands result- ing in large, monotypic stands that lack native plant species (Thompson et al., 1987; Malecki et al., 1993). Established L. salicaria populations persist for de- cades, are difficult to control using conventional tech- niques (chemical, physical, and mechanical), and con- tinue to spread into adjacent areas (Thompson et al., 1987). Purple loosestrife has been declared a noxious weed in at least 19 states. Nature of Damage Economic damage. With the exception of reduced palatability of hay containing purple loosestrife and Figure 1. Purple loosestrife stand. reduction of water flow in irrigation systems in the (Photo by B. Blossey.) West, purple loosestrife does not cause direct eco- nomic losses. Indirect losses accrue due to reductions in waterfowl viewing and hunting opportunities. Ecological damage. The invasion of L. salicaria alters biogeochemical and hydrological processes in wetlands. Areas dominated by purple loosestrife (Fig. 2) show significantly lower porewater pools of phos- phate in the summer compared to areas dominated by Typha latifolia L. (Templer et al., 1998). Purple loosestrife leaves decompose quickly in the fall re- sulting in a nutrient flush, whereas leaves of native species decompose in the spring (Barlocher and Biddiscombe, 1996; Emery and Perry, 1996; Grout Figure 2. Wetland dominated by purple et al., 1997). This change in timing of nutrient release loosestrife. (Photo by B. Blossey.) at a time of little primary production results in sig- Specialized marsh birds such as the Virginia rail nificant alterations of wetland function and could (Rallus limicola Vieillot), sora (Porzana carolina L.), jeopardize detritivore consumer communities least bittern (Ixobrychus exilis Gmelin), and Ameri- adapted to decomposition of plant tissues in spring can bittern (Botaurus lentiginosus Rackett), many of (Grout et al., 1997). which are declining in the northeastern United States 149 Biological Control of Invasive Plants in the Eastern United States (Schneider and Pence, 1992), avoid nesting and for- lations in all but the eastern United States (the oldest aging in purple loosestrife (Blossey et al., 2001a). infested area) still expanding, In the Northeast and Black terns (Clidonias niger L.), once a common Midwest, a significant portion of the potentially avail- breeding species at the Montezuma National Wild- able habitat has been invaded. life Refuge in upstate New York, declined and be- came locally extinct by 1987. The local extinction coincided with a population explosion of purple loos- BACKGROUND INFORMATION estrife from few individuals in 1956 to a coverage of ON PEST PLANT more than 19% of the total area (600 ha), represent- ing 40% of the emergent marsh habitat in 1983 (T. Taxonomy Gingrich, pers. comm.). Another wetland specialist, Purple loosestrife is a member of the Lythraceae (the the marsh wren (Cistothorus palustris Wilson), was Loosestrife family), with highly variable growth form conspicuously absent in purple loosestrife-dominated and morphology. Main leaves are 3 to10 cm long and wetlands but used adjacent cattail marshes (Rawinski can be arranged opposite or alternate along the and Malecki, 1984; Whitt et al., 1999). The federally squared stem and are either glabrous or pubescent. endangered bog turtle (Clemmys muhlenbergi The inflorescence is a spike of clusters of reddish- Schoepff) loses basking and breeding sites to en- purple petals (10 to15 mm in length). Flowers are tri- croachment of purple loosestrife (Malecki et al., morphic with short, medium, and long petals and sta- 1993). mens. Many ornamental varieties have been devel- Purple loosestrife is competitively superior over oped, some through introgression with the native L. native wetland plant species (Gaudet and Keddy, alatum (Ottenbreit and Staniforth, 1994). Until re- 1988; Weiher et al., 1996; Mal et al., 1997). The spe- cently, Lythrum virgatum L. was treated as a sepa- cies is dominating seedbanks, particularly in areas rate species also introduced from Europe but the spe- with established purple loosestrife populations (Well- cies is now considered a synonym of L. salicaria ing and Becker, 1990; 1993).The fact that expanding (Ottenbreit and Staniforth, 1994). Further details can purple loosestrife populations cause local reductions be found in Mal et al., (1992). in native plant species richness has been demonstrated by the temporary return of native species following Biology the suppression of L. salicaria through use of herbi- Purple loosestrife needs temperatures above 20°C and cide (Gabor et al., 1996). However, without the con- moist open soils for successful germination. Seedlings tinued use of herbicides, purple loosestrife re-invades grow rapidly (>1 cm/day) and plants can flower in and re-establishes dominance within a few years their first growing season. Established plants can tol- (Gabor et al., 1996). In areas where the distributions erate very different growing conditions, including of L. salicaria and of the native winged loosestrife, permanent flooding, low water and nutrient levels, Lythrum alatum Pursh., overlap, the taller, more and low pH. Plants can grow in rock crevasses, on conspicuous purple loosestrife reduces pollinator gravel, sand, clay and organic soils. Plants develop a visitation to L. alatum resulting in significantly re- large, laterally branching rootstock with starch as the duced seed set of L. alatum. (Brown, 1999). main form of nutrient storage (Stamm-Katovitch et Extent of losses. Direct losses are difficult to al., 1998). Mature plants can develop rootstocks of quantify due to lack of long-term monitoring pro- heavier than 1 kg and can produce more than 30 an- grams and data. nual shoots reaching a maximum height of more than Geographical Distribution 2 m. Plants are long lived and mature plants may pro- duce more than 2.5 million seeds annually, which re- Lythrum salicaria now occurs in all states of the main viable for many years. Spread to new areas oc- United States, except Florida, Alaska, and Hawaii, curs exclusively by seed, which is transported mainly and in nine Canadian provinces. The abundance of by water but also adheres to boots, waterfowl and L. salicaria varies throughout this range with popu- other wetland fauna. 150 Purple Loosestrife Analysis of Related Native Plants in the Eastern basin (Blossey, 1995b). Additional surveys were con- United States ducted in North America from Maryland to Ne- braska (Hight, 1990). The Lythracea belong to the order Myrtales of which four families (Lythraceae, Thymelaceae, Onagraceae, Natural Enemies Found and Melastomataceae) are native to much of North America. Within the Lythraceae, 12 species (exclud- No native or accidentally introduced herbivores with ing L. salicaria) belonging to the genera Ammannia, the potential for control of L. salicaria were found in Cuphea, Decodon, Lagerstroemia, Lythrum, Rotala, North America (Hight, 1990). More recently, sev- and Didiplis (Peplis) occur in the northeastern Unites eral native pathogens have been evaluated for their States (Gleason and Cronquist, 1991). With the ex- potential as biological control agents (Nyvall, 1995; ception of Didiplis diandra (Nutt.), water purslane, Nyvall and Hu, 1997). Surveys in Europe identified all species of the Lythraceae covered by Gleason and more than 100 different insect species most com- Cronquist (1991) were used in the host specificity monly associated with purple loosestrife (Batra et al., testing (Blossey et al., 1994a, b; Blossey and 1986), but only nine species were evaluated in more Schroeder, 1995). detail (Blossey, 1995b). Host Range Tests and Results HISTORY OF BIOLOGICAL CONTROL Of the nine potential control agents identified in EFFORTS IN THE EASTERN Europe, six species were tested for their host speci- UNITED STATES ficity, against 48 test plant species in 32 genera (for a complete list of test plants taxonomically associated, Area of Origin of Weed associated wetland plants, and important agricultural plants see Blossey et al., 1994b). This selection was Lythrum salicaria has distribution centers in Europe based on literature reports of their specificity, their and Asia. The European distribution extends from distribution and availability in the field, and initial Great Britain across western Europe into central observations of their impact on purple loosestrife Russia with the 65th parallel as the northern distri- performance. The selected species were the root-min- bution limit (Tutin et al., 1968). Purple loosestrife is ing weevil, Hylobius transversovittatus Goeze; two common throughout central and southern Europe leaf beetles, Galerucella calmariensis L. and and along the coastal fringe of the Mediterranean Galerucella pusilla Duftschmidt; a flower-feeding basin. In Asia, the main islands of Japan are the core weevil, Nanophyes marmoratus Goeze; a seed-feed- of the species native range, with outlying populations ing weevil, Nanophyes brevis Boheman; and a gall extending from the Amur River south across the low- midge, Bayeriola salicariae Gagné. lands of Manchuria and other parts of China to Host specificity