Northern Prairie Wildlife Research Center
Shamsi and Whitehead (1974a) described the usual habitats of L. salicari. in Europe as "confined to wet, marshy places, coastal areas, stream banks and ditches." In England, Pearsall (1918) reported purple loosestrife to be frequent in reed-swamps and abundant in mixed fens where its associates were Typha latifolia, Phalaris arundinacea, and Carex elata, among others. Pearsall also listed L. salicaria as locally abundant in open carr where Alnus glutinosa and Salix cinerea were subdominant, decreasing to "frequent" in closed carr where S. cinerea became dominant. Purple loosestrife was also reported "along the margins of bogs near Sphagnum, Erica, Tetralix, Narthecium ossifragum and other calcifuge species" (Shamsi and Whitehead 1974a).
L. salicaria's affinity for wetland habitats in Europe is closely reflected in its invasion of North American sites. Freshwater marshes, open stream margins, and alluvial floodplains are its optimum habitats in the northeastern and north-central United States and adjacent Provinces in Canada. Most of the associates of L. salicaria in Europe and Asia are also abundant in temperate North America. Not surprisingly, they are also its most important associates in North American wetlands. Table 1 was constructed from photo and field notes of the senior author describing 45 sites across North America. All species occurring within a 1-m radius of the L. salicaria plant were listed as associates. Although the data represent inland habitats more than coastal wetlands, a pattern of association is apparent. Typha spp. were the most common associates, but declined somewhat in importance from east to west. Overall, T. latifolia occurred at 47% of the collection sites. Narrow- leaved cattail (T. angustifolia) occurred at only 4 of the 45 sites (9%). Reed canarygrass was overall the second most important associate. It increased in importance from east to west and became the most typical associate of L. salicaria in the Pacific Northwest. Sedges and rushes were the next most common associates in all three regions; Salix spp., Phragmites australis, and Equisetum fluviatile were present at about 9% of all sites.
(N = 15)
(N = 15)
(N = 15)
(N = 45)
|Typha latifolia, broad-leaved cattail||53||40||47||47|
|Phalaris arundinacea, reed canarygrass||27||47||60||45|
|Carex spp., sedge||20||27||13||20|
|Scirpus spp., bulrush||13||7||27||16|
|Salix spp., willow||0||13||20||11|
|Equisetum fluviatile, horsetail||0||0||27||9|
|Typha angustifolia, narrow-leaved cattail||13||13||0||9|
|Phragmites australis, reed grass||13||13||0||9|
|Cyperus sp., flatsedge||0||0||20||7|
|Alisma plantago-aquatica, water-plantain||7||7||0||5|
|Urtica dioica, stinging nettle||0||7||7||5|
|Sparganium eurycarpum, bur-reed||7||0||0||2|
|Agrostis gigantea (alba), redtop bentgrass||7||0||0||2|
|Acorus calamus, sweetflag||7||0||0||2|
|Populus deltoides, cottonwood||7||0||0||2|
|Chenopodium album, pigweed||0||7||0||2|
|Rubus spp., bramble||0||0||7||2|
|Euphorbia spp., spurge||0||0||7||2|
|Impatiens capensis, touch-me-not||0||7||0||2|
|Cornus stolonifera, red-osier dogwood||0||7||0||2|
|Convolvulus arvensis, field-bindweed||0||7||0||2|
|Verbena hastata, blue vervain||0||7||0||2|
|Solanum dulcamara, bitter nightshade||0||0||7||2|
|Eupatorium perfoliatum, boneset||7||0||0||2|
|Solidago sp., goldenrod||0||7||0||2|
|Aster sp., aster||0||7||0||2|
|Cirsium arvense, Canada thistle||0||0||7||2|
|Lactuca scariola, prickly lettuce||0||7||0||2|
The remarkable successes of L. salicaria as a worldwide pioneer is reflected in a combination of attributes that enable it to spread and thrive in disturbed temperate-climate habitats. In addition to an elaborate means of sexual reproduction and prolific seed production, L. salicaria has a wide scope of dispersal mechanisms. Some of these modes are adapted to long-range jumps in distribution (i.e., seeds in plumage of migratory birds); others are well suited to vegetative spread during local perturbations (adventitious shoots and roots from clipped, trampled, or buried stems, Fig. 4). Moreover, L. salicaria's abundant propagules can establish themselves under a wide range of soil conditions, which enables the weed to colonize new surfaces caused by natural- or human-caused perturbations. Last, L. salicaria's ability to make morphological adjustments to changes in its immediate environment (development of aerenchyma on submerged stems; change in leaf morphology with decrease in light level) enables it to adjust to a wide range of seasonal or semipermanent changes in water levels and gives it a competitive advantage against other plants growing under these conditions.
|Fig. 4. In deep-water marsh habitat near Rome, Wisconsin, Lythrum salicaria attempted to spread by adventitious shoots arising from stems that had lodged onto a mat of floating vegetation (Lemna spp.), 15 July 1981.|
The ability of L. salicaria to invade and eventually dominate North American marsh and fen communities is unique among our herbaceous, temperate-zone weeds. If monospecific stands of purple loosestrife were shortlived, the problem would be very much reduced; however, many stands in the Northeast have been self-replacing for more than 20 years without apparent loss of vigor. In similar habitats in the British Isles, purple loosestrife may form dense monospecies stands in areas of disturbance, but these become mixed-species stands within a few years (Shamsi and Whitehead 1974a). L. salicaria's scattered occurrence in mixed-species stands also seems to be the usual situation in central Europe, where Bodrogkozy and Horvath (1977, 1979) studied the dynamics of mixed-species stands on the Tisza River floodplains in Hungary. Similarly, during field surveys in 1979 in seven western and southern European countries, Batra (personal communication) found L. salicaria growing as occasional plants in mixed-species stands. In the summer of 1984, two of the authors traveled for 8 weeks through southern, central, and northwestern Europe and Great Britain. With two exceptions, we observed L. salicaria to be no more than an occasional plant occupying a relict distribution in habitat remnants along ditches and waterways. The first exception was along the coastal marshes of southwestern Sweden, where individual plants became locally abundant along the disturbed upslope edges of vigorous Phragmites australis stands. The second exception was a monospecific stand of purple loosestrife on Weir Wood Reservoir in southeastern England. The willow-shaded borders of the reservoir had been cut in 1981-82 to create an open view of a waterfowl display area. A prolific growth of purple loosestrife dominated the site in 1983, but was already losing dominance to a vigorous resurgence of willow suckers by August of 1984.
In North America, L. salicaria has shown a sharply different pattern of community dynamics in relation to many of its mixed-species associates in Europe. In the late 1950's, a large percentage of 23 small wetland impoundments that had been created in the lower Hudson area of New York had become almost pure stands of purple loosestrife (McKeon 1959). As of 1980, these monospecies stands were still more or less unchanged (G. Cole, personal communication). In 1956, R. H. Smith (1964) began several years of experimental control of purple loosestrife on the Howland Island Game Management Area in central New York. Working in the same area about 20 years later, Rawinski (1982) reported that proposed water level manipulations (to enhance waterfowl habitat) had been hindered because of the threat of continued purple loosestrife infestation. On the northeastern coast of Massachusetts in the mid-1960s', L. salicaria reached sufficient density in freshwater impoundments on Parker River National Wildlife Refuge (NWR) to threaten their usefulness as waterfowl and shorebird habitat (H. J. O'Connor 1968, unpublished report). Despite attempts to reduce the vigor of these stands with salt-water intrusion, mowing, and other treatments, the area was still heavily infested in 1981 (G. Gavutis, personal communication).
Gilbert and Lee (1980) reported an interesting comparison of the population dynamics of purple loosestrife stands near Canberra in southeastern Australia and along coastal British Columbia near Vancouver. Both stands stretched along 500 m of lake or streamside. All plants were tallied each year to obtain style length distributions . Although the results on style length were somewhat confusing, two aspects of this study were noteworthy. First, there was very good evidence that some root crowns did not grow above ground in certain years: "in a dozen cases at Canberra and two at Vancouver, a plant was seen in one year, did not grow above ground in the next, but was present again at exactly the same place and with the same style length in the following year." Second, and germane to our present discussion, the number of plants in the Vancouver stand increased from 46 plants in 1969 to 82 plants in 1979. The results from the riverside near Canberra were altered by catastrophic flooding that reduced the initial population from 108 in 1966 to 6 in 1975. Nevertheless, recolonization returned the population to 28 in 1978, 50 in 1979, and 82 in the last year (1980) of field work.
Although we cannot evaluate their relative importance, several factors can be identified that may account for the stability of North American versus northern European stands of L. salicaria. First, the insect fauna foraging in North American stands is missing many key species that reduce the vigor of European plants (Batra et al. 1986). Second, although the muskrat has spread over much of Europe since its escape in Czechoslovakia in 1905 (Elton 1958), its selective foraging on Typha latifolia probably has a much greater impact in North American marshes where cattails are dominant compared with Europe where Phragmites australis is the dominant emergent aquatic. Last, with very little known of the genetics of European versus North American stocks of L. salicaria, we suggest that more adaptive and vigorous forms may have appeared in the North American populations.