Northern Prairie Wildlife Research Center
We could find no descriptions of L. salicaria control efforts in Europe. Nevertheless, Morse and Palmer (1925) listed purple loosestrife as a semiaquatic weed in Great Britain and recommended the following control procedures for this group of weeds: "Weeds along the margins of ditches and watercourses should be cut regularly, and ditches thoroughly cleared in the autumn." Similarly, Woodford and Evans (1965) listed L. salicaria as an emergent weed along British waterways.
The pioneer work of Louis-Marie (1944) mentioned earlier in this paper was the first report of the control of purple loosestrife in North America. Following a year of preliminary study during 1942, Louis-Marie established an elaborate set of 32 field plots in the floodplain pastures north and east of Pierreville, Quebec, on the south shore of Lac St.-Pierre. In the summer of 1943, he applied mowing, crushing, hand-pulling, flame torch, and chemical treatment to these plot groups. Unfortunately, his efforts were interrupted after the end of the 1943 field season, leaving the critical question (frankly recognized by Louis-Marie) of rootstock recovery in the next growing season a suspended judgment. Nevertheless, his first-year results indicated that repeated mowing would be one of the most effective means of reducing L. salicaria infestations in pastures. His application of fire did not include the usual tactic of igniting dead surface litter in fall or spring. Rather, green plants were hit directly with a flame torch during the growing season. Louis-Marie thought that the flame torch results were parallel to the effects of mowing, but more costly to apply. Similarly, his attempts with crushing and hand-pulling were thought not to be cost effective on large areas. Sodium chloride and chlorate were the chemicals tested; although L. salicaria was more susceptible to desiccation than grasses or sedges, the differences in mortality were not great enough to recommend further efforts.
In addition to his experimental work, Louis-Marie recommended some control methods that were based on the experiences of local farmers and his own field observations. Noting that the most dense stands of purple loosestrife were in areas of soil that were saturated by annual flooding, Louis-Marie recommended that a drainage system be constructed to reduce the vigor of what he referred to as "a vast Salicaria colony of five square miles." He also recommended a schedule of repeated mowing in concert with continuous grazing pressure to reduce the vigor of the colony. In areas of heaviest infestation, he suggested that deep discing followed by harrowing be used to kill rootstocks. Louis-Marie recognized that rootstocks upended by the harrow might make the work of mowing very difficult. The only subsequent mention that we can find of the Lac St.-Pierre area was from Gagnon (1953), who reported that despite many years of control efforts, purple loosestrife was still a serious pasture weed at Lac St.-Pierre.
Although the control methods employed by Louis-Marie were acceptable for agricultural land (drainage, discing, repeated mowing), they would not be compatible with the maintenance of natural areas or with most wildlife habitat objectives, where the need to maintain community integrity recommends minimum impact methods. It is nonetheless unfortunate that Louis-Marie was not able to continue his work. With the possible exception of Rawinski's work (1982), no one has attempted such an ambitious field study of L. salicaria in North America.
By 1960, more than 1,000 small impoundments had been created in New York State to provide new wetland habitat, particularly for waterfowl (Emerson 1961). Unfortunately, by this time L. salicaria was so widely distributed in the uplands of the lower Hudson district that McKeon (1959) reported "a large percentage of marshes in the district have an almost pure stand of purple loosestrife (Lythrum salicaria) which provides little food but does give some cover." McKeon chose a 4.9-ha (12-acre) marsh constructed in 1952 as the site of L. salicaria control studies. By 1955, the central portion of this marsh had become "almost completely dominated by purple loosestrife with a few sedges interspersed." Water level manipulation, burning (in winter), and cutting at surface and subsurface were attempted in sequence, with no success. Six different commercial herbicides (Weedazol, 2,4-D, 2,4,5-T, Karmex 40, 2-4 Dow, and Kuron) were sprayed on trial plots from 1955 to 1958. Although 100% top kill was obtained with three of the sprays, regeneration from rootstocks frustrated control attempts. McKeon concluded that none of the herbicides would be cost effective in management operations. Reports from this district in 1980 (G. Cole, personal communication) indicated that purple loosestrife continued to be a serious problem in the lower Hudson wetlands.
Although L. salicaria was not a problem in the uplands of central New York in the early 1950's, it had been long established along the New York State Barge Canal (Dudley 1886; Wiegand and Eames 1926) and had become troublesome in two wetland management areas bordering the canal. L. S. Smith (1959) reported on problems with purple loosestrife encroachment into impoundments of the Montezuma NWR following maximum drawdown in 1954 and 1955 to achieve carp (Cyprinus carpio) control; he used a truck-mounted sprayer to apply ammonium sulfamate to stands along impoundment shorelines. This resulted in a complete kill of plants within reach of the spray rig; nevertheless, plants in the pool interior could not be reached and the infestation continued to increase on the refuge. Smith considered purple loosestrife to be a serious threat to impoundment manipulation. Meanwhile, in nearby Howland Island Game Management Area, R. H. Smith (1964) began work on loosestrife control in a set of shallow waterfowl impoundments that dropped from a full head of 60 cm in spring to nearly dry soil in August. He was able to obtain partial control with silvex (2[2,4,5-trichlorophenoxy] propionic acid) and mixtures of 2,4-D (2,4-dichlorophenoxyacetic acid) and 2,4,5-T (2,4,5-trichlorophenoxyacetic acid); however, he found it necessary to repeat spraying in subsequent seasons to control resprouting and seedling growth. Smith (1964) concluded that if water levels in shallow impoundments were to be manipulated in the presence of purple loosestrife, spraying of mature plants and seedlings would need to be carried out during drawdown and repeated following reflooding.
Similar problems with L. salicaria in shallow impoundments have been reported from Massachusetts (G. Gavutis, Parker River NWR, unpublished report) and Ohio (R. Kroll, Sandusky Bay, personal communication). Purple loosestrife also escaped into Delta Marsh (H. A. Hochbaum, personal communication; Friesen 1966) where it was difficult to control. More recently, reports have come from the Ada County Weed Control Office at Meridian, Idaho, that L. salicaria invaded a wet pasture bordering the Boise River in 1972 (W. E. Hartman, personal communication).
A common pattern can be seen in these early experiences with the control of purple loosestrife in marsh basins in eastern and midwestern North America. First, a long lapse of time has occurred (50 years or more) between the earliest records of local occurrence and the recognition that a serious infestation had become established. This may suggest a period of acclimation and the selection of ecotypes, or it may simply be a reflection of the relatively slow spread of the plant from local seed sources. Second, all of the serious infestations occurred under conditions that suggest that waterborne seeds or propagules were responsible for dominance of L. salicaria in the local habitat. Third, all of the habitats dominated by loosestrife were occupied by native vegetation that was growing under stressed or disturbed conditions. Drainage of deep-water basins, flooding of shallow basins, or seasonal drawdowns of impoundment pools were all characteristic of the areas of heavy infestation in northeastern North America. Last, once the local infestation was established, all available control measures were inadequate in providing control with a reasonable expenditure of funds and effort.