Northern Prairie Wildlife Research Center
A mature, long-established purple loosestrife plant is often 2 m high and 1.5 m wide. From 30 to 50 herbaceous stems rise from a common rootstock to make the graceful, wide-topped crown that is characteristic of these old-aged clumps. Although lateral growth of the root crown offers new sites of origin for peripheral stems, a natural limitation of about 0.5 m seems to be the maximum diameter obtainable by the root crown. Some confusion exists in the literature as to whether L. salicaria rootstocks can send out rhizomes. Ohwi (1965) described L. salicaria in Japan as "rhizomatous"; however, this may refer to adventitious buds arising on lateral roots. Describing marsh vegetation in England, Pearsall (1918) included L. salicaria as one of three shoreline species that "have tough rhizomes capable of penetrating the interstices of the hard substratum." Morse and Palmer (1925) did not mention rhizomes, but referred to rootstalks as "creeping extensively." We have seen lodged stems that have become buried under 10 cm of silt and sand following an unusually heavy June rainstorm. By late July, new shoots were appearing from adventitious buds on the buried stems, giving a convincing appearance of spread by rhizomes (Fig. 3). We have excavated purple loosestrife root crowns in many parts of its range in the United States and Canada, but have found no evidence of spread by rhizomes.
|Fig. 3. Stems buried by a flash flood gave rise to adventitious shoots of Lythrum salicaria (left foreground) in this riparian pasture near Red Wing, Minnesota, 8 August 1978.|
Bodmer's (1928) thorough work on the developmental anatomy of L. salicaria contained the suggestion that growth of some stem tissue can survive above the rootcrown level for 2 or 3 years. We are unable to interpret these findings and can only comment that most standard botanical works describe L. salicaria as having perennial rootstocks and stout, erect, annual stems. There is evidence that surviving L. salicaria rootstalks occasionally fail to send up stems. Gilbert and Lee (1980) made annual counts of two populations and reported that in 14 instances "a plant was seen in one year, did not grow above ground in the next, but was present again … in the following year."
The trimorphic structure of purple loosestrife flowers attracted the attention of 19th-century naturalists. Darwin (1893) was intrigued by the plant and made some remarkably perceptive experiments and observations. He noted that three kinds of flowers occurred on L. salicaria plants and that three lengths of styles occurred with three combinations of anther lengths. He also noted that the three forms of flowers coexisted in wild populations in nearly equal frequencies. Many workers (East 1927, 1932; Haldane 1936; Levin and Kerster 1973) have since reported on inheritance and population genetics in Lythrum, but from the standpoint of weed ecology, one of the most significant contributions was by Halkka and Halkka (1974). They investigated the frequencies of the three style morphs in L. salicaria on small, isolated island populations in the Gulf of Finland. Despite the small size (0.1 km2) and recent age of some of the habitats (± 100 years) all 16 populations studied contained the three style morphs in roughly 1:1:1 frequencies. They compared their findings with a summary of style morph frequencies from Great Britain, Germany, Switzerland, and the United States assembled by Schoch-Bodmer (1936) and concluded that the nearly identical results over a vast geographic area indicated that the vitality of the various genotypes did not depend on ecological conditions. Moreover, the highly similar frequencies indicated that sexual reproduction was of overwhelming importance in these populations. Although not sure how much of this gene flow could be attributed to pollen dispersal, Halkka and Halkka (1974) stated that seed dispersal seemed adequate "to keep total gene flow efficient." They also noted that their findings were in contrast with the report by Kuusvouri (M.S. thesis, University of Helsinki, Finland, 1960) of dense stands of L. salicaria wherein vegetative reproduction was common. These stands were on relatively fertile shore meadows and were characterized by a high frequency of mid-style morphs. The work of Shamsi and Whitehead (1974a) in Great Britain and our own observations in North America supported the findings of Halkka and Halkka (1974), that is, with the exception of locally disturbed areas (that may account for Kuusvouri's findings), sexual reproduction is of overriding importance in the establishment and spread of L. salicaria. Similarly, if asexual reproduction is of more than local importance, it occurs equally in all stylar morphs.
Heuch (1980) set up a simulation model to test the effect of restricted population size on the rate of random loss of various style morphs in L. salicaria. He concluded that isolated populations of 20 or more individuals would generally be quite stable and postulated that variations in the frequencies of the three forms may be caused by the unequal disappearance of one or more of the forms because of random effects.
The seed output of a purple loosestrife plant is dependent on its age, size, and vigor. Shamsi and Whitehead (1974a) gave the average number of seed capsules for their experimental plants as 900, with a yield of about 120 seeds per capsule. These data probably refer to first-year plants with single stems. Our observations on three mature (4-5 years old) transplants (from Nebraska and Wisconsin wetlands) at Fort Collins, Colorado, showed the mean number of stems to be 30, with 1,000 capsules per stem, and 90 seeds per capsule; the mean number of seeds produced per plant was estimated at 2,700,000. Capsules from early flowers produced ripe seed while the plants were still green. The flat, thin-walled seeds had no endosperm and were about 400 × 200 microns (Fig. 1).
Ridley (1930) noted that L. salicaria seeds sink upon being thrown into water, but rise to the surface following germination. He considered seed dispersal to be largely by means of floating seedlings. We agree with his conclusion on mode of dispersal, but find that not all seeds sink upon falling on water; some dispersal could be by floating, ungerminated seeds. Although Ridley did not include L. salicaria among plants whose seeds were dispersed by wind, Shamsi and Whitehead (1974a) declared that L. salicaria has wind-dispersed seeds. Nilsson and Nilsson (1978) used Sernander's (1901) work to classify L. salicaria as a species that was dispersed by wind over snow and ice. Surely the seeds are light enough (0.5-0.6 mg) to be carried by a strong wind, but we have observed that the densities of seedlings fall off sharply within the first 10 m from the parent plant, suggesting a very limited role for wind dispersal. Moreover, distribution from parent plant is most often downslope rather than downwind. The low mass and small size of L. salicaria seeds make them likely to be spread in mud adhering to aquatic wildlife, livestock, treads of all-terrain vehicles, boots of hunters, or on the hulls of skiffs or barges. They could also survive ingestion by waterfowl and other marsh birds or be carried in the cooling systems of outboard motors. Nevertheless, with the exception of Kerner (1902), there is no evidence to support any of these possibilities. Kerner's studies in Europe included L. salicaria in a list of 21 plants whose seeds he found in mud "obtained from the beaks, feet and feathers of swallows, snipe, wagtails, and jackdaws …"
Correll and Correll (1975) mentioned ducks eating the seeds of Decodon and Lythrum. Torrey (1931) suggested that the relatively rapid, upslope spread of L. salicaria into small riverine wetlands was caused by blackbirds (Icteridae) that had eaten the seeds of plants in the Hudson River lowlands. Last, the senior author has observed the spread of Lemna spp. into adjoining impoundments along the foraging trails of muskrats (Ondatra zibethica). In a similar way, seeds or propagules of L. salicaria could be spread by muskrats, waterfowl, or snapping turtles (Chelydra serpentina), as they move over dikes into adjoining impoundments.
Knowledge of the longevity and viability of purple loosestrife seeds and propagules under field conditions would be very useful to land managers attempting to cope with an infestation. As a test of their experimental procedures, Shamsi and Whitehead (1974a) measured the viability of the seeds used in their work. They compared seeds stored 3 years in dry condition in a refrigerator at 3-4°C with freshly collected seeds. The stored seed began germination 2 days after fresh seed and reached maximum emergence (80 + %) on or about day 13. Fresh seeds reached maximum emergence (90 + %) on or about day 17. None of these differences was statistically significant. To our knowledge, no work exists on the survival and longevity of propagules. The low reserve of stored energy in L. salicaria seeds suggests that drifting germinated seeds would not survive beyond a few weeks; however, this matter needs study.
Shamsi and Whitehead (1974a) reported the critical temperature for germination of L. salicaria seeds to lie somewhere between 15 and 20°C, with no germination occurring below 14°C. They tested germination in short (9-h light, 15-h dark) versus long (16-h light, 8-h dark) day lengths and found no significant differences in percentage or speed of germination. Mitchell (1926) tested the effect of two light levels (diffuse and dark) in germination trials (at 20-25°C) that used natural light and dark intervals for the diffuse treatment. In successive years, she found germination percentages of 84 and 67 in seeds in diffuse light versus 4 and 9 for seeds in dark. L. salicaria can germinate successfully on substrates with a wide range of pH. Shamsi and Whitehead (1974a) indicated germination down to pH 4.0. We observed successful development of young plants (germination implied) at pH 9.1 on the calcareous slope of an infested wetland near Powell, Wyoming.
Floating seeds or propagules must fall or lodge against moist soil to begin a successful establishment. Bodmer (1928) reported that a 3-day-old seedling was about 3 mm long from the tip of the cotyledon to the extremity of the root. By 10 days, the seedling was about 6 mm long, with most of the growth in the primary root. By 20 days, the seedling was about 40 mm; the first true leaves appeared, and lateral and secondary roots had developed. By 25 days, stem elongation began with a 1-2-mm growth of the epicotyl; true leaves and vascular tissue were well developed. At this time, although the aboveground portion of the plant is less than 10 mm high, the young seedling could be well established. Directional growth is clearly indicated in Bodmer's sketches. We have observed this tendency in the elongating root crowns of first-year seedlings and noted that it was still apparent in 2- and 3-year-old plants. This directional growth may be the basis for the puzzling references mentioned earlier to the "rhizomatous" character of L. salicaria.
Most seedling establishment occurs in late spring and early summer when temperatures are high (Shamsi and Whitehead 1974a). These seedlings germinate from seeds or propagules that have survived at least one winter. Shamsi and Whitehead (1974a) found that summer-germinated seedlings in southern Great Britain did not produce more than four or five pairs of leaves and did not survive the following winter.