Northern Prairie Wildlife Research Center
Although we need quantitative measurements of the effects of various stages of L. salicaria invasion on the structure, function, and productivity of North American wetland habitats, the replacement of a native wetland plant community by a monospecific stand of an exotic weed does not need a refined assessment to demonstrate that a local ecological disaster has occurred. Nevertheless, a summary of existing information is needed. McKeon (1959) reported that a large percentage of shallow impoundment marshes in the lower Hudson area of New York had become solid stands of purple loosestrife and were thereby degraded as waterfowl production sites. In upstate New York, L. S. Smith (1959) noted that, with age, invading L. salicaria ultimately dominated wetland habitats with tall, dense, brushlike stands that were impenetrable to boats. These stands also excluded desirable waterfowl food plants (Fig. 13). R. H. Smith (1964) later reported L. salicaria's impact on New York marshes as, "in the wildlife marsh, with average water depths of 12 to 18 inches, purple loosestrife can take over and occupy much water area that should be occupied by waterfowl food plants. "Similarly, Friesen (1966) reported on the replacement of native food plants for ducks and geese in Delta Marsh, Manitoba, by L. salicaria. At the Montezuma NWR in New York, Rawinski and Malecki (1984) used line transects through adjacent stands of emergent purple loosestrife and cattail to measure wildlife usage. They found that muskrats and long-billed marsh wrens (Cistothorus palustris) used cattail stands almost exclusively, whereas red-winged blackbirds (Agelaius phoeniceus) clearly preferred purple loosestrife.
|Fig. 13. Purple loosestrife has displaced native emergent cover in front of this private shooting club in upstate New York, 3.2 km west of north from Savannah, 18 October 1968.|
The history of the encroachment of purple loosestrife into a waterfowl impoundment in central New York offers many insights into the potential impact of this weed in wetland habitats. In 1965, an experimental green-timber impoundment was established on the Montezuma NWR in a tract of muck-hardwood timber lying at the north end of Lake Cayuga (Haramis and Thompson 1984, 1985; Thompson et al. 1968). The forest floors of two adjacent 125-ha units were impounded with spring runoff water to a mean depth of 35 cm to encourage waterfowl nesting. These shallow waters were withdrawn in July or early August to relieve the forest community from possible water stress. Both experimental impoundments were compared with a 125-ha control area that remained undisturbed.
Dikes and ditches that carried water on and off the forest floor were constructed with a shelflike berm that was intended to surround the forest units with a ring of emergent and floating aquatic vegetation. This peripheral zone was to serve as foraging area and escape cover for broods reared in the flooded timber. In the first year of operation (1965), cattails (Typha latifolia) colonized the rims of all of the ditches. In the second year, cattails continued to spread along the impoundment water line, while the shallow berm area developed an ideal mix of open water with floating plants (Potamogeton spp., Utricularia vulgaris, and Lemna minor) and emergent aquatics (Alisma plantago-aquatica and Acorus calamus). Meanwhile, waterfowl—primarily wood ducks (Aix sponsa) and mallards (Anas platyrhynchos)—were nesting in natural tree cavities and on wind-throw mounds in densities of about one pair per hectare. Although early monitoring of the green-timber impoundment units indicated that these forest communities were adjusting to the new water regimen, purple loosestrife appeared along the impoundment rims in the second growing season (1966). The encroaching seedlings took root along the strand of mean full-pool line, suggesting water distribution of floating seeds or propagules. Beginning in 1966, we made biennial estimates of the percentage of green biomass made up by purple loosestrife along the dike slopes facing the east and west impoundments. By 1968, purple loosestrife had become widely established along dike shorelines, but made up less than 5% of total biomass (front cover: left photo). By 1972, L. salicaria was approaching dominance in several areas of emergent-aquatic vegetation, especially where muskrats had foraged among cattail. At this time the senior author estimated that, overall, purple loosestrife made up roughly 30% of the biomass of emergent aquatic and dike-slope vegetation. We were not able to make observations in 1976, but 2 years later, during a brief visit to the study site, the senior author found that L. salicaria had not only dominated the cattail zone, it had spread upward into the dike slopes and outward into the floating aquatic zone and now made up more than 90% of the biomass of the vegetation (front cover: right photo). The desirable mix of native food and cover plants had been displaced by L. salicaria. Figure 14 displays the establishment, expansion, and dominance of purple loosestrife as constructed from our biennial estimates of green biomass. Local establishment of seedlings probably occurred in the first year of flooding (1965); these scattered seedlings went unnoticed in 1965, but overwinter survivors gave rise to the solitary flowering stalks counted in 1966. A vast wetland complex adjacent to the experimental impoundment was infested with purple loosestrife and served as the water source for the green-timber impoundment units. This was doubtless the source of seeds and propagules pumped onto the experimental wetland at the beginning of each growing season. Furthermore, seeding was repetitive each year as pumping continued into the growing season to replace water lost to transpiration from the lush forest canopy. By the fifth year of operation (1970), purple loosestrife was widespread along the water control canals; it had increased to about one plant per 10 m² and made up about 10% of the biomass of dike vegetation. By this time, an accelerated phase of local expansion had begun that displaced more than half of the dike cover biomass within the next 4 years (1974). In another 4 years (1978), the rate of displacement slowed as the dike cover became a virtual monotype of purple loosestrife.
|Fig. 14. Estimated percent of total biomass of dike cover made up by Lythrum salicaria from 1965 to 1978, green-tree reservoir, Montezuma National Wildlife Refuge.|
Haramis and Thompson (1984) emphasized the need for direct and secure routes to accommodate the post-drawdown movements of flightless juvenile wood ducks to marsh areas where they can complete their development. The Montezuma green-timber impoundment layout was weak in this respect; the dominance of purple loosestrife on sites that were designed as brood habitat greatly aggravated the problem. Several (two to five) pairs of Canada geese (Branta canadensis) nested on dikes or in flooded timber from 1965 to 1969. Broods from these nests were reared in low, open dike cover (mixed grasses and forbs) until 1969. By 1978, L. salicaria occupied about half of this foraging area. In addition to the loss of forage, goslings became more vulnerable to their principal predator, the red fox (Vulpes vulpes). Foxes now had enough concealment cover to make close approach to the goslings possible. The value of dike and berm cover for muskrats was also greatly reduced as their principal food (Typha spp.) was replaced by L. salicaria.
Although the infestation and dominance of the green-tree impoundment duck-brood cover was dramatic, it was not as sudden as that experienced following the drawdown of a shallow waterfowl display pool at the Sapsucker Woods Sanctuary near Ithaca, New York. Seedling densities (3-6/m²) on exposed bottom sediments at Sapsucker Woods were similar to those at the Montezuma impoundments, but competition for space with existing native aquatic plants had denied purple loosestrife seedlings at Montezuma the opportunity to take over bare, exposed organic surfaces. The pool at Sapsucker Woods had been formed with shallow dikes that backed water into the edge of a relatively undisturbed hardwood forest. The pond bottom contour gently sloped toward the deeper waters of a waterfowl display pool. In 1955, wild-type purple loosestrife was planted on an open area on the deep side of the pond. Although the bottom contour of the pond was steep enough to prevent the spread of purple loosestrife into the pond, in the next 10 years it spread around the open edges of the pond and into openings in the forest canopy caused by the death of water-stressed trees. Meanwhile, a display-pool feeding program and accompanying high concentrations of migrant and resident waterfowl produced a heavy accumulation of spilled corn and waterfowl feces. The threat of botulism (D. Mills and M. Richmond, 1971-73, unpublished reports of New York Cooperative Wildlife Research Unit) made sediment removal an urgent necessity. Drawdown was begun on or about 20 July 1971; by early September (Fig. 15), the shallow flat in front of the forest edge had been colonized by a vigorous catch of 40-day-old L. salicaria seedlings. Delays with renovation problems extended the drawdown into October, by which time the young loosestrife seedlings were old enough to survive shallow flooding. Within 3 years following reflooding, the shallow pond zone was a solid stand of mature purple loosestrife.
|Fig. 15. Forty-day-old purple loosestrife seedling growing on bottom sediment exposed during renovation of waterbird display pool at Sapsucker Woods Sanctuary, Ithaca, New York, 19 September 1971.|
The impact of the invasion of L. salicaria on North American muskrat and mink (Mustela vison) populations has not been measured, but several interactions are probable. Muskrats themselves frequently hasten purple loosestrife dominance of local habitats through their selective foraging on its principal competitor, Typha latifolia (Fig. 16). In September of 1971, after a winter of muskrat abundance on the Montezuma NWR, the senior author observed muskrats foraging in an infested cattail stand. Typha latifolia stems in the vicinity of the active muskrat houses were heavily exploited, but L. salicaria stems were only partially cut. Many of the latter's stems were cut but not eaten; others were cut into short segments, partially gnawed into the periphery of the outer cortex, and discarded. Each of these partially eaten cuttings was a potential propagator of a new purple loosestrife plant. With rising water levels, many of these cuttings floated out into open water and were driven by wind and current against surface vegetation mats on adjacent shores. The removal of competing cattails, the rafting of cattail debris from muskrat foraging, and the spread of cut L. salicaria stems all suggested a favorable interaction for the dominance of the latter in an area that was largely deep-water cattail habitat. Furthermore, the shift from cattail to purple loosestrife foretold a permanent change in the carrying capacity of the habitat for muskrats and waterbirds. With decreasing prey species, a decline in mink numbers was another probable outcome.
|Fig. 16. Muskrat foraging in broad-leaved cattail has favored the establishment and survival of Lythrum salicaria seedlings, Montezuma National Wildlife Refuge, 30 September 1971.|
Purple loosestrife's ability to usurp wetland habitats is worrisome in view of the vulnerability of threatened native wetland plants and wildlife. Fernald (1940) toured the St. Lawrence flood plain below Montreal and reported that "… the formerly unique and endemic flora of the estuary is being rapidly obliterated by … flowering rush (Butomus umbellatus) and the purple loosestrife (Lythrum salicaria) both … without mercy for the insignificant endemics … " More recently, Coddington and Field (1978) suggested that crowding by L. salicaria may be partly responsible for the decline of a threatened bulrush (Scirpus longii) in Massachusetts. Rawinski (1982) observed that a rare inland population of dwarf spike rush (Eleocharis parvula) in inland New York seemed immediately threatened by invading purple loosestrife. Among vertebrates, several species deserve special attention.
The bog turtle (Clemmys muhlenbergi) is a small, uncommon, and poorly known reptile that is restricted to wetlands and bogs in the eastern United States (Bury 1979). The northern portions of its range have long been colonized by L. salicaria. Shallow, standing water and wet, open fen or meadow with muck substrate are favored habitats for the bog turtle. These are also sites that are readily invaded by purple loosestrife, particularly following disturbance to the habitat from grazing or drainage. The domination of these sites by tall, dense monospecific stands of L. salicaria causes a dramatic change in the physical as well as the trophic structure of the habitat (Kiviat 1978) and further aggravates the tenuous status of the bog turtle in the northern segments of its range.
The continuing invasion of purple loosestrife into glacial basins in the north-central States comes at a difficult time for the black tern (Chlidonias niger). These graceful birds have always been an abundant breeding species in the shallow ponds throughout the northern prairies and Great Plains (Bent 1921). Faanes (1979) recently reported an alarming decline of the black tern in Wisconsin. In addition to the habitat shrinkage described by Faanes (1979), the invasion of shallow pond borders by L. salicaria will probably render typical black tern nest sites unusable. Terns prefer open surface litter, such as the top of an abandoned muskrat house, as nest sites. These would also be likely places for purple loosestrife seeds, propagules, or muskrat cuttings to lodge and take root. Once the species becomes established, the open character of the black tern nest platforms is lost.
The canvasback (Aythya valisineria) has never recovered from the low levels of the great "duck depression" brought on by the disastrous drought of the early 1930's. The species has been nearly simultaneously beset with loss of nesting habitat in the prairie pothole region (Fig. 8) and a gradual decline in the quality of its restricted wintering grounds (Trauger 1974). Its preference for platform nests built over water in cattail, Typha latifolia, or hard-stem bulrush, Scirpus acutus (Stoudt 1982) makes it vulnerable to encroachment by purple loosestrife. Fortunately, L. salicaria seems to be slow in colonizing the prairie pothole region. A plant was collected at Neepawa in 1896 (Stuckey 1980), but the species did not begin to spread into Manitoba wetlands until after 1950. It may prove to be less vigorous than native plants in competing for space under the harsh and widely fluctuating climatic patterns of the northern prairies. The relative isolation of prairie pothole sloughs also makes spread by waterborne propagules a slow process. Nevertheless, it invaded Delta Marsh in 1955 and has required vigorous local control in disturbed areas along the marsh edge (H. Hochbaum, personal communication; Friesen 1966). With drawdown or drought, L. salicaria could threaten the deep-water nesting areas of the canvasback at Delta.
A recent report (Brezenski, personal communication, 31 August 1981) of an infestation on the Platte River floodplain near Elm Creek, Nebraska, indicates that purple loosestrife may become troublesome in Great Plains streams. Brezenski discovered fully developed stands around the margins of several ponds in the Blue Hole Wildlife Area. Two-thirds of the shoreline of the largest pond (Blue Hole) was bordered by a tall and vigorous stand of purple loosestrife that had completely displaced all other cover (Fig. 17).
|Fig. 17. Purple loosestrife infestation around the south rim of Blue Hole—a mitigation pond in sandhill crane Staging Area 2 near Elm Creek, Nebraska, 4 September 1981.|
All of the ponds in the Blue Hole area were constructed as mitigation units to replace riparian habitat destroyed by the construction of Interstate Highway 80. The pond substrates and margins were alluvial sand. Water quality of seepages supplying the ponds was excellent. Fishing was also attractive and bank fishermen trampled purple loosestrife stalks along the shoreline to make room for their sport. Unfortunately, the stalks that were trampled outward into the pond developed aerenchyma along the submerged stems, and sprouted adventitious root and stem buds. By these means, the species had moved outward over a shallowly (1:8) sloping bottom. Unusually high and prolonged flooding in the spring and early summer of 1983 scoured the shorelines of Blue Hole and greatly reduced purple loosestrife abundance; however, by the end of the 1985 growing season, L. salicaria was obviously a hardy and aggressive survivor. It had spread along the scoured outlet of Blue Hole and colonized new habitats on each side of the stream. It had also colonized an adjacent pond where it was displacing a native flatsedge (Cyperus erythrorhizos) along the pond margin (D. Q. Thompson, unpublished field notes).
Although it is too early to assess the impact of the Blue Hole infestation, the potential for habitat degradation is very serious. An ecological study of sandhill cranes (Grus canadensis) along the Platte and North Platte rivers (USFWS 1981) showed that about 80% of North American sandhill cranes use this area as a migratory rest stop enroute to their boreal and subarctic breeding grounds. Flood control and irrigation structures in Colorado, Wyoming, and western Nebraska divert about 70% of the Platte's annual waterflow, resulting in reduced scouring of sandbars and less shifting of alluvial sediments. Thousands of hectares that were formerly emergent aquatic vegetation are now overgrown with woody vegetation and unattractive to sandhill cranes (USFWS 1981). Krapu et al. (1984) estimated that about 0.5 million cranes use shallows and sandbars along 111 km of open river channel as night roosts. By using the estimates of Krapu et al. (1984) we calculated that about 0.4 million of these cranes use night roost sites downstream from the Blue Hole infestation. Management plans for the crane staging areas recognize the need to keep existing roosting sites open and to bring overgrown areas back to earlier seral states. The presence of purple loosestrife in many of these sites seriously complicates these management plans.
Other reports of L. salicaria in Nebraskan riparian habitats come from the Niobrara River along the north-central border. In 1980, the plant was collected at two sites by botanists from the University of Nebraska-Lincoln (R. B. Kaul, letter, 15 February 1982). More recently, K. Menzel (letter, 14 August 1984) reported that L. salicaria infestations are now more or less continuous along 20 km of the Niobrara from south of Norden to Meadville, with another concentration near the bridge north of Newport.
A last example of the successful colonization of a western riparian site comes from Wyoming. The first and, to our knowledge, only report of purple loosestrife from this State is from a field collection made by State botanists in 1979 (Lichvar and Dorn 1980). Visits to this site by the senior author in 1982 and 1983 showed a mature stand of purple loosestrife that had invaded about 1 ha of wetland. This slough had developed in a former dry wash that had become a permanent wetland with the establishment of the Shoshone Reclamation Project in the 1930's. About 30 cm of sediment and organic debris had accumulated in 50 years of seepage and runoff. Despite vigorous stands of Typha latifolia, Scirpus acutus, and Carex spp., L. salicaria had spread throughout the small slough and threatened to dominate the site (Fig. 11).