Rorippa columbiae

Rorippa calycina var. columbiae (S. Watson) Rollins:
United States (North America)

Note: This information is based on publications available through Tropicos and may not represent the entire distribution. Tropicos does not categorize distributions as native or non-native.

Rorippa columbiae (S. Watson) Howell:
United States (North America)

Note: This information is based on publications available through Tropicos and may not represent the entire distribution. Tropicos does not categorize distributions as native or non-native.

United States

Origin: Native

Regularity: Regularly occurring

Currently: Present

Confidence: Confident

Type of Residency: Year-round

Global Range: Endemic to Washington, Oregon, and California, in two geographically distinct regions: (1) gravelly shores of the Columbia River in Washington and Oregon, and (2) seasonally wet areas in the high desert in south-central Oregon and northern California. On the whole, the species' range is predominantly east of the Cascade Range; along the Columbia River, however, it occurs in two segments: the arid Hanford Reach east of the Cascades, and the cooler, wetter Lower Columbia Reach west of the Cascade Crest (WA NHP 2003). Range extent is approximately 50,000 square km if the two population clusters are considered separate and approximately 150,000 square km if the are considered together.

The distribution of Rorippa columbiae appears to be related to the Columbia River lava plateau (Stuckey 1972). According to Stuckey (1972), R. columbiae was part of a very old group of plants that once was widespread and variable when western North America had more wet habitats. As the region became increasingly arid, populations were isolated and diverged morphologically. One line of morphological divergence was in the region of the Yellowstone River (R. calycina), one was in the region of Lake Tahoe (R. subumbellata) and one was on the basaltic lava formation of the Columbia River (R. columbiae). Today, these three species are isolated relics of the once abundant and widespread taxon (Stuckey 1972).

Rorippa columbiae is distinct from other North American Rorippa in having most of the plant, including stems, both leaf surfaces, pedicels, sepals, siliques, and styles, covered with soft, dense, spreading hairs (after Stuckey 1972; Hitchcock et al. 1964). Rorippa columbiae can be distinquished from closely related species by its (1) fruits, which are densely covered with elongated trichomes and which have expanded stigmas, (2) sepals and styles, which are densely pubescent and (3) axillary racemes, which are longer than 5 cm when fully developed and which have ascending pedicels (Stuckey 1972). Mulligan and Porsild (1966) report 16 chromosomes for the closely related Rorippa calycina.

Comments: Grows in damp to wet soils, a habitat preference which it has in common with other Rorippa species (Stuckey 1972). Populations have been observed near all types of bodies of water, including the Columbia River, intermittent snow-fed streams, permanent lakes, snow-fed lakes, internally-drained lakes which may be dry for extended periods of time, wet meadows, irrigation ditches and roadside ditches. A common feature of known sites is inundation for at least part of the year. Individuals are usually found in open, high light habitats, with low vegetative cover. The species grows on a wide variety of soil types including clay, sand, gravel, sandy silt, cobblestones and rocks. 1200-1800 m. Descriptions of habitat at a selection of specific locations follow.

Populations along the Columbia River are currently located on stretches of river that are relatively free-flowing, although flows are impacted on a hourly, daily and seasonal basis by upstream dam operations (Scherer 1991; Pam Camp, personal communication; Lynn Cornelius, personal communication; Gehring, personal observation). Population sites on the lower Columbia below Bonneville Dam experience unpredictable inundation by the river due to (1) dam operations, (2) tides and (3) water flow from the Willamette River at Portland (Scherer 1991; Gehring, personal observation). In most years, they are above water for only two to three months during the growing season (Scherer 1991). At Pierce Island, most of the populations are on cobblestone beaches and banks, although the beach on the north side of the island has a layer of sandy silt on top of the cobblestones. Many of the plants are found in a linear zone of two to three meters, parallel to the river. One population on Pierce Island is growing in sandy silt in a slough that contains water only when the river is relatively high (Cornelius, personal communication; Gehring, personal observation). This population has high vegetative cover (Gehring 1990); the most frequent associated species in 1991 included Deschampsia caespitosa and Equisetum spp. (Gehring, personal observation). Plant species found growing frequently with Rorippa columbiae on the cobblestone banks and beaches of Pierce Island in 1984 included Artemisia lindleyana, Rorippa curvisiliqua, Aster subspicatus, Polygonum persicaria, and Salix fluviatilis (Cornelius 1985). The cover and frequency of Artemisia lindleyana and Helenium autumnale have increased dramatically on the cobblestone beaches since 1984 (Lynn Cornelius, personal communication). Plants at the mouth of Hamilton Creek on Pierce National Wildlife Refuge are located on the rocky north bank of the river. Associated species include Coreopsis atkinsonia, Helenium autumnale, Artemisia lindleyana, Mentha arvensis, and Salix fluviatilis (Harris 1992). Populations have also been observed in open, bare gravelly habitat with little vegetation on Ives Island (Washington Natural Heritage Information System) and on Hamilton Island (Lois Kemp, personal communication). On the river bar at the old mouth of the Sandy River, Rorippa columbiae occurs with Arabidopsis thaliana, Draba arabis, Cerastium nutans, Cerastium viscosum, Scleranthus annuus, Veronica arvensis, Chenopodium botrys, Poa annua, Aira caryophylla, and Marsilea vestita (Oregon Natural Heritage Program Data Base).

There are also numerous populations on the Hanford Reach of the Columbia River. The visual habitat features of populations on the Hanford Reach appear to be similar to those of the populations downstream of Bonneville Dam. Plants only grow below the high water level of the river where vegetative cover is sparse (Sauer and Leder 1985). They are inundated frequently during the growing season (Camp, personal communication; Gehring, personal observation). On the Hanford Reach, Rorippa columbiae grows on wet silty soil beneath a layer of gravel (Sauer and Leder 1985; Camp 1992; Gehring, personal observation), sand with cobbles (Baird 1984; Schuller 1984; Camp 1992), and sandy silt without rocks (Camp 1992; Gehring, personal observation). Associated species include Artemisia lindleyana, Machaeranthera canescens, Salix exiqua, Salix spp., Coreopsis atkinsoniana, Agrostis spp., Helenium autumnale, Plantago major, Melilotis alba, Rorippa spp. (Schuller 1984), Lindernia anagallidea, Veronica spp. (Baird 1984), Apocynum spp. (Camp 1992), Morus alba (Camp 1992; Gehring, personal observation), Phalaris arundinacea, Verbena bractata, Veronica anagalis-aquatica, and Cyperus aristatus (Sackschewsky, personal communication, 1992a).

In south-central Oregon, populations are known from Malheur Lake (Harney County), an internally-drained basin. Water levels at Malheur Lake began increasing in 1982, peaked in 1987 to a height not recorded since settlement by Europeans, and have been rapidly receding since 1987. The lake is expected to be dry in 1992 (Nora Taylor, personal communication). Associated species include Sarcobatus vermiculatus, Artemisia tridentata, Hordeum jubatum, Heliotropium curassavicum, and Potentilla anserina (Kaye and Massey 1991). Populations on the eastern/southeastern shore of Malheur Lake are or were located near the 4100 foot contour line, which was the high water level of the lake in the 1980's (Kaye and Massey 1991). The soil at these sites is now dry (Nora Taylor, personal communication). A number of small populations were found in 1988 after Malheur Lake began receding (Oregon Natural Heritage Program Data Base; Nora Taylor, personal communication); these populations were unknown before lake levels increased in the early 1980's (Nora Taylor, personal communication). Kaye and Massey (1991) were able to relocate only a third of these populations in 1990. The remaining populations were in microsites that retained moisture (Kaye and Massey 1991); for example, plants were found at the bases of large Artemisia tridentata and Sarcobatus shrubs, along the edges of flood debris such as pieces of barn siding, and in open but low-lying sites with standing water, including on an infrequently used road (Kaye and Massey 1991). The plants on the northeastern shore of Malheur Lake are also near the 4100 feet contour line (Kaye and Massey 1991; Oregon Natural Heritage Program Data Base, 1992). The plants are growing on ridges parallel to the lake shoreline; the ridges have clay on the surface (Nora Taylor, personal communication).

In Lake County, Oregon, populations are known from Featherbed and Foley Lakes, which are flat, shallow basins (Kaye and Massey 1991). Featherbed Lake normally contains water year-round, but was dry in 1978, 1990 and 1991 (Kaye and Massey 1991; Ginger King, personal communication). Foley Lake is fed by an intermittent stream from Featherbed Lake and usually has water only in the spring (Ginger King, personal communication). At both lakes, the plants are growing in silty clay and sandy soils and the populations are located in open, high light sites (Ginger King, personal communication). Associated species at Featherbed Lake include Monolepsis nuttalliana (20-25% cover), Artemisia arbuscula (20%), Arnica chamissonis (20%), Camissonia tanacetifolia (4-5%), Sitanion hystrix (trace), Bromus tectorum (trace), and Juncus sp. (trace). The Arnica and Sitanion were largely confined to the upper, drier, more rocky portions of the monitoring plots, and the Camissonia and Monolepsis were restricted to the lower, more moist portions of the plots (Kaye and Massey 1991. At Foley Lake, associated species include Camissonia tanacetifolia and Phacelia thermalis (Kaye and Massey 1991). Plants are also known from Binkey Lake (Kaye and Massey 1991). In 1983, associated species there included Camissonia tanacetifolia, Downingia spp., Iva axillaris, and Muhlenbergia richardsonis. Plants are also known from Silver Lake, from a Scirpus acutus marsh. In 1985, associated species at Silver Lake included Potentilla newberryi, Camissonia tanacetifolia, Epilobium paniculatum, Hordeum jubatum, and Rumex crispus. There were also small populations in a roadside ditch near Paulina Marsh in 1985 (Oregon Natural Heritage Program Data Base). They were observed growing in a wet meadow community with Hordeum jubatum, Camissonia tanacetifolia, Potentilla newberryi, Carex athrostachya, and Ranunculus cymbalaria.

In Klamath County, Oregon, plants occur in an intermittent stream bed on the northeast flank of Stuckel Mountain. The stream flows through a ponderosa pine, juniper and sagebrush community at approximately 5000 feet (Oregon Natural Heritage Program Data Base). The stream is fed by snowmelt and flows in the spring, but if there is little snowpack (as in 1991 and 1992), then it may not flow at all or only for a short period of time (Lou Whiteaker, personal communication). The substrate is sand and gravel on a slope of 1 to 2% (Oregon Natural Heritage Program Data Base).

In California, Siskiyou County populations occur at two lakes in the Shasta-Trinity National Forest: White Deer Lake and Dry Lake. White Deer Lake is a small meadow depression at 4590 feet in elevation that floods in wet years (California Natural Diversity Data Base). The lake has been dry for several years and lodgepole pine is invading the lake bed (Julie Nelson, personal communication). The population occurs in sandy soil in a small artificial pond constructed as a stock pond (California Natural Diversity Data Base; Julie Nelson, personal communication). The plants grow in clusters in an open, high light environment. Rorippa columbiae is found growing near Eryngium aristulatum, Juncus balticus (California Natural Diversity Data Base) and Oenothera tanacetifolia (Julie Nelson, personal communication). Dry Lake, where plants occur on the eastern margin,.is sandy with a hard soil pan near the surface. Normally, it is flooded for three to four months with the water level sometimes above three feet. Associated species include Pinus contorta, Arnica spp., Camissonia spp., Lupinus selinus, Penstemon shastensis, and Juncus balticus (Julie Nelson, personal communication).

Populations are also known from the shores of Clear Lake Reservoir in the Klamath Basin National Wildlife Refuge (Modoc County, California). Clear Lake is at 4480 feet in elevation and water levels decline seasonally (California Natural Diversity Data Base). In 1992, stems of Rorippa columbiae were scattered around the margins of the lake and up Willow Creek, which drains into Clear Lake (Mary Paezel, personal communication). The plants are growing in wet areas near the water's edge, on pumice sand, on higher mud flats of the old lake shore (due to drought, the water levels in the lake have decreased over the last several years) (Mary Paezel, personal communication), and among volcanic boulders in the inundation zone (California Natural Diversity Data Base). In general, plants are found away from competing species in open, high light sites (Mary Paezel, personal communication). Associated species include Psilocarphus spp., Allocarya spp., Polygonum spp., and Juncus balticus (California Natural Diversity Data Base).

Note: For many non-migratory species, occurrences are roughly equivalent to populations.

Estimated Number of Occurrences: 21 - 80

Comments: Most known occurrences are found in Oregon, where there are 25 occurrences last observed 1982-2007 and 15 additional occurrences last observed in 1960 or earlier. In California, 8 occurrences are believed extant (last observed 1987-1997) and an additional 3 are considered historical (last observed 1915-1956). In Washington, 2 occurrences are extant and 1 has been extirpated.

RORIPPA COLUMBIAE is perennial. The plants spread by rhizomes (Stuckey, 1972) and possibly by rooting at the nodes of above-ground stems (Pam Camp, personal communication). The plant is often found growing in clusters of above-ground stems (Gehring 1991, 1992; Kaye and Massey 1991; Pam Camp, personal communication; Lois Kemp, personal communication; Ginger King, personal communication; Nora Taylor, personal communication). Presumably these clusters are clones and represent only one genotype. The clusters range in size and extent of areal coverage from thousands of stems covering large areas in dense colonies to one or only a few above-ground stems.

Growth of RORIPPA COLUMBIAE is largely determined by water availability. On Pierce Island, RORIPPA COLUMBIAE population vigor and size has decreased between 1984 and 1990/1991 on Pierce Island (Lynn Cornelius, personal communication). This trend may be due to weather patterns, since the early 1980's were high runoff years while the late 1980's were low runoff years (Scherer 1991). However, annual growth can be decreased by too much water as well. On Pierce Island, plants growing at lower elevations begin seasonal growth later than those higher on the island, produce few flowers or do not flower at all, mature few if any seeds, and are probably dependent on vegetative spread for population maintenance in most years (Gehring 1990, 1991). At Featherbed and Foley Lakes in south-central Oregon, population frequency has been correlated with annual precipitation since 1983 (Kaye and Massey 1991). In 1991, a large snowstorm occurred in late spring presumably causing population size and density at Featherbed and Foley Lakes to increase; frequency in monitored plots increased from 0.25% in 1990 to 43.3% in 1991 (Ginger King, personal communication).

Most populations found on the Columbia River below Bonneville Dam are relatively small, usually having fewer than 500 individuals but sometimes having only one or two plants. The number of stems in populations of RORIPPA COLUMBIAE on Pierce Island ranges from 50 to 500. At Hamilton Creek in 1992, there were 300-500 stems of RORIPPA COLUMBIAE (Gehring, personal observation). The Sandy River population had approximately 50 clusters of RORIPPA COLUMBIAE stems in May 1992 (Oregon Natural Heritage Program Data Base). At Sandy River there was 5-10% total vegetation cover on 13 May 1992.

The largest known populations of RORIPPA COLUMBIAE are located along the Hanford Reach of the Columbia River, between Vernita and Richland, Washington. The species is abundant along this stretch of the river (Sauer and Leder 1985; Camp 1992; Gehring, personal observation). Population size on the Hanford Reach is probably on the order of magnitude of millions of stems over approximately 40 river miles (Gehring, personal observation).

At Malheur Lake, (Harney County Oregon), reported population sizes vary considerably. Kaye and Massey (1991) reported five groups of plants covering an area over 3000 m2 in 1990. Nora Taylor (personal communication, 1992) described one large group of more than 500 stems and two smaller groups of plants. The records in the Oregon Natural Heritage Program Data Base indicate two large populations with more than 1000 stems each in 1991. Population size is apparently varying from year to year, probably depending on soil moisture.

Population sizes at most of the other Oregon populations are generally small. Approximately 200 stems were growing on the south and east shoreline of Featherbed Lake and in the lake bed near the north shore in 1990 (Kaye and Massey 1991). At Foley Lake, approximately 200 stems of RORIPPA COLUMBIAE were observed growing in the lake bed and in a nearby water hole in 1990 (Kaye and Massey 1991; Ginger King, personal communication). At Binkey Lake in 1983, there were one or two stems of RORIPPA COLUMBIAE growing in sand on the lake shore (Oregon Natural Heritage Program Data Base). At Silver Lake, there were 300 to 400 stems along the southeast shore in 1985 (Oregon Natural Heritage Program Data Base). The two RORIPPA COLUMBIAE populations at Paulina Marsh had approximately 30 stems and 60-80 stems (Oregon Natural Heritage Program Data Base). However, the plants at Paulina Marsh are reported to grow quite large; over two feet in diameter (Oregon Natural Heritage Program Data Base; Ginger King, personal communication). On Stuckel Mountain, the RORIPPA COLUMBIAE population was estimated to have between 100 to 125 plants in 1983 (Oregon Natural Heritage Program Data Base).

In California, the RORIPPA COLUMBIAE population at White Deer Lake was estimated to have between 51-100 plants in 1987 (California Natural Diversity Data Base). By 1991, Julie Nelson (personal communication) estimated there were only 30 to 50 plants. However, some of the plants flowered in 1991 (Julie Nelson, personal communication). At Clear Lake Reservoir in 1987, one population was estimated to have over 1000 plants (California Natural Diversity Data Base). In 1992, the plants averaged 10 cm in height when growing without interspecific competition and reached 2 dm when growing with other species. Some plants at Clear Lake were flowering by mid-June, 1992 (Mary Paezel, personal communication).

The timing of initiation of annual above-ground growth of RORIPPA COLUMBIAE differs between the populations on the Columbia River and in southern Oregon and northern California. Annual growth usually begins in April or May in southern Oregon and northern California, when melting snow begins to provide a source of water for vernal lakes and intermittent streams. In contrast, the Columbia River usually does not drop below the elevation level of most populations until late July or August (Gehring, 1990, 1991) and sometimes as late as September (Sauer and Leder, 1985; Pam Camp, personal communication) so the RORIPPA COLUMBIAE populations do not initiate growth until late summer or even early fall.

Based on herbarium samples, Sauer and Leder (1985) suggested three distinct morphs of RORIPPA COLUMBIAE may have evolved: (1) decumbent, trailing stems less than 30 cm long on the Columbia River; (2) erect, freely branching, spherically shaped, 30 cm in diameter and height on the lava beds of northern California; (3) stems of intermediate size and habit at the south-central Oregon sites. However, it is more likely that morphology is dependent on water availability and regime. For example, plants with little or no stem development have been observed both along the Columbia River (Pam Camp, personal communication) and on lake margins in south-central Oregon (Ginger King, personal communication). Additionally, decumbent, trailing stems have been observed on the Columbia River (Gehring, personal observation) and at Malheur Lake (Nora Taylor, personal communication) while relatively large, erect plants have been collected or observed at Lava Beds National Monument (Sauer and Leder 1982), in roadside ditches in south-central Oregon Ginger King, personal communication), and along the Columbia River (Sackschewsky, personal communication; Gehring, personal observation). One would expect that plants with an abundant water supply (e.g., high water table) and a long growing season could grow larger and branch more than plants which are inundated for much of the growing season or are limited in growth by water availability.

In summary, the distribution of RORIPPA COLUMBIAE indicates that the species requires moist or perhaps wet soil. However, it also seems apparent that annual growth can be decreased or eliminated by too much water during crucial portions of the growing season. The most favorable growing conditions and habitat appear to include spring flooding and a relatively high water table during the remaining portion of the growing season. However, there is no evidence to indicate that actual submergence of the plants is necessary in every habitat or in every year, as long as the soil is wet. A crucial aspect for the growth and long-term survival of RORIPPA COLUMBIAE may be the timing and/or duration of the spring flood, especially if flowering is sensitive to photoperiod.

The amount of sexual reproduction is quite variable, depending on habitat and annual conditions. In many populations of RORIPPA COLUMBIAE, a high percentage of above-ground stems remain vegetative and do not flower (Gehring 1991; Kaye and Massey 1991; Camp 1992; Lois Kemp, personal communication; Ginger King, personal communication; Gehring, personal observation). On the Columbia River, plants have been observed to initiate flowering as early as May 13th (Oregon Natural Heritage Data Base) and as late as September (Sauer and Leder, 1982; Gehring, personal observation). Flowering typically begins in May and June in south-central Oregon (Kaye and Massey 1982; Ginger King, personal communication; Nora Taylor, personal communication). Flowering continues through September and October, apparently on a rangewide basis (Sauer and Leder 1982; Kaye and Massey 1991; Oregon Natural Heritage Data Base; Gehring, personal observation). Several species of insects, including small, native bees, have been observed visiting the flowers and may be acting as pollinators (Kaye and Massey 1991; Mary Paezel, personal communication). However, the plants can set seed in the absence of pollinators, and thus are probably autogamous and self-compatible (Kaye and Massey 1991; Pam Camp, personal communication).

Plants with fruits have been observed from June through November (Kaye and Massey 1991; Oregon Natural Heritage Data Base). In Oregon, the average number of seeds per fruit ranged from 4.9 at Foley Lake to 21.9 at Featherbed Lake in Oregon and the mean number of seeds per plant ranged from 126.9 at Foley Lake to 1346.8 at Featherbed Lake (Kaye and Massey 1991). Fruits can contain a significant number of aborted or immature seeds (Kaye and Massey 1991). Only mature seeds appeared viable (Kaye and Massey 1991). Apparently no one has attempted to germinate the seeds of RORIPPA COLUMBIAE, but the seeds of the closely related RORIPPA SUBUMBELLATA germinate easily (Ken Berg, personal communication).

The plants survive seasonal inundation along the Columbia River (Sauer and Leder 1985; Gehring, personal observation). Evidently clones can remain dormant with no above-ground growth for several years. Populations not observed for at least two consecutive years along the Hanford Reach were subsequently found to cover large areas with a high density of above-ground stems (Camp, 1992). Seemingly, roots and rhizomes remain viable, even when the plants are unable to grow above ground because of high water levels.

Stuckey (1972) suggested that the roots of RORIPPA SINUATA could survive extended dry periods and grow again when wetter conditions returned. This may be the case for RORIPPA COLUMBIAE as well. A population at Feather Bed Lake that previously had only a few stems at low density was very dense with many stems in 1991. The dramatic increase in density of stems was conjectured to be due to a large, late spring snowfall (Ginger King, personal communication). An alternative hypothesis for this phenomenon is that seeds can remain viable in the seed bank and germinate when the soil is wet at a favorable time of the year.

Dispersal of the species probably occurs when plant parts such as stems and rhizomes and seeds float down rivers or across lakes and germinate on wet shores. This is probably how most of the populations at the high water mark of Malheur Lake became established. Dispersal between river drainages or between lakes may occur by seed movement via migrating water fowl (Cheryl McCaffrey, personal communication).

United States

Rounded National Status Rank: N3 - Vulnerable

Rounded Global Status Rank: G3 - Vulnerable

Reasons: Endemic to Washington, Oregon, and California, in two geographically distinct regions: (1) gravelly shores of two segments of the Columbia River in Washington and Oregon (the Hanford Reach and the Lower Columbia Reach), and (2) seasonally wet areas in the high desert in south-central Oregon and northern California. Approximately 35 occurrences are believed extant, and the total population (at least in some years) has been estimated at over 1 million plants. Numerous occurrences have been extirpated as a result of dam/reservoir construction on the Columbia River and agricultural and urban development in other parts of the range. Current threats include hydrological alteration; habitat destruction/degradation (e.g. from road-building, development and recreation); cattle trampling and grazing; and altered interspecific competition dynamics. Of these, hydrological alteration is believed to be the most significant.

Global Long Term Trend: Decline of 30-50%

Comments: Much of the historic habitat along the Columbia River has been eliminated by inundation due to the construction of the dam/reservoir system (WA NHP 2003). Historically, populations were observed at Umatilla, OR and Bingen, WA between the two extant Columbia River clusters, but apparently these plants were destroyed during dam construction and/or were inundated by dam pools. In addition, numerous populations in Oregon have not been relocated since collections were originally made in the late 1800's and early 1900's. Because of their locations, it is likely that most of these populations were destroyed by agriculture or urbanization.

Comments: There are four main areas of concern regarding this species' long-term persistence: (1) water level/availability, (2) habitat destruction, (3) cattle trampling and grazing, (4) interspecific competition. Hydrology is the most important of these factors, because Rorippa columbiae apparently requires moist to wet soil.

Along the Columbia River in Washington and northern Oregon, river flows and water level impact this species' growth and reproduction (Gehring 1990, 1991). Based on a water flow model developed by the U. S. Army Corps of Engineers, the current water levels of the Columbia River from April through August are not higher than historic levels (i.e., flows prior to the construction of dams), but are 20% higher in September (Scherer, 1991). Therefore the growing season may be somewhat shorter now than it was historically for populations along the river. However, the species is seemingly tolerant of, or adapted to, a short growing season. Also, it apparently can survive occasional years of extended submergence. Therefore, the change in water levels may have had only a minor impact. Certainly, the species has persisted with the dams on the Columbia River since 1937, when Bonneville Dam was built.

However, two major changes in the hydraulics of the Columbia River have occurred since dam construction: (1) attenuation of the spring floods and (2) large short-term (hourly and daily) fluctuations in water level throughout the year (Scherer 1991).

A decrease in spring flood flows would presumably decrease scouring of the river banks, which might allow other plant species to invade the river banks. Based on old photographs, plant cover on Pierce Island has increased since the late 1800's and the expected successional trend is evident, but the largest changes have been in the higher vegetation zones (Cornelius 1985). There has been a dramatic increase in vegetative cover at the Rorippa columbiae populations on Pierce Island since 1984 (Cornelius, personal communication) and this type of change, although perhaps initiated because of short-term weather patterns, may be maintained by the elimination of spring floods. However, many R. columbiae populations on the Columbia continue to exist in sites of sparse vegetative cover, and inter-specific competition appears to be a minor component of the habitat at most locations (Gehring, personal observation).

Short-term fluctuations in water levels on the Columbia River are caused primarily by dam operations in response to hydroelectric power demand (Scherer 1991). During the summer, flows are generally low in the evening and at night and high during the day; the practice of changing river flows through the dams to match power demand is referred to as peaking (Scherer 1991). Peaking on the Columbia River was not initiated until the early 1980's and, therefore, may also be a factor in the decline of population vigor on Pierce Island (Scherer 1991; Scherer and Young 1992). In addition, large fluctuations occur during daylight hours (Scherer and Young 1992; Camp, personal communication; Gehring, personal observation). Fluctuations are much larger along the Hanford Reach than below Bonneville Dam (Camp, personal communication) because control of flows on the Columbia River is primarily at Grand Coulee Dam, upstream from the Hanford Reach (Scherer 1991). Peaking results in unpredictable submergence of low elevation populations along the Columbia and presumably decreases the amount of time the plants can effectively photosynthesize. However, water flow in the Columbia River is determined both by weather and water management practices (Scherer 1991). For example, based on water data at Bonneville Dam from the U. S. Army Corp of Engineers, Reservoir Control Center, Portland, a population of Rorippa columbiae located at 9 feet on Pierce Island would have been exposed (i.e., above water) 25% of daylight hours in July 1990 but less than 5% of daylight hours in July 1991 (Scherer and Young, 1992). In August 1990, a population at 9 feet would have been exposed for 60% of daylight hours, but less than 20% of daylight hours in August 1991. The plants' sexual reproduction and, to a lesser extent, growth was correlated with these data: there was a marked decrease in sexual reproduction at the Pierce Island populations in 1991 as compared to 1990 (Gehring 1991). There is concern that the cumulative effects of frequent, short-term inundation during the growing season may depress the vigor of the species, and that current management of the Columbia River may impact its ability to produce seeds (WA NHP 2003).

In south-central Oregon and northern California, water availability at known sites is controlled by meteorological trends. During high precipitation years, the number of populations increases and the size and vigor of existing populations increases as well. This occurred in the mid-1980's. In the 1990s, southern Oregon and northern California experienced a multi-year drought and the size and number of populations declined at Malheur Lake, Featherbed Lake and Foley Lake in Oregon (Kaye and Massey 1991), and at White Deer Lake (Julie Nelson, personal communication) and Clear Lake (California Natural Diversity Data Base; Mary Paezel, personal communication) in California.

Habitat destruction is a rangewide concern. In the past, populations along the Columbia River were lost due to dam construction, and numerous Oregon populations are believed to have been destroyed by agriculture or urbanization. Presently, activities along streams, rivers and lakes that decrease habitat availability or quality, such as road-building, dredging, development and recreation, pose a threat to the long-term survival of this species.

Cattle trampling, and perhaps grazing, is also a rangewide threat. Grazing cattle are present at many of the Oregon and California sites, although some populations are fenced to exclude cattle, at least in California (CNPS 2009). The impact of trampling and grazing by cattle is being monitored at several BLM sites in south-central Oregon. The Sandy River population on the Columbia River was severely trampled by cattle in 1990 (Gehring, personal observation). Plants are commonly trampled and possibly grazed by cattle at Malheur Lake (Kaye and Massey 1991; Nora Taylor, personal communication). At Featherbed and Foley Lakes, the severity of cattle trampling depends on the water level in the lakes. If there is water, the cattle are more attracted to the lakes and consequently trample plants to a greater extent (Ginger King, personal communication). Grazing at Featherbed and Foley Lakes is apparently due more to rabbits than to cattle (Kaye and Massey 1991). In California, there is heavy grazing and trampling by cattle at Dry Lake in the Shasta-Trinity National Forest (Julie Nelson, personal communication) and at Clear Lake in the Klamath Basin National Wildlife Refuges (Mary Paezel, personal communication). At Clear Lake, the U.S. Fish and Wildlife Service anticipates fencing the cattle away from R. columbiae plants (Mary Paezel, personal communication).

This species is usually found growing in open, high light sites with few other species. It is low-growing and probably cannot compete for light if shaded by other plants. In addition, because water is critical for successful growth and reproduction, competing plant species could potentially remove crucial water resources. Therefore, interspecific competition may be considered a potential threat, particularly if water management or cattle grazing are allowing the introduction of weedy, pioneering species into occupied and other suitable sites.

Restoration Potential: Presumably, RORIPPA COLUMBIAE was never a widespread, common species. However, when appropriate habitat is available and the water regime is favorable, large populations can develop. If current habitat is protected from cattle trampling and recreational uses and water levels are allowed to fluctuate in a natural manner, the recovery potential of RORIPPA COLUMBIAE is probably good. However, management of water flows on the Columbia River is a barrier to full recovery of river populations, especially since dam removal is not a reasonable alternative. In addition, the species is only found in wet sites, so short-term droughts or long-term, global climate change resulting in locally dry conditions will cause local population extinction and may threaten the survival of the species.

Preserve Selection and Design Considerations: Primary criteria for locating short-term protection boundaries should be: all of the existing population at a given site, including all contiguous shoreline habitat between extreme high water and extreme low water (historical extremes as habitat defines), and control of the hydrologic regime. Grazing should be excluded and recreation access should be restricted. Columbia River population habitats continue to be subject to major changes in hydrologic dynamics since dam construction on the river began in the 1930's. This has resulted in loss of scouring spring floods, and in artificial fluctuation of the water levels during the growing season. On Pierce Island, WA. (TNC Preserve) where competing vegetation, especially perennial shrubs, is increasing on all occupied habitats removal of competing vegetation could be experimented with to assist in short-term survival of existing individuals until the feasibility of long-term protection strategies can be addressed. Habitat requirements for long-term survival of existing RORIPPA COLUMBIAE populations are currently unknown. Reestablishing some degree of spring flooding and historical growing season water level patterns may be needed for the extensive populations on the Columbia River. More information is needed on successional trends and impacts to the non-Columbia River populations as well. Until more is known about RORIPPA COLUMBIAE habitat requirements, and until long-term protection strategies and goals are better defined and the feasibility of reversing the severe impacts to the Columbia River ecosystem is further discussed, protection strategies for river occurrences will be aimed at short-term protection of surviving populations. More information is needed on successional dynamics and human impacts to the non-river populations before addressing long-term protection strategies at these sites. (Personal communication, Lynn Cornelius, TNC, Washington Field Office.)

Management Requirements: Water management on the Columbia River needs to be affected to reduce unpredictable flooding of the river populations and annual, scouring spring floods need to be reestablished plants on the Columbia River needs to be stopped. In a few populations, interspecific competition should be decreased.

Ideally, water levels should be allowed to fluctuate in a natural manner. However, since removal of the dams on the Columbia River is highly unlikely, the feasibility of affecting water management on the Columbia River needs to be addressed. To reduce interspecific competition, removal of competing vegetation may be needed. Hand removal of competing vegetation by cutting at ground level should be investigated to assist short-term survival.

Monitoring Programs: Monitoring programs have been instituted at (1) Featherbed and Foley Lakes, (2) Malheur Lake, and (3) Pierce Island.

For more information on Featherbed Lake contact Ginger King, U. S. Department of the Interior, Bureau of Land Management, Lakeview District, P.O. Box 151, Lakeview, Oregon. Telephone: 503-947-2177.

The Oregon Department of Agriculture also established permanent transects at Foley Lake in 1990. Similar to the situation at Featherbed Lake, there are two macroplots. One is located within a grazing exclosure. The BLM monitors frequency changes (Kaye and Massey 1991). For more information, contact Ginger King, U. S. Department of the Interior, Bureau of Land Management, Lakeview District, P.O. Box 151, Lakeview, Oregon. Telephone: 503-947-2177.

The BLM set up quadrat frequency monitoring in 1983 at Foley Lake. There is a 200 ft east-west transect with 40 side transects. The side transects are each 20 feet long and are located at 5 feet intervals along the main transect. Adjacent side transects are located on the opposite sides of the main transect. Ten 1 X 1 foot plots are located every other foot along the side transects. This transect was monitored in 1983, 1988, 1989, 1990, 1991 for the presence of RORIPPA COLUMBIAE. In 1991, one-half of the transect was enclosed to prevent grazing (Ginger King, personal communication). For more information, contact Ginger King, U. S. Department of the Interior, Bureau of Land Management, Lakeview District, P.O. Box 151, Lakeview, Oregon. Telephone: 503-947-2177.

At Malheur Lake, Oregon, BLM staff established linear transects through four populations in 1991. Locations of groups of plants are determined by measuring distance along and perpendicular distance away from the transect (Nora Taylor, personal communication). At each location, the number of stems and areal extent of the group is recorded. For more information, contact Nora Taylor, U. S. Department of the Interior, Bureau of Land Management, Burns District Office, HC 74--12533 Hwy 20 West, Hines, Oregon 97738. Telephone: 503-573-5241.

At Pierce Island, The Nature Conservancy, Washington Field Office, established permanent transects at five sites in 1991. Each transect is considered a macroplot, which contains adjacent 0.5 X 0.5 m microplots. Three sites have 40 X 2 m macroplots, one site has two 20 X 2 m macroplots and one site has a 11 X 2 m macroplot (Gehring 1991). For purposes of data analysis, microplots will be chosen randomly from within each macroplot. For more information contact: Janet Gehring, 11322 9th Place West, Apt 1, Everett, WA 98204. Telephone: 206-348-4279. OR Lynn Cornelius, The Nature Conservancy, Washington Field Office, 1601 2nd Ave, Suite 1100, Seattle, Washington 98101. Telephone: 206-343-4344.

In 1992, permanent transects will be established by The Nature Conservancy, Washington Field Office at Hamilton Creek, and on BLM islands in the Hanford Reach. The mapping of populations on the islands is in cooperation with the BLM. A study of water levels and management on the Hanford Reach is also being conducted by The Nature Conservancy, Washington Field Office.

Comments: CONTACT LIST

John Gamon, Washington Natural Heritage Program, Washington Department of Natural Resources P.O. Box 47046, Olympia, Washington 98504-7046; PHONE: (206) 902-1661 (Independent reviewer of this ESA)

Stewardship Overview: The effects of water levels (Columbia River) or annual precipitation (southern Oregon/northern California) on population size, vigor and reproduction should be monitored. In addition, the effects of cattle grazing or trampling and interspecific competition may need to be observed at some sites. Important research goals include (1) determining the level of genetic diversity, (2) seed germination requirements, (3) seed bank information, (4) inventory of known populations, and (5) surveying potential habitat within the range of the species. Existing populations should be protected from cattle, recreational users and development. Feasibility of affecting water management on the Columbia River should be addressed.