Streambank lupine (Lupinus rivularis) COSEWIC assessment and status report: chapter 3

Species Information

Name and classification
Lupinus rivularis Dougl. ex. Lindl. (Lindley’s diagnosis came from cult. plants and Douglas’ notes. (USDA2001)).
Bibliographic citation: Bot. Reg. 19: t: 1595. 1833.
Type: AmericaBoreali-occidentalis, Douglas 263, 1825. CGE. Figures 55 and 56.
Synonymy: L. lignipes Muhlenbergia 8: 66. Fig. 8. 1912. (Type: Eugene, Lane Co., Oregon, Heller 10042 RENO, POM, MIN.)
Common names: streambank lupine, riverbank lupine.
Family:  Fabaceae alt. Leguminosae. Also placed in: Papilionaceae. Edward’s Bot. Reg. 19: t. 1595. 1833.
Sources: Dunn and Gillett 1966; Kartesz and Kartesz 1994

Comment

The genus Lupinus is considered one of the most taxonomically difficult groups in the Pacific Northwest (Phillips 1955, Hitchcock et al.1961, Ceska 2001 pers. comm., Sholars 2001 pers. comm., Ainouche and Bayer 1999, Nicholls and Bohm 1983, and others). Confusion surrounds the taxonomy and nomenclature to the point that a widely varying number of North American species are recognized by different experts: from as few as 100 species to as many as 600 taxa (Nicholls and Bohm 1983). The genus has been explored by several researchers, including Heller (1912), Phillips (1955), Dunn and Gillett (1966), Mikolajczyk (1964), Bisby (1981), Nicholls and Bohm (1983), and Ainouche and Bayer (1999). However, this has not yet allowed a clear picture to emerge. As Nicholls and Bohm (1983, 708) state, “while Lupinus is easily distinguished from other genera, comparatively few characters avail themselves for clear distinctions between species.”

Hybridization and introgression, play a significant role in this confusion, and are commonplace, while “phenotypic plasticity seems the rule rather than the exception” (Nicholls and Bohm 1983, 708). Phillips (1955), who examined lupines of North America exclusive of the southwestern United States and Mexico, indicates that the natural variation in the genus is high, and that this is a result of the considerable amount of hybridization that is taking place, with many of the hybrids producing fertile offspring. Hybridization is increased today in many regions by the active planting of lupines for gardens and highway maintenance. For example, the spread up the Pacific Coast of the highly aggressive Lupinus arboreus, mainly as a result of planting, complicates the picture. This is a species that readily genetically swamps other lupines (Sholars 2001 pers. comm.). All of this has led to some confusion in classification of lupines in general and in our local species.

The taxonomy and nomenclature of Lupinus rivularis does not escape these complexities. In Hitchcock et al. (1961), L. rivularis Lindl. is equated with L. amphibium Suskd. Phillips (1955) considered L. rivularis Dougl. ex Lindl. as synonymous with L. albicaulis, while Dunn and Gillett (1966) treat L. rivularis as a separate species. These last authors mention that it shows morphological affinity to Lupinus arcticus ssp. subalpinus, the two being separated by altitude.Douglas et al. (1998) recognize it as a separate species with no synonymy given. In the US, Riggins and Sholars (1993) describe it as grading into blue-flowered L. arboreus. Sholars (2001 pers. comm), who is presently working on a new treatment of Lupinus for Flora North America, has indicated that in the US it is generally lumped with L. latifolius, although she feels it is a good and separate species. Only one synonym, L. lignipes, is legitimately recognized by Kartesz and Kartesz (1994) for Lupinus rivularis.

Lupinus rivularis is known to hybridize extensively with the aggressive L. arboreus (Riggins and Sholars 1993, Sholars 2001 pers. comm.), and L. arboreus in general is known to hybridize with other lupines species (Rhymer and Simberloff 1996), producing hybrid swarms. In addition to complicating the identification of L. rivularis, this is raising concerns about the survival of L. rivularis and other lupine species (Sholars 2001 pers. comm., Rhymer and Simberloff 1996). Concerns have been raised about the disruption of the gene pool in other lupine species where contact with L. arboreus is occurring (US National Park Service 2001).

Sholars hopes to clarify the situation with North American lupines, including L. rivularis. In the interim, she has indicated that L. rivularis shows strong similarity to L. latifolius, although it is frequently swamped by L. arboreus. She does believe that it is a good species, although pure populations may now be rare, and that it can be separated readily from this and other lupine species by morphological features such as woodiness of the stem, hairiness of the keel, etc., as well as by characteristics such as habitat. She also indicates that they are elevationally distinct.

Given this, in spite of this general confusion with North American lupines, and following preliminary field investigations, specimen examination, and discussion with Sholars and others, we feel that Lupinus rivularis in BC is readily identifiable and fits the description provided by Douglas et al. (1999) (The Illustrated Flora of British Columbia, Volume III). Generally, growth habit, habitat, floral features and elevation of occurrence are distinct and easily separate this species from other coincident lupine species. The work of Kartesz and Kartesz (1994) is followed for nomenclature and synonymy.

Description

Lupinus rivularisis an attractive perennial lupine that reaches heights of between 4 and 6 dm. It is an erect herbaceous plant (though sometimes slightly woody at the base) with lovely lavender flowers that generally takes on a beautiful “bouquet” form (Figure 1). Superficially, the stems and leaves appear glabrous, though small hairs are apparent when magnified. The leaves occur primarily on the stem, and consist of a palmately whorled, delicate-looking, set of 6-9 leaflets that are hairy beneath but usually hairless above, with mucronate tips. The keel of the flower is hairy along most of its length. Seedpods appear black, or with black mottling. Leaves curl slowly inwards on picking.

Figure 1. Lupinus rivularis:Photos by Brian Klinkenberg, illustration courtesy of University of Washington Press.

Figure 1. Lupinus rivularis: Photos by Brian Klinkenberg, illustration courtesy of University of Washington Press.

However, because of the confusion in lupine identification, particularly in our region between Lupinus rivularis, L. arboreus, L. littoralis and others, further elaboration is necessary here. Taxonomically, several authors have described L. rivularis, including Taylor (1974), Dunn and Gillett (1966), Douglas et al. (1999), Hickman (1993), Riggins and Sholars (1993), and others. Illustrations are provided by each of these authors. Although there is some difference in key characters (i.e., Douglas et al. (1999) describe the species as solid stemmed, while Riggins and Sholars (1993) describe it as generally hollow-stemmed), these are likely differences resulting from variability in moisture, etc. (Sholars 2001 pers. comm.). L. rivularisleaflets can also be slightly hairy above, and petiole length can vary (Sholars 2001 pers. comm.). In addition, Sholars (2001 pers. comm.) notes that L. rivularis is generally herbaceous or only slightly woody at the base, while L. arboreus is usually woody.

In our region, L. rivularis appears distinct in the field because of its combined features which include early flowering (May), generally herbaceous hollow stem, delicate looking leaves, erect habit, distinctive lavender flowers, glabrous appearance, and its occurrence in dry gravelly or sandy habitats proximal to river or creek banks at low elevations.

The similarly coloured Lupinus littoralis is easily separated from this species in the field by its abundance of long silky hairs on the stems and leaves, and its sprawling appearance. In addition, while L. rivularis is always found proximal to river or creek banks, L. littoralis is found in coastal areas in the fog zone, on low dunes and sandy substrates, but not necessarily proximal to creek or river banks.

The very similar Lupinus arcticus ssp. subalpinus is also readily separated from L. rivularis by its occurrence at high elevations. Dunn and Gillett (1966) discuss the morphological similarity between these two species.

Hybrids and intermediate forms have been reported between L. rivularis and L. arboreus (yellow bush lupine) in California and elsewhere on the coast (Riggins and Sholars 1993), and hybridization complicates the picture here. We have observed and collected what we believe to be hybrids between L. rivularis and L. arboreus in the Fraser Valley. These plants were more robust than L. rivularis and showed colour gradations from pale yellow to light blue--features more similar to L. arboreus. However, leaf shape and characteristics resembled L. rivularis. Plants with similar intermediate features were observed in the collections we examined from California, Oregon and Washington. We have also found populations of L. rivularis that show some traits of L. littoralis, such as long silky hairs, although in the absence of other hybrid traits, this might fall within the variation of the species.

Hybrization features in lupines, and problems with classifications, are discussed by Mikolajczyk (1964), Nicholls and Bohm (1983), and Bisby (1981). Mikolajczk (1964) has demonstrated that visually large morphological differences in lupines are often the result of a single recessive gene, indicating the difficulty that lies in placing too much emphasis on single traits (Ganders 2001 pers. comm.). And Kazimierski (1961b) has noted that in hybrid plants, the seedpods resemble the ‘paternal’ species in crosses between L. mutabilis and L. douglasi, perhaps providing a clue for tracing parent plants in apparent hybrids.

Ensuring that populations studied here were “pure” rivularis, or close to it, was important in the face of hybridization, particularly hybridization between L. rivularis and L. arboreus, and because of the presence in the region of planted wildflower seed packages that may contain plants from California where hybrid swarms are common.

Although not published at this point in the literature, we have noticed a leaf characteristic in the field that, in combination with other features, we believe readily separates these two species, and may be an indicator of the presence or absence of hybrid genes. There is what we believe to be a major difference in leaf/leaflet behaviour between L. rivularis and L. arboreus that is noticeable both in the field and on herbarium specimens.

When freshly picked, the leaflets of L. rivularis exhibit a distinctive pattern of curling response. That is, the leaflets curl inwards from the tip towards the centre of the leaf, and do so relatively slowly. This is consistent in populations which we feel are good L. rivularis. Contrastingly, leaflets from hybrid plants between L. rivularis and L. arboreus, as well as from pure L. arboreus plants, fold up fan-like when picked, do not curl towards the central leaf point but rather the leaflets fold in towards the mid-vein, and fold up very quickly. This fan–like folding of the leaflets is particularly noticeable when pressing fresh specimens. Those that are L. arboreus, or which contain what we believe are L. arboreus genes, are more difficult to press in a flat manner. L. rivularis leaves readily press flat. We believe this reflects the distinctive evolutionary response to moisture retention that may be typical of L. arboreus, a species that grows in xerophytic environments, while L. rivularis has not developed this response, as its natural habitat is generally more moist. This is a feature that Sholars will explore in her work for Flora North America.

Page details

Date modified: