A New Look at Erythronium Part 2

Revised with “Addenda” added January, 2011.  Copyright 2011 Arthur G. Guppy

An earlier, edited version of this appeared in the Rock Garden Quarterly, Vol. 66 (1), 2008

[ See part 1 here ]

To understand and appreciate the various Erythronium species and their relationships, it helps to have the names of all the species laid out in their natural groupings.  You will find that done below.  Keep in mind that species are not fixed entities that remain the same for all eternity.  They evolve and change, and at any moment in time, what we call a species is in the process of changing, though likely no change would be noticed within the span of several human lifetimes.    In some cases the identity of a species may be clear-cut, but in other cases, likely because the species is in the process of gradually evolving, its identity may be a matter of opinion.  Trying to describe and delimit species can be rather like trying to capture rainbows.  The best I can do is give my impression, and as much as possible follow tradition.

The article is not intended to be a complete identification guide to the species.  It is assumed that anyone growing or identifying Erythronium will have suitable reference books.   The purpose of the article, as the title suggests, is to provide a new way of looking at the genus, and I hope to be able to iron out some of the difficulties with identification and to offer some suggestions about growing the plants.  Making suggestions about gardening is difficult, because gardens elsewhere across the continent will be very different from my garden on Vancouver Island.

The two subgenera divide naturally into smaller groups.

Not only does the genus Erythronium divide naturally into two subgenera, but each subgenus seems to divide naturally into sections.  Here we will run into a problem with the two sections, the Concolorae and the Pardalinae, into which Applegate divided the western species.  Keep in mind that Applegate’s sections were published in 1935, when much that is known today about Erythronium was completely unknown.  Were Applegate writing today, I am sure he would not make that division into sections.  Using the information about Erythronium that is available today, I have found that the western subgenus divides naturally into sections that are very different from those named by Applegate.  Let us look at the entire genus with its natural groupings.  Keep in mind that I didn’t make these groupings.  I am trying to describe what evolution produced over time.

The Myrmecochorous Subgenus

Erythronium sibiricum from east of Novosibirsk

Erythronium sibiricum from east of Novosibirsk

The Eurasian Section

The Eastern North American Section

Erythronium rostratum

Erythronium rostratum, garden grown

The yellow-flowered group: base chromosome number n = 12.

  • Erythronium umbilicatum: diploid (2n = 24): southeastern U.S.
  • E. umbilicatum subsp. monostolum: border of North Carolina and Tennessee.
  • Erythronium rostratum: diploid (2n = 24): Mississippi region.
  • Erythronium americanum: tetraploid (2n = 48): Newfoundland to Tennessee to Wisconsin.
  • E. americanum subsp. harperi: Alabama, Tennessee, and nearby states.
Erythronium mesochoreum

Erythronium mesochoreum

The white-flowered group (tinted pink, purple, or bluish): base chromosome number n = 11.

  • Erythronium mesochoreum: diploid (2n = 22): Illinois to Texas.
  • Erythronium albidum: tetraploid (2n = 44): central and eastern U.S. and Ontario.
  • Erythronium propullans: presumed to be tetraploid (2n = 44) as it can cross with E. albidum: Minnesota.

I have omitted mention of varieties of E. dens-canis.  It seems that commercial bulb-growers are sometimes tempted to concoct names that look like scientific names to cater to the demand by collectors for as many named varieties as possible in their gardens.  I have yet to encounter a named variety of E. dens-canis that seems valid, but perhaps I will meet one yet.

In his book Buried Treasures, Janis Ruksans named two new subspecies of Erythronium sibiricum: E. sibiricum subsp. altaicum and E. sibiricum subsp. sulevii.  I am not yet prepared to express an opinion as to the validity of these subspecies, but they appear to be very interesting plants and I hope to be able to study them in my garden.  There are problems with the way the naming was done.  The type specimens are garden plants, and no adequate descriptions of their wild populations are given.  Also, the name altaicum has been used and is considered a synonym for sibiricum.

Applegate’s Subgenus 

Erythronium grandiflorum, red anthered form

Erythronium grandiflorum, red anthered form

The Grandiflorum Section (diploid, 2n = 24)

  • Erythronium grandiflorum: southern B.C. and Alberta to northern California to Colorado.
  • E. grandiflorum var. nudipetalum: Idaho.
  • Erythronium idahoense: eastern Washington, Idaho, and western Montana.

The Multiform Section

Erythronium tuolummense

Erythronium tuolummense

The group with plain green leaves (diploid, 2n = 24).

  • Erythronium montanum: southern B.C. to northern Oregon.
  • Erythronium klamathense: southern Oregon into northern California.
  • Erythronium purpurascens: southern Cascades and northern Sierra Nevada, California.
  • Erythronium pluriflorum: Madera Co., Sierra Nevada, California.
  • Erythronium pusaterii: Tulare Co., Sierra Nevada, California.
  • Erythronium taylori: Tuolumne Co, California.
  • Erythronium tuolumnense: Tuolumne Co., California.
Erythronium revolutum

Erythronium revolutum

The group with mottled leaves (diploid, 2n = 24).

Erythronium quinaultense from north of Lake Quinault, WA

Erythronium quinaultense from north of Lake Quinault, WA

The group of hybrid origin with only a trace of mottling (tetraploid, 2n = 48).

In the Flora of North America North of Mexico the spelling of the names of two species has been changed from that which has long been in use.  E. multiscapoideum is changed to E. multiscapideum and E. taylori is changed to E. taylorii.  I believe that a scientific name should not be changed unless there is a compelling scientific reason, and consequently I have retained the usual spelling of the two names.   It is true that in 1855, when E. multiscapoideum was not known to be in the genus Erythronium, it was given the name Fritillaria multiscapidea, but the oi spelling has been in use for a very long time.  I have not traced it back to its beginning, but it was used by Applegate in 1935 and by Munz in 1959 and no doubt by every author from then until it was changed in the Flora of North America in 2002.  Changing the spelling now is certain to cause an enormous amount of confusion.  In the case of E. taylori there is a problem with the interpretation of the International Code of Botanical Nomenclature.  The Code requires an ii ending after a consonant unless the name ends in -er, in which case a single i is used.  The dictionaries I have consulted give an “er” pronunciation for both the word tailor and the person’s name Taylor.  As the rule in the Code for an -er ending was made to accommodate pronunciation, I believe it should apply, and the spelling should be taylori.  As that is the spelling in the paper in which the species was named, I have retained it.

Applegate’s sections, the Concolorae and the Pardalinae must be abandoned. 

Above, I suggested a division of Applegate’s subgenus into sections very different from the familiar sections, the Concolorae and the Pardalinae described by Applegate.  He obviously felt that there was an important genetic difference between Erythronium with plain leaves and those with mottled leaves.  It is now evident that the difference in leaf coloration is an adaptation to the difference between snowy winters at high elevations and mild winters lower down, but it otherwise indicates no important genetic difference.  (I will comment in more detail below on the probable adaptive importance of the difference of leaf coloration.)  On  both Mt. Hebo in northwest Oregon and near Lake Quinault in Washington, at some time in the past the plain-leaved E. montanum has crossed with the mottled-leaved E. revolutum to produce a tetraploid hybrid species with only a faint trace of mottling.  Applegate did not encounter either species or he would have had to abandon his sections, for the hybrid species would fit in neither section.

There is other evidence against Applegate’s sections.  In my garden I have crossed E. montanum from San Juan Ridge on Vancouver Island with E. revolutum from a lower elevation, and produced over fifty vigorous hybrid plants that are apparently fully fertile.  Also I have collected seed from an E. montanum in my garden, and to my surprise got two apparently fully fertile hybrid offspring from the seeds.  Those hybrids have leaves that each spring unfold with quite strong mottling, but quickly fade to plain green.  Their white flowers have reddish markings near the base of the tepals which could only have come from either E. oregonum or E. californicum.  As in the year that I collected the seeds, both the E. oregonum and most of the E. californicum in my garden had finished flowering well before the E. montanum bloomed, I am reasonably sure that the pollen parent was E. californicum “White Beauty”, which often flowers quite late.  All the vigorous fertile hybrids in my garden between E. montanum and two mottled species, as well as the occurrence of  tetraploid hybrid species in the wild, demonstrate clearly that the plain-leaved E. montanum is genetically closely related to species with mottled leaves.  Consequently they should not be put in separate sections.

Mottled leaves and plain green leaves must each have an adaptive importance.

It seems probable that the ancestral Erythronium from which all existing species are descended had mottled leaves, as there are mottled species right around the northern hemisphere, while species consisting entirely of plain-leaved plants are confined to western North America.  There, because the plain-leaved species are confined to a large extent to regions with long, snowy winters, it seems reasonable to assume that plain leaves are advantageous where winters are long and snowy, but it is not easy to explain why that is true.  It has been suggested that leaf mottling originally evolved to provide camouflage as protection from grazing animals, but this idea seems contradicted by the fact that species with plain leaves, such as E. grandiflorum, commonly grow in open meadow habitat where the plants are greatly exposed to grazing, while species with mottled leaves are most often in shrubby habitat at low elevations where the shrubs provide some protection.

I suggest the following hypothesis.  At low elevations, where the winters are short and mild, the Erythronium leaf out very early in spring, often in February or March, when the sun is so low in the sky its rays strike the leaves at a low angle and give little heat.  Under such conditions, the plants need all the heat they can get in order to grow, and the brown mottling would be valuable for absorbing heat.  There is a contrasting situation where the winters are long and snowy.  There, the snow often does not melt and the Erythronium leaves cannot emerge until May or June, when the days are long and the sun’s rays strike the leaves from a high angle.  Under such conditions, the leaves are more likely to get too much heat than too little, and shiny green leaves that reflect some of the heat would be advantageous.  That could explain why species with plain green leaves have evolved in regions with long, snowy winters.

We must look at the division of Applegate’s Subgenus into two sections.

At this point an explanation is required for the rather strange division I have suggested of Applegate’s Subgenus into the Grandiflorum Section and the Multiform Section.  Looking at the entire subgenus, I can see only one break between species where there is no indication of a close genetic relationship and no record of hybridization.  E. grandiflorum with its close kin E. idahoense seem to stand in isolation from all the other western species.  Several times I have tried to cross E. grandiflorum with other species in my garden, but always failed.  That proves nothing, as even closely related species can have barriers to hybridization, but observations in nature provide additional evidence.  E. grandiflorum has a range over a huge area, has many different forms, and in a number of places grows in close association with other species, both mottled and unmottled, but no one has ever reported finding a hybrid, except with E. idahoense.  In Applegate’s 1935 monograph there is an account of his encounter with a mixed population of E. grandiflorum and E. klamathense flowering together on the Umpqua-Rogue Divide.  As both species have unmottled leaves, he felt they belonged together in the same section and therefore he expected them to hybridize.  He wrote, rather plaintively I think, “Whether or not they hybridize, I have not as yet been able to determine with certainty, although I am inclined to think they do.”  As E. grandiflorum has yellow tepals and fairly long, recurved stigma lobes (though lobe length depends on the strength of the plant), while E. klamathense has white tepals and a more or less unlobed stigma, Applegate would certainly have been able to spot any intermediates, had there been any present.  I do not want to make the same mistake as Applegate by being too sure of the sections into which the subgenus divides, so I must emphasize that the two sections I have suggested are tentative.  I am certainly not thinking of giving them scientific names.

My suggested sections are based entirely on evidence that species within each section seem likely to be able to hybridize with each other, while there is no evidence of hybridization between sections.  We know that E. grandiflorum and E. idahoense can hybridize with each other, but it is unlikely they can cross with any other species, so they belong together in one small section, which I have called the Grandiflorum Section.  There are so many records of hybridization among the other species that tentatively it seems they all belong together in the same large section, the Multiform Section.

I have included two possibly doubtful species in Applegate’s Subgenus. 

I have recognized two species that are not recognized in the Flora of North America North of Mexico.  That does not make me a “splitter”.  I have merely refused to go along with the “lumping” of two species, and I have done so on the basis of very strong evidence.  E. idahoense has long been lumped with E. grandiflorum as E. grandiflorum subsp. candidum.  An excellent paper by Jane Fritz-Sheridan (1988) shows that subsp. candidum and E. grandiflorum “are two biological species” that in nature can grow in mixed populations without hybridizing.  The paper does not assert that the biological species called subsp. candidum should revert to the name E. idahoense, but very cautiously leaves unanswered the question of whether it is a taxonomic species.  I am not that cautious.  It behaves as a biological species, it is easily identified by its white flowers, and a valid species name exists.  That is enough for me.  To me it is E. idahoense.

E. howellii is a rather different problem, though it also is excluded from the Flora of North America. The author who deals with the western species made the reasonable decision to lump E. howellii in with E. citrinum on the grounds that to her knowledge the two taxa differ only in E. howellii lacking appendages near the base of its inner tepals.  That left E. howellii without even a name as a variety, which is inconvenient because it clearly is a distinctive taxon.  As it needs more study, and even if I had done adequate research with it, I could not name it here as a variety or a subspecies, I am leaving it with its traditional name as a species.

Now let us look through the entire genus, species by species, starting with the Myrmecochorous Subgenus.      

The four species of the Eurasian Section are all rather similar, with E. japonicum being the most distinctive, as is indicated by the elaiosomes on the seeds being distinctly different from those of the other three species.  Identification in the garden must always be slightly uncertain unless the plants were raised from wild-collected seeds; in which case the geographic location always gives positive identification.  With bulbs obtained from commercial sources or grown from garden-collected seeds, one must always rely on the identification supplied with the bulbs or seeds, and hope it is right, while keeping in mind that the plants may be hybrids.  With very rare exceptions, the following descriptions will apply for plants from wild-collected seeds, and with luck for other plants.

E. japonicum has purple or pink flowers and purple to almost black anthers, and it is the only species which has appendages in the form of ridges or fins near the base of both the outer and inner tepals. The other three species have them only on the inner tepals.  E. sibiricum generally has purple or pink flowers with yellow anthers, but white flowers occasionally occur in the wild, and the rather doubtful subspecies “altaicum” named by Janis Ruksans has white flowers.  E. caucasicum has white flowers (sometimes pinkish) with yellow anthers.  E. dens-canis has purple or pink flowers with purple anthers and can be distinguished from E. japonicum by having no appendages on the outer tepals.  In the Balkans there are populations of E. dens-canis with white flowers which are easily identified by their purple anthers.

In the Eurasian Section, E. sibiricum requires our special attention as it has an enormous range and in suitable habitat carpets the ground in spring with its purple blooms.  Possibly it competes with E. grandiflorum of western North America as the world’s most numerous species.  Nevertheless, E. sibiricum is extremely poorly known.  Although we know that the species has two forms, one with mottled leaves and one with leaves that are plain green except for often being tinted reddish when young and often having reddish main veins, no one seems to know what the respective ranges of the two forms are.  Dr. Oleg Kosterin, a lepidopterist who lives near Novosibirsk, just west of the western edge of the range of E. sibiricum, is very familiar with the species.  In connection with his study of Russian butterflies, he has traveled extensively throughout much of Siberia, and he told me that he has seen millions of E. sibiricum, which is his favorite plant, but he has not seen even one with unmottled leaves.  Evidently we can be sure the mottled form is much the most common, and the range of the unmottled form remains a mystery, at least to anyone I have been able to contact.  Nevertheless, the unmottled form certainly exists, for it is available from bulb dealers and I have it in my garden, grown from seed.  The subspecies “altaicum” named by Janis Ruksans seems to include at least some of the unmottled plants, but probably not all of them.  (It has white flowers while others with unmottled leaves have pink or purple flowers.)  Also there is the problem with its name that I mentioned above.  E. sibiricum needs much more study.

The Eastern North American Section is quite well known, but a certain haziness surrounds even those plants because within both the yellow-flowered group and the white-flowered group there are species that are so alike that few people can distinguish between them.

In identifying species in the yellow-flowered group, the first observation might be of the plants’ habit of growth.  Two species, E. americanum and E. rostratum, spread by underground rhizomes (stolons) and consequently form patches of irregularly spaced plants, often with many of the plants not flowering.  The third species, E. umbilicatum, does not produce rhizomes, but spreads by seeds and by offsets tight against the parent bulb, with the consequence that mature plants usually have flowers and often form tight clumps.  Unfortunately, there seems always to be an exception to complicate matters.  E. umbilicatum subsp. monostolum does produce rhizomes, but that causes only a moderate amount of confusion as the subspecies is found only fairly high in the mountains in the border area between Tennessee, North Carolina, and Georgia.  There it forms patches of mainly non-flowering plants.

In the key to the Erythronium species in the Flora of North America North of Mexico, the yellow-flowered species are distinguished by the shape of the outer end of the seed capsules, and by the presence or absence of ear-like appendages (auricles) near the base of each inner tepal of the flowers.  Because the capsule develops from the ovary, the two are very similar in shape, and the description of the shape of the capsule applies fairly well to the ovary.  That means one generally doesn’t need to wait for the capsule to develop to identify the species, though for inspecting an ovary, a magnifying glass is useful.  The Flora places considerable emphasis on whether the capsule is held erect, or is at least off the ground, or is reclining on the ground, but with live plants one notices that the position of the capsule depends to a great extent on its degree of maturity.  In my garden, the capsules of all three species, when fully mature were on the ground, but those of E. rostratum were conspicuously held firmly erect until almost mature.  The amount of vegetation around the plants must make a difference, as it tends to support the capsules.  As all three species are adapted for having their seeds dispersed by ants, one would expect them to have their ripe capsules within easy reach of ants.

For identifying the yellow-flowered species, I place the emphasis on the shape of the outer end of the capsule or ovary and on the size and persistence of the attached style.

Fortunately, identification is much easier in the group with white or whitish flowers.  First, we can eliminate E. propullans.  It is endemic in a small area in Minnesota and is classified as an endangered species.  The flowers are smaller than those of any other Erythronium and often have only four tepals.  It generally does not produce seeds, but has the unique trait that flowering plants reproduce by sending out rhizomes that originate from the stem a short distance above the bulb.  That leaves us with only two species to identify.  E. albidum spreads by sending out underground rhizomes, while E. mesochoreum reproduces mainly by seeds, though it does produce offsets close against the parent bulb.  Leaf-shape provides the other important clue to identity.  E. albidum has leaves that we would think of as being the normal shape for an Erythronium, while E. mesochoreum has long, narrow leaves that are V-shaped in cross section.  Many E. mesochoreum plants have unmottled leaves, though it is my impression that the majority show a trace of mottling when they first unfold in spring.  Mottling, when present, can vary from faint to very strong.

Those who grow eastern Erythronium often find that plants of E. americanum and E. albidum spread so rapidly by fast-growing rhizomes that they seem more like weeds than good garden plants.  That need not be the case.  It depends on what clones you have and the amount of sun they get.  With a good clone of either species in a fairly sunny location, the majority of the bulbs should produce flowers.

In Applegate’s Subgenus, Erythronium grandiflorum comes first to our attention.

In looking at Applegate’s Subgenus, it is logical to start with E. grandiflorum as the most numerous and widespread species.  For those who are unfamiliar with western North American flora, I need to do some explaining.  As I have mentioned, E. grandiflorum has a huge range, and over that range it exists as numerous separate populations.  Because Erythronium, especially plants in Applegate’s Subgenus, lack an effective method for dispersing their seeds for even a moderate distance, populations that are separated from each other by a geographic barrier, even something like a stretch of forest, behave as if they were on islands.  Over time, each population tends to evolve its own characteristics, and in some cases populations are so different from each other it is difficult to believe they are not different species. Some of the different forms of E. grandiflorum have been named as different varieties or subspecies, but there are so many populations with intermediate characteristics, it seems impossible to identify what plant belongs in what named variety, so none, or almost none, of the varieties are recognized by botanists.

In a garden, the different forms of E. grandiflorum often behave very differently.  In general, the forms from high altitudes are a failure in most gardens.  They tend to quickly die out, and if they flower, the flower is a silly, upward-facing thing at ground level.  To the east of the Cascade Mountains, at fairly low altitudes close to rivers and lakes, there are populations that truly merit the name Erythronium grandiflorum.  The plants are large and have large, deep yellow flowers, often with dark red or purple anthers.  In the garden these generally thrive and put up a wonderful show, though they tend to have a regrettable habit of having somewhat curly tepals.  In the same general area, perhaps usually at somewhat higher elevations, there are populations of similar plants with yellow anthers, and there are occasional populations with both red anthers and yellow anthers, with a few pink anthers included.  In my garden I find that often plants with yellow anthers do not do as well as their red-anthered companions, perhaps because at some time in the past they came from a higher altitude.  The form of E. grandiflorum that does best of all for me was raised from seeds from Roman Nose Mountain in western Oregon.  It is smaller and has smaller flowers than the forms I have described above, and its anthers vary in color from cream to yellow, but it makes up for the lack of show by looking perfectly at home in a garden.  Roman Nose is a small mountain with a height of only about 870 meters and it is about half way between the northern and southern boundaries of Oregon, which is quite far south. Plants react to a lower latitude much as they do to a lower altitude, so it is not surprising that the Erythronium from there do well in a garden.

I must explain why in my list of Erythronium I have recognized one variety of E. grandiflorum, but only that one variety.  E. grandiflorum var. nudipetalum, which is endemic in a rather isolated part of Idaho, is a doubtful variety, but I have included it because it was recognized as a variety by C.L. Hitchcock et al (1969) and accounts by field researchers that I have read seem to support that classification.  Applegate named it as a species, E. nudopetalum (a different spelling from that given more recently to the variety), on the basis of its having two characteristics distinguishing it from E. grandiflorum: its stigma is without lobes or is only slightly lobed, and its tepals lack appendages at the base.  In my experience, the lack of stigma lobes often means nothing more than that the plant’s growing conditions are poor, but the lack of tepal appendages may qualify this Idaho form as a variety.  The Flora of North America does not recognize it even as a variety, and suggests it may simply be a depauperate form of E. grandiflorum.  I am treating it as a variety as a compromise, because I have not been able to study it.  The plant’s natural habitat is at a high altitude where winters are long and snowy, and consequently there is virtually no chance I could grow it in my garden.

There are two varieties of E. grandiflorum often encountered in the literature which I have omitted.  Applegate recognized var. pallidum and named subsp. chrysandrum, both on the basis of anther color, but those taxa were not recognized by Hitchcock and are not recognized in the Flora of North America.  They do not merit recognition because with this species there is almost no correlation between anther color and other important taxonomic features.  One finds populations consisting of very similar plants except they have different colored anthers, and it would not make sense to believe that plants growing together and otherwise looking identical are different varieties because they have anthers of a different color.

Mention of var. pallidum is frequently encountered in the literature, and while Hitchcock (1969) does not recognize that taxon as a variety he suggests it would perhaps qualify to be named as a botanical form with a range west of the Cascade Mountains, as it has pale anthers and that is the only anther color which has been recorded in that area..  The problem is that there are several distinctly different forms in the region from the Cascades west, and they cannot reasonably be grouped together as one botanical form or as a variety.  In different populations throughout the area, the anther color varies from white to pale yellow, the leaf shape varies from wide and almost erect to narrow, undulate and almost prostrate, and the plant’s stature varies from moderately tall to almost dwarf.  I am very familiar with the extremes.  From the window above my desk I look out at Mt. Prevost (780 m), a small mountain scarcely a half-hour’s drive from my door.  On its rocky summit the E. grandiflorum are very short and have long, narrow, undulate, almost prostrate leaves and white to pale yellow anthers.  If raised from seed, the plants retain those characteristics, though unfortunately, for unknown reasons, they tend to die out after a few years.  In contrast to the plants on Mt. Prevost, at Botanie Valley, which is a central valley in the mountains at the north end of the Cascades in southwestern British Columbia, the slopes above the valley are clothed with millions of sturdy, erect E. grandiflorum with wide leaves and anthers of various shades of yellow, except that scattered here and there, at a ratio of one in several thousand, are plants with red anthers.  As if to illustrate the difficulty of describing varieties of this species, high above the valley, my son Cris photographed a fine E. grandiflorum plant with pale pink anthers and yellow pollen.  The Botanie plants do not do well in a garden, unlike the similar though smaller plants far to the south on Roman Nose Mountain.  Perhaps some day someone will sort out reasonable varieties from this hodgepodge of different populations, but it has not been done yet.

Before leaving E. grandiflorum, I must mention the closely related E. idahoense, which is often treated as E. grandiflorum var. candidum.  It seems to differ from E. grandiflorum only in having white tepals and somewhat larger flowers (Applegate 1935), but those differences are significant, as Jane Fritz-Sheridan (1988) showed by careful research in western Montana that the two taxa are able to grow together in a mixed population without hybridizing; that is, they are able to behave as separate biological species.  There is an apparently reliable report of an extensive hybrid swarm of the two species south of Coeur d’Alene in northwestern Idaho, but that is not of great taxonomic importance, as hybrid swarms are quite common in the genus and are generally not used as reasons for lumping species together.  It simply means that in some mixed populations the species have evolved barriers to hybridization, while elsewhere barriers have not developed.

Of all the world’s Erythronium, it is the Multiform Section that is the most complex.

Turning to the extraordinary assemblage of Erythronium that I have called the Multiform Section, the plain-leaved group in that section can be thought of as dividing into two subgroups, a northern subgroup consisting of E. montanum and E. klamathense, and a southern subgroup consisting of five apparently closely related species, of which E. tuolumnense is the best known.  From northern Oregon to Mt. Waddington in British Columbia, in suitable subalpine meadow habitat, there are a number of quite dense populations of E. montanum.  These high-altitude plants are generally not successful in a garden, but there are two atypical populations at lower altitudes which can be a source of garden plants.  I have already mentioned the San Juan Ridge E. montanum which are fairly satisfactory in a low-altitude garden, especially if given a little low-nitrogen fertilizer, but they are not first-rate plants in cultivation.  It seems that at present I am the only person who knows of the existence of the other atypical population, though seeds from it will produce excellent garden plants.  I will have much to say about these two atypical populations later in this article.  In southern Oregon and in Siskiyou County, California, at fairly high altitudes, there are populations of E. klamathense, which is a species evidently quite closely related to E. montanum.  It survives in my garden and flowers, but I cannot say it thrives.

In California from the southern Cascades to high in the Sierras in Tulare County, there are five species with plain green leaves, all apparently closely related, and each growing in its own separate area.  E. purpurascens, E. pluriflorum, and E. pusaterii are all high-altitude species which likely will not prove satisfactory in a garden.  In my garden, E. purpurascens has flowered but so far not satisfactorily, and I am still waiting to see what the other two will do.  E. tuolumnense comes from a fairly low altitude, and does so well in the garden that it can be a nuisance by crowding out other plants, but nevertheless, it is a beautiful plant.  Most clones of this species tend to increase rapidly by producing numerous offsets and rarely set seeds, but recently I bought a clone of unknown origin which reliably sets seeds and seems to produce fewer offsets.  E. taylori is a recently discovered species from a somewhat higher altitude.  My plants of the species are still young but a few flowered for the first time this year.  The strikingly beautiful flowers are white, with the central area bright yellow, rather like the flowers of E. helenae, except they have white or cream anthers.  I notice with my young plants that some, instead of flowering, have started to form clumps by producing offsets, which suggests a very close relationship to E. tuolumnense.  This different reproductive behavior of clones within the same species seems to be quite usual in the genus Erythronium.  It is especially noticeable in the Eastern North American Section, especially with E. americanum and E. albidum.

In the group with mottled leaves, E. oregonum comes first to our attention. 

In the Multiform Section, it is in the group with mottled leaves that one finds the greatest complexity.  One species seems to require special attention.  E. oregonum is extremely variable throughout its extensive range from Vancouver Island to northern California, and the many variations seem to group naturally into forms that are somewhat distinct, but are poorly understood and can be confused with other species.  Accompanying this article is a photograph (page    ) for which I posed flowers representative of the two forms that are the most familiar to me.  On the left in the photo is a flower of what I call the northern form, which has white tepals with a showy reddish nectar guide towards the center on the face of the tepals.  This is the form native to Vancouver Island and with many variations to the south through western Washington.  In Oregon I have seen the form only quite high in the Cascades.  In a patch of these flowers, I like to lift the flowers one by one to admire the vivid colors of the nectar guides, for they are amazingly beautiful creations of nature designed not for our benefit, but for the practical function of guiding bees to the nectar so they will pollinate the flowers.  Typically these nectar guides consist of one or two zigzag reddish bars against a yellow background, but they vary greatly from the very showy guide in the flower I selected for the photo to a more modest narrow reddish bar against pale yellow.

On the right in the photo is a form of E. oregonum with very pale yellow or cream tepals and a rather modest orange nectar guide, a form to be found in the Willamette Valley of Oregon, especially in the vicinity of Corvallis.  In Britain, where there are no native Erythronium, but where there are probably more gardens with this beautiful genus than anywhere else in the world, this pale yellow form is called the sulphur form (the spelling in Britain), a descriptive name which I like.  It has been suggested that this is the form that Applegate (1935) named as subsp. leucandrum, but there are problems with that idea.  Applegate stated very clearly that “this subspecies differs from the species only in having white anthers instead of golden-yellow ones.”  The sulphur form near Corvallis has anthers that are almost identical in color to the tepals, or sometimes they are slightly darker.  They definitely are not white, and consequently to me the plants are not subsp. leucandrum.  If one looks at the list of the specimens of subsp. leucandrum collected by Applegate, they are almost entirely in an area from Roseburg on the north to Grants Pass on the south.  Unfortunately, I have not been in that area when the Erythronium were in bloom, so I have not seen subsp. leucandrum, but it seems likely that when I do get to that area at the right time I will find Applegate’s subspecies.

I have now mentioned three more or less distinct forms, one of which has been named as a subspecies.  Still to be considered is the area that was home to Applegate, the Siskiyou region of southwestern Oregon.  Undoubtedly Applegate thought of the E. oregonum of this region as being the typical species.  Unfortunately, it was many years ago that I was in the Siskiyous with the Erythronium in bloom, and my memory of them, though helped by photos I took on the long-ago trip to the region, is a little hazy, but I will do my best to describe how I believe they differ from the other forms.  They are most similar to the northern form, though it seems to me that the tepals were often a less pure white and the anthers were a little paler.  The striking difference — at least for the plants I saw — was that the nectar guide was like a burst of orange at the base of each tepal, rather than being reddish bars as in the northern form.

There is one other group of E. oregonum I must mention, though they may not be distinct in appearance from Applegate’s subsp. leucandrum.  In the now rather outdated A California Flora by Philip Munz (1959), E. oregonum subsp. leucandrum is given as occurring in Mendocino County, and it is described as having white anthers.  I have not seen that population, but I have an hypothesis which may explain the origin of white-anthered E. oregonum both in California and Oregon.  Both E. citrinum and E. californicum have anthers that vary from white to pale cream.  In Oregon, wherever a population of E. oregonum meets a population of E. citrinum, as happens along Eight Dollar Road near Selma, there is a swarm of hybrids.  In all probability such swarms have existed for thousands of years.  In northwestern California no doubt there have been similar swarms, and though I do not know of any contact between E. californicum and E. oregonum, such contacts likely would have happened in the past and produced swarms of E. californicum x oregonum hybrids.  With so many hybrids, many of which would have had white anthers, it is to be expected that there would be introgression of white anthers from the hybrids into the E. oregonum populations.  I hope that someone working with DNA will test this hypothesis.

For those who would like to look for hybrids, I should mention that the ones on Eight Dollar Road are very easy to miss.  E. oregonum and E. citrinum are both more or less white Erythronium that are superficially very similar in appearance.  One must stop along the road and look very closely at the flowers to see that the E. oregonum have wide stamen filaments and long styles with long, recurved stigma lobes, while the E. citrinum have slender filaments and rather short styles tipped by stigmas without lobes or, at the most, with very short lobes.  The hybrids, as one would expect, are intermediates, and often have rather deformed stigmas with the three lobes of different lengths.

Please keep in mind when you read my descriptions of the forms of E. oregonum, that I am generalizing about a species with flowers that vary in almost every imaginable way.  My main point is to warn those who wander about in in the west, especially in Oregon, that E. oregonum has different forms, and they should not be confused with other species.  More than once I have heard someone identify the sulphur form as being E. citrinum because its color seems to fit that name, though actually E. citrinum, in spite of its name, is usually more or less white.

Two very distinctive species scarcely need describing. 

The lovely pink E. revolutum is apparently a close relative of E. oregonum, and like that species, has stamen filaments that are wide toward the base.  A plant which is often sold as “Erythronium revolutum White Beauty” has flowers with narrow filaments and therefore cannot be a form of E. revolutum. (It is obviously a form of E. californicum.)  It is to be noticed that with the flowers of E. revolutum, the stamens tend to cluster closely around the style, rather than spreading apart, as is usual with E. oregonum.  That trait is important in identifying the flowers of the very rare albino E. revolutum, if one is fortunate enough to encounter such a plant.

E. hendersonii is an even more distinctive, and very beautiful species. Where its relationships are I cannot even guess.  I’ll leave that question to the DNA experts.  Its flowers can be described as violet to violet-pink, and vary from very pale to a good, bright color.  At their center is a very conspicuous dark purple nectar guide.  The style is very short (under 8 mm), the stigma lacks lobes, or has very short lobes, the filaments are slender, and the anthers tend to be violet or purple.

There are several species in Oregon and California that can be easily confused. 

Having mentioned the similarity of E. citrinum and E. oregonum, I must mention that in southwest Oregon and northwest California there are three species that not only can be easily confused with each other, but sometimes actually do lack clear-cut morphological differences.   These are all species with mottled leaves, white flowers, white or cream anthers, and narrow stamen filaments.    E. citrinum and E. oregonum can always be distinguished from each other by the different widths of their stamen filaments, but there is often no such clear-cut difference between E. citrinum, E. californicum, and E. multiscapoideum. (See photos of the three species on page __.)  Often the flowers of E. californicum have showy red nectar guides which make identification easy, provided one remembers that E. oregonum, which can have somewhat similar nectar guides, has wide stamen filaments.  It is when E. californicum lacks these red markings, which it often does, that identification becomes difficult.  Generally E. citrinum has a rather short style (6 to 10 mm) with a stigma that lacks lobes or has very short lobes, while E. californium has a longer style (over 10 mm) with well-developed, slightly recurved stigma lobes, but there can be considerable overlap of these sizes, especially with weak plants of E. californicum which have abnormally short styles and short lobes.  Fortunately, a well-developed plant of E. multiscapoideum identifies itself by appearing to have two or more scapes or flower stems from a single bulb, a unique trait of this species.  Instead of the flower stem (peduncle, if you want to be technical) branching high up into flower stalks (pedicels), it branches below ground, often at the junction of the leaf petioles, to produce two or more apparently separate flower stems.  That means that an E. multiscapoideum plant with only one flower may be impossible to identify, as it could be an E. californicum without a red nectar guide.  In summary, it is generally possible to identify E. citrinum by its short style and stigma lobes, and E. californicum by its longer style and lobes, while strong plants of E. multiscapoideum can be identified by what appear to be several flower stems from a single bulb, but there will be some plants that defy identification.  Such difficult plants may best be identified by identifying the populations from which they came.

In the same general area as the above three difficult species and the less difficult E. oregonum, there are two other somewhat similar species and a variety that may need a careful look if they are to be identified correctly.  I have already mentioned E. howellii, which looks very like E. citrinum except for having a different habit of growth and flowers that lack little ear-like appendages near the base of the inner tepals.  Here I can mention an additional reason why I am reluctant to follow the Flora of North America in giving up recognition of E. howellii.  Above, I have described three generally recognized species that tend to lack clear-cut dividing lines between species.  With E. howellii, in spite of its lack of the little appendages being a rather minor difference to distinguish it from E. citrinum, it is at least a clear-cut difference.  The other species to be noticed is somewhat farther south in California.  On and near Mt. St. Helena, one finds E. helenae, which is another white-flowered species with narrow stamen filaments, but in this case it is easily identified by its yellow anthers and a beautiful nectar guide which looks like a  large patch of sunshine at the center of the flower.

Some distance to the west of Mt. Shasta, there is a small area in the mountains with E. citrinum var. roderickii, a remarkable plant that can be readily identified by its white flowers with showy anthers with bright orange-brown pollen.  I have raised this beautiful variety from seed, and though my plants are still very young, I am already impressed with them.  To me there is no question that the variety is thoroughly distinct from var. citrinum.  The anthers of var. roderickii are pale pinkish or reddish with a purple tint, and have bright orange-brown pollen, and as soon as the anthers have dehisced it is that bright orange-brown color that one sees.  Anther color seems much more significant with E. citrinum than with E. grandiflorum because the normal anther color of E. citrinum is pale cream or white and there is no fluctuation from that color except for this one variety.  While my plants are too young for me to reach definite conclusions about them, my impression is that var. roderickii has a distinctly shorter growth habit than that of the typical variety.  It has been suggested that this new variety may be just a hybrid between E. citrinum and E. hendersonii, but I am skeptical.  It tends to have a longer style than var. citrinum, whereas E. hendersonii has a very short style.  Perhaps it has E. hendersonii in its ancestry, but that must have been far back in time, and evolution has been at work.

As if all of the above did not offer enough of a challenge for the person wandering about in northern California or southern Oregon and trying to identify the many lovely Erythronium he encounters, I must mention a very enigmatic population.  In 1993 Jim Robinett collected seed in Trinity County, California from a group of Erythronium which he described as being “in an area where californicum would be expected.  However, they displayed the stem-branching characteristics of multiscapoideum.”  I obtained some of that seed, and the plants I raised from it certainly fit Jim’s description.  They look exactly like normal E. californicum, except they branch exactly like E. multiscapoideum — at least most of them do, though a few branch an inch or so above ground level.  One might think they were simply E. multiscapoideum with a few plants careless about branching in the right place, but aside from being quite far from the normal range of that species, some have the red bar in the nectar guide which distinctively identifies E. californicum! (photo page __)  The probable explanation for the occurrence of this unusual population is that it is an E. californicum x multiscapoideum hybrid swarm, but there is the difficulty that hybrid swarms generally occur where populations of two species meet.  In this case the plants are in a known E. californicum area and a considerable distance to the west of the known range of E. multiscapoideum.  It seems that in this area the range of E. multiscapoideum must extend farther west than has been reported.  If anyone knows where Jim found that remarkable population, I hope he or she will write to me (care of the editor) with that information.   I would very much like to have a good look at it and at the surrounding area.

Two very unusual species require some explaining. 

There are two tetraploid species in western North America and both make excellent garden plants, though neither is well understood.  Both apparently originated from the hybridizing of E. montanum and E. revolutum accompanied by doubling of the chromosomes, and consequently they both tend to have characteristics that are intermediate between the two parent species.  Although E. elegans was named in 1985 and E. quinaultense not until 2001, the latter has been much the more clearly described and therefore is the more easily understood.  I will start with it and come back later to the very enigmatic E. elegans.

E. quinaultense was named by Geraldine Allen, and in the paper naming it she provided a key which identified the species as having leaves with only a faint trace of mottling and having flowers with tepals more or less white near the base shading to pink at the margins and tips. Unfortunately the plants often fail to conform to that description.  There is no great problem with the leaves, though to my eyes they often seem to be plain green.  In the garden even the flowers generally conform fairly well, but some have the outer tepals more strongly colored, especially on the outer surface, which is the description given in the key for E. elegans.  For me the serious problem arose when I visited a population of E. quinaultense in its natural wild habitat at an altitude of about 600 meters on a north-facing bank beside a logging road north of Lake Quinault in Washington State.  It was a rather miserable day with a dense misty rain falling, but I will not soon forget the beauty of  that scene.  There must have been a hundred or so flowers on that bank, and except for two paler ones, they were all a very beautiful shade of pale pink.  Far from having tepals that were white shading to pink at the edges and tips, the flowers were pink with the tepal edges paler.  (See the photo.)  No, I had not strayed to the wrong location.  I was at a site mentioned in Dr. Allen’s paper and the roads were easy to follow.  Furthermore, there is no other species with pinkish flowers and almost plain green leaves except E. elegans, and that is found only in northwest Oregon.

My garden had provided a clue that enabled me to understand the reason for what I was seeing.  My E. quinaultense plants had been grown from seeds very kindly sent to me by Dr. Allen some years before she named the species.  One precocious plant had bloomed a year before the others and its undersized flower was a clear pink.  The next year a number of the plants bloomed, and all of them, including the precocious one, had flowers that were white except for a little pink along the tepal edges.  About then I had to move my garden, and the E. quinaultense bulbs did not get into their new home until March.  They all flowered in spite of the late move, but all the flowers were a beautiful pink.  The following spring they went back to being mainly white.  The obvious conclusion seems to be that stress makes E. quinaultense flowers pink.  The soil on that bank beside the logging road seemed to consist mainly of stones embedded in something rather like clay.  Perhaps that or the spring weather subjected those plants to stress and made their flowers pink.  Such chameleon flowers present a taxonomic quandary.  How can we describe the species when its flower color is changed by its growing conditions and the amount of stress it is under?

Erythronium elegans is a very enigmatic species. 

E. elegans is a much more difficult problem. Its main population is on the rolling, hilly summit of Mt. Hebo (945 m) in northwestern Oregon, where it is spread over an area of some three square miles, though two small populations are reported from mountains 15 and 25 miles to the south.  It was named in 1985 by Paul Hammond and Kenton Chambers, who described it as having flowers varying in color from white to deep rose-pink, and as having leaves varying from plain green to being so heavily mottled as to be almost entirely brown.  When I read their paper, I immediately thought that such a varied population sounded to me more like a swarm of hybrids than a species, and judging from the description, the species involved would be E. montanum and E. revolutum.  I held to that hypothesis until I read that two botanists had counted the chromosomes of a small number of Mt. Hebo Erythronium and found them to be tetraploid.  That forced me to change my hypothesis.  I thought of how E. americanum in eastern North America was believed to have originated as a tetraploid species as a result of the hybridizing of two diploid species, and I decided that one or more of the hybrids on Mt. Hebo must have had a double quota of chromosomes and started a tetraploid population.  Otherwise I clung to my hybrid swarm hypothesis, and guessed there would still be diploids in the population.  There was one weakness in my idea.  I did not know if E. montanum and E. revolutum could hybridize.  It would be unlikely if Applegate’s division of the western species into two sections was correct.  To test my hypothesis, I crossed the two species in my garden and got over fifty beautiful hybrids, as I have already mentioned.

Since then I have spent several days on Mt. Hebo studying the population.  It includes a large number of white-flowered plants which appear to be identical to the E. montanum which grow on San Juan Ridge, Vancouver Island, in a similar habitat to that on Mt. Hebo, and at a similar altitude and distance from the sea.  Both taxa have plain green leaves, white flowers, long, recurved stigma lobes, yellow anthers, and narrow stamen filaments.  I believe, therefore, that the white-flowered plants on Mt. Hebo are E. montanum.  They are more successful in a garden than the San Juan Ridge plants because Mt. Hebo is much farther south and, as I have mentioned, plants react to a lower latitude much as they do to a lower altitude.

For a time I felt rather uncertain of my conclusion that the white-flowered plants on Mt. Hebo were E. montanum, as I was haunted by the fact that a small sampling of the population had been found to be tetraploid.  One would expect E. montanum to be diploid.  I am not equipped to do chromosome counts, so I resorted to what I call a “poor man’s chromosome count”.  Using a white-flowered Erythronium that I had raised from seed from Mt. Hebo, I forced open a flower just before it would have opened naturally, and removed the anthers to prevent self-pollination.  Then I enclosed the flower in a fine-mesh bag, and twice during the next few days, I opened the bag just long enough to dust the stigma with pollen from a San Juan Ridge E. montanum (which I am sure are diploid as they can cross with E. revolutum).  As the plant produced a full capsule of seed, I feel sure it is diploid, which is what one would expect an E. montanum to be.

The largest proportion of the Mt. Hebo plants have flowers of some shade of pink, and these are very difficult to explain.  In the paper in which she named E. quinaultense, Geraldine Allen accepted the name E. elegans for the Mt. Hebo population, but considerably changed the description of the species from that given in the paper by Hammond and Chambers.  She concluded that the species is tetraploid and must have originated in much the same way as had E. quinaultense.  In her paper she gave a key to distinguish between the two species, in which E. elegans is described as having “tepals more or less white to pinkish, the outer ones generally more strongly colored, especially on the outer surface.”  In contrast, E. quinaultense is described as having “tepals more or less white near the base, shading to pink at the margins and tips.”  She made no specific mention of the white-flowered plants in the E. elegans population.

Unfortunately, I have found that the descriptions in Dr. Allen’s key do not reliably apply to the plants in nature or in the garden.  I have already mentioned the wild population of E. quinaultense in which almost all of the flowers fit the key description for E. elegans.  Even a few plants of E. quinaultense in my garden have the coloration supposed to belong to E. elegans.  On Mt. Hebo I found the pink-flowered plants to be extremely varied, and could find no distinct group of plants that seemed to fit the description of E. elegans in the key.  To complicate matters, I found that often with plants with two flowers, the most recently opened flower would be almost white, while a flower that was a few days old would have taken on a strong pink coloration.  As I have often observed with E. revolutum x oregonum hybrids, it is a usual characteristic of such hybrids that the pink coloration deepens rapidly after the flowers open.  That rapid color change makes it very difficult to describe the flowers, as they are not the same from one day to the next.  No one can precisely describe the color of the flowers of a species, when various influences can change that color, and without a reliable description of E. elegans, there is no way of knowing whether all the pink-flowered plants in the Mt. Hebo population are that species, or whether some are simply diploid E. revolutum x montanum hybrids, like the ones in my garden.

I am back almost to where I started my investigation of the Mt. Hebo population.  I now know it must include some tetraploids we can call Erythronium elegans, but I have only a vague idea what they look like.  I continue to suspect the population includes diploid hybrids, but I cannot be sure.  I have made some progress, because I now know there are a large number of E. montanum in the population, and they are a form of that species that does well in the garden, and that is very useful knowledge.  Nevertheless, for me the Mt. Hebo population remains the great Erythronium enigma.

Hybrids are of great interest and give us information about the parent species. 

A byproduct of investigating the Erythronium on Mt. Hebo was that I raised from seed a beautiful group of E. revolutum x montanum hybrids.  I had long been interested in Erythronium hybridization, and especially in a phenomenon I had become aware of as a result of reading Applegate’s 1935 monograph.  In connection with the hybrids that occur in nature where two Erythronium species meet, he wrote “Usually there are unmistakable indications of hybridization.  These consist of color changes and various malformations of flower parts.”

About the time I read those perceptive lines written by Applegate, I learned about a very interesting hybrid swarm of E. revolutum x oregonum at Skutz Falls, not far from where I am now living on Vancouver Island.  At that time it was almost a two-hour drive from what was then my home, but reaching it was well-worth the drive, and I spent much time studying those plants.  Near the river falls in one direction there was hilly, dry-in-summer rocky land with some impressive patches of E. oregonum, and in the other direction there was a moist river flood plain with many E. revolutum.  Near where those two very different habitats met, at some time in the past someone had cleared and leveled a large patch of land, and then apparently abandoned it.  Various wild shrubs were growing up on the cleared patch, and among them were hundreds of E. revolutum x oregonum hybrids.  There was no doubt that Applegate had been right.  I found all sorts of interesting malformations among the hybrids.  The most remarkable was a hybrid with a rather large, upward-facing, pale pink flower that had the stamen filaments fused to the tepals so the anthers seemed to be growing on the tepals.  As Murphy’s Law dictated, I found that remarkable flower on a day with pouring rain and was unable to take a photograph.  A very conspicuous and quite common aberration was a flower with extra tepals.  On an average day I would see about half a dozen of those.  Less conspicuous and less common were flowering plants that had one very large leaf instead of the usual two.  There was one extraordinary flower with two pistils, a large one with four stigma lobes, and a small one with one lobe.  None of these extreme malformations occurred year after year in the same plant, though the plants that displayed abnormalities tended always to be somewhat abnormal.  Upward-facing flowers were the most common aberration of these abnormal plants.  I must confess that I was somewhat disappointed when a plant which had a showy eleven-tepalled flower one year had a perfectly normal flower the next year.  In spite of Applegate’s suggestion that malformations were to be expected in hybrid groups, I was a little surprised at the number I encountered at Skutz Falls.  However, that swarm of hybrids had obviously been there for a considerable length of time, and there had been much pollen exchange between hybrids and some backcrossing to the parent species, and I learned from raising my own groups of hybrids that it is just such a situation that produces the greatest number of abnormal plants.  Groups of first-generation hybrids that I raised from crossing the two parent species included very few plants with abnormalities.

For the person interested in taxonomy, the occurrence of numerous malformations in a group of hybrids is much more than an interesting natural phenomenon.  In the past there have been taxonomists who have suggested that E. revolutum and E. oregonum are merely two different color forms of the same species.  The occurrence of a large number of abnormal plants in the Skutz Falls population absolutely proves that they are two different species with different chromosomes that do not pair up properly in the hybrids.

Erythronium hybridization can be of special interest to the gardener. 

For those who wish to produce interesting hybrids in their gardens, the Myrmecochorous Subgenus has little to offer.  The four species of the Eurasian Section probably could be crossed with each other, but the species are so similar that the hybrids would not be very exciting.  In the Eastern North American Section there are four different chromosome numbers among six species, and as there is little chance of  a successful cross between species with different chromosome numbers, there are few possibilities for hybridization in the section.  E. umbilicatum likely would cross with E. rostratum, but with both species so alike, one would scarcely know the offspring were hybrids.  E. propullans is known to be able to cross with E. albidum but apparently the offspring are very weak and do not survive for long.  In the wild the two species grow together, but there seem to be no reports of groups of hybrids.  A report published in 1993 by Thomas Morley, a researcher at the University of Minnesota, described research in which he crossed the two species, collected the seeds, and planted them, but got very poor germination and very poor survival of the offspring, with only two living for any worthwhile length of time.  He did get flowers and was able to confirm that the plants were hybrids, but they soon died.  Clearly no one is going to do much hybridizing in the eastern section.

Those who want to do exciting hybridizing must go west to Applegate’s Subgenus.  Even in that subgenus, the Grandiflorum Section has little to offer, but by accident I did get one  splendid plant from an E. grandiflorum cross, though it was not from the crossing of two different species.  The beautiful plant was the result of a cross between two of the very different forms of E. grandiflorum that some call varieties.  I was trying to grow a naturally small form of E. grandiflorum with cream anthers from the Washington Cascades, but I had not been very successful.  One plant did flower and produced seeds, and in the hope of having better success with new plants, I collected and planted the seeds.  The plants I got from those seeds were a pleasant surprise.  Evidently an enterprising bee had been at work and had carried pollen from one of the large-growing plants with dark red anthers and had deposited it on the stigma of the little pale-anthered plant.  I lost many of the seedling plants in the recent chaos of moving my garden, but those that survived have started to flower and have turned out to be intermediates between the two very different forms.  One of those offspring of the little pale-anthered plant gives promise of being as large as its pollen parent, and its first flower had anthers of the most brilliant red one could imagine.  If it is happier in garden conditions than its seed parent, it should be a very showy addition to my garden.

The Multiform Section will always be the happy hunting ground for hybridizers.  The possibilities seem almost unlimited.  It would seem possible that any two diploid species in the section could cross with each other, except that some species have evolved protective barriers to hybridization which may sometimes prevent that happening.  I have found, for example, that when crossing E. revolutum with E. oregonum, when I used E. revolutum as the seed parent, I got full capsules of seed, but every attempt failed when I tried to use E. oregonum as the seed parent.  It seems that at least some forms of E. oregonum have a barrier to hybridization, and it is possible that even some populations of E. revolutum may have evolved a barrier.  Hitchcock (1969) reported that he had grown E. revolutum and E. oregonum intermixed for nearly 20 years and had seen no intermediates.  My experience with the two species has been very different.  I have already mentioned my observations at Skutz Falls, and every spring I visit the beautiful large garden of friends, where under the trees there are masses of white oregonum and pink revolutum and a delightful blending of the two produced by thousands of hybrids of various shades of pink.  In addition there are small patches of violet E. hendersonii and yellow E. tuolumnense, and here and there hybrids of each of those with E. oregonum.  It is interesting that Hitchcock, in commenting on his E. revolutum and E. oregonum not hybridizing, wrote “for which reason I am inclined to regard them as distinct species rather than geographic races of one.”  Had he seen  the numerous aberrations among the hybrids at Skutz Falls, he would have known that the presence of those aberrations was proof the two taxa were distinct species, even though some populations hybridize readily.

Almost all my hybridizing efforts have been part of taxonomic experiments, so I have much to learn about the possibilities for the activity.  I have hybrids in my garden, but most are either the byproduct of an experiment or arrived by accident.  I was rather fond of an accidental E. hendersonii x oregonum cross but I fear I lost it in the chaos of moving my garden.

The accidental hybrids that have turned up in my garden seem insignificant compared with three that I was shown in a friend’s garden.  My friend had owned the garden for years, and it was a well-established garden when she bought it, so the origin of the Erythronium in it is unknown, but it has great numbers of E. oregonum and of an especially bright pink form of E. revolutum.  Recently three especially large Erythronium had caught my friend’s attention, and she pointed them out to me.  They were all somewhat different, but it was obvious from their bright pink flowers with reddish markings near the base of the tepals that they were all E. revolutum x oregonum hybrids.  They were tall, with two or three flowers each, and one of them had the largest leaves I have ever seen on an Erythronium.  I wondered if they could be polyploids, and decided to test them with a “poor man’s chromosome count” by applying pollen from each of them separately to three E. quinaultense in my garden.  All three produced full capsules of seed, so I believe my friend’s three remarkable Erythronium are all tetraploid.  I look forward to seeing what I can grow from the seeds from the three test capsules.  The offspring should have genes from E. oregonum, E. montanum, and a double allotment from E. revolutum.  It is an interesting thought that what happened in my friend’s garden is the same kind of event, except for one different parent species, as must have happened at some time in the past near Lake Quinault and on Mt. Hebo.  If the three remarkable garden plants were in a suitable bit of wild forest land, they could pollinate each other and start a unique new species.

Being an amateur botanist can be a dangerous activity. 

I have made one attempt to use hybridization to produce a new garden plant.  I had read of a number of hybrid cultivars produced by crossing E. tuolumnense with a white-flowered species, and I even had two such cultivars in my garden.  I decided to join the throng, but to produce something just a little different, so I crossed E. tuolumnense with E. helenae.  The cross didn’t take very readily, and I got very few seeds, and only one germinated.  However, that one grew up, and now many years later, has formed a large clump of bulbs.  It won’t win any prizes, as the flowers, though a nice bright yellow, are rather small and are widely spaced on tall, lanky stems.  Nevertheless, I am fond of my creation, as in early March the large leaves, which are mottled with bright green veining on brown, form the showiest plant in my Erythronium garden.  The plant had to have a name, and surely it deserved to have a name as absurd as that of any cultivar on the market.  Remembering that one of its parents was E. helenae, I decided to call it “Erythronium Helen’s Baby”.  The fact that E. helenae was the pollen parent added a nice extra touch of absurdity.  It was when I was writing this article that I realized I was on dangerous ground.  The natural habitat of E. helenae is on Mt. St. Helena in California, and it seems likely that Applegate, who named the species, had that in mind when he chose the name.  Obviously the name I had given my plant could get me into serious trouble.  I rushed to a botanical Latin reference book, and to my great relief found that I was safe.  The book says that a plant named after Mt. St. Helena must have the name sanctae-helenae.  Applegate, whether he knew it or not, had not named the species after the mountain.  He had named it after a woman or girl named Helen.  No saint was involved!

I have found that the most difficult, and certainly the most hazardous, part of being an amateur botanist is getting published.  I wrote an article on my discovery that the genus Erythronium divides naturally into two subgenera, and in the hope of getting it published, sent it to a botanist who has his own publication.  He sent it to another botanist who is senior author of at least three papers on Erythronium.  The first botanist told me bluntly that my discovery was of absolutely no importance, and he published my article only after deleting all references to subgenera.  After I had spent some fifteen years doing the research, that was rather crushing.  The second botanist commented on my article in an e-mail to someone else, which was passed on to me.  In it he wrote that he “found the article of interest” and that he was “totally unaware of the role of ants in moving seeds for some members of this genus.”  Coming from a person with extensive knowledge of Erythronium, I found that very encouraging.

Apparently the botanist who published a shortened version of my article believed that a subgenus does not exist until it has been given a scientific name.  As in practice scientific names are generally published in professional botanical journals, which usually do not publish the work of amateurs, he evidently believed that the subgenera could not exist because as an amateur I could not name them, and therefore I should not have written that I had discovered them.  By publishing my article with my conclusions deleted, he had in effect invited any professional botanist in need of an important topic for a paper, to name the subgenera based on my information, which would mean that the professional botanist would get credit for the discovery.  I contend that with Erythronium, evolution has so clearly separated the two groups, there can be no doubt they are subgenera.  Also, as I mentioned near the beginning of this article, I am not the first person to suggest that the two groups are subgenera.  Shevock et al.(1990) described the two groups as “perhaps corresponding to subgenera”, so I was simply proving what had previously been considered a possibility.

I was especially interested in the experienced botanist’s statement that he had been totally unaware of the role of ants.  It confirmed my belief that an amateur botanist with a garden — providing he has basic botanical knowledge and considerable expertise at gardening — can make important contributions to botanical knowledge.  No botanist working in a laboratory could have learned what I have learned in a garden, and it would have been virtually impossible for a botanist working in the field to find all the various Erythronium species at the right time to observe ants collecting the seeds.  My point is that professional botanists working at universities are not the only people who can contribute to botanical knowledge.  Those who raise plants in a garden can contribute too.  I am confident that in the future every author writing about the genus Erythronium will divide it into the two subgenera that were discovered in my garden.  That is my garden’s contribution to botanical science, though it remains to be seen who gets the credit.

References Cited in Part 2 

Allen, G.A. 2001. Hybrid speciation in Erythronium (Liliaceae): a new allotetraploid species from Washington State. Systematic Botany 26:263-272.

Applegate, E.I. 1935. The genus Erythronium: a taxonomic and distributional study of the  western North American species. Madroño 3:58-113.

Flora of North America Editorial Committee. 2002. Flora of North America north of Mexico.  Vol. 26. Oxford Univ. Press, New York, NY.

Fritz-Sheridan, J. K. 1988. Reproductive biology of Erythronium grandiflorum, varieties  grandiflorum and candidum (Liliaceae). Amer. J. Bot. 75:1-14.

Hammond, P.C. and K.L. Chambers. 1985. A new species of Erythronium (Liliaceae) from the Coast Range of Oregon. Madroño 32:49-56.

Hitchcock, C.L., A. Cronquist, M. Ownbey and J.W. Thompson. 1969. Vascular plants of the Pacific Northwest. Part 1. Univ. Wash. Press, Seattle, WA.

Morley, T. 1993. Apparent hybrids between Erythronium albidum and E. propullans grown from wild seed – Final Report. Minnesota Plant Press 13:1-2.

Munz, P.A. 1959. A California flora. Univ. Calif. Press, Berkeley and Los Angeles, CA.

Parks, C.R. and J.W. Hardin. 1963. Yellow Erythroniums of the eastern United States. Brittonia 15:245-259.

Shevock, J.R., J.A. Bartel, and G.A. Allen. 1990. Distribution, ecology, and taxonomy of     Erythronium (Liliaceae) in the Sierra Nevada of California. Madroño 37:261-273.


About Art Guppy

Art spent over 70 years studying and writing about native plants of the Pacific Northwest from BC to California, especially the genus Erythronium and related plants. This site is a compilation of his work for the benefit of naturalists everywhere.
This entry was posted in Erythronium, Taxonomy and tagged , , . Bookmark the permalink.

One Response to A New Look at Erythronium Part 2

  1. Pingback: Ants and a New Look at Erythronium Part 1 | Art Guppy, Naturalist

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s