Circular - Volumes 77-110 - Page 3 (1944 )
Penstemon in Your Garden

(1) Glenn Viehmeyer is Assistant Professor of Horticulture. North Platte Experiment Station, University of Nebraska.

In 1946 a small group of gardeners formed the American Penstemon Society. The organization is dedicated to bringing this showy genus of wild flowers into cultivation and promoting its use as an ornamental

Had the organizers known the difficulties inherent in bringing a genus of wildings into cultivation, the society might not have come into being. Early attempts to hybridize species native to the plains and mountains were, with few exceptions, unsuccessful. As early as 1928, the author attempted a series of crosses of native western species without success, and abandoned the effort as an "impossible" task. In that attempt, not a single seed resulted from several hundred pollinations involving several species mated in all possible combinations. A. C. Hildreth cites similar results at the Cheyenne Horticultural Field Station.

Fortunately, the members of the new society were not aware that breeding and domesticating penstemons was "impossible." They sent an expedition into the wilds to collect plants and seed, members became collectors, and even more important, a seed exchange was established that provided members with access to a wide range of materials. Species that do not meet in nature were placed in intimate contact in gardens, and the stage was set for the occurrence of natural hybrids. Here curious amateurs began attempting crosses, not knowing that the task was an "impossible" one.

(2) Viehmeyer, G. 1958. "Reversal of Evolution in the Genus Penstemon," American Naturalist, Vol. XCII, 129-137.

At Columbia, Missouri, Fred Fate succeeded in crossing Penstemon grandiflorus and P. myrrayanus to produce the Fate hybrid. At Cook, Nebraska, Lena Seeba discovered a second hybrid in a progeny of P. grandiflorus seedlings that became the progenitor of the Seeba hybrid. Near Flathead Lake, Montana, George Murray collected and Anna Johnson introduced the Flathead Lake hybrid which was destined to play a leading role in the domestication of penstemon (1). Viehmeyer (2) has used the Flathead Lake hybrid in an extensive crossing program. In the gardens of Society members, species from many places were brought together and a few natural hybrids began to appear. These, in many cases, were not recognized as hybrids. They were more vigorous and better adapted to garden culture, and this was reason enough to collect their seed and send it to the seed exchange from whence it went out to other members of the Society. These people were not botanists, and botanists, and did not know about penstemon classification.

They planted the seed and called the resulting seedlings by the name written on the seed packet. As they became better versed in classification, they began to recognize species and to discover evidence of hybridization. The stage was set for the development of a whole new race of ornamentals for American gardens.


Penstemon is truly an American plant. The genus is confined to North America, with the exception of a single species that has crossed the Bering Strait to the Asian mainland. Few American genera have wider geographical or ecological distribution. Penstemon species are native from the Atlantic to the Pacific and from Alaska to Central America. They are found growing in alpine meadows, on talus slopes, rock ledges, sandhills, prairies, canyons, mountain tops, flood plains, deserts, in fact every sort of habitat except swamps.

Morphologically, penstemon species vary as greatly as do their ecological preferences. They range from tiny rock plants less than an inch tall to herbs, taller than a man. Among the 300 species, subspecies and forms are herbaceous plants, woody evergreens, and even a climber.

Unfortunately, most penstemon species are highly specialized and each species is adapted to a particular habitat. Many of them fail in the garden unless the gardener simulates conditions that prevail in their native homes. A number of species are easily grown. Some of these are listed in Table 1.

In Nebraska many species can be grown with average garden care. The main difficulty is with disease, which is discussed in another section of this bulletin. Species from the southwest deserts are particularly susceptible to leaf diseases and soil-borne organisms that often make them short-lived. The woody evergreens of subgenus Dasanthera may be grown only when they are given protection from winter sun and drying winds. They too are susceptible to diseases that must be controlled.

These "difficult" species have little place in the average garden. They are of interest to specialists and to breeders who want to use them in breeding programs. They can be grown in the Great Plains area if the grower is willing to invest the necessary time and effort. For the average gardener, hybrids are the most satisfactory. They are less sensitive to environment, have greater vigor and often live longer. Hybrid penstemons now available are the result of wide crosses that “breed true” for plant type.. They do not breed true in the genetical sense, but are uniform enough for garden use, even in advanced generation material. Colors are clear and there is little of the muddiness that appears in volunteer populations of such ornamentals as petunia, which "runs out," unless constant selection is practiced to keep the variety "pure."

Perhaps it would make the story clearer to be a little more specific and follow a cross through the first and second generations.

Penstemon barbatus Nutt. is a native of the plateaus and deserts of the southwestern United States, and ranges southward to Mexico. In this habitat it has, through thousands of generations, adapted itself to drought and heat and produces its brilliant scarlet flowers season after season. To the north of the range of P. barbatus, the blue-flowered P. strictus Benth. grows in high mountain valleys and on mountainsides. The two species cross rather readily in the garden to produce hybrids better adapted to garden culture and more ornamental than either parent.

The first generation of P. barbatus x P. strictus is intermediate between the two parents and much more vigorous than either. The flowers are not red as are those of P. barbatus, nor are they the blue of P. strictus; instead they are purple. If these purple flowered hybrids are grown in isolation and allowed to produce seed, plants grown from this seed will have red,purple and blue flowers in the ratio of 1 red : 2 purple : 1 blue. This is Mendelian segregation for the single factor pair, red and blue.) Less obvious is the segregation that occurs for characters that determine adaptiveness, which also follows the rules of Mendelian segregation, though in a far more complicated manner. Going back to the P. barbatus and P. strictus parents we find that P. barbatus has evolved in a desert where, through millenniums of residence, it has accumulated a genetic complex that fits it to live through intense drought and heat, while P. strictus evolved on high plateaus and mountainsides, where there are both more moisture and lower temperatures than occur in the habitat of P. barbatus.

When the two species cross, the first generation is a combination of germplasms, one of which is adapted to heat and drought and the other to cold with moisture. The first generation hybrids combine characters for tolerance to heat, cold, drought and moisture. They are intermediate in their environmental requirements. When seed of these first generation hybrids is planted, segregation for adaptation occurs, but it is a far more complicated process than the simple segregation for color. Tolerance for heat, cold, drought and moisture is controlled by many factors (genes) that interact to produce the end results. (Geneticists call such segregation multifactorial.) In the second and subsequent generations, multifactorially controlled characters recombine to form a multitude of variant individuals. Actually, in a species cross such as the one being discussed, there are potentially millions of possible gene combinations. Each of these may react to the environment in a different manner. At the one extreme of the hybrid population (if it is large enough) one might expect to find a few individuals that are more tolerant of heat and drought than the desert species P. barbatus and at the other extreme, individuals that are more cold-moisture tolerant than the montane P. strictus.

To the breeder this means that he will encounter in large advanced generation populations, segregates reasonably well adapted to any environment physically intermediate between the desert and the mountain-top. By choosing the individuals best adapted to his garden and interbreeding them, he may expect to approach the highest level of adaptation (to conditions prevailing in his garden) inherent in the combined germplasms of the parental species. For practical purposes, the following broad principles may be stated:

  1. If two species, which occupy different environments in nature, are crossed, the first hybrid generation will require an intermediate environment for optimum performance.
  2. Advanced generations, if large enough, will contain individuals that exceed the tolerance of the parental species to limiting factors of the new environment.
  3. The breeder, working with a segregating population in any intermediate habitat, may expect to select (if the population is large enough) lines fairly well adapted to the particular habitat in which he is growing his advanced generation population.
  4. The "raw hybrids" that are the result of interspecific hybridization offer the amateur breeder excellent material for penstemon improvement. By selecting and interbreeding those segregates best adapted to his garden he can isolate strains that perform well and consistently. To do this he needs only the most rudimentary instruction in plant breeding. The process is almost automatic if he chooses his breeding material from that part of the segregating population best adapted to his breeding plots.

The amateur or professional breeder should be aware that the breeding behavior of the interspecific hybrid is considerably different than that of intraspecific hybrids in that characters are inherited in entire chromosomes or blocks of chromosomal material (1). As a result, genes are linked in inheritance and do not assort independently. Such linkage may persist for generations.

Examples of this are found in the Fate hybrid of P. grandiflorus x P. murrayanus. In the Fate hybrid the narrow corolla, perfoliate bract and elongate inflorescence of P. murrayanus has persisted through 15 or more generations as have the inflated corolla, non-perfoliate but clasping bracts and more compact inflorescence of P. grandiflorus. The two forms are interfertile but the three characters retain the original association without recombining. (4).

In the case of the Flathead Lake complex, the "shark head" corolla shape of P. barbatus and P. labrosus dominate flower shape to the exclusion of the spreading lobes of all blue flowered parental species. Only by repeated backcrosses to a species having spreading lobes are the erect upper and reflexed lower lobes replaced by the character of spreading lobes.

Female fertility in F1 species crosses is often much reduced. This reduction ranges from about 60 to 100%. However, female fertility is rapidly recovered in advanced generations. Data on male sterility have not been studied.

These high levels of sterility may account for the failure of whole classes of phenotypes to appear in the F2 and subsequent generations. One may suspect that embryos bearing certain combinations of germ plasm are not viable.

There appears to be a cumulative effect in certain segregates which makes the individual hybrid less affected by conditions that might hamper or destroy the species parent(s). Uhlinger and Viehmeyer, working separately at North Platte, Nebraska, are selecting segregates with better garden survival and greater disease resistance.

Meyers (10), hybridizing within the subgenus Dasanthera, states: "Penstemon barrettiae hybrids (with other Dasanthera species) do so much better in my garden than the Columbia River plants do that l would bet that they would do better in the more difficult areas." P. barrettiae is endemic in the Columbia River gorge(7).

The F2 and advanced generations of a two-species cross may be depauperate in phenotypes as reported for the Fate hybrid above. Double crosses involving germ plasms of several species (multispecies hybrids) can be expected to yield greater numbers of morphologically distinct segregates as well as segregates with wider adaptation and greater disease resistance.


Since the discovery of the role of intermediary parents in penstemon breeding (Viehmeyer, 2), the genus is no longer a difficult one with which to work. The materials represented by the hybrids now available are sufficient to engender a whole race of new ornamentals. Yet these hybrids represent only a fraction of the species of the genus Penstemon. Other sections and other species may offer even greater possibilities to the hybridism.

By following the suggestions given below, the amateur or professional breeder may venture into unexplored areas and, perhaps, bring entirely new ornamentals into being. Plant breeding is no longer only the vocation of the expert trained in the science of genetics, rather, it has also become an avocation for hundreds of amateurs. These amateurs may well become the greatest single force for the improvement of ornamentals. Their activity is even now overshadowing the effort of the few professionals, simply because of their numbers.

Dr. Frank E. Palmer (11), onetime director of the Ontario Horticultural Experiment Station, a man who has made plant breeding his hobby and his lifetime profession, has this advice for the amateur: "There is no particular need to worry about genetics, although such knowledge can be useful. Generally leave genetics to the geneticists. Their knowledge is their livelihood. You're a hobbyist out to enjoy yourself—-part of the attraction of plant breeding is that it is an art rather than a science."

To Palmer's comment we might add "Plant breeding is creative work within the reach of anyone who is willing to spend a few hours studying a text on breeding methods or visiting with a plant breeder." You need not be a highly trained scientist to be an effective plant breeder. Remember that genetics is one of the tools you will learn to use as you work. If you are interested you will soon pick up the things you need to know. The following paragraphs give the techniques of penstemon breeding.


The flower of the penstemon is a "perfect flower" in the sense that it has both male and female parts. The male parts, the anthers, are attached to the corolla. The female parts are represented by the stigma, style and ovary. The latter organ develops into the capsule. The flowers of the penstemon are borne in a multiflowered spike or panicle and in order to facilitate the work of breeding, it is desirable that the number of individual flowers in the cluster be reduced. This is illustrated in Figure 1, where A and B show a cluster of penstemon buds before and after thinning.

The next step in penstemon breeding is to emasculate the flowers of the plant chosen as the female parent. This must be done before the pistil is mature. A bud in the proper stage is shown in Figure 1, C. Actually, the operator has considerable latitude and the bud may be emasculated as much as three or four days before the flower would normally open.

Emasculation is performed by grasping the lower side of the bud and pulling outward as indicated in Figure 1, D. The way in which the corolla is pulled away is important. By grasping the lower side of the bud, injury to the developing pistil is avoided. With a little practice anyone can emasculate a flower each second. Danger of pulling the entire flower bud off the plant is practically eliminated if the operator supports the bud by holding the calyx between the thumb and forefinger of one hand and removes the corolla and its attached stamens with the other hand. After the corolla has been removed there is no need to cover the flower to avoid contamination. None of the insects that normally pollinate penstemon will visit a flower so mutilated, and the probability of accidental pollination by midges or flies is too remote for consideration.

Pollen Collection

Anthers of the selected male parent are collected just as the flower is opening or just after it has opened. The former is preferable since it eliminates the possibility of visiting insects contaminating the anthers with foreign pollen. After anthers are collected, they should be spread on a pane of glass and dried in a warm room out of the wind and sun. As they dry, they split and pollen falls to the glass where it can be separated from the dry anthers and used at once or stored in glass vials for short periods. If only a few crosses are to be made, the breeder may prefer to squeeze the pollen from a mature anther onto a fingertip and apply it to the pistils of the chosen female parent.


When the penstemon flower is emasculated in the bud stage, the enclosed pistil will be immature (Fig. 1 E). It will not be receptive to pollen until the tip turns downward and a sticky knob forms as is shown in Fig. 1 F. If the flower is not pollinated, the pistil remains receptive for several days. It can be fertilized at any time during the period of receptivity. This is an advantage where large numbers of crosses are being made or if other work interferes with crossing.

As soon as a flower has been successfully pollinated, the style begins to wither and the ovary begins to swell (Fig. 1 G). Within three weeks the capsule reaches full size and during the fifth and sixth weeks seed is mature enough to harvest (Figure 1, H).

It is desirable to harvest the capsules while they are still high in moisture and before they begin to split. If harvested early, they should be stored in a warm place and dried rapidly. Unless care is taken, the capsules may mold or heat and the seed will be lost.

Keeping Records

As a plant breeder, keep records of your crosses and hybrid progenies. Whenever a cross is made, attach some sort of identifying tag. Small cardboard price tags with strings attached are excellent. The cost is low and they will last until seed is ready to harvest. When a flower is pollinated, a tag identifying both parents is attached to the stem. At harvest time the tag is harvested with the seed and remains until the seed is threshed. At that time the tag data are transposed to the packet in which the clean seed is stored.

In recording crosses always list the female or seed parent first. This is the standard method of writing pedigrees in plant breeding, exactly the reverse of animal pedigrees.


The penstemon breeder has a wide choice of material from which to choose his parents. He may choose to work with the species hybrids already existent; he may wish to bring the germplasm of new species into the breeding complex; or he may wish to try to make new species crosses.

For the beginner, working with existing hybrids is surest. Many of these are available and there is a wide range of compatibility between hybrids. They cross readily and they are "raw hybrids" that have not been purified through generations of selection. Seed of such hybrid material will produce a wealth of new types with minimum effort. From such hybrid populations the novice can choose those individuals best adapted to his own garden and make them the basis of his breeding program.

The Flathead Lake Complex, which includes direct descendants of the original Flathead Lake hybrid collected by Murray, its first generation hybrids with P. alpinus, P. cobea, P. glaber, and P. strictus, and the complex hybrids that have resulted from intercrossing these, will provide the most successful female parents for further hybridization. Individual hybrids vary widely in ability to accept foreign pollen. There are the expected incompatibilities between hybrids, and between hybrids and species to contend with but the overall aspect is one of a high degree of compatibility. Any single individual may not accept a given pollen but if the breeder uses a series of such hybrids as female parents he is almost certain to find one that will produce viable hybrid seed.

In Table 2 are listed the sections and species of the genus that can be expected to cross with the Flathead Lake material.

In addition to the species listed in Table 2, it is certain that species belonging in the sections Fasciculus and Peltanthera will be brought into the breeding complex represented by the Flathead Lake Complex. Breeders are working with other sections of the genus to produce entirely different hybrid populations. At present, penstemon breeding is a trial and error proposition. Only a small part of the total possible matings have been attempted. Successes have been frequent enough to encourage additional exploration.

As the beginner gains experience he may wish to bring additional species into the breeding complex. Here he can expect to encounter the difficulties presented by the isolating mechanisms that separate species in nature. He is advised to use hybrid forms as female parents and pollinate with the pollen of the species he hopes to bring into the breeding program. Sometime this is easy. In other cases it is difficult, and he may have to pollinate a whole series of hybrid females to get a few seeds. Some crosses have proved impossible thus far, e.g., crosses between existing hybrids and the species belonging to subgenera Dasanthera and Saccanthera as well as crosses between species belonging to sections Spermunculus and Ericopsis have failed.

Johnson: Flathead Lake Penstemon (1951)