Ann. Mo. Bot. Gard., 15(3): 241-332. (1928)

The problem of species in the northern blue flags, Iris versicolor L. and Iris virginica L.
Edgar Anderson


The problem of species is a dual one. It asks two questions: What are species? How have they originated? In studying the nature of species we are first of all concerned with the extent to which the species of the orthodox taxonomists reflect actual divisions among plants and animals. Do such species, in the main, rest upon discrete natural groups of individuals or do they represent merely an artificial cataloguing device, ordering up, as best they may, the myriad forms occurring in nature.

It might be supposed that upon so fundamental a problem biologists would have come to some working agreement. Even the most casual survey of recent literature will show that this is far from being the case While several of those who have committed themselves to print on the subject may use such phrases. as "modern biologists recognize this to be the case," or, "all geneticists agree, etc., etc.," it is a noteworthy fact that no two of them are of the same shade of opinion. At one extreme are those who believe with Lotsy ('16) that, "The species is merely a conception of the human mind, and a very primitive anthropocentric conception in the bargain.* * * The species of the taxonomist is the comparative insignificant rest of large swarms of individuals which arose from the cross of two parents. * * * Species in the present taxonomic sense do not exist." On the other hand, we have such statements as Harper's ('23) "Linnean species do, by and large, constitute recognizable groups of more or less freely inter-breeding individuals. Only extremists deny the possibility of segregating and recognizing such units."

The disagreement as to species is just as extreme in regard to their size as in regard to their nature. If groups corresponding to our species do occur in nature, do they fit most readily into the species of the orthodox taxonomists (what we may call Linnean species for want of a better name) or would some other unit better express the relationships which we actually find among individuals? What are we to do with the vexed question of true-breeding sub-groups (variously termed micro-species, Jordanons, iso-reagents) whose existence has been demonstrated for numerous Linnean species? Are we to retain them as significant sub-groups within the Linnean species, are we to treat them as of even greater importance, or are we to cast them out altogether?

It should be a relatively simple matter to answer these questions for any one species. A comprehensive survey of variation within the species over its entire range would show whether it were an independent unit or whether it merged into other forms; whether there were recognizable sub-groups within it; and what interrelationships obtained between the individuals which go to make up these sub-groups and the species itself.

However, such an intensive study should be able to contribute to other questions besides the nature of species. It should offer valuable evidence for the more popular question of their origin. During the last quarter of a century an attempt has been made to study evolution experimentally. In the course of these experiments two types of mutations have been observed in our laboratories and breeding plots. They have been thought to be the same sort of changes which, operating in the past, have brought about the evolution of our present-day forms. The first type of mutation has been shown to be due to changes which take place at a particular point in the germ-plasm. Since only a single gene is affected, such changes have been called gene-mutations.

The second type of mutation has been shown to be due to re-alignments of the germinal material, to duplications of chromosomes or whole sets of chromosomes. The morphological results of these two types of mutation are quite dissimilar. An intensive and extensive survey of variation within a single species should therefore be able to demonstrate which of these two processes has been most active in causing progressive changes within that species. Such a study should, in other words, enable us to evaluate the evolutionary importance of gene-mutations and chromosomal re-alignments.

The present study is just such an intensive and extensive survey of two closely related species. It is an attempt to present a fairly complete picture of the variation within two natural groups of individuals over their entire range. It has as yet been almost purely morphological in scope. Though the morphological differences between individuals and groups of individuals undoubtedly rest upon basic physico-chemical ones, our knowledge of these physiological differences is as yet too incomplete, among the flowering plants, to possess much phylogenetical significance.

So far as is known, no such complete survey of variation within a species or group of species has ever before been made with plant material. It is an ambitious attempt for a single individual unless the problem be made as simple as possible. If we are to learn anything about the ultimate nature of species we must first of all reduce the problem to the simplest possible terms and study a few easily recognized, well-differentiated species. The group to be studied should therefore possess few sub-groups and intergrading forms. Unfortunately those species which have been selected for intensive study in the past have been chosen by reason of their very complexity.

The northern blue flags (Iris versicolor of the seventh edition of Gray's 'Manual') were accordingly chosen for the study, primarily, because they were a comparatively simple, stable, and well-marked group. They possessed certain other features which have materially reduced the labor involved in locating and studying a large number of individuals. They are common, conspicuous, colonial, and perennial. Though the merits of these characters are practically self-evident, it may not be out of place to call attention to the colonial nature of Iris versicolor L. and its close relatives. Plants of these species are seldom found as isolated individuals but usually occur in colonies of a few to several hundred plants. Nor are these colonies merely the vegetative offspring of a single plant as is the case with so many colony-forming species. The individuals vary so strikingly in the size, shape, and color of their flowers, that at blooming time the limits of a single clone can readily be determined. These seldom exceed a few square feet, though exceptionally large clones do sometimes occur, as will be described below. It was the truly colonial nature of the blue flags which was particularly useful in the present study, making possible the examination of large numbers of plants in each locality with a minimum of effort.

An attempt was made to visit as much as possible of the range of the northern blue flags during their flowering season. Numerous colonies were studied in detail and measurements made on twenty to fifty individuals of the characters which had been selected for study. Representative plants were sent back from each of the colonies and established in an experimental plot at the Missouri Botanical Garden, where they were made the subject of genetical, morphological, and cytological studies.

In 1923 collections were made in central Michigan. In the spring of 1924 a trip was taken through northern Missouri, eastern Iowa, southern Wisconsin, central Michigan, western New York, and central New Hampshire. In 1925 collections were made in central Missouri, northwestern Arkansas, northern Ohio, including the Bass Islands, western and central New York, and southern Vermont. Fruiting material was collected that fall from central and northern New Hampshire and eastern Michigan. In the spring of 1926 a trip was taken through southern Illinois, southern Missouri, eastern Arkansas, western Tennessee, and central Kentucky. In June, collections were made in Missouri, Michigan, northern Vermont, Ottawa, Canada, and at Lake Timagami, in northern Ontario, Canada. In the early spring of 1927 another southern trip was taken, principally to collect material from Mississippi and Alabama. In June of that year collections were made in Ohio, Maryland, Pennsylvania, New York, Ontario, northern Michigan, and Wisconsin. In 1928 North and South Carolina and eastern Tennessee were visited in early May. With the exception of the extreme north the territory has now been quite thoroughly covered as shown by the map in fig. 1.

Fig. 1. Localities at which I. versicolor and I. virginica have been studied.


Section 2