Carl Epling

THE BREEDING GROUP AND SEED STORAGE: A STUDY IN POPULATION DYNAMICS

This is a report on the yearly frequencies since 1941 of blue and white flowered plants of Linanthus parryae, a conspicuous annual of the Mojave desert. Like most desert annuals, its abundance and dispersion varies greatly from year to year and from place to place in response to weather. The size of the breeding group would be expected to vary with these expansions and contractions. An initial survey (Epling and Dobzhansky, 1942) suggested that the inheritance of color depends on one pair of alleles. Having these natural markers, the species appeared to be a useful subject with which to study the effects of selection and genetic drift. The initial survey sampled a naturally delimited area of about 600 square miles. Linanthus was extraordinarily abundant in 1941 and the population was almost literally continuous throughout the area (fig. 1) . For the most part it consisted of white flowered plants. Three mixed areas were found, however, in which the color frequencies were locally highly variable. Only occasional blue individuals or small patches were observed in some parts of the white area. Subsequent sampling has shown that these mixed areas are persistent and that their limits have not materially changed. Four years sampling of the westernmost mixed area by the coarse method used in the initial survey also indicated that color frequencies tended to remain constant in very local populations. Permanently marked experimental plots were accordingly constructed within this area, which have permitted a systematic and detailed sampling of dispersion and flower color each year since 1944. The exact mode of inheritance of flower

Supplementary notes on American Labiatae-VII

Notes on the taxonomy and distribution of various species of Labiatae of Mexico, and Central and South America, are presented, as part of a continuing series of supplements to the authors’ earlier monographs. Eight new species are described in the generaHyptis, Salvia, Satureja, andScutellaria. Two new combinations are made, one each inHarlanlewisia andSatureja. A key and revision are given for the six species ofLepechinia sect.Parviflorae.

INCREASE OF THE ADAPTIVE RANGE OF THE GENUS DELPHINIUM

The genus Delphinium is closely allied to Aconitum. Species of both genera resemble each other closely in vegetative habit, and certain Eurasian species of both approach each other in details of floral structure. Both genera are similar cytologically. The chromosome number is the same in both. The karyotype of Delphinium is apparently similar throughout the genus, except for ploidy (Lewis et al., 1951). But Schafer and La Cour (1934) have shown that the karyotype of Aconitum is variable from species to species, and within species, in respect of chromosome length and spindle attachment. Of particular interest here is the fact that a choice of certain chromosomes from the variants shown by them (p. 702, chromosomes ABDEIJKM) would provide a genome scarcely distinguishable from that common to Dephinium species. These similarities of structure and karyotype strongly suggest therefore that both genera are divergent members of the same phylad.!

DELPHINIUM GYPSOPHILUM , A DIPLOID SPECIES OF HYBRID ORIGIN

The purpose of this paper is to examine evidence that Delphiniitni gypsophilumtn had its origin by hybridization at the diploid level without change in chromosome number and by a more direct process of recombination than that ordinarily referred to as introgression. This hypothesis was first stated in an abstract (Lewis and Epling, 1946), and was discussed later by Epling (1947) and by Stebbins (1950) after more data were at hand. We shall now discuss the total evidence that has accrued since our abstract appeared. The genus Delphiniumti is mostly associated with plant communities that have been derived during later Cenozoic time from the Arcto-Tertiary Geoflora of the New and Old Worlds. The species we have studied are associated with present derivatives of the Madro-Tertiary Geoflora: the oak woodland, chaparral, and to some extent, the shrubby communities of the Sonoran desert. They have generally independent ranges with a considerable overlap in some cases, and some are included within the ranges of others. Whether their ranges overlap or are included, the species that grow in the same vicinity generally occupy ecologically different sites. Repeated observations indicate a considerable restriction of migration from colony to colony, either by pollen or by seeds. For example, blue or white flowered colonies of the same species persist unchanged for indefinite periods within short distances of each other and in similar habitats (Epling and Lewis, 1952). Thus, each colony has a considerable permanence of its adaptive attributes. The colonies of two species are sometimes contiguous and may be composite, providing ample opportunity in either case for gene exchange between them. The vectors of pollen are bees, especially Bombus. There is no apparent selectivity on their part for flower color or conformation, and visits from plant to plant in mixed populations appear to be at random. The species with which we shall deal are self compatible, but the behavior of the bees results mostly in cross pollination because of protandry. Barriers to interspecific hybridization are often lacking, not only because of the pollinating process, but also for genetic causes. A full seed set occurs after most first crosses. The seeds are viable for the most part and generally produce normally developed and vigorous F1 plants. The fertility of F1s when selfed is sometimes high and sometimes low but is reasonably consistent for each interspecific cross. The seed set is generally higher after backcrossing. Even though fertility is low for a particular cross, some individuals, even triploids, will set occasional viable seeds. With these opportunities for gene exchange one might expect to find hybrid swarms wherever two species intermix. This is seldom the case. On the contrary, hybrid individuals in flower are few and seldom seen. Mixed colonies are mostly divisible into two taxa of which neither differs substantially from its relatives throughout their main ranges. Nevertheless, a considerable number of hybrid plants may be present in the vegetative state in these colonies and reach flowering only in response to particular conditions (Epling and Lewis, 1952). These hybrids can and do persist without flower-

NATURAL HYBRIDIZATION OF SALVIA APIANA AND S. MELLIFERA

Reference is made with increasing frequency in plant taxonomic literature to the occurrence in nature of interspecific hybrids: Ceanothus (McMinn, 1942), Sctlvza (Epling, 1942), Balsamorhiza (Ownbey and Weber, 1943), Clematis (Erickson, 1945), Wyethia (Weber, 1946), and Aquilegia (Munz, 1946), to cite some of the more recent. In these examples, so far as the reader is able to judge, F1 hybrids, backcrosses and even hybrid swarms may occur. Because of the weak intrinsic barriers between species. which characterize these and many other plant groups, the assumption might logically be made that gene flow between the species concerned would obliterate or blur the specific morphological differences if long continued. Yet the taxonomic treatment in such papers suggests that the constancy of the species concerned has not been materially affected. The observations which follow seem to confirm this latter view and to indicate that gene flow from species to species does not ordinarily occur, or if so, at a level difficult to distinguish from expected intraspecific variation, even though continued hybridization may seemingly provide a channel.

ON THE ROLE OF INVERSIONS IN WILD POPULATIONS OF DROSOPHILA PSEUDOOBSCURA

The salivary gland chromosomes of Drosophila provide an exceptional means to study the role of inversions in a wild population because their frequencies can be established with a high degree of accuracy. This is particularly true of Drosophila pseudoobscura, in which species Dobzhansky and Sturtevant (1938), and Dobzhansky have described fifteen or more inversions, and the latter has established their geographical distribution throughout its range from British Columbia to Guatemala (Dobzhansky in Dobzhansky and Epling, 1944). Certain regularities of occurrence soon became apparent during the course of these investigations. Because some inversions overlap each other, Dobzhansky found it possible to arrange the whole number in two phyletic series and then found that one series was centered in the western United States and the other in Mexico. Second, he found that each inversion type has a definite range. Only one occurs throughout most of the species area and the others are variously restricted and some are apparently quite local. Third, he found that the frequencies of some of the more abundant types vary along regular geographical clines. These clines do not coincide; hence each region is characterized by different frequencies of its principal types. Other regularities led to the belief that the inversions might be an instrument with

A revision of Agastache

Agastache occurs throughout most of the United States and extends to Central Mexico and into Canada. One species also occurs in Japan, Manchuria and Eastern China. The genus comprises two well defined sections, each of which has a separate geographical area. The first, Chiastandra, is northerly in distribution and occupies the mesic areas of the Mississippi drainage and the Pacific slope. It recurs in Eastern Asia. Its species are eight in number. They are relatively well defined and present but little intraspecific variation in comparison with the species of the second section. That section, Brittonastrum, is found in the arid regions of the Southwestern United States and Mexico. Its twelve species, as delimited here, not only present a greater range of difference than those of Chiastandra, they are also more variable and the extremes of one are often similar to the extremes of another. Hence they are difficult to delimit.

A synopsis of the tribe lepechinieae (labiatae)

The New World genera Lepechinia and Chaunostoma comprise a closely knit generic group clearly set off from all others of the family. Chaunostoma is monotypic and is found only in mesic montane forests of Chiapas and Guatemala. Lepechinia ranges in the New World f rom about 40 ° S. lati tude to 40 ° N. I t is also fou~ad on R6union Island and in Hawaii. Its continental distribution is predominantly cordilleran. 1

CHANGES IN AN INVERSION SYSTEM DURING A HUNDRED GENERATIONS

An almost continuous record has been made since 1939 of the frequencies of different gene arrangements in populations of Drosophila pseudoobscura in the San Jacinto Mountains near Los Angeles (Dobzhansky, 1943, 1947; Epling, Mitchell and Mattoni, 1953, 1957). If we assume that a generation of this species has an average life span of about a month, this period of observation covers not less than 100 generations. More than 40,000 third chromosomes have been examined and a comparable number of X chromosomes, these being the two chromosomes in which inversions occur in detectable frequencies. The samples have been made at monthly intervals as far as possible and the record reflects the changes that have occurred each year for a number of years. The basic analysis of each gene arrangement has been published by Dobzhansky (Dobzhansky and Epling, 1944). Three other arrangements are being described in this paper. They bring the number known in the third chromosome to twenty, of which twelve occur in the populations referred to. Two arrangements are known in the X chromosome.

FREQUENCIES AND ALLELISM OF LETHAL FACTORS WITHIN AND BETWEEN GENE ARRANGEMENTS. I. SAN JACINTO MOUNTAINS

This report will subsume the differences in frequency of lethal chromosomes among certain inversion rearrangements of the third chromosome of Drosophila pseudoobscura, together with their degrees of allelism in different environments. This first paper will deal chiefly with arrangements found in the San Jacinto Mountains, southern California. A second will deal chiefly with arrangements found in Guatemala. A third will compare mutation rates between certain arrangements. Populations of Drosophila pseudoobscura are known to carry a considerable load of recessive autosomal lethals (Wright, Dobzhansky, and Hovanitz, 1942). Furthermore, the third chromosome of this species has undergone a series of inversions which have resulted in 22 different gene arrangements. Crossing over between the heterozygotes of these arrangements is virtually suppressed (Dobzhansky and Epling, 1948; Levine and Levine, 1954). A change of arrangements may also significantly change the frequencies of crossing over at certain loci in different structural homozygotes (unpublished data). Thus, each arrangement tends to be preserved in its population as a discrete and different linkage group. Finally, the several arrangements occur at different frequencies in different environments within the species range (Dobzhansky and Epling, 1944). Each arrangement fluctuates in frequency at a given collecting site, but the pattern of their relative frequencies has remained not dissimilar for many generations at the sites that have been