Section 9 — Population Genetics 



hypothesis of dominance and the hypothesis of 

 overdominance-are not mutually exclusive, 

 but the causes of heterosis with which they deal 

 may act, and as a rule do act, simultaneously. 

 The heterosis effect cannot be explained by a 

 single genetic cause, by one type of interaction 

 of hereditary factors, but is the result of the total 

 similar action of various genetic processes. Not 

 one of the proposed hypotheses determined for 

 any one type of interaction of hereditary factors 

 can be accepted as a general theory of heterosis, 

 although certain of them, and, in particular, 

 the two hypotheses mentioned, are in good agree- 

 ment with determined experimental data. They 

 contain the elements of exact knowledge and 

 can be considered as fragments of the general 

 theory of heterosis. 



It is represented as extremely probable that 

 the concept capable of helping to unite these 

 fragments into a single whole and of supplying 

 the missing elements, taking into account the 

 types of interaction of hereditary factors (allelic 

 and non-allelic) and also the role of conditions 

 of the surrounding medium in the development 

 of characters in hybrids, is the theory of genetic 

 balance. 



The value of any character in each parent 

 variety (line) is a result of a determined balance 

 (equilibrium) worked out in the course of selec- 

 tion, in the variously directed action on this 

 character of many hereditary factors and condi- 

 tions of the surrounding medium, in which the 

 development of the organism takes place. In 

 the cross-breeding of parent varieties (lines) 

 which differ in their heredity there takes place 

 in hybrid offspring a change in the genetic ba- 

 lance with relation to a larger or smaller propor- 

 tion of the characters which can cause a deviation 

 in the value of one or another character in the 

 direction of an increase or decrease by compari- 

 son with the parent forms (positive or negative 

 heterosis). 



Departing from what has been said above, 

 an attempt is made to interpret the causes of 

 heterosis and in particular the role in this process 

 of various types of allelic or non-allelic inter- 

 action of hereditary factors on the basis of the 

 theory of genetic balance. 



meiotic chromosome pairing. Up to the present, 

 the following 9 cytogenetic types have been 

 identified in natural populations. Plants of the 

 constitution AA (2/2 = 16), BB (2« = 18), and 

 AA BB (2// = 34) are propagated sexually by 

 seeds as well as vegetatively by bulb multiplica- 

 tion. Plants of the constitutions BBB(2« = 27), 

 BBBB (2w = 36), AAB(2« = 25), ABB (2// = 26), 

 ABBB (2n = 35), and AABBB (2« = 43) set some 

 seeds, instable sexually, but propagate by clon- 

 ing by bulb multiplication. 



These various cytogenetic types are found 

 usually in juxtaposition in a population. An 

 index of homogeneity of the populations as to 

 these cytogenetic types ranged from 0.82 to 

 0.22, from population to population, where 

 1 .00 indicates a pure stand of a single cytogenetic 

 type. The indexes were computed as follows. 

 Let a be frequency of a type A, b that of another 

 type B, c that of another type C, and so forth. 

 Simple computations (a 2 + b 2 + c 2 + . . .)/ 

 (a + b + c + . . .) 2 give the index values 

 ranging from 1 to certain values approaching 0. 



9.22. Population Structure and Dynamics in a 

 Complex Scilla scilloides (Liliaceae). Tutomu 

 Haga and Shozo Noda (Fukuoka, Japan). 



In a perennial plant Scilla scilloides Druce, 

 two basic genomes A (x = 8) and B (.v = 9) are 

 clearly distinguishable by karyotype and by 



9.23. Progress From One Generation of Selection in 

 Nicotiana tabacum. Dixie Bright 244 X Coker 



139. D. F. Matzinger, T. J. Mann, and C. 

 Clark Cockerham (Raleigh, U.S.A.). 



Within the F 2 generation of a cross of 2 na- 

 turally self-fertilizing flue-cured tobacco va- 

 rieties, Dixie Bright 244 x Coker 139, individual 

 plants were self pollinated and crossed to random 

 members of the population. Only the cross- 

 bred material was evaluated and estimates of 

 additive genetic and dominance variances were 

 obtained assuming no epistasis. Parental plants 

 high in per cent total alkaloids were identified 

 on the basis of mean cross-bred performance. 

 From remnant seed, selfs of these superior plants 

 were intercrossed to form a new population. 

 In this first selected cycle, both self and cross- 

 bred families were evaluated. Estimates of 

 additive genetic, dominance, and additive x 

 additive epistatic variances were obtained. 



In both the initial and selected populations, 

 significant estimates of additive genetic variance 

 were obtained for cured leaf yield, per cent 

 total alkaloids, days to flower, number of 

 leaves, and plant height. Additive genetic 

 variance was not significant for leaf value in 

 either population. The only significant estimate 

 of dominance variance was for leaf value in the 

 original population. None of the estimates of 

 additive x additive variance were significant 

 in the selected population. Selection for increased 



150 



