prilagođeno pretraživanje po punom tekstu

ŠUMARSKI LIST 1-2/1966 str. 35     <-- 35 -->        PDF

whole complex of characters, could be made by means of covariance and
multivariate analyses of phenotypical data collected from plus trees and from
a number of surrounding or adjacent sample trees, especially if the relative
economic value of each character was known, and the economic weights of all
traits were taken into account. A further step towards improving the selection
effect and there by the gain of selection, would be to construct indexes (of the
typo suggested by Smith, 1936, Hazel and Lush, 1942, and Hazel,
1943) for a number of base populations of each species. For the compilation
of such indexes the following data must also be known: Heritabilty (narrow
sense) of each tree character as well as, the genetic correlations between these
characters in different selection models (see Ster n and Hattemer , 1964),
and under different environmental conditions (see, e.g. T o d a, 1963, Le Roy,
1960, Stern, 1961 and 1963, and Zobel, 1963). There is no doubt that
selection indexes of this type are needed (cf. Ander s son, 1963 and 1965,
Ply m Forshell , 1963); but as long as information is lacking on the
requisite genetic parameters of the populations, and on the relative economic
values of the characters to be selected, it is not possible to construct the type
of seelction indexes, for e.g. Scots pine and Norway spruce, which has been
described by, e.g. Smit h (1936), and Haze l (1943). Theoretically, the
selection index method is more effective than the method of independent culling
levels; but in Drosophil a it has been found that the reverse order of
merit may occur for index selection and independent culling, when the

characters are controlled by pleiotropic genes (see Rasmuson , 1964).

If, phenotype control has been hitherto, the primary control of plus trees,
then clone tests (see Syrac h Larsen , 1947) and one-parent progeny tests
are other forms of the control of single trees. In the Swedish programme for
testing plus trees, such experiments have been carried out also on both conifers
and decidous trees. The clonal trials have been made for various purposes: 1) to
obtam useful preliminary information on trees that have either been included
in seed orchards already, or have not, i.e. information on stem straightness,
stem form, growth rate, branch habit, resistance to diseases of a parasitic
nature, variations in technical properties of wood, hardiness, and relations
between the phenotype of original trees and the corresponding clones; 2) to
study the interactions between clones and different environments in different
localities, e.g. in respect of varying quantities of fertilizer; and to study the
relations between spacing and the quality development of the clones, and
between spacing and clonal flowering; 3) to study intensity of clonal flowering,
flowering time, seed production and seed quality in different environments

(e.g. in north, central, and south Sweden); 4) to study the influence of various
root-stocks on the flowering and growth of the clones; 5) to study the
development of primary grafts in relation to secondary grafts; and 6) to study the
development and flowering of the grafts when scions have been taken from
different parts of the tree crown.
The one-parent tests are of some interest for studying general combining
ability, variation in quality factors and parasitic attacks, ect, within a population,
and especially when the plus trees are pollinated by a large number of
trees; but these tests are far from being accurate.

The most important control of plus trees must occur in controlled progeny
tests. This method involves artificial crossing. The ideal case is that, when all
clones in a seed orchards are crossed diallelly. The method, however, is very