J. J. Hammond

Screening for low grain cadmium phenotypes in sunflower, durum wheat and flax

Cadmium (Cd) level in nonoilseed sunflower (Helianthus annuus L.), flax (Linum usitatissimum L.), and durum wheat (Triticum turgidum L. var. durum) grown on uncontaminated, alkaline soils has exceeded limits established in Northern Europe. Separate field experiments were conducted to investigate variability of grain Cd levels among sunflower, durum wheat and flax germplasm, and to seek an efficient screening method for future breeding. There were large variations in leaf Cd concentration among 200 sunflower lines. These lines performed more consistently for Cd uptake at the R5 stage than at the V8 stage across 4 locations with markedly differing soils. Cd concentration in V8 leaves was not related to Cd in grain. The positive correlation between R5 leaf Cd and kernel Cd level was obtained from nonoilseed hybrid (Sigco 954) (R2; = 0.74**), and 200 lines (R2 = 0.44**) tested over 4 locations in 2 field trials, respectively. This indicates that an efficient and low cost screening method can be developed for genotype selection, but plants must be grown to the R5 stage. A preliminary evaluation of 30 durum wheat and 74 flax lines indicated large variations in grain Cd level of durum wheat and flax. Grain Cd concentration ranged from 0.11 to 0.34 mg Cd kg-1 DW for durum wheat, and 0.14 to 1.37 mg Cd kg-1 DW for flax, respectively. This variability indicates that breeding for low grain Cd in durum wheat and flax should be feasible.

Inheritance of reduced height in sunflower

SummaryBreeding to improve stem strength is a major objective of researchers of sunflower (Helianthus annuus L.). This study was undertaken to investigate genetic factors controlling reduced plant height and increased stem diameter in three sources of sunflower, DDR, Donsky, and Donskoi 47, crossed with a conventional height line, HA 89. As these two characters may lead to improved standability, knowledge of their inheritance will assist researchers in utilizing proper breeding methods. Estimates of additive, dominance, and epistatic genetic effects controlling reduced height indicated that the additive component was most important in two of the three crosses with the additive and epistatic component nearly equal in the third cross. Breeding efforts to reduce height of sunflower hybrids utilizing these lines in crosses could be effective due to the magnitude of additive effects. The dominance component of genetic effects controlling stem diameter was the most important for two crosses, with both dominance and additive components important for the third cross. Epistasis was present, but minor, for controlling stem diameter. The high relative importance of the dominance component indicates that testcross evaluation of lines in early generations could identify lines for producing increased stem diameter in hybrids. Even though the three sources of sunflower with reduced height were different in morphologic and agronomic characteristics, they had similar genetic control of plant height and stem diameter. Each could be utilized in a breeding program to develop lines with reduced height and larger stem diameter.

Variable selection in multivariate multiple regression

A procedure for selecting variables in Multivariate Multiple Regression is given. The procedure combines the notion of adequate subsets advanced by McKay (1979) and the branch and bound algorithm due to Hocking and Leslie (1967). An example is also given along with the algorithm implementation in the Statistical Analysis System.