Volume 24, Number 4, 1993
|Page(s)||355 - 364|
Schätzung von Populationsparametern für verschiedene Merkmale bei der Honigbiene (Apis mellifera carnica)A. Willam and A. EßI
Institut für Nutztierwissenschaften, Universität für Bodenkultur A-1180 Wien, Gregor-Mendel-Straße 33, Österreich
Abstract - Estimation of population parameters for several colony traits in the honey bee (Apis mellifera carnica)
Considering the reproductive biology of the honey bee which deviates from that of other livestock species population parameters were estimated for production and behavioral traits. Data were sampled from a total of 2 864 performance-tested colonies belonging to the Institute of Bee Research in Lunz am See/Austria and 4 private beekeepers close by over a period of 15 yr (1975-1989). Traits recorded were honey yield (HL), the subjectively judged behavioral traits quietness (RU), a combination of aggressiveness and calmness, and swarming (SN) (score ranging from 1 = worst to 4 = best), and, though not available with all observations, the number of combs covered with bees both after hibernation (VA) and at the beginning of main honey flow (BT). The traits VA, VT and BT were considered to be auxiliary traits for HL. The procedure of performance testing was in accordance with the recommendations described by Ruttner (1972). Table I shows the numbers, means, SDs, minima and maxima of the measured colony traits. Population parameters were estimated by an adaptated half-sib analysis based on queens using restricted maximum likelihood (REML) methodology followed by a BENDING-procedure (EßI, 1991). The average relationship between 2 randomly chosen females (queens) of a colony depends on the number of drones per queen and the number of drone-producing queens on the mating station. Bienefeld et al (1989) described a method to compute this average relationship for the island mating stations in northern Germany. Depending on the structure of the mating stations in Lunz, Bienefeld's method had to be adapted and the average relationship between 2 queens was estimated to be ≈ 0.38 for this population. The estimated heritability for HL after BENDING was 0.24. This corresponds to many performance traits of other livestock species, and successful breeding can therefore be expected. The respective values for RU and SN were 0.08 each. Phenotypic correlations between these 3 traits were low (+0.02 to +0.06). Genetic correlations between HL and the behavioral traits RU and SN were +0.07 and +0.15, whereas a negative relationship (-0.17) was found between the behavioral traits (table II). The negative correlation has to be interpreted very carefully because genetic correlations between traits with low heritabilites estimated from a small sample (n = 2 064) are affected by large sampling errors. The estimated heritabilities for the auxiliary traits BT, VT and VA were somewhat lower (0.18, 0.21, 0.19 respectively) than for HL. The phenotypic correlations between BT, VT and HL were 0.47 and 0.50. The correlation between HL and VA was remarkably lower (0.28). This could be explained by VA being a trait for hibernation ability rather than for honey yield. The estimated genetic correlations have to be interpreted carefully due to the small sample size (n = 1 082). They were all distinctly positive and the correlation between HL and VA (0.49) was surprisingly high (table III).
Key words: population parameters / heritability / honey yield / behavior