01640nam a2200145 a 450000100080000000500110000800800410001910000210006024501600008126001420024152010480038365000150143165300230144670000250146919066882011-11-22       2011    bl uuuu     u00u1 u    #d1 aGRATTAPAGLIA, D.  aTwenty years of Eucalyptus molecular breedingbfrom discrete marker-trait associations to whole-genome prediction of complex traits.h[electronic resource]  aIn: IUFRO CONFERENCE ON IMPROVEMENT AND CULTURE OF EUCALYPTS, 2011, Porto Seguro. Proceedings... Piracicaba: ESALQ, 2011. p. 33-36.c2011  aMarker Assisted Selection (MAS) has been the main justification underlying several large-scale forest tree genome projects worldwide in the last 20 years. The long generation times needed to complete a breeding cycle and the low heritability of most target traits were identified as the hurdles that MAS strategies could help overcome. However 20 years have gone by and no report exists on the actual use of QTL or association genetics (AG) information for operational tree breeding in general, irrespective of species, in spite of the large volume of QTLand AG mapping information published. Reasonsfor this derive mainly from the undomesticated nature of forest trees that leads to a rapid decay of LD in tree genomes, such that marker-trait associations detected in specific mapping families do not hold in unrelated pedigrees. Furthermore, not surprisingly, several reports in recent years have shown that many more QTLs with small and variable effects across backgrounds and environments typically underlie complex traits in forest trees.  aEucalyptus  aMolecular breeding1 aRESENDE, M. D. V. de