03777naa a2200229 a 450000100080000000500110000800800410001910000190006024501190007926000090019830000130020749000370022050001800025752027430043770000180318070000210319870000230321970000200324270000220326270000170328477302460330114668222007-07-27       2004    bl uuuu     u00u1 u    #d1 aMARTINS, A. S.  aAnalyzing the grographic positions of variety trials of soybean breeding programs in the central region of Brazil.  c2004  a244-245.  a(Embrapa Soja. Documentos, 228).  aEditado por Flávio Moscardi, Clara Beatriz Hoffmann-Campo, Odilon Ferreira Saraiva, Paulo Roberto Galerani, Francisco Carlos Krzyzanowski, Mercedes Concordia Carrão-Panizzi.  aLocations to be used in a variety trial network of a plant breeding program must be chosen in such a way that the region, for which the program is conducted, is well represented. This study was carried out in order to analyze the geographic position of variety trial locations of Embrapa Cerrados' soybean breeding programs against spatial distribution of soybean harvested area in the central region of Brazil. The variable analyzed, soybean harvested area in each county, was picked up from IBGE tables, in 2001, for every county of the following states: Goiás, Mato Grosso, Mato Grosso do Sul, Minas Gerais, São Paulo, Bahia, Piauí, Maranhão e Tocantins. There are 3074 counties in these nine states. In only 660 counties (21%) the soybean harvested area was different from zero. The soybean production is highly concentrated in few counties: in just 29 counties (less than 5%) it was gathered 50% of the total area planted with soybean; in 64 (10%) counties it was gathered 70% and in 153 counties (23%) it was gathered 90% of the total area planted with soybean. A geographic information system, SPRING, release 4, was used to link Brazil's county network map with the table containing soybean harvested area per county. Following this step, three maps were drawn; each one showing the regions that concentrate 50%, 70% and 90% of the total area panted with soybean, respectively. Positions of the trial network locations were marked on these three maps. The two locations, Senador Canedo-GO and Planaltina-DF, where many segregating populations are conducted from F3 to F6 generations, are outside of the three regions with the greatest concentrations in soybean harvested area. As bulks have been assayed as a hole, instead of assaying each plant, the occurrence of natural selection is expected to be happening, contributing for genotypes more adapted to theses two locations. Preliminary tests take place in seven locations. Six of them are outside of the region that concentrates 50% of soybean harvested area, and three of them are outside of those regions responsible for 70% and 90% of the soybean harvested area. Important regions in Mato Grosso and Bahia states don't have any preliminary test locations. The final trial network includes 36 locations. Twenty-one of them (61%) are outside of the region that concentrates 50% of soybean harvested area; fifteen locations (42%) are outside of the region with 70% of the harvested area, and ten locations (almost 30%) are outside of the region with 90% of the total soybean harvested area. Taking into account the importance and the expense of conducting final field trials, this study suggests a necessary change in the position of these ten final field trials locations.1 aDUARTE, J. B.1 aNUNES JUNIOR, J.1 aOLIVEIRA, P. M. F.1 aSOUZA, P. I. M.1 aASSUNÇÃO, M. S.1 aSILVA, S. A.  tIn: WORLD SOYBEAN RESEARCH CONFERENCE, 7.; INTERNATIONAL SOYBEAN PROCESSING AND UTILIZATION CONFERENCE, 4.; CONGRESSO BRASILEIRO DE SOJA, 3., 2004, Foz do Iguassu. Abstracts of contributed papers and posters. Londrina: Embrapa Soybean, 2004.