Soybean breeders need to evaluate alternative strategies for increasin
g yield, protein, and oil content, The correlated response for yield w
hen selection is conducted for increased protein among unreplicated pl
ant rows is unknown, Our objective was to evaluate selection for incre
ased protein content at an intensity of 50% using unreplicated plots t
o determine the correlated response for yield, gross value per hectare
, oil content, and the direct response for protein content. Ten popula
tions that included 30 lines each were evaluated for yield, protein, a
nd oil content at each of seven environments. Lines were selected sole
ly for high protein content with one replicate at each of three select
ion environments and based on the average of two replicate in three se
lection environments, Additionally, selection was conducted on the bas
is of the population mean in the selection environment(s). Selected an
d unselected lines were evaluated in four validation environments that
were not the same as the selection environments, Selection among indi
vidual lines solely for increased protein content decreased yield 70 k
g ha(-1) and increased protein content 7 g kg(-1), when lines were sel
ected on the basis of the mean of two replicates and three environment
s. Selection among populations for increased protein content, decrease
d yield 110 kg ha(-1) and increased protein content 5 g kg when select
ion was based on the population mean averaged across two replicates an
d three environments. On average, selection for increased protein cont
ent resulted in decreased yield.