It depends on how related (genetically) these "lines" used for cross are.

crossing is the one of the main technique of generating variability in any crop plants. you can cross distant, diverse parents in the crossing programme for the characters you wanted to develop the more variability.

First, what is the purpose of your breeding program? You always can find germplasm contrast or similar in the desired traits. Usually, at least one parent must be with good commercial traits, and second - can be a donor of desired feature. In worst case it can be G. soja or semi-cultured accessions - it is difficult to save wild traits (such as resistance) in progeny of such crosses.

It depends on the trait ,you are working on.If it is a major gene,you can backcross the progeny from donor and the desired genotype used as recurrent parent.If you are working on quantitative traits,you can improve your trait by crossing diverse genotypes. G.sja is good source of some resistance genes.But you need a large number of backcrossing to recover recurrent parent genome content.Recovering seed size is big problem,when crossed with soja.

As said earlier by the Chetana,Alex and Anita. Crossing of two parent can be used to create variability depending on the objectives and materials used for crossing. If inbred lines are used are as the parent, then variability is more obvious in F2 generation due to the segregation of the alleles. If non inbred lines or wild species are used, then variability can be seen as from F1 especially in cross pollinating crop species

Of course yes if your 2 parents are different for the trait. Variability appears from F2 generation. In F1, no variability could be observed because your descendants will be uniform.

@ Roland,

1. The statement ("in F1, no variability could be observed") is only ture when the parents (P1 and P2) are 'pure lines'. Hybrid F1 can also be produced through P1 (inbred) x P2 (inbred). If these parental inbred lines are produced through a few generations of 'inbreeding' from a outcross species, 'heterozygosity' might still exist in a (very) small portion of the loci in the inbred genomes. And this can lead to 'minor' variations among the individuals in the F1 population.

2. F1 plants can still have phenotypic variation. These variations are due to environments or variation attribute to GxE.

@ Ermias,

It depends on which source you use for cross, and what kind of populations you want to create. For example, open-pollinated population or F1 hybrid population you want to create through crossing? Open-pollination can cause a greater amount of variation within plant populations, and conserve the genetic variation. Although F1 hybrids have their benefits such as high vigor, hybrid breeding can cause genetic diversity erosion.

I attach a paper for your reference:

Title: Understanding crop genetic diversity under modern plant breeding

Why not? You will get F1 hybrid that may show heterosis or not then you will obtain F2 that have higher variability due to segregation. 

If you if the parents are pure lines you can change the environment conditions so the genetic background could change. Vp = VG + VE+ VGE 

If you cross two identical pure lines, you virtually create no genetic variation.

Dear Ermias

There is two types of crossing : the first is an inbreeding system which means mating between relatives or inside group , and as a result of this cross increase homozygous genotypes and reduce the heterozygous ( reduce genetic variability ), inbreeding leads to low viability and fertility ( low fitness ).

The second type is out breeding system which means a cross between pure lines or between strains , and the offspring result from this crossing will be heterozygous and increase genetic variability, and maybe get heterosis and high fertility and viability.

Good Luck

Dear  Ermias,

Yes of course, but parents which you will use, must be with difference for more than one character.

And it also depends on what kind of 'genetic variability' you want to create in the progeny. Some, you cannot just randomly cross two lines to obtain. For example, if you want to generate a 'disease-resistance' trait, you first need to identify the germplasms which possesses this characteristic for crossing. If there is no such gene in the gene pool, it will be very hard to accomplish your goal by conventional cross breeding.

Of course crossing is a feasible and most natural mode of creation of variability among others like mutation and biotechnological approaches. but care must be taken that cross must be effected between two genotypes which are extremely contrasting for at least the trait of your interest.

yes.. crossing is the first step, can create only one type variability (in F1 ) other than the parents (except under some conditions like maternal effect, epigenetic etc). Crossing of F1 s with either of parents (back crossing) or any other pureline would create more variability while selfing of the F1 would create huge variability if the parents were distinct by more alleles like ...cultivated x cultivates, cultivated x wild accessions, cultivated x related wild, cultivated x exotic lines..

Yes, genetic variabilty can be created through crossing in soybean. To serve that purpose

1. Take the Inbred lines as parents which are contrasting for the different characters i.e. containing diverse alleles for different characters

2. Make the crosses to attain F1

3. Either attempt back crosses to attain the variability in the progeny or self the F1 to get F2 progenies which will show maximum variability for different traits under study.

4. Select the progenies exhibiting high variaibilty and utilize for breeding programme

This way maximum variabiliy would be obtained as a result of recombination from both male and female gametes followed by segregation and independent assortment of alleles. Even one can get the transgressive segregants if the character is governed by polygenes.

Good Luck

@ Yuan-Yeu Yau,

The Pure lines may itself becomes genetically variable if undergo natural mutations, natural hybridization with unknown pollen parent OR even through unintentional mechanical mixture. Thus one pure line (act as female parent) is grown in a separate plot from the pollen parent for controlled hybridization. During this period if there is natural mutation OR hybridization in the plants growing in these two plots, there will be a genetic variations even when we are crossing TWO identical pure lines.