Suppose, I want to find a marker for drought tolerance in maize which nobody has identified yet. So, how can I develop that marker so that in future it can be used by others to screen populations for drought tolerance?
I recommend that you read this review of marker-assisted selection (MAS) by Collard and Mackill (2008). It explains the basic process of QTL mapping and MAS.
Xu Y, Crough JH (2008). Marker-assisted selection in plant breeding: From publications to practice. Crop Sci., 48, 391-407.
In addition, as the draft genome for maize is available for some time and resequencing as well I would look at genome structural differences in the area of interest to develop molecular markers e.g. SNP markers and test them.
There is an old strategy. If you have the desired trait segregating in more or less monogenic way, you may collect DNA samples from F2 hybrids, then make pools of DNA from 10 plants of one phenotype and from 10 of another and then make a PCR with any randomly amplified fragments (RAPD, ISSR, IRAP, REMAP etc.). You use 10-plants sample as a single DNA sample. If you find any fragments which are present in one spectre and absent from another, this may mean that this marker is tightly linked with desired gene. Then you need to excise it from gel and sequence, this can allow you to design this fragment-specific primers.