Although there is not a definite answer to this question, because it depends on the allele frequencies of the population, one can see this problem as an allele sampling case and try to give a number.
Check this paper where we found 30 plants to be a good number to sample the alleles of maize accessions. It was developed from a perspective of information theory.
I suppose 'consanguineous depression' is another (or related) term for 'inbreeding depression' (ID). Number of plants (sample size n) with cross pollinated crops like maize needs to be relatively large to avoid inbreeding depression. The answer to the question "How many plants?" will depend on what level of ID exists and what maximum level of ID you wish to tolerate. I guess it may perhaps be difficult to avoid ID 100% - one could I guess only try to minimize it to an acceptable level as may be appropriate for the intended purpose. Sorry I do not have a definite answer to what n should be. I hope other colleagues provide some insights / references - which I too am keen to look forward to.
Hello, it depends on the type of maize population do you want to regenerate, for example if you have a very variable population developed for breeding through recurrent selection, an open pollinated variety or a landrace, CIMMYT's genebank standards suggest planting plots with 256 plants and obtain at least 100 families or ears. If the germplasm is an homozygous genotype or a source of specific genes you don't need more than 20 plants or ears. It's important to have a proper isolation of the plot and good pollination technics to avoid contamination from other maize genotypes. Taba, S., M. Van Ginkel, D. Hoisington, and D. Poland. 2004. Wellhausen-Anderson Plant Genetic Resources Center: Operations Manual 2004. El Batan, Mexico: CIMMYT.