The genome size of human is 3.000.000.000 bp and for Arabidopsis it is 255.000.000 bp. I know these numbers because their genomes have been published. Which methods are used by scientists to estimate the genome size before they sequence that genome?
In the absence of molecular sequencing data the most accurate estimates are based on reassociation kinetics, more commonly known as Cot curves. Here is a classic paper from 1974 that describes the basic approach. While pulsed field electrophoresis can be used to resolve the chromosomes of species with relatively short chromosomes (like baker's yeast), there are really no suitable size markers that you could use to construct a standard curve for size estimation. C. elegans has six chromosomes, but they are all fairly close to each other in length, and the genome size of 100 Mb is the same as a single long human chromosome.
Gel electrophoresis is a basic lab method to get size info about nucleotide fragments, although one might need to chop up larger DNA to be able to apply this method. The wikipedia page covers the topic well, thus I'd suggest reading it. I'll just give a short outline:
Nucleotide fragments are loaded into a gel (agarose or acrylamide).
The gel is put into a static electric field
- Nucleotide fragment move to the positive end, because they have negative charge
- larger fragments migrate slower than smaller fragments
- standardized DNA ladder can be used to get approx size.
UV fluorescent dye is used to stain the fragments.
Another way, is to check closely related species. Among closely related species it is a fair assumption that they have similar genome size. You can also hybridize the unknown DNA to a DNA of such close relative to see how much of them are similar and you can also make estimates based on the size of the hybridized regions.
I guess you could estimate the length various ways. Pulse field electrophoresis could show DNA length, but you need a standard. When you combine Karyogram and FISH, you would could make a map. By genetical approaches, you could estimate distance on the map by centimorgan. Then you could know how many centimorgan chromosomes are. One centimorgan would be 1 million base pairs.
Oh, I forgot the simplest way. For example, take 1 million cells and estimate how many grams of DNA there.