If I am reading the same thing you are, it doesn't say a tablespoon of
melted agar-agar, but rather "Melt 1/2 tablespoon agar-agar with 1/2 cup
running buffer..." The point I'm making is not the typo of 1/2 tablespoon
vs 1 tablespoon, but that you are measuring the solid agar, and so it
shouldn't vary that considerably. If you want to be more precise though,
you would measure the agar by mass and not volume.
What you really want to do for reproducability is keep the percent agar
(versus buffer) the same for similar gels. It's most common to measure this
percentage as w/v (weight agar / volume buffer). You can then vary the
percentage agar of your gels based on the length of DNA you are trying to
separate. For example, with pure agarose, you would use a 0.3% gel to get a
good separation for molecules 5-60 kb, 0.9% for 0.5-7 kb, 2.0% for 0.1-3 kb,
etc. I don't know exactly what percentages would be best for agar (this is
something we should measure and publish unless someone can dig something
up), but I suspect they would be similar.
BTW those percentages I gave are for grams of solute (agar in this case) per
100 ml of solvent, as is typical for w/v percentages. I could never deal
with measuring things in tablespoons or cups :)
So if you are measuring your agar by mass, and buffer by volume, the only
thing you would really need to worry about is boiling away buffer when
melting the gel (thus increasing your concentration), or variability in the
composition of the agar (since it is not purified agarose). But since you
normally run a ladder lane, it's not like you have to be totally exact, you
just want to get the gel percentage roughly right for what you're trying to
separate. What you're doing when you adjust the gel concentration is
adjusting the pore size of the gel, you don't want it to be too big that all
the DNA molecules pass right through, or too small that they all get stuck.
-Josh