Whether it comes as a powder or a liquid, by law the values on a fertilizer label are in weight percentages, including the modified ones for P & K.
When I was devising my
fertilizer calculators, I was focusing on
liquid fertilizers, so I simply used the water standard of 1 milliliter volume
weighing 1 gram as the "master" conversion. I recognized that the
densities of different formulas will vary depending on the minerals used to make
up the fertilizer formula, but considering that the mass of nutrients in
solution is so small, that when measuring liquids, using the 1g/ml conversion is
probably close enough.
As an example, Dyna-Gro "Grow" formula is around 18 weight percent nutritional
elements. For 100 ppm N, the calculation suggests that 1.41ml/l is needed, so
that means that a liter of solution contains:
0.18 x 1.41/1000 = 0.00025 grams of
nutritional elements.
If my "concentrate" density is off by 5% (that is, 0.95 or 1.05), then my final
solution will contain between 0.00024 and 0.00027 grams, respectively, an
insignificant difference.
Obviously, those factors are likely to vary even more when using powders, and
they truly should be handled by weight, not volume, unless you know the bulk
density of the fertilizer powder you're using.
I was considering taking an example fertilizer and back-calculating the
makeup from the formula and the minerals used to make it up, throwing in the
bulk densities of the components, and seeing where that got us, but then I
remembered that the bulk density of powders varies by not only what the material
is, but how finely it is ground, what shape those particles are, etc. As
an example, the bulk density of ordinary silica sand, with it's fairly uniform,
quite rounded particles, runs about 1.5 g/cc. If you have the fine particle size
silica used as a thickener in everything from paints to cosmetics to ketchup,
its bulk density is about 0.05 g/cc.
Figuring that most water soluble, powdered fertilizers contain more-or-less the
same commodity chemicals, and they are probably similar in their ground
properties (there are a few exceptions), the range of bulk densities is probably
reasonably narrow.
FWIW, the bulk densities of most of the major components are about 1 g/cc,
according to some large-scale mineral suppliers I have already contacted,
suggesting that our volumetric approach might not be all that bad. Some
research into the physical properties of the chemicals states 1 g/cc +/- 20%,
but it pretty much averages out when their ground to powders.
I spoke to technical folks at a number of different fertilizer companies. While
having come at the answer from multiple directions, they all pretty much came to
the same conclusion.
Areas of absolute agreement:
While that's fine for someone like me who uses several pounds to make up many
gallons of concentrate to be metered by a device into the water stream, it is
unlikely that most hobby growers have the capability of measuring out a gram or
two or fractions of an ounce of powder to make up a gallon of fertilizer.
There were basically three responses to that:
Considering the inaccuracy in measuring powders ("is that level or slightly heaped?"), any other inaccuracies coming from metering devices like hose-end sprayers or siphons, etc., and the very small concentrations of dissolved minerals in out fertilizer solutions, it looks like a straight 1 g/ml density is reasonable for both powders and liquids, so feel free to use that measuring spoon for either!