I suspect that you want to do work in pots or the lab. Spike in mg P/kg soil. The usual salt is KH2PO4, but you may need to change the salt depending on the purpose of the study and the properties of the soil.

If you are concerned about distribution, you can dissolve the salt, distrubute the soil and drip the solution onto the distributed soil then tumble it

Dear NA, The idea of getting a P response related to the high Arsenic would be favored if you could use 0, 10, 100, and 1,000 ppm P which will be better able to generate a more complete response curve. Paul M. has good insight into getting uniformity in the application. Since the idea is to get the response take good data on growth and yield parameters and complement these with atomic analysis of both P and Arsenic and better yet if you can have a complete atomic analysis.

Paul Milham , thank you for your advice, actually I usually use Na2HPO4*12H2O as phosphorus source, and pH of soil is acid, I am going to determ available P, water-soluble As and acid phosphatase, after spiking soil with P. So I should calculate weight of the soil how much im going to use in my each experiment? determing As desorption and acid phosphatase need to add different volume of P liquid, so I am not sure how to calculate it

I am understanding that you may be doing an As/P competition study in acidic soil. Is that correct?

Is it entirely in the lab without biota or will you measure some 'biotic availability' ?

If it's chemistry only in the lab do you in fact want to suppress biotic activity?

These factors are important in deciding what to do and how.