This paper reports a procedure for determining the content of strongly fixe
d NH4+ in soil. The procedure consists of a Kjeldahl digestion followed by
an acid attack of the residue with a 5 M HF:1 M HCl solution. Distillations
after each of the two treatments recover different forms of NH4+. The proc
edure was tested on fine earth (<2 mm) and skeleton (>2 mm) fractions of tw
o forest soils developed on sandstone parent material. In both soil fractio
ns we evaluated three different forms of NH4+-N: (i) Kjeldahl, (ii) nonexch
angeable and (iii) micaceous. The last is located in the interlayer of mica
flakes larger than 50 mu m that resist the Kjeldahl digestion and is consi
dered strongly fixed. The total NH4+-N content of a soil is obtained by the
summation of the Kjeldahl and the micaceous NH4+-N. In the soils under con
sideration, the micaceous form prevails in the skeleton because this fracti
on is richer in micas of sand size (>50 mu m). Following the proposed proce
dure, we found that micas (muscovite and biotite) contain about 3000 mg kg(
-1) of NH4+-N in the interlayer. The presence of micaceous NH4+-N in soil i
s generally ignored because the skeleton is usually excluded from analyses,
and the micas larger than 50 mu m cannot be dissolved by the Kjeldahl trea
tments. The micaceous NH4+ is the least extractable form of NH4+-N, and we
infer that it is the least available to plants.