The rotation of crops with either legumes or natural regrowth of native veg
etation is known to supply N to crops in tropical land-use systems, but lit
tle is known about the effect of such rotations on P supply to crops. An ex
periment was conducted on two soils in Kenya, an Oxisol (very fine, isohype
rthermic Kandiudalfic Eutrudox), and an Alfisol (mixed, isothermic Kandic P
aleustalf), to determine the effect of unfertilized, organic-based land-use
systems on fractions of soil inorganic P (P-i) and organic P (P-o). The la
nd-use systems involved growth of three maize (Zea mars L,) crops (18 mo to
tal duration) after 17 mo of either (i) Sesbania sesban (L,) Merr, tree gro
wth (sesbania fallow), (ii) natural regrowth of vegetation without cultivat
ion (natural fallow), (iii) three crops of unfertilized maize (maize monocu
lture), or (iv) bare uncultivated soil (bare fallow), Soil (0-15 cm deep) w
as collected before post-fallow maize and after the three post-fallow maize
crops. Sesbania and natural fallows increased the amount of P in light fra
ction soil organic matter (OM) (>150 mu m, <1.13 Mg m(-3)) and macroorganic
matter (250-2000 mu m) on each soil and increased chloroform extractable P
on the Oxisol, Land-use systems had no effect on extractable P-i (bicarbon
ate, Bray-1, iron-oxide impregnated paper), except for resin P on the Oxiso
l. Unfertilized sesbania and natural fallows cycled soil P through plant re
sidues, resulting in increased P in soil microbial biomass immediately afte
r the fallows and increased P in light fraction soil OM for three post-fall
ow cropping seasons.