Treatment effects on levels of soil carbon and nitrogen in four long-term experiments at Samaru are reviewed in relation to an expected change in the future away from a bush-fallowing agricultural system towards an intensive mechanized system of permanent agriculture.

Where different rates of farmyard manure were applied to the soil annually for nearly 20 years marked differences in soil organic matter contents developed. The mean soil carbon (0·82 per cent) of plots that received 12·5 tonnes ha-1 annum-1 is now almost four times as great as the mean content (0·22 per cent) of control-plot soils. Phosphate applications tended to encourage higher levels of soil organic matter, but where mineral fertilizers (N, P and K) were applied with farmyard manure lower soil organic matter levels and C:N ratios than those of manure-only plots resulted. Apparent retention in the soil of carbon and nitrogen applied as farmyard manure was of the order of 12–15 and 30% respectively. A groundnut shell mulch, applied for 9 years, produced significantly higher levels of soil organic matter than control plots.

In a rotation experiment in which three years of cropping alternated with either 2, 3 or 6 years of planted Gamba grass (Andropogon gayanus) fallow, large differences in soil organic-matter status and rate of turnover existed between the three rotations. In particular the 3-year fallow showed an annual rate of soil organic-matter increase over twice that of the 2-year fallow. Soil nitrogen build-up exceeded 50 kg ha-1annum-1 in the 3-year fallow.

Shortage of farmyard manure and the relatively poor return from systems that include unproductive fallows suggest that in future soil organic matter must be maintained by the return to the soil of all crop residues.