Page 317
Soil analysis.
One of the first
af? &f 47 Department was to secure nafticdr data for the
isks
soils to be developed, particularly in respect of acidity (pH), and available P2O, N, Ca and K. A considerable volume of analytical work was accordingly carried out, using indicators for pH and a rapid extraction method for "avail- able" plant foods. The method used, which is due to Truog, suffers from the common defect of all such methods, that its results are only of use in so far as the method is calibrated by field experiments. It had, however, proved of use in the rapid detection of phosphate deficiency in British soils, and on this basis it was adopted. The important thing was to make a start with a technique that could be used at once to give general impressions, rather than to delay the work until more complicated methods (of equally unknown validity) could be adopted. More recently, of course, standard laboratory methods, including the glass electrode for pH measurement, have been introduced, and the majority of the soils have been analysed again, at least for Ca.
A detailed summary of the results of these analyses is presented in section 11. It shows that all the Kongwa soils are low in available P and high in available K. The characteristic pH values for the red soils and the upland pallid soils at Kongwa are in the neighbourhood of 6.0 and 6.5 respectively. The mean available Ca content for each group is about 85 mgm. available Ca per 100 gm. of soil. The valley soils are highly calcareous, having a mean available Ca content of 220 mgm. per 100 gm. of soil, and consequently these soils have generally higher pH values (6.6 to 8.0) than those of the other two groups. Available Mg content is roughly equal to available Ca content in red and upland pallid soils, but is only about one half of available Ca content in the valley soils. The Kongwa soils thus exhibit the usual characteristics of the soils of semi-arid regions.
Very few soil samples from the Urambo Region and the Southern Province have so far been analysed, but the soils in these areas may prove to be deficient in P but to have adequate K. At Urambo most of the grey soils have very low available Ca contents, although they are only slightly acid.
Soil fertility experiments
Four types of experiments were designed to elucidate the main require- ments of the groundnut crop, on which fertility studies in the year under review were concentrated. Each of these types was repeated eight times in the production area at Kongwa, once on grey and once on red soils at Urambo and once on a somewhat atypical heavy red loam at Namanga in the Southern Province (Block A), making 44 experiments in all. These types were follows:-
General exploratory experiments
as
These were designed to test the effects of N, P2O5, K2O and Ca, and the placement drilling of N and P2Os. The rates of application of plant foods
were:
0-1 cwt. N per acre as } cwt. of sulphate of ammonia. 0.5 cwt. P2Ōs per acre as 24 cwt. of superphosphate. 0.6 cwt. K2O per acre as 1 cwt. muriate of potash.
8 cwt. per acre of ground limestone from Sagara, Kongwa district.
The detailed results of these experiments, at all centres, appear in section 11. The most important features of the results at Kongwa are shown in the following table, in which the figures for the eight experiments are averaged :-
Table 1: Averaged results of 8 exploratory experiments, groundnuts, Kongwa,
1947-48. Yields in kernels, lbs. per acre.
No N or P205
N alone
P2O5 alone
N and P20, together
Broadcast
714
777
778
Placed
745
846
833
...
Mean of Broadcast and placed
698
730
812
806
Effect of placement
31
68
55
106