Tentative general conclusion on soil fertility in the three Regions
The evidence gained in 1947-48 would appear to indicate that the soils of the Kongwa Region are not deficient in lime but tend to deficiency in phosphate. At Urambo the red soil is sharply deficient in phosphate but not acutely so in lime; while the grey soil, at least at the lower levels, needs liming, but is not deficient in phosphate. It would, however, be dangerous to apply this conclusion to all the grey soil at Urambo, since the site of the 1947-48 grey soil plot was on the edge of a drainage line. The red loam at Namanga is acutely phosphate deficient and acid, and may well prove to be calcium deficient as well.
These conclusions are broadly in accordance with the expected effects of the soil-forming factors operating in the Regions. The Kongwa soils are typical semi- arid steppe soils, while the Namanga soils and the ancient plateau soils of Urambo have been formed under conditions of heavy leaching leading to removal of lime, with acidity and low availability of phosphate.
The extraction for two minutes with 0.3N HC1 has proved of some general use in indicating these conditions, at least where soil types have been repeatedly sampled and a general impression has become possible. It is not, however, claimed that the conditions of a particular site can be forecast with confidence from one or two samples.
12. GROUNDNUT PRODUCTION
The general results of experiments in this Section have been covered in fair detail in Section 6. This section presents the more important experimental data in full with the minimum of discussion. Unless otherwise stated, all experiments were machine-planted.
Plant population
An experiment was set out on a fairly typical medium red loam at Kongwa Experiment Station, to examine the relationship between plant population and yield. Population was varied by reducing the spacing between rows, the planter spacing within the row being kept constant at 4 in. The design also incorporated factorial N, PO, and limestone treatments, broadcast on the plots and harrowed in shortly before planting. The layout used was a confounded randomised block design in four blocks of eight plots each, plot size being .035 acre.
Owing to heavy rain during planting, three of the 14 in. row plots were not planted until some days later than the rest, under less favourable conditions, resulting in a lower percentage establishment and somewhat poorer growth. The vegetative development of plants in both 14 in. and 18 in. rows was noticeably less than at the wider spacings throughout the season. This is reflected in the performance values.
The results were as follows:-
Table 19.-Results of Spacing, Lime and Fertiliser Trial, Groundnuts, Kongwa Experiment Station, 1947-48 Variety: Valencia
I. Plant population data:
Spacing between rows, inches Theoretical population, 000/acre Actual population, 000/acre Per cent. establishment
14
18
28
36
Mean
99.6
77.4
49.8
38.7
58-3
52.9
39.0
27.7
58-5
68-3
74.4
71.5
II. Yield data, lbs. kernels per acre:
Mean yield
±48·1
781
925
703
674
771
Response to fertilisers:
N
119
59
-154
174
-10
P
±96.3.
123
38
56
63
70-48.1
Ca
118
64
-100
19
Performance values (yield/1,000 plants)
13.4
17.5
18.0
24-3
Fertilisers have in general proved ineffective, no fertiliser response being significant, although the constantly positive response to phosphate is perhaps worth noting The main yield differences are those corresponding tofpopylation
137
[
IL
»
f
No comments yet.
Private notes are available after approval.