CAB129-37 — Page 364

Page 364

Table 8.-Differential Responses in Mean of Eight General Exploratory

Experiments, Groundnuts, Kongwa Area, 1947-48

I. Mean of broadcast

N

P

K

191

L

and placed.

Response

Mean responses

to

absent present absent present absent present absent | present

ZAKL

N

95

öwüü

13

114

76

21

-15

10

19

1

2175

32

-6

31

-5

99

91

122

222

4

68

-15

37

17

28

-8

II. N and P

broadcast,

N

P

8

72

180

80

21

16

0

64

III. N and P

placed.

N

P

118

1983

17

148

888

47

-13

The figures in section v of table 7 were obtained using the following methods. pH values were determined by means of the glass electrode. P, Ca and K were determined in 2 minute 0.3 N HCl extracts. P was determined by reduction of phosphomolybdate and measurement of blue colour on an electrophotometer. Ca was determined by a rough turbidimetric method for experiments 5, 11, 12, 16 and 7a, and by an oxalate-permanganate method for the remainder. K was determined by a rough turbidimetric cobaltinitrite method.

The most general and significant result is the response to phosphate. This is most marked where the fertiliser was placed and where nitrogen was not given. There is some connection between response to phosphate and extractable soil phosphate. Nitrogen has been ineffective, and the NP interaction is in most cases small or negative. There is only one significant positive NP interaction, on a soil very low in lime and phosphate, although not acid. Lime is effective only on this soil (5), which bears out the general result of the liming experiments set out below (page 134). The effect of potash is significant on one soil only (11) of low potash analysis. A second soil (34) of even lower extractable potash, did not respond to potash fertiliser but it was an alkaline soil of high calcium content, so that analytical results may not be reliable.

At Urambo one experiment of this type was conducted on red soil and one on grey. The experiments were both planted late on somewhat uneven ground. Owing to wet conditions at planting time, the placement machinery could not be used on the grey soil and nitrogen and phosphate were applied to place- ment plots by hand along the rows.

At Namanga a single exploratory experiment, on heavy red loam, was set out. A very poor stand was established, and the local staff, in an endeavour to save the experiment, planted up the gaps with fiwi bean (Dolichos lablab). At harvest, groundnuts and fiwi bean were taken separately, the latter being weighed green. Owing to a series of mischances, no data for groundnuts can be presented, but the fiwi bean figures give some indication of the likely deficiencies of the soil. For obvious reasons they cannot be treated rigorously and no standard error has been computed.

1

The results of these experiments, together with relevant soil data are presented in Table 9. Page 364 of 1097

Page 364

129

M

AL

- ૬,

Ek

Ľ

41

I t

Π

1

!

| 1

וד

11

1+ 1

ון

16