NOTES ON SOME MANURIAL EXPERIMENTS IN CANTERBURY AND OTAGO.

New Zealand Journal of Agriculture, Volume XXX, Issue 1, 20 January 1925, Page 24

NOTES ON SOME MANURIAL EXPERIMENTS IN CANTERBURY AND OTAGO.

F. W. HILGENDORF,

D.Sc., Canterbury Agricultural College, Lincoln.

In past years the Department of Agriculture made . a great number of manorial and variety experiments in co-operation with farmers in many localities. The following notes refer to, a series conducted in the South Island, chiefly by Mr. A. Macpherson, during the years 1911-16. A uniform plan of operations was followed until the war made certain manures scarce, and then the modification was the least possible. The trials were made- on about a hundred, farms, chiefly in Canterbury and Otago, and they lasted, as already indicated, for five seasons. By, this means some very valuable figures were obtained, and any mistakes or inaccuracies due to soil or seasonal variations will have been smoothed out by the large number of the trials averaged in the following tables. By the use of tables of probability, too, any abnormal variations are taken account of, so that great confidence may be felt that the results of these trials will be reflected on the average . of a series of years on a farm having the average soil and climate of the districts in which the trials were made. Records of the experiments were published in various issues of the Journal during the period in question, and. are indexed in , the halfyearly volumes, also separately in the Consolidated Index for 1910-20 under the general heading of “ Co-operative Field Experiments/’ There is introduced into the tables that follow a column that requires some explanation. It is called “ Odds in Favour of Significance.” In any experiment repeated trials will give various results. For example,. if one tosses six coins there may come up four heads or two heads instead of the expected three. The average of the first ten throws may be 3-5 or 2-5 heads, and yet the coins be quite normal'. If this is so, then'differences between 2-5 and 3-5 in an average of ten coin-tossing experiments do not mean anything, are merely chance, or, to use the common term, are not “ significant.” If, however, we tossed the six coins a thousand times and found then that on the average 3-5 or 2-5 heads had turned up at each, throw, we would be justified in saying that there was a real difference between the chances of heads and tails turning up — the difference between the number of heads and tails in the trials meant something, or was “ significant ” ; and inspection would probably prove that some of the coins were heavier on one side than on the other. On observations such as these there has been built up a Theory of Probabilities, by which one can measure the . chances that the difference shown between two series of experiments is a real one due to the different manure or. variety, or only an accidental one due to variations ’ in soil, seeding, weighing, recording, &c. It is these chances that have been calculated and entered in the column referred to. Odds of over 30 to 1 are regarded as practical certainty.

TURNIP MANURIAL TRIALS.

Those under . consideration were conducted on five farms in Canterbury ‘ and forty-nine’ in Otago in the years 1911-16. In the cases where the same ■ farm was used in ' two different ~ years it is counted as two farms. In every case considered six plots were tried on the same farm, every series containing one unmanured plot, one with 11 cwt. superphosphate, and so on. The results are given in tons per acre of roots alone, and the method of computation was that known as “ Student’s,”* each other plot being differenced against 11 cwt. super, as that manuring may be considered standard practice. Plot 6 is, however, contrasted with Plot 5 for a reason obvious on examination.

The great increase due to the use of cwt. super, or, rather, the great loss from not using super, is in accord with universal experience. The 2 cwt. guano gives less than super, and the chances are 5 to 1 .that the difference is a real one. The 3 cwt. super gives an increase on 1J cwt., but the chances in favour of significance are not. large enough to allow a recommendation to adopt the practice. It was thought that by taking, the results out separately for' the wetter districts the 3 cwt. might show to greater advantage, but the information regarding the rainfall during the growing season was not obtainable, and computions from average yearly rainfalls gave no result. A dressing of 2 cwt. super plus f cwt. bonedust gave an increase of 1 ton per acre over cwt. super, and the chances are 24 to 1 that the difference is not due to the chance variations of the experiment, but is a real one that would be repeated under similar circumstances. Such odds in its favour cause this experiment to be worth very serious consideration. Of course it is impossible to say whether the increase is due to .the added super or to the bonedust, but the absence of marked effect from the 3 cwt. super should cause the bonedust to be regarded with attention. The indications are that of the manures tried in this series i| cwt. to 2 -cwt. super plus | cwt. of a slow-acting phosphate, such as' bonedust or guano, will give the best results on turnips in Otago. It is a question for each individual farmer whether the extra ton of turnips per acre will pay for the increased manure. The addition of potash gave no increase in yield. A future article will deal with the manurial trials on other crops.

* An explanatory statement on this method by Dr. Hilgendorf is appended /next page), more particularly for the information of other experimenters or instructors.—Editor.

Manure. Number of Trials. Yield. Gain or Loss on 11 cwt. Super. Odds in Favour of '■ Significance. (i.). No manure 54 12-8 9-0 tons loss Thousands to i (2.) 2 cwt. guano 54 . . 20-6 i-2 tons loss 5 to 1 (3.) i| cwt. super .. .54 21-8 (4.) 3 cwt. super 54 22'4 o-6 tons gain 9 to 1 (5.) 2 cwt. super, plus f cwt. bonedust 54 22-8 i*o tons gain 24 to 1 (6.) No. 5, plus | cwt. sulphate of potash 54 . 22-7 o-i loss on No.5 I-J- to I