Rit Landbúnaðardeildar : B-flokkur - 01.12.1961, Síða 35

33
of nitrogen and potash remain unchanged for all plots. Graph 2 shows that the rate of in-
crease of phosphorus uptake is similar for all experimental sites, whereas the total uptake at
any particular level of phosphorus application varied markedly. At the Sámsstadir experiment
station, for instance, grass on peat soil (Sámsst. (70 N)) took up far greater phosphorus than
on silt loam soil (Sámsst. (120 N)), in spite of the fact that the latter plots received about
70 per cent more nitrogen. The peat soil gave larger yields and also grass with somewhat
higher phosphorus content (see Tables 11 and 12).
Tables 5, 6 and 7 and graph 3 show that the „apparent phosphorus balance" of the soil
dropped rapidly with increased phosphorus fertilization, in particular for low rates of appli-
cation. This means that in relation to the amount applied, the quantity taken up by the grass
was by far greatest when the phosphorus applications were low. Thus in three of the present
trials, at the annual application of 13,1 kg P per hectare, the total quantity taken up by the
grass was 15 to 45 per cent greater than that applied to the soil (the „apparent phosphorus
balance“ is 115 to 145 per cent). When the phosphorus application was doubled, to 26,2 kg
P/ha, the „apparent phosphorus balance" dropped to about 70 to 80 per cent. This efficient
utilization of low rates of phosphorus applications was somewhat unexpected, yet a pleasant
surprise. The relatively low requirements and efficient utilization of phosphorus was unex-
pected for soils which when plowed and harrowed in a virgin condition are usually so totally
deficient in phosphorus that no plant will thrive without a phosphorus application.
Two questions relative to phosphorus are of primary interest to the farmer: (1) What is
the minimal quantity of phosphorus required to avoid yield reduction due to phosphorus defi-
ciency? (2) Is the supply of available phosphorus in the soil enough to ensure grass or fodder
with sufficient phosphorus content from a nutritional standpoint? These questions are treated
in another paper (Jóhannesson, 1959), and graph 4 is from that paper. This graph indicates
that a small annual applicatiion, or about 30 kg P2O5 per hectare, was sufficient for the re-
spective fields, and other evidence presented in the mentioned paper indicated that approxi-
mately 40 to 45 kg P2O5 sufficed along with 120 kg N or less pr. hectare. Analytical data listed
in the paper also indicated that grass from plots which receive annually 13,1 kg P (30 kg
P2O5) pr. hectare is sufficiently rich in phosphorus from a nutritional point of view.
It is obvious that more soil phosphorus accumulates as the annual applications increase. Table
13 lists some analytical data on the amount of extracted soil phosphorus, and shows an in-
creased quantity of „available“ phosphorus with increased rates of phosphorus fertilization.
These data, however, are not absolute measures of the quantity of soil phosphorus that may
become available to the grass, and therefore they cannot be used as a basis for any quantative
calculations. Experimental results of the agronomic experiment stations (Arni Jónsson, 1947—
1958) and some scattered fertilizer trials (Jóhannesson, 1956), however, have shown that very
appreciable residual effects from phosphorus fertilizer (triple superphosphate) are observed
in permanent grass fields. Yet, available evidence strongly indicates that for permanent grass
fields it is more profitable to top-dress with relatively small quantities of phosphorus fertiliizer
annually than with larger quantities at greater intervals.
Comments relating to Tables 8 and 9.
These Tables are included because they throw some further light on the orders of magni-
tude of phosphorus content in grass and of the „apparent phosphorus balance“ of the soil or
phosphorus utilization. Table 13 lists some characteristics of the respective soils.
It is difficult to make comparisons between the four experiment stations as regards yields and
chemical composition, since climatic or growing conditions usually are significantly different
in any one as well as in different seasons. The present information is insufficient to state in
quantitave terms the climatic influence or to eliminate its varying effects. Thus climatic