SPRINKLER IRRIGATION OF CLAY SOILS IN SOUTHERN FINLAND IV . The effect of repeated applications of water and nitrogen fertilization on spring cereals

In 1967 and 1968, in the month of June, irrigation of silty clay soil in southern Finland by a single sprinkling, which involved the application of 30 mm of water, increased the yields of spring wheat by an average of between 880 and 970 kg per ha, or 37—51 %, and the yields of barley by 1140—1340 kg per ha, or 37—42 %. Repeating the irrigation after about one week further increased the wheat yields on the average by from 670 to 800 kg per ha and the barley yields between 810 and 860 kg per ha. The effect of the irrigation on the size of the yields was virtually of the same magnitude both years in spite of the fact that in 1967 both June and July were extremely dry months, whereas in 1968 the dry June was followed by a rainy July. The sprinkling affected the ripening of the grain, on the other hand, differently each year: slightly retarding the ripening in 1967 but considerably hastening it in 1968. In the latter year, the irrigation prevented detrimental late tillering and thereby augmented the hectoliter weight and 1000-grain weight of the crops. In response to the one-time sprinkling, the protein content of the grains decreased in the case of the wheat by an average of 1.0—1.3 and of the barley by 0.1—0.7 %-units. The effect of two sprinklings was corresponding decreases of 1.9—3.1 and 0.8—1.0 %-units. The application to the seedlings as surface dressing immediately before irrigation ofcalcium nitrate containing 60 kg/ha of nitrogen increased the protein content of the wheat on the average by 1.6 and of the barley by 1.9 %-units. The nitrogen fertilization thus prevented excessive lowering of the protein content by irrigation and, furthermore, considerably intensified the salutary effect of the irrigation in augmenting the yield. A disadvantage of abundant nitrogen fertilization was a slight retardation of ripening. The placement of basal dressing with a fertilizer drill at a depth of 9 cm proved effective, particularly in 1967, when both the wheat and the barley yields increased 22 per cent in comparison with the effect ofsurface dressing. Use of the fertilizer drill, moreover, promoted the ripening process. Sprinkling did not appear to reduce the placement effect of the fertilizer; rather did these two methods form an exceedingly favorable combination. In 1964 a series of investigations was started with the aim of thoroughly ascertaining the effects of sprinkler irrigation on clay soils in southern Finland. In the years 1964

in the case of the wheat by an average of 1.0-1.3 and of the barley by 0.1 -0.7 %-units.
The effect of two sprinklings was corresponding decreases of 1.9-3.1 and 0.8 -1.0 %-units.
The application to the seedlings as surface dressing immediately before irrigation of calcium nitrate containing 60 kg/ha of nitrogen increased the protein content of the wheat on the average by 1.6 and of the barley by 1.9 %-units.The nitrogen fertilization thus prevented excessive lowering of the protein content by irrigation and, furthermore, considerably intensified the salutary effect of the irrigation in augmenting the yield.A disadvantage of abundant nitrogen fertilization was a slight retardation of ripening.
The placement of basal dressing with a fertilizer drill at a depth of 9 cm proved effective, particularly in 1967, when both the wheat and the barley yields increased 22 per cent in comparison with the effect of surface dressing.Use of the fertilizer drill, moreover, promoted the ripening process.Sprinkling did not appear to reduce the placement effect of the ferti- lizer; rather did these two methods form an exceedingly favorable combination.
In 1964 a series of investigations was started with the aim of thoroughly ascertaining the effects of sprinkler irrigation on clay soils in southern Finland.In the years 1964 66, a single sprinkling of 30 mm of water at the right time increased the yields of spring cereals between 25 and 50 per cent (Elonen, Nieminen and Kara 1967).The investigators observed that the effect of the sprinkling on the soil moisture conditions was of relatively short duration; it was therefore reasoned that more frequent sprinklings would probably have produced even better results.It was likewise surmised that the efficacy of the irrigation could have been increased by more abundant application of nitrogen fertilizer, which would at the same time have prevented the excessive reduction of the protein content of the grains caused by the irrigation.The validity of these assumptions was investigated in the following two-year period, 1967-68.The results arrived at are set forth in this publication.

Field experiments
Treatments and their symbols.The field experiments were laid out in a split-plot design (Steel and Torrie 1960), and they included four factors: as whole plots, four levels of irrigation (I); as subplots, two different crops; as sub-subplots, two levels of nitrogen fertilization (N); and as subsub-subplots, two fertilization depths (D).There were four replications; accordingly, the trials embraced 128 experimental plots.The experimental design was in principle similar to the one described in detail and illus- trated by a diagram in a previous publication.(Elonen, Nieminen and Kara 1967).
The irrigation was performed at night by rotary sprinklers with a radius of 12 ±2 m.
Drainage ditch water with apH of about 7 and electrical conductivity of < 0.1 mmho per cm was applied at a rate of smm an hour, the total single application amounting to 30 mm.It was endeavored to perform the irrigation at a time when, in the light ofprevious experience, plants were most likely to benefit from it (Fig. 1).The irrigation dates were The experimental plants were spring wheat (Svenno variety) and barley (Ingrid variety).The crops were cultivated in the same way under identical growing conditions side by side.
The basal dressing (N x ) was applied in conjunction with the sowing May 12, 1967 and May 10, 1968.In the former year, 60 kg of N, 43 kg of P and 56 kg of K per hectare were applied in the form of compound fertilizer.In the latter year, the corresponding figures were 75, 44 and 62.The fertilization was carried out with a fertilizer drill, the coulters being spaced 16 cm apart; on the one hand, a placement depth of 9 cm was used (D,), and on the other, the application was on the surface (D 0) with the plastic ferti- lizer tubes outdrawn from the coulters.
The application of the additional nitrogen (N 2) was performed with the same machine on the surface to the sprouts immediately before the initial irrigation.Nitrogen was applied as calcium nitrate in the amount of 60 kg N per ha; thus the nitrogen levels N x and N 2 in 1967 were 60 and 120 kg N per ha, and in 1968, 75 and 135 kg N per ha. (upper, fertilizer broadcast or placed) and 1968 (lower, fertilizer placed).
Experimental soils and their moisture conditions.The experimental plots were located in fields of silty clay on the farm maintained by the Re- search Foundation for Agricultural Machinery near Helsinki.The experimental soils were very much alike both years.The top soil contained an average of 50-52 per cent of the clay fraction (< 2 pm) and 3.4-4.0per cent of organic carbon.The pH of the soils (in 0.01 M CaCl 2 ) varied between 5.4 and 6.1, and their nutrient condition was fairly good.
The subsoils were heavy clays (clay content > 60 per cent).
The moisture condition of the soil was investigated by Bouyoucos' (1954) gypsum block method.Immediately after the sowing, 48 blocks were dug to three depths in 16 barley plots.The results of the measurements are given in Table 1 as averages for five-day periods.
As is well known, the method has certain drawbacks, so the results could scarcely correspond to the actual amounts of available water in the ground.The readings probably do, however, reflect fairly accurately the changes in the moisture condition of the soil and the effect of irrigation on them.
In 1967 both June and July were exceptionally dry months.The amount of »available water for plants» was reduced close to zero to a depth of 10 cm as of June 21, to a depth of 20 cm about a week later and to a depth of 40 cm some three weeks later.After that, there were no measurable amounts of available water for plants until August, when fairly heavy rainfalls moistened the surface layer of the ground.
The effect of a single 30-mm sprinkler irrigation was of rather short duration only about one week.By June 21 the ground had dried out to such an extent that 30 mm of water could not suffice to moisten the soil to a depth of 20 cm.Watering of the surface did appear, however, to retard the drying of the soil at a depth of 40 cm.The moisture condition in twice-irrigated soil was evidently fairly beneficial to plant growth throughout the month of June, but by the beginning of July the top soil dried nearly to the wilting point.
In 1968 June was once more dry, and the top soil dried close to the wilting point.At the end of June and the beginning of July, the total precipitation was 70 mm during a three-week period; thanks to the generous rainfall, the ground was fairly moist through- out the month of July.The crops evidently utilized the first two irrigations rather com- pletely, but after the third watering, performed on June 24, dry weather continued for only two days longer.The thrice-watered experimental plots were moister than the other plots for the rest of the summer.
The growth of the plants.In 1967 the grain crops emerged evenly.The beneficial effect of the placement of fertilizer could be perceived clearly after as short a time as two weeks after the emergence of the seedlings.Even more conspicuous was the effect of sprinkler irrigation.The watered crops were distinctly higher and denser than the unwatered ones.The barley-was gathered with a combine harvester on August 24 and the wheat on August 30.
The rains that came right after the sowing in 1968 destroyed the porous structure of the soil surface and weakened the sprouting of the crops.The rains that followed in the wake of the dry month of June caused harmful late tillering of the unirrigated crops, which did not have time to ripen.The irrigated crops, by contrast, ripened evenly.The crops were harvested on September 5.
Table 1."Available soil moisture" (%) at various depths in the barley plots.Mean values during the periods offive days.

Results
Grain yields (Table 2).The sprinkler irrigation on June 15, 1967, increased   the wheat yields an average of 970 (1900-2870) ± 430 kg per ha, or 51 23 %, and 680 kg per ha, or 42 i 24 %.The effect of the sprinkling carried out six days later was approximately of the same magnitude: the wheat crop increased 920 kg per ha, or 48 %, and the barley crop kg per ha, or 40 %.
Two applications of water caused the wheat yields to increase by no less than 1640 kg per ha, or 86 %, and the barley yields by 2010 kg per ha, or 71 %.Repetition of the sprinkler irrigation (I 3 -Ij) thereby augmented the wheat crops by an average of 670 kg per ha, or 35 %, and the barley crops by 810 kg per ha, or 29 %.
The effect of the watering was noticeably greater in the N 2 than Nx plots.A single watering of the Nx plots caused the wheat yields to increase by an average of 720 kg per ha, or 39 %, and the barley yields by 890 kg per ha, or 33 %, whereas the corresponding figures for the N 2 plots were 1190 kg per ha, or 61 %, and 1450 kg per ha, or 49 %.Cor- respondingly, two irrigation rounds in the N x plots augmented the wheat yields by an average of 1300 kg per ha, or 71 %, and the barley yields by 1550 kg per ha, or 58 %; but in the N 2 plots, the wheat yield increased as much as 1990 kg per ha, or 102 %, and the barley yield 2460 kg per ha, or 83 %.
Calcium nitrate applied to the shoots increased the average wheat yield by 520 dz 140 kg per ha, or 21 dz 6 %, and the barley yields by 810 dz 310 kg per ha, or 23 dz 9 %• The effect of additional nitrogen depended significantly on the sprinkler irrigation.With- out irrigation, the additional application of nitrogen increased the wheat yields by only 110 kg per ha and the barley yields by 300 kg per ha.The corresponding figures for once- watered plots were 580 and 860 kg per ha and for twice-watered plots 800 and 1210 kg per ha.Sprinkler irrigation thus enhanced the efficacy of nitrogen fertilization many times over.
The placement of basal fertilizer to a depth of 9 cm augmented the wheat yields by an average of 540 dz 80 kg per ha, or 22 dz 3 %, and the barley yields by 770 dz 160 kg per ha, or 22 dz 5 %.The placement effect of fertilizer on the wheat yields was an increase of 470 kg per ha without irrigation, 560 kg per ha with a single watering and 540 kg per ha with two waterings.The corresponding figures for barley were 930, 690 and 750 kg per ha.Sprinkler irrigation did not therefore significantly decrease the placement effect of the fertilizer.The placement of basal fertilizer augmented the wheat yields of the N x plots by 580 kg per ha and of the N 2 plots by 490 kg per ha.The corresponding figures for barley were 830 and 700 kg per ha.The application of additional nitrogen to young crops thus tended to decrease the placement effect of the basal dressing, but the interaction was not statistically significant.
In 1968 a single sprinkling augmented the wheat yields by an average of 880 i 410 kg per ha, or 37 ± 17 %, and the barley yields by 1340 i 580 kg per ha, or 37 ± 16 %.
Two waterings augmented the wheat yields, in comparison with unirrigated crops, by 1680 dz 410 kg per ha, or 71 d: 17 %, and the barley yields by 2200 dz 580 kg per ha, or 60 dz 16 %.Repetition of the sprinkler irrigation (I 2 -I x ) thereby augmented the wheat yields by 800 dz 410 kg per ha, or 25 d: 13 %, and the barley yields by 860 d: 580 kg per ha, orl?dzl2%.A third sprinkling tended to increase the yields further, but the effect was no longer statistically significant.
The effect of dressing with additional nitrogen was noticeably slighter than the year before: the wheat crop increased an average of 410 dz 150 kg per ha, or 12 £5 %, and the barley crop 390 i 380 kg per ha, or 8 ± 8 %.The positive interaction between nitrogen dressing and sprinkler irrigation was likewise noticeable only to a very slight degree: nitrogen dressing increased the wheat yields of plots 10, I 0, T x , I 2 and I 3 by 290, 300, 470 and 600 kg per ha and the barley yields, correspondingly, by 160, 230, 210 and 970 kg per ha.The significance of the placement of the basal dressing was likewise substantially slighter than in the previous year and statistically significant only in the case of the barley plots, where the yields increased on the average by 280 i 150 kg per ha, or 6 ± 3 %.
As in the previous year, sprinkler irrigation and additional nitrogen applied to the seedlings did not appear to have decreased the placement effect of the basal dressing.Moisture of grains at harvest (Table 3).In 1967 the irrigation significantly affected the ripening of the cereal crops only in the barley plots.The moisture of the grains at threshing time was 2.5 d: 0.9 %-units higher in the twice-irrigated than in the unirrigated plots.The additional nitrogen also slightly retarded the ripening, increasing the moisture content of the wheat by 1.0 dz 0.5 and that of the barley by 1.4 ± 0.3 %- units.Placement of the basal dressing, on the other hand, had the effect of promoting the ripening process: the moisture content of the wheat was 1.2 ± 0.4 and that of the barley grains 2.2 dz 0.3 %-units lower in the experimental plots treated with placed fertilizer than in the plots treated with surface dressing.In 1968 the differences in the moisture content of the cereal crops were very great, owing to late tillering, and the sprinkler irrigation had an exceedingly beneficial effect on the ripening of the grain: once sprinkled, the wheat was 5.4 3.3 %-units, and twice sprinkled, 8.3 i 3.3 %-units drier than unirrigated wheat.With respect to the barley, the corresponding differences were even greater 10.6 5.5 and 15.3 ± 5.5 %-units.
Differences in moisture content of such magnitude were partly due to the fact that the unirrigated barley crop included 4.8 %, the once-watered crop 3.0 % and the twicewatered crop only 0.8 % of green grains.Even a third application of water seemed to promote further the ripening of the crops.As in the previous year, the application of additional nitrogen retarded the ripening of the cereal crops: the moisture difference in the case of the wheat was 3.1 dz 0.8 and that in the case of the barley 2.3 dz 0.8 %-units.
The placement of basal dressing, on the other hand, had no effect on the ripening process.
Protein content of grains (Table 4).In 1967 a single sprinkling reduced the protein content of wheat (5.7 X total N) on an average by 1.0 i 0.7 and two sprink- lings by 1.9 dz 0.7 %-units.The protein content of barley (6.25 X total N) was reduced by a single sprinkling by 0.1 dz 0.5 and two sprinklings by 0.8 dz 0.5 %-units.Sprinkler irrigation carried out once did not therefore significantly cause a decrease in the protein content of barley.Additional nitrogen applied to young crops markedly increased the of the grains was slight and significant only in the case of the twice-watered wheat plots, in which the irrigation raised the weight of 1000 grains on an average by 1.3 ±0.9 g.
The effect of the irrigation was most marked in the case of those wheat plots in which basal dressing had been placed and which had received additional nitrogen.Both the placement of basal dressing and the application of additional nitrogen tended to increase the weight of the wheat grains; but they had no effect on the weight of the barley grains.
In 1968 a single sprinkling raised the weight of 1000 wheat grains on an average by 2.6 rb 1-4 and of 1000 barley grains by 2.5 rb 2.4 g.A second or third application of water no longer caused any increase in the weight of the grains but tended, on the contrary, to lower it.As in the year before, the placement of basal dressing and the application of additional nitrogen slightly increased the weight of the wheat grains but had no effect on the weight of the barley grains.
Hectoliter weights (Table 6).In 1967 sprinkler irrigation had no significant effect on the hectoliter weight of the wheat, but that of the barley was lowered as a result of two waterings on an average by 1.2 ± 0.6 kg.The application of additional nitrogen reduced the hectoliter weight of wheat on an average by 1.0 rb 0.4 and that of barley by 0.5 rb 0.3 kg.The placement of basal dressing had no significant effect.
In 1968 sprinkler irrigation distinctly improved the hectoliter weight of both the wheat and, in particular, the barley.The increases brought about by one, two and three sprinklings in the case of the wheat were 1.0 ± 0.8, 1.3 ± 0.8 and 1.1 ± 0.8 and, cor- respondingly, in the case of the barley, 1.9 ± 1.1, 2.8 Jr 1.1 and 3.2 ±l-1 kg.The application of additional nitrogen reduced the hectoliter weight of the wheat on an average by 0.6 ± 0.2 kg but it had no significant effect on the barley.The placement of basal dressing had no significant effect.

Discussion
The experimental years constitute quite an interesting pair.Common to the two years was a dry June up to the month's 26th day.Even after that date, in 1967, the dry spell stretched out all the way to August, and the ground became parched.In 1968, how- ever, there was fairly abundant rainfall at the end of June and the early part of July, with the effect that the ground was moist all through the middle of the summer.
In spite of the difference between the midsummer period of the two years, the effect of the sprinkler irrigation carried out in June on the size of the harvest was highly similar both years.A single sprinkling augmented the wheat yields on an average by 880-970 kg per ha, or 37-51 %, and the barley yields by 1140-1340 kg per ha, or 37-42 %.Two sprinklings increased the wheat yields on an average by 1640-1680 kg per ha, or 71- 86 %, and the barley yields by 2010-2200 kg per ha, or 60-71 %.Repetition of the watering thus further augmented the yields quite markedly, although, to be sure, the increase brought about by the second sprinkling was in the case of the wheat on an average 1 1 4 smaller and in the case of the barley 1/3 smaller than the crop increase obtained through the first sprinkler irrigation.
The effect of the sprinkler irrigation on the protein contents of the crops was also very much the same both years.A single sprinkling lowered the protein contents of the wheat on an average by 1.0-1.3 and of the barley by 0.1-0.7 %-units.Two sprinklings reduced the protein contents correspondingly by 1.9-3.1 and 0.8-1.0%-units.It therefore looks as if sprinkler irrigation reduces the protein content of wheat more drastically than that of barley.
In the drought year of 1967, the unirrigated cereal crops ripened, as if under pressure, earlier than normally; thus irrigation ostensibly delayed the ripening, though only very slightly.The following summer, the rainy midsummer season brought about detrimental late tillering of unirrigated crops and thereby delayed ripening.Under such conditions, the irrigation had the contrary effect, one of preventing belated tillering of the crops and promoting the ripening process.
The findings of the investigation prove conclusively that the weather conditions prevailing in the month of June have a decisive effect on the development of spring cereal crops in the areas of clayey soil in southernFinland.If the month of June is short on rain, crops are incapable of effectively utilizing the nutrients contained in the parched earth and growth of plant life is weak.If also the middle and late summer seasons are dry, unirrigated soil is liable to have left over an abundance of unutilized plant nutrients (Kaila and Elonen 1971).If a dry June is succeeded by a rainy July, the abundant nitro- gen reserves and sparse growth stimulate belated tillering of plants which to some extent improves the harvest but at the same time delays its ripening.
By sprinkler irrigation, the nutrient reserves in the ground can be made effectively available to plants as early as June (Kaila and Elonen 1970), a dense plant association develops and it ripens evenly regardless of the weather conditions prevailing during the middle and late summer seasons.It has been pointed out in previous studies that the most critical season for spring cereal crops is about three or four weeks long midway between the emergence of the plants and the appearance of the ears.The effect of irrigation at this time remains virtually unchanged provided the dry spell continues (Elonen, Nieminen   and Kara 1967).The findings of the present investigation confirm, as far as they go, the results of earlier research: the irrigations carried out on June 15 and June 21, 1967, proved   by and large equally beneficial.Calcium nitrate containing 60 kg N/ha, when applied to the seedlings, increased the protein content of the wheat on an average by 1.6 and of the barley by 1.9 %-units.The excessive reduction of the protein content induced by irrigation could thereby be fore- stalled, besides which the effect of the watering on the size of the harvest improved notice- ably.The amounts of nitrogen of the N 2 -level, 120 and 135 kg N per ha, seem to be highly suitable to the fertilization of a wheat field awaiting irrigation.The amounts of fertilizer applied to the malting barley proved to be excessive, however, for in spite of the sprinkler irrigation the protein contents exceeded 12 per cent.
The placement of basal dressing to a depth of 9 cm improved its utilization by the crops quite effectively, especially in 1967.Once again, the same astonishing result was arrived at as in previously reported investigations: irrigation did not diminish the placement effect of fertilizer (Elonen, Nieminen and Kara 1967; Kaila and Elonen 1970).Plants grow around the rows with placed fertilizer thick root clusters, which thirstily suck the ground dry and the watering of which is probably particularly advantageous (Kähäri and Elonen 1969).After the irrigation, the ground dries up once more, first at the surface and only later at the placement depth; accordingly, irrigation is likely to improve for a longer time the utilization of placed nutrients than of those applied onto the surface.It is known, furthermore, that in clayey soil sprinkling does not to any noteworthy extent carry downward other macronutrients than nitrate (Aura 1967).This also em- phasizes the importance of fertilization by placement.
Quite favorable results have been obtained with the placement of fertilizer in Finland (Larpes 1967, Nieminen, Kara and Elonen 1967, Kivi and Hovinen 1969, Pessi et ai.1970, Paulamäki and Luostarinen 1971).It seems clear that this method could be profitably applied also in irrigated fields.It is possible that also additional nitrogen, which in the present investigation was given to seedlings, could be advantageously applied in immediate conjunction with sowing by the placement technique.

Fig. 1 .
Fig. 1.The stages of development of barley on the dates of irrigation in 1967(upper, fertilizer broadcast or placed) and 1968 (lower, fertilizer placed).

Table 3 .
Moisture of grains at the harvest, %.