Maataloustieteellinen A ikakauskirja Vol. 61: 361—370, 1989 Effect of different rates of P fertilization on the yield and P status of (he soil in two long-term field experiments MARKKU YLI-H ALLA Kemira Oy, Espoo Research Centre, Luoteisrinne 2, SF-02271 Espoo, Finland Abstract. Two field experimentson P fertilization were conducted on clay soils in South- ern Finland. The rates of P applied yearly in granular NPK fertilizers were 0, 13/16, 26/32, 47/56 and 60/72 kg P/ha in 1974—82/1983 —85. Oats, barley, spring wheat and winter wheat were grown, in two years also oil seed crops. In one experiment, the maximum yield of cereal grains in the first nine years (4 460 kg/ha) was reached at the P rate of 13 kg/ha, but there- after at 32 kg P/ha. The average difference between the maximum yields and the ones ob- tained without P fertilization was 470 kg/ha (12 %) in 1974—80, but during the last four years the difference increased to I 360 kg/ha (40 %), owing to the depletion of P in the plots not fertilized with P. Also in the other experiment, in which the maximum yield of cereal grains (4 790 kg/ha) was obtained at the P rate of 26/32 kg/ha, the response to P fertilization in- creased towards the end of the trial, the mean response during the last three years being 570 kg/ha (12 %). Phosphorus fertilization, up to the P level at which the maximum yield was reached, decreased the moisture content of cereal grains at harvest. The quantity of P extracted with 0.5 M NH 4-acetate-0.5 M acetic acid (pH 4.65) decreased in the plots not fer- tilized with P, from 5.8 mg/l to 2.2 mg/l and from 6.2 mg/l to 1.8 mg/1 in the course of the two trials. The original level of acetate-extractable P was somewhat maintained but not elevated by P rates of 26/32, 47/56 and 60/72 kg/ha. Residual P was recovered mainly in the fractions extractable with NH 4F (“Al-P”) and NaOH (“Fe-P”). Index words: phosphorus fractionation, soil analysis, long-term experiments Introduction Plants take up phosphorus from the fer- tilizer being applied currently as well as from the previous reserves contained in the soil. Field experiments conducted in the USA and Australia recently showed that only 9—23 % (Sharpley 1986) or 16 % (McLaughlin et ai. 1988)of P taken up by wheat plants originated in the P fertilization applied simultaneously with sowing, and that the rest came from the P reserves of the soil. The results of an exten- 361 JOURNAL OF AGRICULTURAL SCIENCE IN FINLAND https://www.c-info.fi/en/info/?token=aZXgV2oQhnkm1ALK.mayAnmIX1UxcyoMyX7Cppg.39gezT0ErgQeOBJM9YiUiyy7AtWxdn2MWhPrwSrnA1vhX7xRwYLtVOyg6mke1Qtkaj5a2deAunR9Vcy_RXGX8SHKbvb46DdKUx-KMnAf3uwlzJ1UCCJndzXMwTmCCJPQYmm7Q1CxWDmF5cigd8PPkXVzP1QrFNw_yr4KDUOxxQ sive series of fied experiments performed in Finland (Saarela 1989) showed that less re- sponse to P fertilization was obtained in soils moderately high in easily soluble P than in soils low in easily soluble P. Those results im- ply that in soils high in easily soluble P, cur- rent fertilization has only a minor contribu- tion to the P supply to the plants. Besides soil properties, the quantity of plant-available P in the soil depends to a large extent on the ap- plications of fertilizer P given to the soil in the past (Jankovic and Nemeth 1979, Har- tikainen 1989b). In order to assess the ap- propriate rate of P fertilization in the long run, experiments on P fertilization need to have a sufficient duration so that also the labilepool of P in the soil has time to be modi- fied to the size determined by the rate of P fertilization. Further, long-term field experi- ments are needed for the calibration of soil fertility tests. The present paper is based on the results of two field experiments which lasted 11 or 12 years. The aim is to contribute to the deter- mination of the level of P fertilization at which the supply of P to the crop does not lim- it yield formation but, on the other hand, does not lead to a build-up of unnecessarily high levels of easily soluble P in the soil. Therefore, in addition to the quantity and quality of the yield, the P status of the plots receiving re- peated applications of fertilizer P was exam- ined. Materials and methods Two field trials, called here experiment A and B, were set up in spring and autumn of 1974, respectively, and continued until 1985 at Kotkaniemi experimental farm in Vihti, Southern Finland. The trials were laid out on clay fields according to the randomized block method with four plots in experiment A and three plots in experiment B. The size of the plots was 3 by 25 m and 5.5 by 20 m in ex- periment A and B, respectively. Initially and nearly every year in the course of the trials, the plots were sampled and analyzed for P and pH. Phosphorus was extracted with 0.5 M NH 4-acetate 0.5 M acetic acid at pH 4.65, abbreviated AAAc (Vuorinen and Mäkitie 1955), and the pH was measured in water sus- pension using the soil-to-water ratio of 1:2.5. At the end of the experiment, a composite sample of about 25 kg was taken from each plot and analyzed for the particle size distri- bution and organic carbon. The samples were also analyzed for poorly crystalline Fe and Al, extractable with 0.05 M NH 4 -oxalate at pH 3.3 (Hartikainen 1982) as well as for water- soluble P (Hartikainen 1982). Some proper- ties of the experimental fields are listed in Ta- ble 1. Further, inorganic P of the samples taken at the end of the trials, was fractionated according to a slightly modified version of Chang and Jackson’s procedure (Hartikai- nen 1979). The granular compound fertilizers for the experiments were manufactured at the pilot plant of Kemira Oy Espoo Research Centre. The fertilizers had increasing concentrations of P (Table 2). Nitrogen fertilization together with sowing was 100 kg (N/ha in 1974—1982 and 120 kg N/ha in 1983—85. Potassium fer- tilization was 80 and 96 kg K/ha, respective- ly. The rates of P were 0, 13, 26, 47 and 60 Table 1. Some properties of the experimental fields. Experiment Clay Organic C pH Paaac Al % % mg/1 ...mmol/kg A: mean 37 3.3 5.4 5.4 84 71 range 27—44 2.6—4.0 5.3—5.6 4.5—6.2 72—100 64—77 B: mean 31 3.4 6.0 6.0 64 85 range 25—40 2.1—5.1 5.8—6.1 5.8—6.1 52—86 76—91 362 Table 2. Composition of the experimental fertilizers. Table 3. Crops and varieties grown in the field experi- ments. Year* Experiment A Experiment B Crop Variety Crop Variety 1974 Spring wheat Tähti 1975 » » Winter wheat Aura 1976 » » » » 1977 Oats Ryhti » » 1978 » » Oats Ryhti 1979 » » » » 1980 Barley Aapo Turnip rape Torch 1981 » » Spring wheat Tähti 1982 » » Spring rape Lergo 1983 Spring wheat Tapio Spring wheat Tapio 1984 » » » » 1985 » Ruso Barley Aapo The year of harvest kg P/ha until 1982 and 0, 16, 32, 56 and 72 kg P/ha in 1983—85. Dividing the quantity of P applied to the plots during the whole peri- od by the number of experimental years gave average P rates of 0, 14, 28, 49 and 63 kg P/ha. The fertilization together with sowing was performed according to the placement method. Spring wheat, barley and oats were cultivated (Table 3) in experiment A, but in experiment B also winter wheat and oil seed crops were grown. Winter wheat received 55 kg N/ha in spring, in addition to the NPK fer- tilization applied in the previous fall together with sowing. After harvest, the grain or seed yields were weighed and their moisture con- tent was determined. The grains and seeds were analyzed for P by a vanado-molybdate method (Saari and Paaso 1980). The straw was ploughed under. Results The yields of cereal crops varied from year to year, mainly because of weather conditions. In the plots fertilized yearly with 26/32 kg P/ha, the grain yields ranged from 3 270 Fig. I. Grain and seed yields in the two experiments at P rates of 0 and 26/32 kg P/ha P fertilization Concentration k g/ha n P K % % % 0 19 0 16 13/16 15 2 12 26/32 15 4 12 47/56 15 7 12 60/72 15 9 12 363 Table 4. Grain and seed yields in the plots fertilized with different quantities of P*. kg/ha to 5 360 kg/ha and from 2 570 kg/ha to 5 800 kg/ha in experiment A and B, respec- tively (Fig. 1). In 1977 the low yield of winter wheat in experiment B was caused by winter damages, and in 1981 the yields were low in both trials because the weather was wet and exceptionally cool. The coefficients of varia- tion (s/x) for grain yields were 15 % and 26 % in experiment A and B, respectively. Ac- cording to analysis of variance and the Stu- dent-Neuman-Keuls’ test (Steel and Torrie 1980), P fertilization increased the average yields of cereal crops statistically significant- ly in both experiments (Table 4). The increase was greater and more consistent in experiment A (F = 38.54***) than in experiment B (F = s.so***). The oil seed crops, cultivated in two years in experiment B, did not respond to P fertilization as far as the size of the yield was concerned. The effects of P fertilization on the yields of single experimental years were also studied. Fig. 2. Response of the yields of cereal crops to repeated applications of different rates of P fertilization Fertilization Grain yield Oil seed yield kg P/ha —Z : ~TZ Z ' , D uu Experiment B### Experiment A# Experiment B## kg/ha kg/ha kg/ha 0 3 750b 4 430 b I 940' 13/16 4 470' 4 660'b 1880" 26/32 4 440" 4 790" 2 160" 47/56 4 440' 4 550" 1 950" 60/72 4 550' 4 810' 2 030' # Each column has been tested separately. Means with a common letter are not different at the 95 % of statistical probability. # 12 years # # 9 years # # # 2 years 364 In experiment A, P fertilization increased grain yields in 10 out of 12, the nonexistent response occurring in the 2nd and 3rd ex- perimental years. In experiment B, the yields obtained from the plots fertilized with P were statistically significantly greater only in the sth, 9th, 10th and 11th experimental year. Based on the magnitude of the response to P fertilization by grain crops, the experiments were divided into two periods (Fig. 2). Dur- ing the first part, consisting of seven years (1974 —80) in experiment A, the maximum yield was reached at the P rate of 13 kg/ha, at which 470 kg/ha (12 %) more grain was ob- tained, compared to the plots not fertilized with P. In the latter period (1981 —85), the yield increase, attributable to P fertilization of 13/16 kg/ha, was as much as 1 060 kg/ha (33 %). Differing from the earlier years, the maximum yield during the last three seasons (1983 —85) in experiment A was reached at the P rate of 32 kg/ha, at which the yields were 1 360 kg/ha (40 %) greater than those ob- tained without P fertilization. In experiment B, the maximum yields were obtained at the P rate of 26/32 kg/ha throughout the ex- perimental period. During the earlier part (1975 —81) of experiment B, the average yield increase of cereal grains, attributable to P fer- tilization, was 250 kg/ha (6 %), but during the latter part (1983—85), it was increased up to 570 kg/ha (12 %). According to the paired t test (Steel and Torrie 1980), the average in- creases in yields of cereal crops were also statistically significantly greater during the lat- ter period of the trials, the test values being t = 39.35*** and 4.56* in experiment A and B, respectively. When P fertilization was increased from 0 to 13/16 kg P/ha, the moisture content of cereal grains at harvest decreased by 1.8 and 1.0 percentage points on average in experi- ment A and B, respectively. In the six wet growing seasons, when the moisture content at harvest was more than 30 %, the cor- responding difference was as much as 3.5 per- centage points, which indicates that P fertili- zation had speeded up the development of the crop. The difference was smaller in dry years. Additional decrease in the moisture content of the grains did not occur as a result of P rates beyond the level at which the maximum yield was reached. The phosphorus concentration in the cereal grains did not depend on the size of the yield or on the cereal grain species cultivated, and it was only slightly elevated by P fertilization (Table 5). The effect of P fertilization on the P concentration was more prominent in the oil seed crops, which also had a considerably higher concentration of P than did the cereal grains. The differences in P concentration of the cereal grains were much greater between the yields obtained in different years than the ones attributable to P fertilization within a given year. The uptake of P by the crop was enhanced by P fertilization in both experi- ments, mainly owing to the increase in the size Table 5. Phosphorus content of the grains or seeds receiced from plots fertilized with different quantities of P*. * Each column has been tested separately. Means with a common letter are not different at the 95 % level of statisti- cal probability. It Cereal crops Fertilization P in dry matter Uptake of P kg P/ha 1Exp A# Exp. B# Oilseeds Exp. A# Exp. B# Oilseeds g/kg g/kg g/kg kg/ha kg/ha kg/ha 0 3.6 h 3.7" 7.8d 11.7‘ 14. F 13.1" 13/16 3.6" 3.8" 8.61 13.8" 14.9" 14.1" 26/32 3.6" 3.8" 8.9" 13.9" 15.4"" 16.7“ 47/56 3.7“ 3.8" 9.2“ 14.4“" 14.6"c 15.5“ 60/72 3.7“ 3.9“ 9.3“ 14.6“ 15.8“ 16.3“ 365 of the yields. Considerable quantities of P were mined from the soil by the crops grown without P fertilization. In the last experimen- tal year, the crop in experiment A was still able to extract as much as 11.6 kg P/ha from the plots not fertilized with P in 12 years al- though, at that point, the yield formation was already seriously limited by the poor supply of P, as indicated by the yield increase of 32 % in the plots receiving 32 kg P/ha. The P balances of the plots were calculated as differences between the quantities of P giv- en in fertilizers and those removed via the crops (Table 6). The plots receiving no P fer- tilization were depleted by 141 and 153 kg/ha in experiment A and B, respectively. The plots receiving 13/16 kg P/ha yearly did not gain or lose P, and P was accumulating in the plots receiving higher rates of P. Residual P had ac- cumulated in forms extractable with NH 4F (“Al-P”) and in experiment B also with NaOH (“Fe-P”) (Table 7). The fraction ex- tracted with H2S0 4 (“Ca-P”) had increased slightly in experiment A. In absolute amounts, both “Al-P” and “Fe-P” were equally af- fected by the fertilization treatments, but due to the smaller quantity of “Al-P”, the rela- tive changes in this fraction were more pro- nounced. Assuming the depth of the plough layer to be 25 cm, the recovery of residual P Table 6. Inputs and outputs of P during the experimental period of 12 years (experiment A) and 11 years (experi- ment B) as well as the P status of the plots at the end of the trials. * Each column and experiment has been tested separately. Means with a common letter are not different at the 95 % level of statistical probability. Table 7. Fractions of inorganic P at the end of the experiment.* * Each column and experiment has been tested separately. Means with a common letter are not different at the 95 % level of statistical probability. Fertilization Total input Carried away Balance P AAAc * P w* kg P/ha kg P/ha in the yields in the soil mg/1 mg/kg kg P/ha kg P/ha experiment A 0 141 —l4l 2. l b 2.5» 13/16 166 165 I 3.8» 3.6» 26/32 333 166 167 4.3» 5.1 b 47/56 583 171 412 5.1» 7.3» 60/72 650 176 574 4.3“ 5.5 b experiment B 0 153 —137 1.7» 5.1» 13/16 154 162 —8 3.2b» 6.7b 26/32 309 171 138 5.2» b 9.4» 47/56 541 162 379 5.1» b 12.3b 60/72 696 174 522 6.9“ 13.7» Fertilization P (mg/kg) extracted with Sum of kg P/ha fractions NH 4F NaOH H 2 S04 Experiment Experiment Experiment Experiment AB AB AB AB 0 69' 42' 151“ ISO 104h 230' 324' 422' 13/16 9lb 46' 158“ I58b' 107b 236“ 356b' 440 b 26/32 109“b 77 b 165" I95 b H2 b 216“ 386“b 494 b 47/56 125“ 95“b 173" I98b 121“ 236“ 419“ 559“ 60/72 108“b 120“ 162“ 241“ 121“ 242“ 391“b 605“ 366 367 Table 8. Apparent recovery of residual P in the plough layer. in each plot was calculated as the quantity of total fractionated P from which the quantity of P in the plots not receiving P during the experiment was subtracted (Table 8). These quantities were compared to the theoretical differences created by fertilization and P up- take of the crop and calculated from the data in Table). In terms of inorganic P, the plots fertilized and not fertilized with P differed from each other less than expected, and the recovery of residual P was no more than 24— 71 % of the theoretical quantities. The soil analyses performed during the ex periments (Fig. 3) showed that in experiment B, the quantities of P AAAc were continuously decreasing in the plots receiving no P or 13/16 kg P/ha yearly. The ones fertilized with 26/32 kg P/ha or 49/56 kg P/ha (not shown in Fig. 3) somewhat maintained the original level of P AAAc. Even in the plots receiving 60/72 kg P/ha yearly, no marked increase of P AAAc was measured. In experiment A the soil ana- lyses did not show equally clear trends in any of the various P treatments. Nevertheless, at the end of the experimental period the level of P AAAc was the higher the more P had been applied to the plot, with the exception of plots receiving 60/72 kg P/ha (Table 6). Accord- ing to the paired t test, the differences in P AAAI . extracted from the plots receiving the same fertilization in the two experiments were practically nonexistent at the end of the ex- periments (t = 0.86ns ). However, extraction of P with water (P w) revealed considerable differences between the two experiments. The quantities of P w were higher in experiment A when the corresponding plots of the two ex- periments were compared (t = 4.71 **). Fig. 3. Development of the quantities of P extracted with 0.5 M NFl 4 -acetate- 0.5 M acetic acid at pH 4.65 ( p aaac) 'n plots fertilized with different rates of P. Fertilization Recovery of residual P kg P/ha T : 7 ~~ ~ Experiment A Experiment B kg/ha % kg/ha % 13/16 80 56 45 35 25/32 155 50 165 60 47/56 238 43 268 52 60/72 168 24 468 71 Discussion In a long-term experiment on P fertiliza- tion, residual P accumulates in the plots fer- tilized with P; in plots not receiving this nu- trient, P is continuously depleted. Repeated P treatments thus lead to a continuous differentiation of the plots in terms of the P status of the soil, a phenomenon disclosed in both of the present experiments. Therefore, the response to P fertilization in such experi- ments must be considered a function of cur- rent P fertilization as well as of residual P ac- cumulated in the soil since the beginning of the trial. The most important outcome of the present experiments may indeed be the documentation of the phenomenon that the apparent response to P fertilization was in- creasing over time, as was also reported by Li and Barber (1988). However, it must be pointed out that the phenomenon was due to the depletion of the plots receiving no P fer- tilization, which was used as the reference level. The overall response to P fertilization in experiment A was as much as double the response obtained by Saarela (1989) in a se- ries of field experiments in which cereal grains were cultivated for 10 years in soils which in terms of P AAAc were in the same range as the present trials. In experiment A, the response to P fertilization was of the same magnitude as reported by Saarela (1989). The limited material does not allow conclusions to be made about the response of various cereal grain species to P fertilization. During the first few years, greater response to P fertilization was obtained in experiment A than in experiment B. This was in agree- ment with the smaller quantities of P AAAc ex- tracted from the plots receiving no P in ex- periment A. However, towards the end of the experiments the plots to which no P was ap- plied did not differ from one another in the two experiments in terms of P AAAc . The soil analyses performed in the latter period of the experiments thus contradict the yield in- creases, which were much greater in experi- ment A. This result suggests that despite the equal PAAAc, the soil of experiment A had a poorer P status than that of experiment B. The difference between the experimental soils was, however, disclosed by the water extraction, performed at the end of the trials. This ob- servation is in line withother studies in which water extraction has been superior to the AAAc method in predicting yield increases ob- tained by P fertilization in field experiments (Sippola and Saarela 1986) or P uptake by plants in pot experiments (Aura 1978, Sippo- la and Jaakkola 1980). It may be suggested that an AAAc extraction in which the soil: so- lution ratio is only 1:10 gives an estimate for P intensity. In turn, it was shown by Schachtschabel and Beyme (1980) that Pw , obtained at the wide soil:solution ratio of 1:60, also reflects the capacity of the soil to supply the plants with P. Further, an impor- tant observation was that even excessive P fer- tilization could not elevate the level of P AAAt. in either of the experiments. It remained an open question whether the residual P was real- ly converted in the soil into forms unavaila- ble to plants or whether the current result was an indication of the extensive P buffer capac- ity of the two soils rich in poorly crystalline A 1 and Fe. Due to the fact that the soil samples taken at the beginning of the experiments were not available for the fractionation of P, conclu- sions about the influence of P fertilization on the fractions need to be made by studying the differences in P fractions within an experi- ment at the end of the experiment, and assum- ing that the plots did not differ from one an- other in terms of P fractions at the beginning of the experimental period. Despite these shortcomings, the current results were in ac- cordance with those of Hartikainen (1989a) who found that residual P seemed to accumu- late mainly in the fractions extractable with NH 4F (“Al-P”) and NaOH (“Fe-P”). Low recovery of residual P, observed in the pres- ent study, has earlier been reported by Bar- ber (1979), who discovered that only half of the calculated quantity of residual P was reco- vered when total P was determined at the end 368 of a 25-year field experiment. Among the several processes which contribute to the low recovery of fertilizer P, major effects by ero- sion can be excluded in the present experi- ments due to the flatness of the area, and leaching of P seems unprobable in soils rich in poorly crystalline A 1 and Fe. Net transfor- mation of P into organic phosphates, sug- gested to be important by Wagar et al. (1986), or occlusion of P into inorganic forms not extracted in the fractionation procedure, may explain part of the low recovery. An ad- ditional explanation may be exchange of soil between the plots. In a Danish study by Sib- besen (1986) it turned out that the net import of soil richer in P from the adjacent plots had more than compensated the export of P in the harvested crop in the plots not receiving P fer- tilization in two 90-year field experiments. The plots which according to the experimental de- sign got no P fertilization, have probably gained soil from other plots richer in P also in the present trials. Supplement P entering these plots along with the imported soil ma- terial evidently increased the quantities of in- organic P measured upon fractionation. Since the recovery of residual P was calculated as the difference in the quantities of inorganic P measured in the plots fertilized and not fer- tilized with P, import of soil inevitably results in low apparent recovery of residual P. In experiment B, the difference in yields ob- tained from the plots fertilized and not ferti- lized with P exceeded 10 % only in the last few years. This indicates that major accumu- lation of plant-available residual P had oc- curred in the soil prior to the experiment. In experiment A, the pool of labile P was less ex- tensive, since a response to P fertilization was obtained from the very first experimental year. Even though the fertilization with 13 kg P/ha was sufficient for most of the ex- perimental period, the yields at that P rate were obviously produced at the expense of the labile P reserves of the soil. Fertilization with 13/16 kg P/ha was not able to prevent those reserves from decreasing below the critical level. This was indicated by the fact that dur- ing the last few years in experiment A, the maximum yields were obtained at a higher P level, 32 kg/ha. The present experiments demonstrate that if the soil contains consider- able reserves of plant-available P at the be- ginning of the experiment, the results of the first few years prompt recommendations for P fertilization that are too low in the long run. The long-term requirement of P fertilization in the cultivation of cereal crops in the cur- rent experiments was approximately double the export of P in the grain yields. At that P rate, it was possible to maintain both the max- imum yields and the intensity of P in the soil. Acknowledgment. The author wishes to thank Mr. Auvo Leskelä, former head of Kotkaniemi experimental farm, for the design and management of the field experi- ments. References Anon 1987. Viljavuustutkimuksen tulkinta peltoviljelys- sä. Viljavuuspalvelu Oy. 70 p. Aura, E. 1978. Determination of available soil phospho- rus by chemical methods. J. Scient. Agric. Soc. Finl. 50: 305—316. Barber, S.A. 1979. Soil phosphorus after 25 years of cropping with five rates of phosphorus application. Commun. Soil. Sci. PI. Anal. 10: 1459—1468. Hartikainen, FI. 1979. Phosphorus and its reactions in terrestrial soils and lake sediments. J. Scient. Agric. Soc. Finl. 51: 537—624. 1982. Water soluble phosphorus in Finnish mineral soils and its dependenceon soil properties. J. Scient. Agric. Soc. Finl. 54: 89—98. —1989 a. Effect of cumulative fertilizer dressings on the phosphorus status of mineral soils. I Changes in inorganic phosphorus fractions. J. Agric Sci. Finl. 61: 55—59. —1989 b. II Comparison of two phosphorus testing methods. J, Agric. Sci. Finl. 61: 61 —66. Jankovic, M. & Nemeth, K. 1979. Einfluss langjährig gesteigerter K- und P-Diingung auf die K- und P- Dynamik im Boden sowie auf den Ertrag, Landw. Forsch. 32: 283—291. 369 Li, R.-G. & Barber, S. A. 1988. Effect of phosphorus and potassium fertilizer on crop response and soil fer- tility in a long-term experiment. Fertilizer Res. 15: 123—136. McLaughlin, M.J., Alston, A.M. & Martin, J.K. 1988. Phosphorus cycling in wheat-pasturerotations. I. The source of phosphorus taken up by wheat. Austr. J. Soil. Res. 26: 323—331. Saarela, I. 1989. Fosforilannoitus taloudelliseksi ja ym- päristöystävälliseksi. Käytännön Maamies 38; 3: 36—39. Saari, E. & Paaso, A. 1980. Mineral element composi- tion of Finnish foods. II Analytical methods. Acta Agric. Scand. Suppl. 22: 15—25. Schachtschabel, P. & Beyme, B. 1980. Löslichkeit des anorganischen Bodenphosphors und Phosphatdiin- gung. Z. Pflanzenernähr. Dting. Bodenkd. 143: 306—316. Sharpley, A.N. 1986. Disposition of fertilizer phospho- rus applied to winter wheat. Soil. Sei. Soc. Amer. J. 50: 953—958. Sibbesen, E. 1986. Soil movement in long-term field ex- periments. Plant and Soil 91: 73—85. Sippola, J. & Jaakkola, A. 1980. Maasta eri menetelmil- lä määritetyt typpi, fosfori ja kalium lannoilustarpeen osoittajina astia- ja kenttäkokeissa. Maatalouden tut- kimuskeskus. Maanviljelyskemian ja -fysiikan laitos. Tiedote 13: 24—41. & Saarela, I. 1986. Some extraction methods as indicators of need for phosphorus fertilization. Ann. Agric. Fenn. 25: 265—271. Steel, R.G.D. & Torrie, H.J. 1980. Principles and procedures of statistics. A biometrical approach. 633 p. 2 nd Ed. Singapore. Vuorinen, J. & Mäkitie, O. 1955. The method of soil testing in use in Finland. Agrogeol. Pubi. 63, 44 p. Helsinki. Wagar, 8.1., Stewart, J.W.B. & Mom, J.O. 1986. Changes with time in the form and availability of residual fertilizer phosphorus on chernozemic soils. Can. J. Soil Sci. 66: 105—119. Ms received August 31, 89 SELOSTUS Fosforilannoituksen vaikutus saloon ja maan fosforillaan kahdessa pitkäaikaisessa kenttäkokeessa Markku Yli-Halla Kemira Oy 02271 Espoo Kemira Oy:n Kotkaniemen Koetilalla tehtiin v. 1974—85 hiesusavimaalla kaksi kenttäkoetta (Koe A ja B), joissa tutkittiin erisuuruisten P-lannoitusmäärien vai- kutusta pääasiassa viljakasvien satoon ja maan P-tilaan. Koejaksonalussa oli maassa ammoniumasetaattiliuokseen (pH 4.65) uuttuvaa fosforia 5.5 mg/l (Koe A) ja 6.2 mg/1 (Koe B). Viljavuusanalyysin tulkintaohjeen (Anon 1987) mukaan kokeen A P-tila oli ’’välttävä” ja kokeen B P-tila ’’tyydyttävä”. Fosforilannoitusmäärät olivat v. 1974—82 0, 13, 26, 47 ja 60 kg/ha ja 1983—85 0, 16, 32 56 ja 72 kg/ha. Fosforilannoitus annettiin sijoituslannoi- tuksena koetta varten valmistettuja NPK-lannoitteita käyttäen. Kokeessa A saatiin maksimisato aina yhdek- sänteen koevuoteen asti P-tasolla 13 kg/ha ja viimeisinä koevuosina P-tasolla 32 kg/ha. Kokeen alkupuolella (1974—80) oli maksimisato keskimäärin 470 kg/ha (12%) suurempi kuin ilman P-lannoitusta viljellyiltä ruuduilta saatu jyväsato. Kokeen loppupuolella (1981 —85) ilman P-lannoitusta viljeltyjen ruutujen P-varat ehtyivät siinä määrin, että P-lannoitus tuotti sadonlisäystä keskimää- rin 1360 kg/ha (40 %). Kokeessa B P-lannoituksella saa- tiin pienempiä sadonlisäyksiä, jotkakuitenkin kasvoivat kokeen loppua kohti ollen enimmillään (1983—85) 570 kg/ha (12 %). P-lannoitus alensi puintikosteutta erityi- sesti märkinä vuosina. Ilman P-lannoitusta viljellyillä ruu- duilla asetaattiliuokseen uuttuneet P-määrät alenivat jyr- kästi ollen kokeiden lopussa 1.8 mg/1 (Koe A) ja 2.2 mg/l (Koe B). Asetaattiuuttoisen P:n määrä maassa ei kohon- nut edes runsaimman P-lannoituksen saaneilla ruuduil- la. Maahan jäänyt P oli kertynyt pääasiassa raudan ja alumiinin sitomiin fraktioihin. Optimaalisen P-lannoitus- tason määrittämiseksi on tehtävä riittävän pitkäaikaisia kenttäkokeita. Vain muutamia vuosia kestävien kokeiden perusteella saadaan lannoitussuosituksia, jotkaovat pit- kän päälle liian alhaisia. 370