RESIDUAL VALUE of N; BULLETIN 322 JANUARY 1960 sAGRICULTURAL EXPERIMENT STATION Auburn, Alabama A AU BU RN E. V. Smith, Director U NI V ERS IT Y CONTENTS Page RESULTS OF EXPERIMENTS----------- -- 4 - Residual Effects of Superphosphate on Hartsells Fine Sandy Loam-Residual Effects of Various Sources Residual Effects of Superphosphate Applied to Soils in Cement Bins- - - - - -- GENERAL DISCUSSION---------SU M M A RY - - - - - - - - - - - - - - - - -- - - - - - - - - - - - - - LITERATURE CITED---- --14 --18 --19 --20 AC KNOWLEDGEMENT Field experiments to study the residual value of applied phosphates have been in progress for nearly 30 years.. In preparing this bulletin, the author summarized results of many workers. Studies on the Experiment Fields were conducted by J. T. Cope, Jr., F. E. Bertram, Fred Glaze, T. Williamson, 2 J. R. Taylor,' and R. W. Taylor.' Studies the Substations were conducted by John Boseck, R. C. Christopher,' S. E. Cissendanner, Fred Stewart, 3 and J. P. Wilson. 2 1 Resigned 2 Deceased J. W. Richardson, J. F. Segrest, Jr.,' J. on 3Retired COVER PHOTO. These plots at the Sand Mountain Substation in 1957 show effects of residual'phosphorus on growth of cotton. Plot at left had received no phosphorus since 1930. One at right got applications of superphosphate from 1930 until 1955. FIRST PRINTING 5M, JANUARY 1960 RESIDUAL VALUE of PHOSPHATES L. E. ENSMINGER, Soil Chemist ADDED PHOSPHORUS accumulates to some extent in soils. This occurs because only a small percentage of phosphorus applied in fertilizers is removed in the harvested portion of crops or lost by leaching. Extent of accumulation depends on such factors as amount of phosphorus added, amount lost by erosion, crop grown, and how much of the crop is removed from the land. Loss of phosphorus by erosion and runoff has been recognized as an important factor in reducing the accumulation of applied phosphates (1, 2, 6, 10). In a study of nutrient losses from a 3-year rotation on Dunmore silt loam, Rogers (6) found that the eroded material from corn land was richer in nitrogen and phosphorus than the original soil. Removal of large tonnages of hay or silage may result in the removal of as much as 50 or more pounds of P - 0 5 per acre per season. The amount of P 20 5 removed depends on the crop and the amount of growth removed, as shown by the following examples: Crop Cropounds Cotton (1,500 pounds seed cotton) Corn (60 bushels grain) Alfalfa (4 tons hay) Coastal Bermudagrass (6 tons hay) P205 removed, per acre 18 23 46 54 The residual value of applied phosphates has been studied by a number of investigators (5, 10, 12, 13). Volk (10) found that cotton yields declined on a Hartsells fine sandy loam when phosphorus fertilization was discontinued after various rates had been applied over a 5-year period. However, there was a residual effect that was in proportion to the amount that had been added. 4 ALABAMA AGRICULTURAL EXPERIMENT STATION Data reported by Weeks and Miller (13) showed that crop yields declined with time on plots that had formerly received superphosphate, whereas there was no decline on plots that had previously received 4 times as much phosphorus from rock phosphate. The availability of accumulated phosphorus for succeeding crops is of practical importance. Soil tests can be used to determine the amount of residual phosphorus in soils and from these tests phosphorus needs can be predicted. Since available phosphorus levels of soils vary a great deal because of past management, a soil test is the only practical method of evaluating residual phosphorus present in soil of a particular field. Now that many soils have accumulated rather large quantities of residual phosphorus, phosphorus fertilization should be based on this fact. Chemical fixation of phosphorus by soils has been credited with the low-efficiency often obtained (7, 9). Fixation, which is the reversion to a less soluble form, is undoubtedly responsible for some loss in availability of phosphorus. However, yield data and radiophosphorus uptake have shown that fixed phosphorus is fairly available to plants (3). This bulletin summarizes pertinent data obtained in Alabama showing residual effects of applied phosphorus. RESULTS OF EXPERIMENTS Residual Effects of Superphosphate on Hartsells Fine Sandy Loam FIRST PERIOD (1930-55). This experiment was started in 1930 at the Sand Mountain Substation to determine residual effect of rates of superphosphate as measured by yields of cotton in continuous culture. The experiment is still in progress in revised form and is perhaps the oldest residual phosphorus test in existence. Rates of P 20 5 of 0, 30, 60, 90, and 120 pounds per acre were applied annually to certain plots during 1930-34 and none thereafter. Figure 1 shows the residual effects of these rates as measured by yields of seed cotton. Yields gradually declined after phosphorus was discontinued in 1934, but residual effects were evident for 21 years and were in proportion to the amount of phosphorus that had been applied. For the last 6-year period (1950-55) the treatment that received 120 pounds of P 2 0 5 an- RESIDUAL VALUE of PHOSPHATES 5 nually during 1930-34 and none thereafter averaged 640 pounds of seed cotton more than the treatment that had not received any phosphorus. However, yields from the 120-pound residual treatment averaged 320 pounds less than the one that received 60 pounds of P20 5 annually. SECOND PERIOD (1956-57). The experiment was revised in 1956 so that residual effects of rates of superphosphate applied anYield of Seed Cotton, Pounds per Acre - 160C- p (1) (2) 60 60, 1400- I00 - (1)90,(2)0 800 -(1)60, 600.()30,(2) 400 Lbs. P2 05per acre S applied annually =1930 -34 (1) Check (2)0 (2)=1935-55 200- 1930-34 1935-39 1940-44 Years 1945-49 1950-55 FIGURE 1. Response of cotton to residual phosphorus on Hartsells fine sandy loam is shown above by yields of seed cotton. 6 TABLE 1. ALABAMA AGRICULTURAL EXPERIMENT STATION RESIDUAL EFFECTS OF SUPERPHOSPHATE COTTON AND EXTRACTABLE AS MEASURED BY YIELDS OF HARTSELLS VERY SEED PHOSPHORUS, FINE SANDY LOAM P20 5 per acre applied as superphosphate 1930-34 1935-55 1956-57 Lb. 0 0 0 0 0 30 60 90 120 Lb. 0 30 60 90 120 30 60 90 120 Lb. 0 0 0 0 0 30 60 90 120 Seed cotton yields 1956-57 Lb. 1,081 1,555 1,710 1,850 1,845 1,668 1,770 1,812 1,802 P20 5 extracted from soil samples Dilute acid Neutral NH 4F HC-- Resin exNHF change p.p.m. 6 17 34 68 96 17 38 65 102 11 38 76 126 193 37 77 132 224 p.p.m .p.m p p.p.m. 21 65 66 174 120 264 454 258 378 605 169 63 135 280 280 451 429 662 nually for as long as 21 years could be studied. Yield data for the first 2 years after revision are given in Table 1. Plots that received as much as 60 pounds of P 2 0 5 annually for 21 years and none thereafter produced as much cotton as plots that continued to receive 60 pounds annually. The soluble phosphorus content of soil samples collected at beginning of the second phase of the residual study in 1956 was determined by four methods of extraction, Table 1 and Figure 2. Although the amount extracted by the various solutions varied considerably, any one of them could be used as a measure of available soil phosphorus if calibrated in terms of yield response to residual phosphorus. RESIDUAL VALUE of PHOSPHATES 7 PPM Po 0 Extracted Pounds per Acre P2 0 5 Applied Annually 1930-55 FIGURE 2. Shown is extractable phosphorus content of Hartsells fine sandy loam at Sand Mountain Substation when treated with varying amounts of superphos- phate. Residual Effects of Various Sources CORN-COTTON ROTATION WITHOUT WINTER LEGUMES. A sources- of-phosphorus test was conducted at four locations from 1930 to 1945. Sources of phosphorus along with rates applied and locations are given in Table 2. Phosphorus treatments were discontinued after 1945 to study residual effects. Yield data are presented for the last 4 years of the period during which phos- TABLE 2. INCREASED YIELDS OF SEED COTTON FROM VARIOUS PHOSPHATES FOR THE LAST 4 YEARS OF THE PHOSPHATING PERIOD AND FOR THE 4-YEAR RESIDUAL PERIOD IN A ROTATION OF COTTON AND CORN Source of phosphorus Increased yield of seed cotton P 05 2 per acre Norfolk si, Magnolia fsl, Decatur si, Decatur ci, Average of applied Wiregrass Monroeville Tennessee ValAlexandria 4 locations Expt. Field ley Substation Expt. Field to cotton Substation 1930-45 1942-45 1946-49 1942-45 1946-49 1942-45 1946-49 1942-45 1946-49 1942-45 1946-49 r~I. .SI- Superphosphate------------------- Superphosphate Conc. superphosphateBasic slag---------------- --------------------- - Lb. 24 48 48 48 Lb. 199 216 205 157 Lb. 104 174 29 249 Lb. 499 480 363 420 Lb. 378 461 171 Lb. 367 360 846 437 Lb. 174 265 Lb. 520 652 Lb. 860 427 Lb. 896 427 Lb. 254 332 0 Ppt. trical. phos.----------Ammo-phos. A.------------Colloid. phos------------. Rock phosphate----------Rock phosphate Superphosphate'---2 --------------.-- 48 48 48 48 96 24 173 -471 134 42 107 233 169 -223 99 58 133 136 411 -225 348 358 338 432 571 312 -102 155 284 285 261 196 372 624 438 409 407 345 210 252 170 814 327 219 266 209 119 230 224 -399 --486 679 662 -417 -281 418 435 _ -22 ----264 359 413 __ 89 185 266 264 cc c F CC cm ir - Superphosphate ----------Average yield of checks.---- 24 0 243 1,029 256 503 504 849 518 870 439 890 332 1,011 626 777 474 727 453 886 395 778 m m ' In addition to superphosphate, rock phosphate applied at rate of 2,000 pounds per acre in 1930, 1936, and 1942. 2 In addition to superphosphate, basic slag applied at rate of 2,000 pounds per acre in 1930, 1936, and 1942. z -I z m C r- TABLE 8. INCREASED YIELDS OF SEED COTTON FROM VARIOUS PHOSPHATES FOR THE LAST 4 YEARS OF THE PHOSPHATING PERIOD AND FOR THE 4-YEAR RESIDUAL PERIOD IN A ROTATION OF COTTON AND CORN WITH WINTER LEGUMES Source of phosphorus P 2 05 per acre to cotton and to winter Increased yield of seed cotton per acre Norfolk sl, Wiregrass Substation Greenville fsl, prattville Experiment Field 1946-49 Lb. 60 78 67 180 160 -370 172 54 96 82 109 1,192 Decatur sl, Tennessee Valley Substation 1942-45 1946-49 Lb. Lb. 405 255 429 283 328 188 582 457 332 201 331 265 270 145 219 128 249 341 884 244 506 329 1,026 973 3Averageof r" a 0 *I CA m > g legumes 1930-45 1942-45 Lb. Lb. 24 302 376 48 48 238 804 48 48 273 48 -3829 48 254 48 112 96 188 24 430 24 416 0 927 1946-49 1942-45 Lb. Lb. 157 183 179 161 95 209 382 155 220 228 -56 -416 98 160 124 62 151 215 206 206 352 227 519 1,632 1942-45 Lb. 288 322 258 347 278 -138 207 131 227 340 883 1,195 1946-49 Lb. 166 180 117 340 194 -161 159 102 187 177 263 895 17 0 "H Superphosphate Superphosphate Cone. superphosphate Basic slag Ppt. trical. phos. Ammo-phos. A Colloid. phosphate Rock phosphate ------Rock phosphate Superphosphate Average yield of checks 1 2 Superphosphate' 2 In addition to superphosphate, rock phosphate applied at rate of 2,000 pounds per acre in 1930, 1936, and 1942. In addition to superphosphate, basic slag applied at rate of 2,000 pounds per acre in 1930, 1936, and 1942. 10 ALABAMA AGRICULTURAL EXPERIMENT STATION phates were applied as well as for the residual period, Table 2. Average yields for the four locations show a response to only 24 pounds of P20 5 from superphosphate for the last 4 years of the phosphating period. Rock phosphate did not produce as much cotton as did 24 pounds of P2 0 5 from superphosphate even when applied at the rate of 96 pounds of P2 0 5 per acre. During the residual period, increased yields of cotton were less in most cases than during the phosphating period, but basic slag showed the greatest residual effect. In the residual period 24 pounds of P2 0 5 from superphosphate produced about the same amount of cotton as 24 pounds of P20 5 from superphosphate plus periodic applications of rock phosphate. CORN-COTTON ROTATION WITH WINTER LEGUMES. A sources-ofphosphorus experiment similar to the preceding one was conducted at three locations from 1930 to 1945. In this experiment winter legumes were grown after cotton as a green manure crop for corn. Phosphorus was applied to winter legumes as well as to cotton. The phosphorus treatments were discontinued after 1945 and residual effects of sources were studied. Since corn did not respond much to phosphorus, corn yields are not given. Increased cotton yields for the last 4 years of the phosphating period show that basic slag produced the highest yields, superphosphate the next highest, and raw phosphates considerably less that the other two sources, Table 3. Yields declined during the residual period, but slag showed the greatest residual effect of any of the sources, Figure 3. Blue lupine was used as the winter legume on the Norfolk sandy loam soil at the Wiregrass Substation. Since it did not respond to phosphorus, yields of winter legumes from this location are not presented. Yields of vetch are given for the other two locations in Table 4. Yields for the last 4 years when phosphate was being applied show that 24 pounds of P2 0 5 from superphosphate produced more vetch than 96 pounds of P20 5 from rock phosphate. Also 24 pounds of P20 5 from superphosphate produced more vetch than 48 pounds from basic slag, tricalcium phosphate, or colloidal phosphate on Greenville fine sandy loam. On Decatur silt loam, basic slag and tricalcium phosphate produced about the same amount of vetch as an equivalent amount of phosphorus from superphosphate. However, basic slag showed a greater residual effect than other sources. On the Greenville sandy loam, some of the other less soluble sources, RESIDUAL VALUE of PHOSPHATES -vx P .. Fla: to r -' phi, piolus on Gr-cnviilc fine sandy loam at Prattville Experiment Field is shown in the photos mode March 21, 1950. Plot at top had received no phosphate since 1930. One at bottom got 48 pounds P 0. per acre annually from basic slag during 1930-45 and none thereafter. such as tricalcium pho~sph~ate anid colloidal phosphate, produ ced more v etch than an eqivxalent amount of phosphorus from superphosphate. Treatments of 24 pounds of PA)0 froml sulperphosphate plus peiodic applications of b~asic slag or rock phosphate showed rather high residual effects oii 1both soils. Radiophosphiorus uptake hy cotton plants andi extractab~le phosphoris content of soil samfples were used to measure the extent and ax ailahility of accuulated phosphorus. The radiophosphorus tiptake data were used to calculate A v aluies (4). Since "tagged" suiperphosphate was used as the standard, A v aliues indicate the anmount of soil phosphorus that was as av ailab~le as superphosphate. Based on1 these v alues, Table 5. all rates and sources 12 TABLE 4. ALABAMA AGRICULTURAL EXPERIMENT STATION 4 INCREASED YIELDS OF VETCH FROM VARIOUS PHOSPHATES FOR THE LAST YEARS OF THE PHOSPHATING PERIOD AND FOR 4-YEAR RESIDUAL PERIODS IN A ROTATION OF COTTON AND CORN WITH WINTER LEGUMES Source of phosphorus P 20 5 per acre to cotton and winter Green weight of vetch, per acre Greenville fsl, Prattville Experiment Field 1943-46 1947-50 1951-54 Lb. Lb. Lb. 3,981 6,024 4,809 9,232 8,019 1,629 8,834 5,528 7,931 8,846 10,168 8,211 Decatur sl., Tenn. Valley Substation 1943-46 1947-50 Lb. Lb. 5,626 7,768 7,065 8,050 7,042 7,396 4,847 2,696 5,155 7,865 10,073 1,735 2,117 4,665 3,018 7,628 3,644 3,287 3,179 2,578 3,496 6,224 6,235 3,518 legumes 1930-45 Lb. 5,307 24 Superphosphate 48 6,118 Superphosphate Cone. superphosphate ......48 4,997 Basic slag 48 3,431 Ppt. trical. phos................. 48 4,845 Ammo-phos. A. 48 4,200 Colloidal phosphate------48 4,597 Rock phosphate .____________. 48 3,943 Rock phosphate ............... 5,067 96 1 Superphosphate 24 6,078 Superphosphate2 .24 5,000 Average yield of checks 0 4,444 4,073 -................ 5,666 -............... 4,703 7,284 8,133 1,331 7,384 4,541 7,361 6,278 8,766 7,432 1 In addition to superphosphate, rock phosphate applied at rate of 2,000 pounds per acre in 1930, 1936, and 1942. 2 In addition to superphosphate, basic slag applied at rate of 2,000 pounds per acre in 1930, 1936, and 1942. resulted in an increase in available phosphorus for both soils. In case of the Greenville soil, A values showed that superphosphate had a greater residual availability than rock or colloidal phosphate applied at same rate of P20 5 but not as great as basic slag. A values for the Decatur soil indicate that superphosphate had about the same residual availability as tricalcium phosphate and rock phosphate at same rate but less than basic slag. Residual availability of rock phosphate as measured by A values increased with increasing rates applied. The extractable phosphorus data on samples collected at end of the residual period, Table 5, show that all sources applied at 48 pounds of P20 5 or more per acre annually from 1980 to 1945 resulted in some accumulation over unphosphated plots. Even the 24-pound rate of superphosphate resulted in some accumulation in the Greenville soil. Amount of phosphorus extracted by the three solutions varied with source. For example, neutral NH 4F solution extracted more phosphorus from the soil that had previously received 48 pounds of P 20 5 from superphosphate than did dilute acid solution. However, the reverse was true for soils that had previously received an equivalent amount of P2 0 5 from RESIDUAL VALUE of PHOSPHATES 13 o 00 ~~00c0 0 0c11 z 0 / d 4) 4)1~ U~ z -t0 co q1 - v,) 0 00 'to 0.cbofc ZQ)I~ - 4 ~ ~ in4) Co o w0 H C4 o~ ~CIt0 -40C0 cO0I -- 0 tcco0I i4 - 7 r-,-r--, N-O E ol C F4 V U, W A a 00C0000 0 4 ocSo CO CD OOl10 10 cq1404) o ) 14) 0b) ci00000 U, r4 00-~c' r-000 O -oIr P4 e +; QUrIC~ 4 l P- CT ]C I:I 44 O: :I ):I 4 0 A, 0P4 4) 4-JO ifiC Ph p4 4c 3 O UC) C J . P O 4 QP4P144 -d P 14 ALABAMA AGRICULTURAL EXPERIMENT STATION rock phosphate. Evidently, much of the rock phosphate remained in the soil in unreacted form and was easily acid-soluble. Thus, values based on acid extractants are likely to be high in terms of residual availability where rock phosphate has been used. Soil samples were analyzed for HC1-NH4F-soluble phosphorus, since previous data (1) showed that such an extraction gave a good measure of the quantity of phosphorus that had accumulated in sandy soils. This extraction indicates that only 56 per cent of the phosphorus from the 48-pound rate of superphosphate applied to Greenville soil had accumulated. Loss of phosphorus by erosion (1) is undoubtedly a big factor in lowering the efficiency of applied phosphorus. Residual Effects of Superphosphate Applied to Soils in Cement Bins Surface soil from Norfolk sandy loam, Eutaw clay, and Cecil sandy loam was brought to Auburn in 1934 and placed in 11- X 6-foot cement bins to a depth of 8 inches. From 1984 to 1947 the bins were used for rates of phosphorus and residual studies with vegetable crops (12). This was followed by a study of residual value of phosphorus for cotton from 1948 to 1950. During 1954 and 1955 the bins were used to study the relationship between TABLE 6. RESIDUAL EFFECTS OF SUPERPHOSPHATE AS INDICATED BY YIELDS OF SEED COTTON GROWN IN CEMENT BINS, 1948-50 P205 per acre a p p li ed annually Average per acre yields of seed cotton Norfolk Eutaw cay Lb. 1,318 1,869 1,975 2,055 2,172 2,125 1,829 1,935 2,019 1,870 2,155 269 191 Cecil loam Lb. 326 1,058 1,433 1,588 1,642 1,659 810 1,275 1,425 1,851 1,685 475 3386 Dilute acid-soluble P205 per acre Norfolk loam Lb. 42 67 96 148 176 304 62 82 102 152 254 Eutaw clay Lb. 32 76 214 360 502 548 29 62 114 188 309 Cecil loam Lb. 18 53 116 221 8337 478 34 79 141 208 296 1934-42 1948-50 19447 Lb. Lb. Lb. 0 20 40 60 80 80 0 0 0 0 0 0 0 80 40 160 80 240 120 320 160 3202 160 80 0 160 0 240 0 320 0 3202 0 L. S. D. 0.01 0.05 1 2 loam Lb. 353 653 871 1,109 1,219 1,508 473 605 766 832 1,378 358 254 Norfolk soil received only half as much P 20 5 for the first 2 periods. Limed in 1934 and in 1939. Norfolk received 2,995 pounds of lime, Eutaw 5,580, and Cecil 2,617. RESIDUAL VALUE of PHOSPHATES 15 soluble phosphorus as determined by soil test methods and Ladino clover yields (14). COTTON YIELDS. Average yields of seed cotton for the three soil types are given in Table 6. All three soils showed an appreciable response to applied phosphorus. Residual effects from superphosphate applied from 1934 to 1942 were also appreciable and were in proportion to the amounts that had been applied. The highest residual phosphorus treatments produced 68, 82, and 86 per cent as much cotton as an 80-pound annual P20 5 treatment for Norfolk, Cecil, and Eutaw, respectively. Where lime had been applied, yields from residual phosphorus were 91, 100, and 100 per cent for Norfolk, Cecil, and Eutaw, respectively. It should be pointed out that the Norfolk received only half as much phosphorus as the other two soils before phosphorus was discontinued. LADINO CLOVER YIELDS. In the fall of 1958, 1 ton of lime per acre was applied to the Cecil and Norfolk bins and 2 tons to the Eutaw bins. Four bins of each soil that had received the highest rate of superphosphate during the period 1934 to 1950 received 160 pounds per acre P20 5 annually from superphosphate in 1953 and 1954. Yields from the phosphated bins represent 100 per cent yields and all other yields are relative to those from the phosphated bins. Ladino clover was seeded in October 1953 and clippings were made for 2 years. Relative yields of Ladino clover, Table 7, showed appreciable TABLE 7. RESIDUAL EFFECTS OF SUPERPHOSPHATE AS INDICATED BY RELATIVE YIELDS OF LADINO CLOVER GROWN IN CEMENT BINS, 1954-55 Relative yields (phosphated bins = 100)2 applied annually Norfolk Eutaw Cecil 1934-42 1943-47 1948-50 sandy claw clay loam clay loam Lb. Lb. Lb. Pet. Pct. Pet. 1 194-42 1948-50 194-47 0 40 80 120 0 0 0 0 0 0 20 40 60 0 0 0 0 0 P20O per acre by soil test Norfolk Eutaw sandy Eutaw loam clay Lb. Lb. 64 104 183 258 77 121 190 248 287 18 43 138 260 33 64 131 203 153 Cecil clay loam Lb. 17 44 98 229 29 57 115 191 229 9,394 0 80 160 240 80 160 240 320 320 1 46 65 86 97 54 54 76 80 95 18 68 79 88 44 75 78 84 95 4 40 68 78 25 52 69 74 88 Norfolk soil received only half as much P2 0 for first 2 periods. 100 per cent yields are: Norfolk = 9,954, Eutaw - 10,691, and Cecil pounds of dry matter for the 2-year period. 2 16 ALABAMA AGRICULTURAL EXPERIMENT STATION residual effects of past applications of superphosphate. In general, residual effects as measured by clover yields were greater for the Norfolk and Eutaw soils than for the Cecil. On the Norfolk soil almost as much clover was produced by the high residual phosphorus treatment as by the treatment that received 160 pounds P 2 0 5 from superphosphate. SOIL TEST VALUES. For residual phosphorus to be used efficiently, it is necessary to have a rapid chemical method of assaying availability of accumulated phosphorus. If chemical extraction data are to have any value in this regard, they must be correlated with yield response. The extractable phosphorus data presented in Tables 6 and 7 show that past phosphorus treatments are reflected by the amount of phosphorus extracted and yields of cotton and clover are related to past applications of phosphorus and to extractable phosphorus. Soil samples from the clover experiment were analyzed for extractable phosphorus by three different solutions. The results were related to relative yields of clover as shown in Table 8. The amounts of soil test P20 5 required to give specified relative yields were calculated for five classes. The 0.03N NH 4F + O.1N HC1 solution extracted much more phosphorus than either of the other two solutions. However, the degree of correlation between TABLE 8. DIVISION OF SOIL TEST VALUES FOR PHOSPHORUS INTO CLASSES ON THE BASIS OF SOIL TEST REQUIRED TO GIVE SPECIFIED RELATIVE YIELDS OF LADINO CLOVER Soil type Relative 25 Amount of P20 5 /A. required as determined by: NaHCO, HC1+ H 2SO4 NH 4F + HCi 32 33- 77 78-189 190-817 > 317 15 G 29 144 Cecil 26-50 51-75 76-90 >90 25 30- 69 70-167 168-282 > 282 - 145- 312 313- 678 679-1081 > 1081 - Eutaw 16 80 26-50 51-75 76-90 90 e25 16- 41 42-112 118-203 203 32 33- 65 65-128 129-193 > 193 17- 43 44-116 117-210 210 - 81- 198 199- 491 492- 846 846 190 36 Norfolk 26-50 51-75 76-90 > 90 37- 75 76-157 158-246 > 246 191- 365 366- 701 702-1037 > 1037 SFor Cecil, Eutaw, and Norfolk, 100 per cent yields were 9,394, 10,691, and 9,954 pounds of dry matter per acre for the 2-year period, respectively. RESIDUAL VALUE of PHOSPHATES 17 Relative Yields Pounds P2 0 5 Per Acre By Soil Test FIGURE 4. Division of soil test values into classes on basis of values required to give specific relative yields of Ladino clover is presented in the graph. Data are from Table 8 for Cecil clay loam and 0.05N HCI + 0.025N H2SO solution. 4 amounts extracted and relative yields is the important thing. All three solutions gave a high degree of correlation with relative yields. This points out the necessity of correlating chemical soil test values with yields if chemical tests are to have value in making fertilizer recommendations. A graphical presentation of the classes for the 0.05N HCl + 0.025N H 2SO4 extraction for the Cecil soil, taken from data in Table 8, is given in Figure 4. These data show that residual phosphorus influences clover yields and that yields are related to chemical soil tests. However, for a complete evaluation of soil 18 ALABAMA AGRICULTURAL EXPERIMENT STATION test data, response of clover to rates of phosphorus on soils with various residual phosphorus levels should also be known. With this information at hand, it would be possible to predict the amount of phosphorus that should be added to raise relative yields to the desired level. GENERAL DISCUSSION The phosphorus status of Alabama soils has changed considerably as a result of past fertilization. Most soils of the State were deficient in available phosphorus when first cleared and as a result phosphorus fertilization was generally needed. Now the soils range from low to very high in available phosphorus, depending on past treatment. This means that a general recommendation will not fit as many of the soils now as it did earlier. Even though processed phosphates become chemically fixed when applied to soils, yield data have shown that the accumulated phosphorus is of considerable value and that it should be considered in making fertilizer recommendations. Since it is impractical to determine crop response on every field, soil testing appears to be the most practical way of making best use of residual phosphorus. Numerous extraction methods can be used to give an index of available phosphorus if properly calibrated against yield response in the field. Soil testing should be considered as a means of extending the usefulness of field data. It is doubtful if any soil testing method will give an exact measure of available phosphorus in a soil. However, to know that a soil is either low, medium, or high in available phosphorus is helpful in making fertilizer recommendations. It is emphasized that it is seldom advisable to discontinue phosphorus fertilization for most crops even though the soil test for phosphorus is high. It is often suggested that raw phosphates may have a greater cumulative or residual effect than processed phosphates. Results from experiments conducted for a period of 16 years do not indicate any appreciable cumulative effect for raw phosphates. Results for a 4-year residual period varied with location. In the case of a Greenville fine sandy loam, the residual effects of raw phosphates in some instances exceeded those of an equal amount of P2O5 from superphosphate. On a Decatur silt loam, superphosphate was superior to raw phosphates during the phosphating RESIDUAL VALUE of PHOSPHATES 19 period as well as during the residual period. The residual effect of basic slag was high on both Greenville and Decatur soils. Since basic slag is also a liming material, it is difficult to determine how much of the residual effect resulted from phosphorus and how much from lime. SUMMARY Numerous field tests have been conducted since 1930 to determine the residual value of phosphorus in terms of crop yields. In many cases extractable phosphorus was correlated with yields for calibrating soil test methods. Radiophosphorus has also been used to evaluate the availability of accumulated phosphorus in soils. Results of residual phosphorus studies to date are summarized as follows: 1. Soil analysis data showed that applied phosphorus accumulated in soils and the extent of accumulation was in proportion to the amount applied. 2. Yields of cotton, vetch, and Ladino clover showed residual effects that were directly related to past phosphate fertilization. Where moderate amounts of phosphorus had been applied, crop yields usually decreased when application was discontinued. Where high amounts of phosphorus had accumulated, yields were not reduced much when phosphorus was discontinued. 3. Considerable residual effects were obtained from all sources studied as measured by crop yields. For any particular source, the residual effect was directly related to the amounts that had been added. Basic slag gave the greatest residual effect of any of the sources. 4. Even though accumulated phosphorus is chemically fixed by soils, yield data show that it is of considerable value in crop production and should be considered in making fertilizer recommendations. 5. Extractable phosphorus content of soils was directly related to yield response to residual phosphorus. This relationship is the basis for making phosphorus fertilizer recommendations by soil test. LITERATURE CITED (1) L. E. Loss of Phosphorus by Erosion. Soil Sci. Soc. Amer. Proc. 16:338-842. 1952. (2) ENSMINGER, L. E. AND COPE, J. T., JR. Effect of Soil Reaction on the Efficiency of Various Phosphates for Cotton and on Loss of Phosphorus by Erosion. Jour. Amer. Soc. Agron. 39:1-11. 1947. ENSMINGER, (3) ENSMINGER, L. E. AND PEARSON, R. W. Residual Effects of Various Phosphates as Measured by Yields, p-32 Uptake, and Extractable Phosphorus. Soil Sci. Soc. Amer. Proc. 21:80-84. 1957. (4) FRIED, M. AND DEAN, L. A. A Concept Concerning the Measurement of Available Soil Nutrients. Soil Sci. 73:268-271. 1952. (5) JONES, U. S. Phosphate Fertilizers. Miss. Agr. Expt. Sta. Bul. 508. 1953. (6) ROGERS, H. T. Plant Nutrient Losses by Erosion From a Corn, Wheat, Clover Rotation on Dunmore Silt Loam. Soil Sci. Soc. Amer. Proc. 6:263-271. 1942. (7) SCARSETH, G. D. The Mechanism of Phosphate Retention by Natural Alumino-Silicate Colloids. Jour. Amer. Soc. Agron. 27:596-616. 1985. (8) SCARSETH, G. D. AND CHANDLER, W. V. Losses of Phosphate from a Light-Textured Soil in Alabama and Its Relation to Some Aspects of Soil Conservation. Jour. Amer. Soc. Agron. 80:361-374. 1938. (9) SCARSETH, G. D. AND TIDMORE, J. W. The Fixation of Phosphates by Clay Soils. Jour. Amer. Soc. Agron. 26:152-162. 1984. (10) VOLK, G. W. Response of Residual Phosphorus of Cotton in Continuous Culture. Jour. Amer. Soc. Agron. 87:330-40. 1945. (11) (12) (13) (14) WARE, L. M., BROWN, OTTO, AND YATES, HAROLD. Residual Effects of Phosphorus on Irish Potatoes in South Alabama. Proc. Amer. Soc. Hort. Sci. 41:265-269. 1942. L. M. AND JOHNSON, W. A. Phosphorus Studies with Vegetable Crops in Different Soils. Ala. Agr. Expt. Sta. Bul. 268. 1949. WEEKS, M. E. AND MILLER, H. F. The Residual Effects of Phosphates Used in Long-Term Field Experiments. Soil. Sci. Soc. Amer. Proc. 183:102-107. 1949. WARE, WELCH, L. F., ENSMINGER, L. E., AND WILSON, C. M. The Correlation of Soil Phosphorus with the Yields of Ladino Clover. Soil Sci. Soc. Amer. Proc. 21:618-620. 1957.