LONG-TERM -, FERTILITY EXPERIMENTS ON COTTON, CORN, SOYBEANS, SORGHUM, AND PEANUTS, 1929 - 1982 BIUt1F II 561 ALGdS1 1984 - L 4 \L ABAM \ AGRICU . IURAL EXPE RIMENT S"IATION AUBURN UNIVFRSITY ( \L F A. B1 UHANAN , DIRECTOR AUBURN t NIVERSITY, ALABAMA CONTENTS Page INTRODUCTION.............................................3 Soil-Test Calibration Research........................ PROCEDURES...............................................5 4 Two-Year Rotation Fertilizer Experiments (TYR)........ 5 Rates of NPK Experiments (NPK) .................... 6 RESULTS AND DISCUSSION............................ ............ 7 SAND MOUNTAIN SUBSTATION, HARTSELLS FSL ........ 8 Cotton............................................ 8 Soybeans.......................................... 12 Corn ............................................. 13 Wheat ............................................ 13 Grain Sorghum and Sweet Sorghum................... 14 BREWTON EXPERIMENT FIELD, BENNDALE FSL ... .......... .11 Cotton................................................18 Soybeanls..............................................18 Corn .................................................. 19 Wheat, Grain Sorghum, and Sweet Sorghum............. 19 WIREGRASS SUBSTATION, DOTHAN FSL ........ .... .. ...... .20 Cotton................................................20 Peanuts...............................................24 Corn................................................. 24 Soybeans ............. ................................. 25 .. ..... Wheat and Sweet Sorghum ........................... MONROEVILLE EXPERIMENT FIELD, LUCEDALE: FSL .......... Cotton................................................29 Soybeans..............................................29 Corn ................................................. Wheat and Grain Sorghum............................. PRATT VILLE EXPERIMENT FIELD, LUCEDALE SCL ....... ..... Cotton..........................................30 Soybeans........................................ Corn................................................. Wheat and Grain Sorghum............................. TENNESSEE VALLEY SUBSTATION, DEWEY SIL..... .... ....... Cotton................................................35 Soybeans..............................................39 Corn ................................................. 25 25 29 30 30 34 35 35 Wheat and Grain Sorghum ........................... UPPER COASTAL PLAIN SUBSTATION, SAVANNAH 39 39 40 SCL ......... Cotton ........................................... Soybeans.......................................... Corn.................................. SUMMARY AND CONCLUSIONS ... ...... .... .... .... ........... ........... .. 40 40 43 43 Responses to Nitrogen................................. Responses to Phosphorus............................... Responses to Potassium................................ Responses to Lime.................................... Responses to Magnesium anld Sulfur..................... Rejuvenating Wornout Soils........................... FIRST PRINTING 47 48 50 51 51 52 ACKNOWLEDGMENT........................................54 5;M, AUGUSTT 1984 LONG-TERM FERTILITY EXPERIMENTS ON COTTON, CORN, SOYBEANS, SORGHUM, AND PEANUTS, 1929 - 1982 J. T. Cope' INTRODUCTION A MAJOR ACTIVITY of the substation and experiment field system established in Alabama in 1928 has been research on response of crops to fertilizers. Funds provided by the 1928 legislature provided for the establishment of five substations and 10 experiment fields on the major soil types in the State. These stations concentrated on research with fertilizers during the 1930's and 1940's. The number of substations has been increased to 11 and experiment fields reduced to three over the years. Fertility research continues to be a major part of programs on many of these stations. Fertility research on experiment stations and cooperative experiments on farmers' fields served as the basis for general fertilizer recommendations for the different soil associations through the 1930's and 1940's. It was recognized, however, that soil fertility was changing during this period due to fertilization, crop removal, and management practices. General fertilizer recommendations based on soil type were no longer accurate because phosphorus (P) tended to build up in the soil under high rates of fertilization while nitrogen (N) and potassium (K) , which were applied at lower rates, were often depleted. Much research was done in field experiments and in laboratories to develop chemical soil tests that could be used to measure soil fertility and on which to base fertilizer recommendations. This resulted in establishment of the Alabama soil testing program in 1953. 'rofessor Department of Agronomy and Soils. 4 ALABAMA AGRICULTURAL EXPERIMENT STATION Soil-Test Calibration Research Fertilizer recommendations from soil tests in Alabama are based on numerous experiments on substations and experiment fields of the Alabama Agricultural Experiment Station and on farmers' fields. Both yield-response data and soil-test data from experiments are necessary for calibration. Best data for soil-test calibration have been found to come from long-term experiments where treatments were repeated on the same plots over a period of years. This was also necessary to accurately determine rates of individual nutrient elements to which specific crops would respond. These experiments also make it possible to determine effects of the treatments on levels of fertility in the soil. Experiments with rates of fertilizer-nutrient elements have therefore been maintained over long periods to gain reliable soil-test calibration and yield-response data. Data from two sets of long-term fertility experiments which have been conducted at six and seven locations are reported here. The oldest of these is known as the Two-Year Rotation Fertilizer Experiment (TYR). It was started in 1929 and has run continuously with many revisions in treatments and in crops used in the rotation. The second set of experiments, known as the Rates of NPK Experiment (NPK), was started in 1954 at the same six locations plus the Upper Coastal Plain Substation at Winfield. This experiment has been in cotton, corn, soybeans, and sorghum at different periods from 1954 through 1982. Data from both experiments were published in 1970 in Experiment Station Circular 181. Procedures used in both experiments are described briefly, and yield results and soil-test data are discussed by location, beginning with soils that were most responsive to fertilization. Data from the two experiments at each location are presented in separate tables. For easier comparisons among the treatments, yields are given as percentages of that of the standard treatment, which received what was considered to be adequate or optimum rates of all nutrient elements. Yields of the standard treatment are presented in bushels, pounds, or tons per acre. To convert percentages to actual yields for the other treatments, multiply percentages shown by yield of the standard treatment for the different crops. Yield data have been subjected to the analysis of variance procedure to determine which differences are significant at the 5 percent level. Percentages marked by an asterisk are different from LONG-TERM FERTILITY EXPERIMENTS 5 the next lower or higher treatment with which they should be logically compared at the 19 to 1 probability level. This means that the odds are 19 to 1 that the differences were due to the treatment and not to commonly occurring variation among plots. Soil-test values for P and K are also presented. For the TYR, soil-test and index values from samples taken in the fall of 1982 are presented in the last two columns of each table. Original soiltest values from samples taken in 1928 and stored in glass jars until recently analyzed are presented in the footnotes for each location. For the NPK, soil-test values from 1954 through 1982 are presented for each location. Dates of sampling indicated are spring dates for samples taken the previous fall or winter as recommended. These data show the effects of rates of P 2 0 5 and K20 applied on soil fertility. They indicate build-up or depletion and are expressed in pounds per acre or PP2M of P or K extracted by the Mehlich I extractant used in the Auburn University Soil Testing Laboratory. Index values in the last column are ratings and percent sufficiency of the most recent samples. Ratings shown in tables for both experiments are those used for corn and soybeans. Ratings for cotton and some legumes would be lower. These soil-test data have been used to calibrate the soil-test recommendations and keep them based on up-to-date research information since the soil testing laboratory was established in 1953. PROCEDURES Two-Year Rotation Fertilizer Experiments (TYR) These experiments were started in 1929 on a cotton-winter legume-corn rotation, and numerous revisions to keep them up to date are still in progress. Each consists of two adjacent 34-plot tiers with each plot 21 feet by 69.2 feet, or 1/30 acre. Both summer crops are grown each year with two replications of each treatment on both crops, making four replications for soil-test data presented. Winter legumes were used in the rotation except for plots in the N rates comparison until it was changed to a cornwheat-soybean rotation in 1968. Wheat was continued in the three-crop rotation through 1978. Grain sorghum was substituted for corn in 1982. Soil is turned in alternate directions every other year to minimize movement of soil from plot to plot. Harvest for yield and soil samples are taken from the two center rows, which are still 6 ALABAMA AGRICULTURAL EXPERIMENT STATION representative of the treatments after 54 years of cropping. Fertilizers are broadcast annually prior to planting, except on nonlegumes where two-thirds of the nitrogen is sidedressed to minimize leaching and runoff losses. Soil samples are taken from all plots every 2 years as needed and to determine effects of the treatments on soil-test levels and to indicate when lime should be applied to maintain pH at recommended levels. The treatments include four rates of N on plots that did not have winter legumes prior to 1968. Nitrogen rates varied among crops as indicated by footnotes under the yield tables. The standard treatment, which is assigned a value of 100 percent yield in the tables, received 120-60-60 in pounds per acre of N-P 20 5 -K 20. The no-P treatment had no P since 1957, but received 30 pounds per acre of P 20 5 from 1929 through 1957. The K rates have been continuous since 1958, with the no-K treatment not having received any K since 1928, if ever. The no-lime treatment has had no lime since 1928. The no-Mg treatment has been limed as needed with calcitic lime while other treatments were limed with dolomitic lime. The no-sulfur (S) treatment has had no S since 1977, prior to which it received S from superphosphate or gypsum as did all other treatments. The micronutrient (ME) treatment received a mixture of adequate amounts of zinc (Zn), boron (B), manganese (Mn), copper (Cu), and molybdenum (Mo) broadcast before planting corn every other year, beginning in 1959. The untreated plots shown at the bottom of each table received no fertilizer or lime from 1929 until the last one listed (treatment 17) was limed and fertilized with the standard 120-60-60 beginning in 1979. Soil-test values of original samples stored in 1929 are shown in footnotes under each yield table. Data on build-up or depletion of P and K during 50 years of fertilization at all locations were published in the Soil Science Society of America Journal, Vol. 45, 1981. Rates of NPK Experiments (NPK) Experiments with six rates of N, five rates of P 20 5, six rates of K 20, and two rates of lime were started at seven locations in 1954. Plots were 21 feet wide and 35 feet long with four replications in randomized blocks. The crop was cotton from 1954 through 1961, corn from 1962 through 1964, and cotton again from 1965 through 1969. Yield data for these periods were published in 1970 in Experiment Station Circular 181. LONG-TERM FERTILITY EXPERIMENTS 7 Nutrient rates were varied one at a time with other nutrients and lime held constant and adequate. The standard rates were 120-100-100 in pounds per acre of N-P 2 0 5-K 20 for cotton and corn. The standard rates were 0-100-100 for soybeans and peanuts. Soil samples were taken from all plots every 2 or 3 years for use in soil-test calibration. Soil pH was maintained at 5.8 to 6.5 except for the unlimed treatment. Gypsum was applied annually to all plots to supply 30 pounds per acreS except for treatment 6, which was made a no-S treatment in 1970. Boron (B) was applied at 1.5 pounds per acre for peanuts. All locations were put in residual study of P and K for cotton from 1970 through 1972. Nitrogen rates were applied as usual, but no P or K was applied except to the standard treatment, No. 5, which continued to receive 100 pounds per acre of P 20 5 and K2O. Rates of P 20 5 and K20 were resumed at Brewton, Monroeville, Wiregrass, and Sand Mountain in 1973, 1974, and 1975, when the crop was soybeans or peanuts. Rates of P 20 5 and K 20 were discontinued again at these locations in 1976 and 1977. Experiments at Prattville, Upper Coastal Plain, and Tennessee Valley locations were in residual study for P and K continuously from 1970 through 1977. Rates of P and K were resumed at all locations in 1978. Crops were changed to get response and calibration data on as many important crops as possible. After 3 years in soybeans from 1973 through 1975, experiments were planted in corn from 1976 through 1979 or 1980. In 1980 the crop was changed to sweet sorghum at Sand Mountain, Brewton, and Wiregrass as a part of the Experiment Station research effort on biomass for alcohol production. The other locations were returned to soybeans in 1980. Peanuts were grown again at the Wiregrass Substation in 1981 and 1982. RESULTS AND DISCUSSION The data from both experiments are presented by location, beginning with the Sand Mountain Substation, on which the Hartsells fsl soil is generally one of the most productive and most responsive to fertilizers of all Alabama soils. Because of the volume of data, only brief discussions are presented by crops. The levels of N, P, and K to which responses were found are indicated and significant responses to other nutrients are pointed out. Re- 8 ALABAMA AGRICULTURAL EXPERIMENT STATION lationships between soil-test levels and response are summarized along with the effects of treatments on soil-test values. Readers should study the tables for more details than those mentioned in the brief discussions. Results from the last 24 years of the 54-yearold TYR will generally be discussed first. Data from the most recent 18 years of the 29-year-old NPK experiments will follow. Poor yields due to drought, poor stands, late spring freezes, heavy rains, insect or disease damage, or other causes were experienced some years at all locations. Data from such years are generally not reliable for use in making fertilizer recommendations. Therefore, average yields shown do not include data from years when such problems were encountered, and years from which data were used are indicated in the tables. SAND MOUNTAIN SUBSTATION, HARTSELLS FSL (TABLES 1, 2, AND 3) Cotton Response to N. The Two-Year Rotation (TYR) experiment included a few plots which did not have vetch in the rotation from 1959 to 1967. These were used to compare 0, 60, and 120 pounds per acre of N for cotton and corn, table 1. The data show a response of 900 pounds per acre of seedcotton [(92-58), percent x 2,650] to 60 pounds per acre of N, with a response of only 210 pounds per acre [(100-92) percent x 2,650] to an additional 60 pounds per acre of N. The NPK experiment, table 2, showed a 100-pound-per-acre increase from 90 N over 60 N in 1965-69. The small decrease from 150 N over 120 N indicates that the 90 N rate would have been adequate for cotton. The increase from 120 N over 90 N in 1970-71 is not valid because no K 20 was applied during this residual period, except to the 120 N rate, and K deficiency limited the yield response to N and P. Response to P. The TYR, table 1, shows a 19 percent, or 500 pounds per acre, increase from 60 pounds per acre P 20 5 at low soil P. The NPK also shows a response to 60 over 40 P 2 0 5, indicating that 60 P 20 5 was the best rate for cotton, table 2. This is the rate presently recommended for Medium soil-test levels by the soil testing laboratory. No response was found from the 100 P 2 0 5 rate over 60 P 20 5 because this rate had raised the soil-test level to High by 1965, table 3. During the residual period 1970-71, the TABLE 1. RELATIVE YIELDS (PERCENT) OF CORN, GRAIN SORGHUM, COrloN, SOYBEANS, AND WHEAT IN THE 2-YEAR ROTATION FERTILIZER EXPERIMENT ON HARTSELLS FSL AT THE SAND MOUNTAIN SUBSTATION, 1959-82 Treatment', 2 No. No Cottons, 6 of 9 r. av., 95-6 1959-67 -r v Soybeans4 7 of Variable, lb.ace 11 years, 3-yr. 1979 r v, av., 82 lb/acre 1968-78 105 103 100 - 6 of 9 a. a., 1959-67 yr. Corn3a 7 of 11 av 1968-78 46 years, 88* 2of3 , 1979-81 35 47 years 87* srahm srhm. , Wheat6 7ofll1 Soiltest,1983 Lb./acre Index 1982 tas 1968-78 ears 16....No N 11.....30, 60, or 90 N 14.. .60, 90, or 120 N 10.....120 or N 12.. No P 6.. No K 7...301K20 14....60 K20 15.....120 K20 4....No lime 5..... No magnesium 58 92* 100 150 98 94 100 94 20 69* 100* - - 71 100 81 85 56 84* 100* - - - 100* - 81* 24 89* 2,650*lb. 98 89* 97 - 71* 39 87* 38*bu. 103 55* 87* - 67* 37 81* 48*bu. 98 58* 85* 90* 96 73* 61 95* 110 bu. 105 77* 76* 47 100* 92 67* 39 Ca 990 - H 200 - 78* 91 P K 13 39 L 60 L30 89* 131*hu. 100 56* 86* 83*bii 93 41* 54 105 bu. 92 16* 98 45 hu. 96 58* 59 140 172 Ca260 60 H 90 H 100 pH-5.0 102 - 98 - 8..Plus 1..Untreated 17..Untreated untill1979 'Original 2.....No sulfur since 1977 micronutrients 97 82* 82 83 87 83 96 - Mg 38 - L - 80 101 97 91 95 98 13 23 30 42 45 35 94 8 10 20 24 21 80 13--107 soil-test values in 1929 were: pH - 6.2, P - 32 lb./ acre (M), K- 100 lb./acre (M), Mg - 42 2Final soil-test values of the 120-60-60 treatment, no. 14, were: pH - 6.2, P lb./acre (VH) , K (H), Ca - 990 lb./acre (H). -118 lb./acre (L), -140 lb./acre Ca- 550 (H) , Mg -161 lb./acre (H). lb./acre 3N rates for cotton 1959-67 were 0, 60, and 120 lb./acre. For corn, 1959-78 rates were 0, 6(1, and 120 11/acre, and 1979-81. were 0, 90, 120, and 150 lb./ acre. 4N rates for soybeans 1968-82 were 0, 30, and 60 lb./acre. 5N rates for grain sorghum in 1982 were 0, 60, 90, and 120 lb./acre. 6N rates for wheat 1968-78 were 0, 50, and 80 lb./ acre. *Indicates significant difference at 5 percent level. 'FABLE 2. RELATIVE YIFLDS (PERCENT) OF Corr'oN, SOYBEANS, CORN. AND SWEET SORGHUM FROM 1)IFFRENi' RASFo ON THE SAND MOUNTAIN SUBSTATION, 1965-82 Seedcotton CornSweet 1976-79(4) N, P, AND K Treatment no. Lb./ acre applied 1970-72(2) 1656'6691 '65,'6,'6,'69residue N2 7o,,71 Soybeans,Consrhm 1973-75(3) 1980-82(3) Nitrogen 43 84* 96* 100 100 97 43 74* 95* 99 100 29 75* 91* 104* 98 2,460 2 3 4 5 6 7 8 9 10 5 11 12 13 14 15 5 16 6 S.... S.... . . . . . . . . S.... . . . . 0 30 60 90 120 150 180 39 72* 81* 85 100* 89 Phosphorus 46 71* 99 101 98 27 89* 100 100 98 74 91* 104* 101 100 84 93* 100 103 108 37 bu./acre 74* 104 11. 9 60 85* 89 97* 100 81 93* 106* 103 10'1 112 b)u./acre 73* 100 18 12 51 90* 96 100 100 p 2 0. 5 S.... .. . . . . . . . . . 0 20 40 60 100 82* 87 100* Potassium 19 47* 64* 77* 67 88* 82 97 100 72 82* 86 94 94 22.1 tons/acre 62* 96 13 K2 0 0 20 40 60 80 100 No lime No sulfur since 1969 Cv LSD (05) lb./acre 86* 2,560* 78* 16 9 84* 12 9 lb.f acre 9 'Numbers in parenthesis are years included in the average. 2Only the 30- and acre rates of N were used on soybeans. *Indicates a significant difference from the next lower or higher rate at the 5 percent level. 60-lb.! 0 r- Z 0 m TABLE 3. SOIL-TEST VALUES FROM 1)IFFERENT RATES OF P AND K ON HARTSELLS FSL AT THE SAND MOUNTAIN SUBSTATION, 1954-83 Rates of P 2 0.5 or K 2 0,1 Values, pounds/acre r Soiltest 1983 index, m lb. 1954 1965 1969 residue 1975 1980 1983 rating Soil-test phosphorus (P) x 10 1 2 0 20 40 16 7 13 19 11 22 33 15 22 32 13 17 26 10 97 68 Low 40 m 0 6'0 28 58s 51 37 4 4Mdu 100 81 106 Soil-test potassium 107 (K) 96 106 High 210 z -1 H' 4075 10122 Soil pH, no lime 5.4 5.2 110 152 4.8 100 190 4.9 116 185 5.0 99 180 5.0 74 147 4.9 High 7 90 iRates of P2 0 5 and K2 0 applied annually except in 1970, 1971, 1972, 1976, and 1977. applied each year since 1954. The 100-lb/acre rate of P 2 0 5 and K-)O was 12 ALABAMA AGRICULTURAL EXPERIMENT STATION low P treatments continued to produce at the previous levels but 40- and 60-pound-per-acre rates were limited by K deficiency. Response to K. This soil produced low cotton yields without K fertilization. The two experiments agree closely in response of cotton to K. They showed average increases of about 1,900 pounds per acre of seedcotton from 60 K20 when soil-test K was Low. No increase was found in either experiment from more than 60 pounds per acre of K 20. The 60 pounds per acre rate increased the soil-test K level to High in both experiments. The residual study showed that this High level is temporary without K fertilization and that where none is applied, soils should be sampled annually to prevent loss in yield as was found in 1970-71. Other nutrients. Cotton showed average decreases of 290 and 540 pounds per acre of seedcotton where lime was not applied in the two experiments. Soil pH dropped to about 4.8 to 5.0 and leveled off, but drastically reduced cotton yields. Use of calcitic instead of dolomitic lime reduced the soil test Mg to Low 80 but did not significantly reduce cotton yield. The micronutrient mixture of B, Zn, Mn, Fe, and Cu in the TYR did not increase cotton yields. Soybeans Response to N. Both experiments included rates of 30 and 60 pounds per acre N on soybeans. No response to N was found. This supports the recommendation that N not be applied to soybeans on soils where soybeans have been grown previously. Response to P. Responses of 11 and 16 bushels per acre were found for the two periods in which the TYR was in soybeans. This was expected on this soil which was Low in P. The NPK experiment produced a 9-bushel-per-acre response to 40 pounds per acre of P 20 5 in 1973-75. The 40-pound-per-acre rate produced top yields and maintained soil P at Medium. Higher rates were required to raise this soil to High, but the higher rates did not result in higher yields. Response to K. Soybeans responded to K in both experiments. The increase from K ranged from 5 to 9 bushels per acre for 40 to 60 pounds per acre of K 2O. The 60-pound-per-acre rate of K 2 0 raised the soil-test level to High in both experiments. Other nutrients. Where lime was not applied and pH dropped to 5.0, yields were reduced 10 to 20 bushels per acre. Allowing Mg to drop to Low from use of calcitic instead of dolomitic lime LONG-TERM FERTILITY EXPERIMENTS 13 in the TYR reduced soybean yields 5 to 7 bushels per acre. No response to the micronutrient mixture in the TYR was found. There was a 5-bushel-per-acre response to S in the TYR but none in the NPK. Corn Response to N. Corn yields for the standard 120-60-60 treatment in the TYR averaged 116 bushels per acre for the 15 years reported and 112 bushels per acre for 4 years in the NPK. The 120-pound-per-acre N rate was best in both experiments, although rates went to 150 pounds per acre in the TYR and 180 pounds per acre in the NPK in recent years. This supports the present statewide recommendation of 120 pounds per acre of N for corn. The soil-test recommendation includes a comment, that corn on sandy soils, such as the Hartsells, may respond to 150 pounds per acre of N. Since there is no satisfactory soil test to determine the capacity of Alabama soils to supply N for growing crops, recommendations are based on the crop to be grown. The organic matter content of all Alabama soils is low and they do not vary much in their capacity to supply N. These Sand Mountain soils have generally been the most responsive to N of any soils in the state. Response to P. Corn has produced good responses to 60 pounds per acre of P2 0 5 in both experiments. This rate has raised the soiltest P level to Very High in the TYR and to High in the NPK. This soil has the highest P requirement of any of the soils included in these experiments. Response to K. Corn responded to 60 pounds per acre K 20 5 in the TYR and to 40 pounds per acre in the NPK. The 60-poundper-acre rate raised the soil-test level to High in both experiments. Other nutrients. Corn yields decreased in successive periods without lime to only 41 percent of normal yield in the TYR at pH 5.0. Yield was reduced 24 bushels per acre for 1976-79 at pH 5.0 in the NPK experiment. Response of 15 bushels per acre to S was found in 1979-81. Magnesium and micronutrients have not increased corn yields in the TYR. Wheat The N rates for wheat in the TYR in 1968-78 were 0, 50, and 80 pounds per acre, with 20 pounds per acre applied at planting and the remainder topdressed in the spring. The average yield of treatment 14 for 7 years was 45 bushels per acre, The 80 N rate 14 ALABAMA AGRICULTURAL EXPERIMENT STATION produced 7 bushels per acre more than 50 N. Wheat produced a 10-bushel-per-acre response to P and 8-bushel-per-acre response to the first 30 pounds per acre of K 20, with no increase from higher rates. Allowing soil pH to drop to 5.0 decreased wheat yields 19 bushels per acre. Use of dolomitic instead of calcitic lime did not significantly increase yield even though soil Mg levels were Low. Grain Sorghum and Sweet Sorghum Grain sorghum for the first time in TYR in 1982 showed indications of increase in yield from N, P, K, lime, Mg, and S. Since the experiment includes only two replications, data were not analyzed statistically. This crop is highly responsive to N, K, and lime and better information on rates will be available after the experiment is repeated or from averages among locations. Sweet sorghum was grown in the NPK experiment in 1980-82 to determine its response to fertilizers and lime and to evaluate it as a biomass crop for alcohol production. The entire crop was removed from the area. Fresh weight yields showed large responses to 60 N, 60 P 2 0 5, 60 K 20, and to lime. Soil samples taken in the fall of 1982 showed that 100 pounds per acre of P 20 5 and K 20 were required to maintain High soil-test levels of P and K where such large amounts were removed in the crop. The present recommendation of 60 pounds per acre of P 2 0 5 and K 20 for corn or sorghum silage on Medium P and K soils was adequate to produce top yields and maintain the Medium level through this 3-year period, table 3. BREWTON EXPERIMENT FIELD, BENNDALE FSL (TABLES 4, 5, AND 6) This Benndale fsl soil is a gray terrace soil with cation exchange capacity (CEC) of about 4 milliequivalents per 100 grams. There is little clay in the subsoil. This soil is the coarsest and least fertile of all soils used for these experiments. The subsoil is Very Low in P with no movement below the plowed layer in over 50 years of fertilization. The subsoil is low in K and has a pH of 5.0 at 3-foot depth, even where lime has been applied to the surface for 50 years. This soil has low moisture holding capacity but makes good yields when well-fertilized and rainfall is adequate. TABLE 4. RELATIVE YIELDS (PERCENT) OF COTTON, SOYBEANS, CORN, G;RAIN SORGHUM, AND WHEAT IN THE 2-YEAR ROTATION FERTILIZER EXPERIMENT ON BENNDALE FSL AT THE BREWTON FIELD, 1959-82 Treatments, 2 Variable, lb./acre 16..... No N 11.....30, 60 or 90 N 14.....60, 90 or 120 N 10.....120 or 150 N 12..... No P 6..... No K 7..... 30 K20 14.....60 K20 15.....120 K20 4..... No lime 5..... No magnesium 2 ..... sulfur since 1977 No Plus micronutrients 8 ..... 1.... Untreated 17 .... Untreated until 1979 Cottons, 6 of 9 yr. av., 1959-67 60 107* 100 97 38 95* 2,280 lb. 98 88* 96 110* 13 15 Soybean 3-vr. av., 7 of 11 years, 1968-78 92 90 100 72* 64 100* 39 bu. 103 38* 78* 107 23 26 100 100 100 97 84* 42 97* 38 bu. 100 29* 76* 97 105 18 84 4 6 of 9 yr. av., 1959-67 32 90* 100 - Corn 7 of 11 years, 1968-78 64 102* 100 87* 53 97* 97 bu. 99 53* 97 100 8 19 2 of 3 sorghums, years, 1982 1979-81 69 104* 100 97 79* 22 91* 77 bu.* 108 16* 99 101 105 1 77 74 89* 100 115 88* 65 112* 68 bu. 97 21* 101 96 103 12 82 Wheat, of 11 years, 1968-78 46 104* 100 75* 92 100* 24 bu. 96 71* 105 96 - S Lb./acre Soil test, 1983 Index, pct. -- - 94 85 101* 81 bu. 107 94 106 106 15 14 Ca 730 P 12 K 34 66 122 142 Ca 160 Mg 23 - H 240 - VL 50 L 60 M 80 H 110 H 120 pH 5.0 L 100 - 12 'Original soil-test values in 1929 were: pH - 5.9, P - 28 lb./acre (M), K - 40 lb./acre (L), Mg - 28 lb./acre (H), Ca - 560 lb./acre (H). 2Final soil-test values in 1983 of the 120-60-60 treatment, no. 14, were: pH - 6.4,P - 124 lb./acre (VH). K - 122 (H), Mg - 122 (H), Ca - 730 (H). 3N rates for cotton in 1959-67 were 0, 60 and 120 lb./acre. For corn, 1959-78 rates were 0, 60. and 120 lb./acre, and in 1979-81, 0, 90, 120, and 150 lb./acre. 4N rates for soybeans in 1968-82 were 0 30, and 60 lb./acre. 5N rates for grain sorghum in 1982 were 0, 60, 90, and 120 lb./acre. 6N rates for wheat in 1968-78 were 0, 50, and 80 lb./acre. *Indicates significant difference at 5 percent level. TABLE 5. RELATVE YIELDS (I RCFINT) OF- COTTON, SOYBEANS, CORN, AND SWEET SORGHUM FROM DIFFERENT RATES OF N, P, AND K ON THE BREWTON FIELD, 1965-82 Seedcotton Treatment no. Lb./ acre applied , 196569(''70, '65, '66, '68, '69 1970-72(2) '71 residue1908"3 17-53 Soybeans, 1976-94 Corn,Swe sorghum, N'2 1 2 3 4 5 6 7 8 9 10 5 11 12 13 14 15 5 16 6 S.... S.... . . . . . . . . S.... . . . . 0 30 60 90 120 150 50 81* 94* 103* 100 97 88 96* 100 99 100 45 82* 94* 100 103 2,530 lb. 98 11 7 II Nitrogen 66 82* 94* 96 100 90* Phosphorus 105 105 104 30 90* 97 100 101 81 84 93* 90 100 77 87 103* 95 102 128 bu. 52* 102 14 12 64 76* 100* 100 104 P205 S..... .. . . . . . . . 0 20 40 60 100 91 103* 88 100 100 Potassium 78 90* 95 102 100 83 90* 99* 101 94 40 bu. 74* 94 8 6 I II 84 90 101* 102 100 63 72 102* 101 110 17.0 ton 57* 110 21 16 c- r c: mI K20 0 20 40 60 80 100 No lime No sulfur since 1969 53 74* 90* 94 102 2,250 lb. 103 11 R m Cv LSD (05) IIIII 9 I z -H 'Numbers in parenthesis are years included in the average. 2 Only the 30- and 60-lb./acre rates of N were used on soybeans. *Indicates a significant difference from the next lower or higher rate at the 5 percent level. TABLE 6. SOIL-TEST VALUES FROM DIFFERENT RATES OF P AND K ON BENNDALE FSL Rates of P2 05 or K 2 0.1 Values, pounds/acre 1972 1969 residue 1975 Soil-test phosphorus (P) ATHE BREWTON FIELD, 1954-83 Soil1980 1983 test rating 1983 index, pct. 1954 1965 0 20 40 18 20 32 42 22 37 5841 16 32 57 112 55 65 90 100 123 5.6 20 26 4838510 15 27 68 130 48 70 118 110 148 5.5 22 32 60 121 52 69 86 102 110 5.3 Low Medium High 70 80 120 240 70 80 9 100 100 60 100 0 2073 6098 80 100 Soil pH 5.6 64 61 82 52 80 111 65 76 105 115 120 5.4 57 102 61 78 107 150 146 5.6 Soil-test potassium (K) Medium High 104 110 5.5 applied each year since 1954. 'Rates of P 2 0 5 and K20O applied annually except in 1970, 1971, 1972, 1976, and 1977. The 100-lb./acre rate of P2 0 5 and K2 0 was 18 ALABAMA AGRICULTURAL EXPERIMENT STATION Cotton The best rate of N for cotton in the two experiments was 90 pounds per acre, which is the rate recommended. Yield in the TYR was increased more than 1,000 pounds seedcotton by 60 N with no further increase from 120 N. In the NPK, 90 pounds N increased yield 900 pounds per acre and was the best rate over the 6 years reported. No response to P was found in the TYR when soil-test P was Medium. In the NPK, yield was increased by 20 pounds per acre of P 20 5 when soil-test P was Low. No further increase was found from the higher rates. Application of 20 pounds per acre of P2 0 5 raised soil P to Medium and 40 pounds raised it to High. During 3 years without P in 1970-72, the residue from the 20-pound P 20 5 rate produced top yields. Cotton produced large responses to K in both experiments. The best rate was 60 pounds per acre of K 20 in both cases with no response from higher rates. The average increase from 60 K 20 was 1,400 pounds per acre of seedcotton. All rates applied increased the K level in this sandy soil, and the residue maintained yields at previous levels for 3 years, 1970-72. The soil-test K ratings indicated in table 6 are those used for corn and soybeans on which 81 pounds per acre of soil-test K is rated High. The higher K requirement for cotton to reach index values greater than 100 requires 121 pounds per acre of soil-test K for a High soil-test rating. When pH was allowed to drop to 5.4, yield was reduced 230 pounds per acre of seedcotton. Cotton showed a 10 percent increase from the micronutrient mixture. This was most likely from B, which is recommended on all soils. Soybeans Soybean yields from the standard treatments in the two experiments averaged 39 bushels per acre for the 13 years reported. No response to N was found in either experiment. Response to P of 11 bushels per acre was produced in the TYR from 1968-78 after soil-test level of the no-P treatment dropped into the Low range. The best rate of P in the NPK was 40 to 60 pounds per acre of P2 0 5 when the soil was Medium in P. This supports the present recommendation of 40 pounds per acre at Medium. The first 30-pound-per-acre increment of K 20 increased soybean yields 14 to 22 bushels per acre at Low soil-test K in the TYR. No further increase was found from 60 over 30 pounds per acre of K20. The best rate of K 20 in the NPK was 40 pounds per LONG-TERM FERTILITY EXPERIMENTS 19 acre at Medium soil-test K. This shows that on this sandy soil producing good yields of soybeans, the present recommendation of 40 pounds per acre K 2 0 at Medium is adequate and that higher rates increased the soil-test K level to High. No response to K was found when soil-test levels were High. Large responses to lime and to Mg were produced when the soil pH was not maintained by liming with dolomitic lime. Corn When corn was rotated with cotton in the TYR, 120 pounds per acre of N increased yield 8 bushels per acre over 90 N. When rotated with soybeans in 1968-81 and following soybeans in the NPK, the best rate of N for corn was 90 pounds per acre. Such data have led to the present comment used with recommendations for corn that when following soybeans, N rates for corn may be reduced by 30 pounds per acre. Response of corn to P in the TYR increased with time as the P level of the no-P treatment dropped from Medium to Low. The data indicate that the recommended rate of 40 pounds per acre P 20 5 annually or 80 pounds in alternate years is adequate for Medium P soils in a corn-soybean rotation. Plots in the TYR that have not received any K since 1929 are severely deficient and produce very low yields. Applications of 40 to 60 pounds per acre of K 20 annually are needed on this sandy soil to produce top yields and to maintain the desirable level in the soil. As the pH of unlimed plots has dropped to around 5.0, yields of corn have been severely limited. Corn plants will not produce satisfactorily on this soil under such acid conditions. Magnesium deficiency from use of calcitic lime continuously instead of dolomite has not limited corn yields on this soil, although soil-test Mg dropped to 23 pounds per acre by 1983. Wheat, Grain Sorghum, and Sweet Sorghum Wheat yields at this location were lower than on the other stations, averaging only 24 bushels per acre for 7 years. The data indicate that 50 pounds per acre of N, 30 pounds per acre of P 20 5 at VL50, and 30 pounds per acre of K 20 at L60 were adequate for wheat on this soil. Wheat also produced a response to lime at pH 5.0, but not to Mg at H 110 or to micronutrients. Grain sorghum yield in 1982 in the TYR was limited to 68 20 ALABAMA AGRICULTURAL EXPERIMENT STATION bushels per acre by drought. Response was found to 90 pounds per acre N, to P, to 30 pounds per acre K 20, and to lime. Sweet sorghum in the NPK in 1980-82 responded to 60 N, to 40 P 20 5, to 40 K20, and to lime. Soil-test K dropped between 1980 and 1983 because of the high rate of removal where 17 tons of green sorghum was removed from the plots annually. WIREGRASS SUBSTATION, DOTHAN FSL (TABLES 7, 8 AND 9) This gray fine sandy loam soil is representative of a large acreage throughout the Lower Coastal Plains of Alabama. It has a sandy clay loam subsoil at about 10-inch depth which accumulates K, Mg, and Ca as they leach from the surface soil. The surface soil on the Wiregrass was High in soil P due to previous fertilization when the TYR was started in 1929. The area used for the NPK was also High in P when it was started in 1954. Summer crops have never responded to P on this substation. Response to P was found on winter legumes when they were used in rotations to supply N for summer crops. Quick perusal of the relative yield data from the no-P treatments in both experiments, tables 7 and 8, will show that crops on this soil still do not respond to P after 54 and 29 years of continuous cropping without P fertilizer. Therefore, discussion of response to P will be omitted for the different crops. The soil-test data in tables 7 and 9 show that soil-test P dropped to Medium in the TYR and to Low in the NPK in 1983 where none was applied. These experiments should begin to show response if continued. Response at Medium is less likely to occur when the level is dropping than when it is rising from annual applications. Cotton Both experiments produced about two bales of cotton from 60 pounds per acre of N for the 10 years of data reported. Higher rates did not increase the yield in either experiment, and the 120 N rate decreased the yield in the TYR. Rates from 60 up to 150 pounds per acre N produced about the same yield in the NPK for the 4 years reported. Yield of seedcotton in the TYR was increased 1,100 pounds per acre by 30 pounds per acre of K20, with no further increase from TABLE 7. RELATIVE YIELDS (PERCENT) OF COTTON, SOYBEANS, IPEANUTS, CORN, GBRAIN SORGHUM, AND WHEAT FERTILIZFR EXPERIMENT ON DOTHAN FSL AT THE WIREASS SUBSTATION, 1959-82 IN THE Two-YEAR ROTATION Treatmentl,2 Cotton3, 6 of 9 1959-67 Soybeans4 7 of 11 Peanuts, years 3 years, 1982 1968-78 1979-81 92 92 97 103 97 103 35 bu. 100 89* 104 100 6 of 9 years, 1959-67 84 110* 100 - CornG 7 of 11 years, 1968-78 65 99* 100 - 3 r 1979s81 62 rhun 1982 53 Wheat6, 7 of 11 xears, 1968-78 41 Soil tests 1983 Lb/acre - 16.....No N 11.....30, 60, or 90 N 14.....60,90, or 120 N 10.....120 orl5ON 12.....No P 6.....No K 7. 30120 73 107* 100 97 58 91 100 100 100 106 95 100 103 104* 100 102 105 99* 100 118 92 88 100 - - - Ca 980 - H 330 - 106* 2,610 lb. 102 90* 107 107 14.....60 K20 15.....120 K20 4..... No lime No Mg 2.....NoS 5 ..... 97 112 88 94 97* 97 33 bu. 3,220 lb. 94 91 85* 89* 97 104 88* 100 94 86 97* 79 bu. 93 81* 101 94 102 93 99 100 bu. 97 91* 99 99 98 P 42 M 90 86 96 74 101 100 97 K 76 144 M80 H 120 H 160 H 180 78 bu. 100 95 bu. 114 41 bu. 94 193 221 77* 100 18* 101 95 98 93 - Ca280 Mg 46 - pH5.1 H 180 - 8..Plus ME since 1977 100 96 86* 103 89* 107 17..Untreated until 1979 Untreated .... 22 22 69 57 48 73 85 95 24 22 28 33 38 74 13 99 - - 17 "Original soil-test values in 1929: pH - 5.8; P - 66 lb./acre (H); 2Final soil-test values of the no. 14, 120-60-60, treatment: pH (H); Ca - 980 lb./acre (H). 3N rates for cotton in 1959-67 were 0, 60, and 120 lb./acre. For 0, 90, 120, and 150 lb./acre N. 4N rates for soybeans and peanuts in 1959-67 were 0, 30, and sN rates for grain sorghum in 1982 were 0, 60, 90, and 120 6N rates for wheat in K - 90 1b.f acre (H); Mg - 35 lb./acre (H); Ca - 340 lb./acre (H). 6.2; P - 129 lb./acre (VH); K 193 lb./acre (H); Mg - 10 lb./acre corn in 1959-78, rates were 0, 60, and 120 lb./acre N, and in 1979-81, 60 lb./acre. lb./acre. *Indicates significant difference at 5 percent level. 1959-67 were 0, 50, and 80 lb./acre. TABLE 8. RELATIVE YtELDS (PERCENT) OF COTrON, PEANUTS, CORN, AND SWEET SORCHUM FROM DIFFERENT ON THE WntREAss SUBSTATION, 1965-82 RATES OF N, P, AND K Seedcotton Treatment no. Lb./acre applied G5, 17Cor2) 1965-69(2)'1170,'722)7Peanuts, PensSweet '66residue Nitrogen 62 88* 100 102 100 103 Phosphoru 98 98 97 98 100 94 97 101 '1976-79(3) 7,'8,'9sorghum, mPeanuts, 1981-82(2) N2~ 2 .... .... .... o 30 60 90 120 150 82 83 102* 90 100 98 43 88* 98 100 95 3 5 4 11 6i 95 105 125 100 125 98 96 99 ...... . . . 98 95 98 96i 90 20 105 94 40 97 99 60 96 96 100 100 100 100 K2 0 Potassium 80 54 0 85 37 95* 103* 78* 96* 20 105 104 95* 40 99 102 98 106 91 60 15 .. . . . . . 103 96 103 80 93 10 .. . . . . . 91 bu. 3,920 lb. 100 2,650 lb. 2,420 lb. 11 .. . . . . . 33* No lime 63* 38* l & 17 . . . . . No sulfur 6 .. . . . . . 99 100 since 1969 15 19 9 12 CV 15 16 LSD (05) 10 7 "Numbers in parenthesis are years included in the average. 2Only the 30- and 60-lb./acre N rates were used on peanuts. the next lower or higher rate at the 5 percent level. Indicates a significant difference 12 14 13 16 .... .... .. . . . . . .. . . . . . 117 123 119 99 100 86 122 109 105 112 20.7 tons 0* 106 21 NS 104 102 103 102 100 94 102 99 103 106 4,470 lb. 36* 101 7 9 from 0 z Q -I TABLE 9. Rates of '20;; and K2 01 SOIL-TEST VALUES FROM DIFFERENT RATES OF P AND K DOTHAN FSL AT THE WIRE.RASS SUBSTATION, 1954-83 Values, pounds/acre 1954 195 199 resldue Soiltest rating r- in 1983 index, pct. 7d1975 1981 1983 Soil-test phosphorus (P) 59~9 38 45 45 1222 Lw7 20O 58 64 60 44 2 88 x 66 109 85 82 154 182 4.5 89 91 84 88 135 154 4.9 40 86 81 80, 60470Meim9 60 100 0 20110 80 100 Soil pH 88 84 111 79 112 124 84 120 180 200 5.4 100 112 84 100 175 200 5.3 107 111 62 81 175 200 5.2 z High 180 90 90 110 130 Soil-test potassium (K) Medium High 5.6 130 134 5.5 'Rates of P2Orj applied annually except in each year since 1954. 1970, 1971, 1972, 1976 and 1977. The 100-lb./acre rate of P205 and K20 was applied 24 ALABAMA AGRICULTURAL EXPERIMENT STATION 60 and 120 pounds per acre. The 20-pound-per-acre rate of K2 0 produced 96 percent yield in the NPK in 1965-66, but dropped to 78 percent when none was applied in 1970-72. This shows that where low rates are used, they must be applied every year on these sandy soils. When soil-test levels are increased to High by application of 60 pounds per acre K 20 or more, the residual effect may last for several years. Cotton produced good response to lime in both experiments. Yield was also increased by the micronutrient mixture, presumably from the 0.5 pound per acre of B in the mixture. Peanuts Peanuts are known to be quite unresponsive to direct applications of fertilizer and it is generally recommended that they be grown in rotation with other well-fertilized crops. Peanuts were grown in the NPK experiment in 1973-75 and again in 1981-82. They were used in the TYR in 1982 for the first time in 54 years. The greatest response in both experiments was to lime. Increases from lime of 1,690 and 2,860 pounds per acre of peanuts were found in the two periods of the NPK. The unlimed treatment in the TYR produced 89 percent yield, or 2,870 pounds per acre of peanuts, where no lime had ever been applied and where the pH was 5.1. The untreated plot, which had received no fertilizer or lime for 54 years, produced 85 percent yield. A small response to the first increment of 20 or 30 pounds per acre of K 20 was found at Medium soil-test K levels in both experiments. No increase was produced by more than 30 pounds per acre of K 20. Peanuts did not respond to N, P, Mg, S, or micronutrients in these experiments. These data show that the recommendation that peanuts be grown in rotation with well-fertilized crops is sound. This is advisable because they remove large amounts of nutrients from the soil and will reduce fertility to such a low level that following crops will not grow satisfactorily if these nutrients are not maintained at satisfactory levels. Corn Data from both experiments show that 90 pounds per acre of N was adequate for yield levels produced at this location. Drought stress at some period during most growing seasons kept average yields below 100 bushels per acre. If all years were included, the averages would be lower, Other soils in the Coastal Plains that LONG-TERM FERTILITY EXPERIMENTS 25 have higher water holding capacity or are located where summer rainfall is more dependable will produce higher yields and respond to rates of N up to 120 pounds per acre. Corn produced about 12 bushels per acre response to 20 and 30 pounds per acre K 20 with no additional response to higher rates. Application of 30 to 40 pounds per acre of K 20 raised soil-test K to High in both experiments. The K level in both experiments remained at the Medium level for corn, peanuts, and soybeans without applications of K20 for 54 and 29 years in these experiments. Corn responded to lime and S but did not show response from Mg or micronutrients. These data show that present recommendations from the soil testing laboratory should be adequate for corn on sandy soils of the coastal plain. Soybeans Soybeans showed little response to fertilizers in the TYR from 1968 through 1981. There was about a 3-bushel-per-acre average increase from 30 pounds per acre of N, but this difference was not statistically significant. There were also small increases from 30 K 20, from lime, and from S in the 1979-81 period. These data show that soybeans have a lower fertility requirement than cotton and that present soil-test recommendations are more than adequate for this crop. Wheat and Sweet Sorghum Wheat yields in the three-crop rotation averaged 41 bushels per acre from the standard treatment. It responded to 80 pounds per acre N but did not respond to P, K, lime, Mg, or iicronutrients. Sweet sorghum in the NPK in 1980 had a poor stand and responded only to 20 K 20 and to lime. Sorghum is so sensitive to low pH that the seedlings did not survive at pH 4.5 in 1980. MONROEVILLE EXPERIMENT FIELD, LUCEDALE FSL (TABLES 10, 11, AND 12) This reddish fine sandy loam topsoil and its clay loam subsoil have considerable capacity to retain added nutrients and supply them to future crops. Soil-test P and K data in table 12 show that there was little change in the levels of these nutrients in the sur- TABLE 10. RELATIVE YIELDS (PERCENT) OF CORN, GRAIN SORGHUM, COTTON, SOYBEANS, AND WHEAT IN THE 2-YEAR ROTATION FERTILIZER EXPERIMENT ON LUCEDALE FSL AT THE MONROEVILLE Treatment , 2 Variable, No. S6 b.acre 16..... No N 11..... 30,60 or 90 N 1.4..... 60,90 120 N 1 FIELD, 1959 82 Grain sorghum, 19-811968-78 1982 61 101* 100 102 Cotton3, Soybeans4 Corn 6 of 9 7 of 11 6 of 9 7 of 11 yr. av., years -yr.av., yr. av., years, 6719 1968-78 1958-781959-67 1 897981 9yr. av., years,9 80 107 100 - 3 Wheat, i of 11 Soil test 1983 Lb/acre years, 62 104* 100 - pct.Lb./acre or 89 95 100 - 95 105 100 - 46 87* 100* - 66 100* 100 - 77 87 100 88 - - Ca 1420 - H 470 - 10..... 120 or 150N 12..... No P 6.....N'o K 7.....30 K20 14.....60 K20 1.5.....120 K20 4 ..... lime No 5 ..... magnesium No 2 ..... No sulfur 8 ..... Plus micronutrients 1.... Untreated 17..... Untreated until 1979 108 65 101* 2,020 lb. 100 83* 94 I15 29 34 95 90 97* 39 bu. 103 59* 92* 103 36 33 105 59 95* 39 bu. 105 79* 90* 95 103 54 87 94* 92 100* 77 bu. 101 75* 98 97 27 27 92* 84 100* 90 bu. 101 66* 98 100 36 40 96 82 104* 113 bu. 97 39* 105 84* 104 30 75 81* 61 87 90 bu. 100 22* 90 94 82 21 86 96 100 108 26 bu. 100 88* 97 100 23 P K 17 53 83 155 182 Ca 350 Mg 55 - L 60 L 60 M 70 H 100 H 110 pH 4.9 H 110 - - - - 'Original soil-test values in 1929 were: pH - 5.8, P - 30 lb./acre (M), K - 84 lb./acre (M), Mg - 60 lb./acre (H), Ca - 580 lb./acre (H). 2Final soil-test values in 1983 of the 120-60-60 treatment, no. 14, were: pH - 6.1, P - 118 lb./acre (VH), K - 155 (H), Mg - 280 (H), Ca - 1,420 (H). 3N rates for cotton in 1959-67 were 0, 60, and 120 lb./acre. For corn in 1959-78, rates were 0, 60, and 120 lb./acre N, and in 197981, were 0, 90, 120, and 150 lb./acre N. 4N rates for soybeans in 1968-82 were 0, 30, and 60 lb./acre. 5N rates for grain sorghum in 1982 were 0, 60, 90, and 120 lb./acre. 6N rates for wheat 1968-78 were 0, 50, and 80 lb./acre. *Indicates significant difference at 5 percent level. TABLE 11. RELATIVE YIELDS (PERCEF.NT) OF CorroN, SOYBEANS, AND CORN FROM 1)lFFERENT RATES OF N, P. AND K ON THE MONROEViLLE FIELD, 1965-82 Soybeans, Corn,Soeas 1976-80 (2 ) Tratet no. Treament Lb.Seedcotton1970-72 (2) bacre1 applied 196569()1'0'17,7767 '65,'6,'6,'69residue 1973-75 (2) Sybas 1981-82 (2) Nitrogen N2 70 55 0 1................. 81* 88* 30 2............... 96* 102* 3................ 60 107 102 4..........90 100 100 5................. 120 105 150 103 6................. P2 0 5 Phosphorus 7 ....... 0 89 91 96* 106* 20 8................ 104 9................. 40 97 10................. 60 100 101 5................. 100 100 100 K2 0 Potassium 68 59 11................. 0 12................ 20 91* 88* 94 99* 13................. 40 14................. 60 101 102 15................ 80 101 116 5................. 100 1,950 lb. 1,910 lb. 92 16................ No lime 99 6................ No sulfur since 1969 CV 11 12 LSD (05) 7 11 'Numbers in parenthesis are years included in the average. 2Only the 30- and 60-lb./acre N rates applied on soybeans. 94 10093 - 51 84* 921009 774 95* 93 103 100 91 102 98 94 94 88 his. 56* 93 23 21 96 95 99 73 93* 91 98 100 79 89 90 92 95 29 bu. 76* 97 12 11 89 104* 101 102 100 86 94* 102 96 99 33 ho. 81* 99 10 8 *Indicates a significant difference from the next lower or higher rate at the 5 percent level. TABLE 12. SOIL-TEST VALUES FROM DIFFERENT RATES OF P AND K ON LUCEBALE FSL AT THE MONROFVILLE FIELD, 1954-83 Rates of P2 0 5 and Values, pounds/ acre 1972est Soil97 1975 1981 1983 rating 1983 index, Kzl1954 K0194 1965 16 1969 199 1972e residue pct. Soil-test phosphorus (P) 0 29 211 24 16 22 2 5Lw7 20 40 33 51 46 60 - 30 40 36 5043510 33 54 130 83 115 153 174 235 5.3 35 75 148 78 120 159 18610 233 4.8 Medium High 90 150 800 70 80 100 130 60 100 0 2088 60121 80125 100 Soil pH each year since 1954. 68 94 75 68 80 110 80 91 122 131 148 5.3 52 100 70 77 111 121 170 5.3 67 123 79 98 173 207 220 5.2 Soil-test potassium (K) Medium High 5.4 141 5.3 'Rates of P2 0 5 applied annually except in 1970, 1971, 1972, 1976, and 1977. The 100-lb/acre rate of P2 0 5 and K2 0 was applied LONG-TERM FERTILITY EXPERIMENTS 29 face soil where none was applied under continuous cropping from 1954 through 1982. This shows that this soil supplied substantial amounts of these nutrients without much drop in fertility. Applications of 40 pounds per acre of P 2 0 5 and K 20 were adequate to raise the level of both nutrients to High for corn and soybeans. Cotton Data in table 10 show that 60 pounds per acre of N was adequate to produce 2,160 pounds of seedcotton with no further increase from the higher rate. No response to P was found at Medium soil-test P in the TYR, while a small increase from 20 pounds per acre P 20 5 was indicated in the NPK where the level was Low. Cotton produced large increases from 30 to 40 pounds per acre of K 20, with no further increase from higher rates, which increased soil-test levels to High in both experiments. Small increases from lime were produced at a pH of about 5.3. The micronutrient mixture which included B increased yield about 300 pounds per acre in the TYR. During 2 years without P or K in 1970-71, plots continued to produce at the previous level. Soybeans During the first few years when soybeans were used in the TYR, they showed about a 4-bushel-per-acre increase from 30 pounds per acre of N. This did not occur in later years or in the NPK, so it is doubtful if this was a real difference. They did not respond to P 20 5 in the TYR where the soil-test level was Medium, but did respond to the first 20 pounds per acre of P 2 0 5 in the NPK during both periods. The soil-test level was Low on the noP treatment. Response to 20 and 30 pounds per acre of K 20 was found in both experiments where the K level was Low to Medium. Large response to lime was found in both experiments. Lack of Mg reduced yields about 4 bushels per acre, or 10 percent. No differences were found for S or micronutrients. Corn The best rate of N for corn in the TYR was 90 pounds per acre, which averaged 95 bushels per acre for the 9 years reported from 1968-81. In the 2 years reported for the NPK, 120 pounds per acre N increased yield 7 bushels per acre over 90 N. Increases in yield from the first increment of P were found in both experiments. In 30 ALABAMA AGRICULTURAL EXPERIMENT STATION the NPK which was Low in soil-test P, 20 pounds per acre of P 2 0 5 increased yield 18 bushels per acre and raised the soil-test level to Medium. No further increase was found from 40 P 2 0 5, which raised the level to High. The first increment of 20 and 30 pounds per acre of K 20 produced top yields and maintained the Medium soil level. Use of 60 pounds per acre of K 20 increased soil levels to High in both experiments, but did not increase yield over the lower rates. Corn responded to lime and to S but not to Mg, which was High in this soil, or to micronutrients. Wheat and Grain Sorghum Wheat averaged only 26 bushels per acre for the years reported and responded to only 50 pounds per acre of N, 20 pounds of which was applied in the fall and 30 pounds in the spring. It also responded to lime but not to P, K, Mg, or micronutrients. Grain sorghum produced 90 bushels per acre in 1982 and produced large responses to 90 N, 60 P 2 0 5, 60 K 20, lime, and possibly Mg. The two replications for only 1 year are not adequate to determine significance of small differences. Grain sorghum has higher fertility requirements for P, K, and lime than do corn and soybeans. PRATTVILLE EXPERIMENT FIELD, LUCEDALE SCL (TABLES 13, 14, AND 15) This soil is similar to the Lucedale fsl at Monroeville but has more clay and a higher cation exchange capacity of about 6 milliequivalents per 100 grams. The soil at Prattville, like the Dothan soil on the Wiregrass Substation, had been built up to a High level of P by fertilization before the experiment field was acquired by the Agricultural Experiment Station in 1928. It has not responded to P on summer crops. This was the case for all crops in both of these experiments. The soil-test P in the no-P plot of the TYR dropped to Medium by 1983 but has not yet shown response. In the NPK, the P level remains High after 29 years without P application, table 15. Both experiments have produced response to 20 to 30 pounds per acre of K 20 on most crops. Cotton A large response to 60 pounds per acre N was found in both experiments, with small insignificant increases from the next higher I- 0 TABLE 13. RELATIVE YIELDS (PERCENT) OF CORN, GRAIN SORGHUM, CO'TrON, SOYBEANS, AND WHEAT IN THE 2-YEAR ROTATION FERTILIZER EXPERIMENT ON LUCEDALE SCL AT THE PRATTVILLE FIELD, 1959-82 Cottons, 6of9 Soybeans4 7of11 2of4 Corn Grain 7of117of11 years, 3-yr.av.,1sorghum.,eIndex, 3 z Treatment No. 1 2 6of9 Wheat6 yers 33 Soil test 1983 9*1U Variable,sogu Vareyr. lb./acre 1959-67 57 av., 1968-78 92 years, 1979-82 90 years, 1959-67 yr. av., -r 196878 60 1979-81 60 182 L/ar - 16....No N 11.....30, 60, or 90 N 14.....60, 90, or 120; N 10.....120 or150N 53 93* 100 97 41 95* 2,480 lb. 100 90* 100 - 103 100 - 103 100 - 79* 100* 95 83 98* 93 bu. 100 96 102 - 87* 100* 103 84 101* 97* 100 99 103 82* 100* 95 97 100* - 79* 97 - - 6in Ca 1180 P 38M H 390 - x m 12.....No P 6.....No K 14.....60 K20 15.....120 K20 4..... Nolime 5.....No magnesium 2.....No 7...30 K 2 0 100 77 97* 39 bu. 95 62* 97 - 93 83 100* 29 bu. 97 62* 110 97 90 76 bu. 97 91* 97 - 101* 91 106 87 79 86* 100 103 K 67 120 M M 70 80 67 bu. 88 58* 106 36 97 21* 112 100 97 78 bu. 87 39 bu. 100 79* 102 - 230 340 Ca 300 Mg 57 - 1130 H 190 pH4.8 H 110 - z -4 HA 8. Plusmicronutrients sulfur 102 103 93 100 105 103 - Untreated .... 17..Untreated untill1979 'Original 20 21 44 23 - 55 90 5.7, P - 18 15 62 41 33 (H) , K - 105 (M) , 27 Mg - 15 90 lb./acre - - soil-test values in 1929 were: pH lb.! acre (H). 2Final soil-test values in 1983 of the 120-60-60 treatment, no. 14, were: pH - 6.2, P'- 169 lb./acre (NH) , K - 230 (H) , Mg - 187 (H) , Ca - 1,180 (H) . 3N rates for cotton in 1959-67 were 0, 60, and 120 lb./acre. For corn in 1959-78. rates were 0.60, and 120 lb./acre N, and in 197981, 0, 90, 120, and 150 lb./acre N. 4N rates for soybeans in 1968-82 were 0, 30, and 60 lb./acre. rates for grain sorghum in 1982 were 0, 60, 90, and 120 lb./acre. 6N rates for wheat in 1968-78 were 0, 50, and 80 lb./ acre. *Indicates significant difference at 5 percent level. lb./acre 120 lb./acre (H) , Ca 960 5N TABLE 14. RELATIVE YIELDS (PERCENT) OF COT'TON, SOYBEANS, ON THE PRATITVILLF AND FIELD, 19655-82 COiRN FROM DIFFEREN-r RATES OF N, P, AND K S Treatment no. ' Lb./acreCornSoybeans, Seedcotton 197072(3) Soybeans, residue 1973-75(3) Corn 1976-80(5) applied N2 '65, '66, '68, '69 1965-69(4)1 residue 1981-82(2) 95 95 96 Nitrogen 41 70 96* 0 2 3 4 5 6 7 8 9 10 5 11 12 13 14 15 5 16 6 S.... S.... . . . . . . . . 60 90 120 150 51 73* 94* 103* 97 105 93 51 94* 99 100 105 100 109 101 107 100 100 97 93 101 103 51 bu. 100 98 16 15 100 100 101 98 100 98 101 100 91 99* 98 106 100 2,560 1b). 91* 100 101 Phosphorus 101 96 98 94 100 Potassium 76 89* 102* 101 98 2x,660 101 100 95 95 106 108 28 bu. 86* 78 7 100 103 101 101 100 P20 5 S.... S.... . . . . . . . . 0 20 40 60 100 107 102 96 103 100 103 100 99 101 107 34 bu. 81* 97 16 12 K20 0 20 40 60 80 100 No lime No sulfur since 1969 11 10 7 9 'Numbers in parenthesis are years included in the average. 2Only the 30- and 60-lb/acre N rates applied on soybeans. Cv LSD (05) "Indicates a significant difference from the next lower or higher rate at the 5 percent lev el. TABLE 15. Rates of SoIL-TES VALUES FROM DIFFERENT RATES OF P AND K ON LUGEDALL SGL AT PRATTVILLE, 1954-83 Values, pounds acre Soil1977 residue (P) test 1983 index, P205 and K201 1972 1954 1965 1969 residue 1975 residue 1981 1983 rating pct. Soil-test phosphorous 0 20 93 98 133 100 110 88 95 90 94 82 94 73 39 86 H 8 40 60 100 162 178 213 142 150 184 126 148 200 116 142 197 114 128 230 111 105 207 129 131 176 V High Medium 260 350 Soil-test potassium (K) 20179 40205 80267 10284 Soil pH 5.6 5.5 164 207 2 282 5.4 135 170 217 290 5.3 104 136 168 260 5.3 129 140 202 270 5.2 13 172 232 280 5.0 173 24013 284 30917 5.2 High V High 100 160 'Rates of P2 0,5 applied annually except in 1970 through 1977 when only the 100-lb./acre rate was applied. 34 ALABAMA AGRICULTURAL EXPERIMENT STATION rate. Therefore, it appears that somewhere between 60 and 90 pounds per acre of N would be the best rate on this soil. No response to P was found in either experiment at High soil-test levels. The first 30 pounds per acre of K 20 in the TYR increased yield 1,340 pounds seedcotton per acre with no significant increase from the higher rates. Less response to K was found in the NPK where the soil-test level was higher than in the TYR. A 10 percent increase from lime was found in both experiments, but no increase from Mg occurred because this soil was High in Mg where none had been applied since 1929. Soybeans An increase of about 3 bushels per acre of soybeans from 30 pounds per acre of N was found in both periods reported. This increase was found at several locations. It was most likely due to the fact that soybeans followed corn or wheat which received different rates of N and produced widely divergent yields of grain and crop residues. The soybeans that received no N followed corn that received no N. Where soybeans received no N and followed well-fertilized corn or wheat, this difference did not occur. The organic matter content and perhaps other soil factors were adversely affected by years of cropping without nitrogen fertilization. No response to P on soybeans was found in either experiment. Good response to 30 K 20 was found in the TYR at Medium soil K, but none in the NPK where the soil-test K level was High. Soybeans responded to lime in both experiments but not to Mg, S, or micronutrients. Soybeans produced higher relative yields on the old untreated plots than did the other crops, and yields returned to 90 percent of maximum within 3 years after liming and fertilization were resumed following 48 years of cropping without treatment. Corn Corn has generally responded to 120 pounds per acre of N at this location. A series of droughts occurred in recent years and 90 N was adequate where yields were limited by moisture deficiency in both experiments. Corn, like cotton and soybeans, has responded to 30 pounds per acre of K20 in the TYR but not to the other nutrient elements applied. LONG-TERM FERTILITY EXPERIMENTS 35 Wheat and Grain Sorghum Wheat responded to 80 N and to lime in the three-crop rotation but not to P, K, Mg, or micronutrients. Grain sorghum responded to 90 N and to K. It produced a tremendous response to lime at pH 4.8. TENNESSEE VALLEY SUBSTATION, DEWEY SIL (TABLES 16, 17, AND 18) This red Dewey silt loam is of limestone origin, has a cation exchange capacity of 10-12 milliequivalents per 100 grams, and is the most fertile of the soils used in these experiments. It is rated by the soil testing laboratory on different scales for both P and K from the other soils. It requires lower levels of soil-test P and higher levels of soil-test K to reach Medium and High ratings and is therefore classified as a Group III soil, which requires CEC greater than 9.5 milliequivalents per 100 grams. Cotton Data from both experiments indicate that 60 pounds per acre N is adequate for cotton on this soil. The general recommendation for cotton on all soils since 1970 was 90 pounds per acre N until the rate was reduced to 60 N for these Group III soils in 1982. These and other data led to this change after the Tennessee Valley area became the most extensive part of Alabama for cotton production. Many growers on Dewey and Decatur soils similar to this one were having problems with too much growth from 90 N, which increased problems with insect and disease control and sometimes reduced yields. Use of P increased yield of cotton about 10 percent in both experiments. This was less increase than was anticipated because both areas were Low in P. Soils rated low are expected to produce less than 75 percent of maximum yield without P fertilization. The soil-test P level remained about the same where none was applied for 54 and 29 years in the two experiments. Only 20 pounds per acre of P 20 5 was required to produce top yields in the NPK. All rates applied increased the soil level. These data indicate that more P is being recommended than is producing response because at Medium and High levels no response was found, TABLE 16. RELATIVE YIELDS (PERCENT) OF CORN, (;RAIN SORGHUM, COTITON. SoxYBEANS, AND WHEAT IN THE 2-YEAR ROTATION FERTILIZER EXPERIMENT ON DEWEY SIL AT THE TENNESSEE VALLEY SUBSTATION, 1959-82 Cotton', Soybeans4 CornGa yr.-6 av, 1959-67 v.of Treatment', 2 6of99 7of11 year Grain 2-yr.av., sorghums,/ Wheat6 7of Soil test 1983 Index, No. Variable, lb./acre yr. av., P a. 1959-67 85 93 100 - yas 1968-78 109 106 100 - 4-yr. av., 197982 98 92 100 - yer, 1968-78 yeors1 99-1 192 yer, 1968-78 40 b/ar - pct. - 16.....No N 11.....30, 60,or 90 N 14.....60, 90, or 120 N 10.....120 orl5ON 54 87* 100* - 68 92* 100* - 43 93 96* 100 85 91* 98 99 100 104 122 101 93* 100 - - - Ca2010 - H 400 - -Yb. a 12..... No P 6.....NoK 7...301K20 14.....60 120 15.....120 K2 0 4..... No lime 5..... No magnesium 2.....No sulfur since 1977 90 93 102* 2,500 lb. 99 108 106 - 91* 91 100* 32 bu. 100 72* 103 - 8..Plus micronutrients 1..Untreated 17..Untreated untill1979 98 57 50 97 50 34 92* 90 98* 50 bu. 100 84* 96 94 100 58 86 89* 98 99 80 bu. 98 100 105 - 96 94 95 103 bu. 100 104 101 - 89* 102 P 13 K116 L 70 M70 100 27 16 98 43 45 102 99 bu. 96 103 100 95 102 33 70 95 101 bu. 104 120 94 88 127 60 87 96 45 bu. 98 96 91 - 155 372 369 Ca 1550 Mg 150 - M 80 H 150 H 150 pH5.4 H 330 - 000 9I- 104 - 18 'Original soil-test values in 1929 were: pH - 5.7, P - 14 lb./acre (L), K - 220 lb./acre (H), Mg - 150 lb./acre (H), Ca - 1600 lb./acre (H). 2Final soil-test valtues in 1983 of the 120-60-60 treatment, no. 14, were: pH - 6.2, P 100 lb./acre (VH), K - 372 (H), Mg - 386 (H), Ca - 2,010 (H). 'N rates for cotton in 1959-67 were 0, 60, and 120 lb./acre. For corn in 1959-78, rates were 0, 60, and 120 lb./acre N, and in 1979-81, 0, 90, 120, and 150 lb./acre N. 4N rates for soybeans ill 1968-82 were 0, 30, and 60 lb./acre. 5N rates for grain sorghum in .1982 were 0, 60, 90, and 120 lb./acre. 6N rates for wheat in' 1968-78 were 0, 50, and 80 lb./acre. *Indicates significant difference at 5 percent level. z -4 -I 0 z TABLE 17. RELATIVE YIELDS (PERCENT) OF COTTON, SOYBEANS, AND CORN FROM DIFFERENT RATES oF N, P, AND K ON THE TENNESSEE VALLEY SUBSTATION, 1965-82 Treatment no. Lb./ acre api Seedcotton 1965-69(4)1 '65,a '66,d'68,'69 1970-72(3) Soybeans, residue 1973-75(3) 1976-80(3) Corn, Soybeans, 1981-82(2) residue '76,'78,'7919 1 2 3 4 5 6 7 8 9 10 5 11 12 13 14 15 5 16 6 S.... S.... . . . . . . 02 30 60 90 120 150 02 20 40 60 100 Nitrogen 57 87* 101* 103 100 101 92 97 103 103 100 69 106* 100 102 92* Phosphorus 102 100 104 101 100 94 100 39 80* 101 * 100 104 98 104 96 99 100 100 99 106 102 100 107 bu. 103 102 14 18 . . 103 96 107 106 S.... . . . . S.... .. . .. . . . . . . . . . . . . . 95 99 95 101 100 91 97 95 104 104 46 bu. 92* 96 10 8 103 100 100 99 Potassium K20 103 93 0 106 97 20 103 99 40 100 101 60 102 101 80 100 2,760 lb. 2,800 lb. No lime 104 99 No sulfur (since 1969) 14 10 CV 12 8 LSD (05) 'Numbers in parenthesis are years included in the average. 2Only the 30- and 60-l-b.! acre N rates applied on soybeans. 98 98 105 107 110 50 bu. 96 107 13 NS *Indicates a significant difference from the next lower or higher rate at the 5 percent level. TABLE 18. Rates of P2 0 and K201 SOIL-TEST VALUES FROM DIFFERENT RATES OF P AND K oN DEWES" SIL AT THE TENNESSEE VALLEY SUBSTATION, Values, pounds! acre 1954 1965 199 residue residue 1981 Soil- 1954-83 1983 1983 test rating index, pct. Soil-test phosphorus (P) 0 20 11 12 13 18 15 19 1370 16 7168 40 60 100 16 21 33 24 29 48 24 26 ,_66 20 22 40 21 23 38 18 20 36 Medium High 80 80 120 Soil-test potassium (K) 40194. 60 80 100 Soil pH 229 248 240 5.6 5.6 183 240 252 280 5.5 190 230 262 300 5.4 200 210 230 260 5.4 301 290 338 390 5.7 21810 229 212 277 5.6 High 100 100 120 'Rates of P 2 0 5 applied annually except in 1970 through 1977 when. only .the 100-lb../ acre rate was applied. LONG-TERM FERTILITY EXPERIMENTS 39 Cotton responded to only 20 or 30 pounds per acre K20 in the two experiments. Considerable variation in soil-test K was found among seasons. Soil-test K fluctuates from one season of the year to another or among years much more than does soil-test P. This is due to differences in rainfall and in crop growth and is greater in soils testing High than when K levels are lower. This demonstrates why growers should keep soil-test records so they can evaluate changes between samplings. It also demonstrates why the fertility index is more useful to growers than are pounds per acre figures, which fluctuate much more on these soils that retain large amounts of K. Cotton did not respond to lime on this soil at pH 5.4. This demonstrates why these Group III soils are allowed to drop below pH 5.6 before lime is recommended. The sandy soils of groups I and II receive lime recommendations for most crops when they drop below pH 5.8. Cotton did not respond to Mg or micronutrients. Soybeans Soybeans produced about a 10 percent yield response to both P and K in the TYR, and to K in the NPK. The lowest rates of 20 to 30 pounds per acre of P2 0 5 and K2 0 were adequate to produce top yields and to increase soil-test levels of both nutrients. The data indicate that present recommendations based on soil tests are more than adequate for this soil. Response to lime was found in both experiments, but none to the other nutrients. Corn The best rate of N for corn varied from 90 to 120 pounds per acre N. This supports the present recommendation of 120 pounds per acre for this crop since no further increase was found from 150 N in either experiment. Response of 8 to 10 bushels per acre to P was found in the TYR but not in the NPK. Little or no response to K was found in either experiment, even after 54 years without K application. Corn did not respond to lime at pH 5.4, or to Mg, S, or micronutrients on this soil. Wheat and Grain Sorghum Wheat yields averaged 45 bushels per acre for the 7 years reported. It responded to 80 pounds per acre N and to P, but not to 40 40_ALABAMA ALABAMA AGRICULTURAL AGR A.a UL .. u rAL mmwW FXT EXPERIMENT f. STATION . a I\ K or lime. The 91 percent yield without Mg indicates a response, but this is doubtful because this soil is much higher in Mg than any of the other soils. Grain sorghum in its first year, 1982, produced well, but there was much variation between the two replications and differences are of doubtful significance. The high yields without P, lime, and with micronutrients do not agree with previous data on other crops. UPPER COASTAL PLAIN SUBSTATION, SAVANNAH SCL (TABLES 19 AND 20) This substation was not established until 1946 so does not have the TYR experiment. The NPK was established in 1954 along with the other locations. This gray soil has a CEC of about 8 milliequivalents per 100 grams, which is about the same as the two Lucedale soils. The site was originally Low 70 in P and Medium 90 in soil-test K, and was therefore expected to produce some response to both nutrients. The recommendation for a cornsoybean rotation would be 120-160-80 pounds of N-P 20 5-K 20 for corn with no fertilizer to the soybeans. For cotton, the recommendation would be 90-100-50 annually. Cotton Yields at this location were lower than at most other locations and data showed considerable variation. The best rate of N for cotton was 60 to 90 pounds per acre. Response to 20 to 40 pounds per acre of P 20 5 and K 20 are indicated. Cotton also produced responses of about 200 pounds per acre of seedcotton to lime at pH 5.1. Soybeans The first time this experiment was in soybeans, 1973 through 1975, the P and K treatments were in a residual study and received no P or K from 1970 through 1977 except for the standard treatment, which received 100 pounds per acre of each every year. Yields averaged 47 bushels per acre for 3 years with no response to N. The residue from 20 pounds per acre of P 20 5 was adequate, although the soil-test level of these plots was Medium in 1971 and dropped slightly into the Low range by 1975. When rates were re- TABLE 19. RELATIV'E YIELDS (PERCENT) OF COTrON, SOYBEANS, AND CORN FROM DIFFERENT ON THE UPPER COASTAL PLAIN SUBSTATION, 1965-82 RATES OF N, P, AND K Seedcotton Treatment Lba7.cre nople N'2 l ' .... . . .. . . . . Soybeans,, 10 72 71, on 178( 956() '65, '66, '68, '69 residue '72 1973-75(3 residue '76, 52 79 Soybeans, 18-22 3 4 5 6 S.... c .... . . . . 0 30 60 90 120 150 59 67* 90* 85 100 92 83 88 87 94 100 P2 0. .... . . . . . . . . . 8 9 I0 5 11 12 13 14 .. 0 20 40 60 100 K2 0 0 85 20 94 40 93 60 94 80 92 100 1,930 lb. 16 No lime 88* 6 No sulfur since 1969 CV 14 LSD (05) 12 'Numbers in parenthesis are years included in the average. 2Only the 30- and 60-lb./acre rates applied on soybeans. *Indicates a significant difference from the next lower or higher rate at the Nitrogen 70 84* 100* 92 100 94 Phosphorus 91 97 95 99 100 Potassium 60 75* 93* 91 94 2,330 lb. 91* 100 99 96 99* 109 100 95 90 111 101 120 100 103 102 103 104 112 73 bu. 91 93 21 32 92 1ll 109 91 98 102 105 100 88 92 93 97 98 47 bu. 92* 102 8 8 5 percent level. 70 100 98 104 100 98 97 95 102 110 27 bu. 102 97 21 22 I TABLE 20. SOIL-TEST VALUES FROM DIFFERENT RATES OF P AND K ON SAVANNAH SCL AT THE UPPER COASTAL PLAIN SUBSTATION, 1954-83 Rates of P2 0 5 and K2 0' Values, pounds! acre 1975 1972 1969 residue residue Soil-test phosphorus (P) 0 30 16 17 12 10 1 1Lw7 Soil1981 1983 test rating 1983 index, 1954 1965 pct. I- 20 27 29 24 22 3 59 4026 60 53 52 29 2 3253912 39 124 107 174 182 260 5.0 42 106 102 18010 201 254 6.0 Medium High Medium 90 210 80 A 1087871 018 60198 80 100 Soil pH 5.6 221 268 5.1 1786 165 196 243 5.1 75 140 160 249 5.0 70 Soil-test potassium (K) 104 165 152 291 5.0 2-4 C f0I- C 110 140 m x m m 'Rates of P2 0 5 applied annually except in 1970 through 1977 when only the 100-lb./acre rate was applied. z FERTILITY LONG-TERM 5 opimm a a a 'M vio- 0 anm EVS EXPER IMENTS muffff &'ll I sumed in 1978, the level increased to Medium again by 1981. Application of 60 pounds per acre of P 20 5 was required to increase the level of this soil to High. Soybeans were less responsive to K than was cotton. The residue from 40 to 60 pounds per acre of K 20 maintained the soil-test level at High throughout the 8 years without application and produced 45 bushels per acre of soybeans. The data indicate that 20 pounds per acre K 20 annually would have been adequate to produce top yields and maintain a High soil-test level. When the plots were returned to soybeans in 1981 and 1982, there was response to 20 pounds per acre P2 0 5 but none to K20. These were the same responses as found on corn in 1976 and 1979. Soybeans in the first period produced an 8 percent, or 4 bushels per acre, response to lime at pH 5.0. These plots were apparently limed by mistake in 1981, for the pH increased to 6.0 and no response was found in 1981-82. Corn Corn suffered from drought during most years in the period 1976-80 and yields from the best 2 years averaged only 73 bushels per acre. It responded to 60 pounds per acre N, with no response to the higher rates. The no-N plots produced 38 bushels per acre so the increase from N was limited by drought to 35 bushels per acre. This helps explain why 60 N was adequate, for much research has shown that 11/ to 2 pounds per acre N is adequate to increase yield 1 bushel per acre. Corn responded to 20 pounds per acre of P 2 0 5 at Low soil test P. It did not respond. to K at Medium although no K was applied from 1970 through 1977 except on the standard treatment. A 10 percent increase in yield from lime was found at pH 5.0. SUMMARY AND CONCLUSIONS Average yields and soil-test values for all locations of both experiments are presented in tables 21, 22, and 23. Average response to rates of N are useful in making N recommendations because they do not vary much among locations. All of these soils are low in organic matter and there is not a satisfactory soil test to determine the N-supplying capacity in individual soils. Therefore, results of these N rate studies should agree with present N recommendations for the crops involved. TABLE 21. AVERAGE RELATIVE YIELDS (PERCENT) OF CORN, GRAIN SORGHUM, EXPERIMENT COTTON, SOYBEANS, AND WHEAT IN THE 2-YEAR ROTATION 1959-82 FERTILIZER AT Six LOCATIONS, Treatment, 2 Cotton3, 6 of 9 Variable, lb./acre yr. av., 1959-67 (36) Soybeans 7 of 11 years, 4-yr. av. 1979-82 1968-78 4 6 f 9 Corn 7 f 11Grain 3 Vheat6 2-yr. av., 1979-81 yr. av., 1959-67 years, 1968-78 (42) (42) 96 98 100 - (18) (36) ((6)(42) 55 98 100 96 sorum 66 89 100 100 7of11 Soiltest,1983 years, 1968-78 Av.of 6 locations 16..... No N 11..... 30, 60, or 90 N 14.....60, 90,or 120 N 10.....120 or150N 61 100 100 - 95 99 100 - 45 87 100 - 62 95 100 - 43 92 100 -- Ca 1,220 P 22 K 66 104 202 238 Ca 480 Mg 64 - 12..... No P 6..... No K 7.....30 K2 0 14.....60 K 2 0 15.....120 K 2 0 No 4..... lime 5 ..... magnesium No 2..... No sulfur 95 53 99 2,420 lb. 100 91 100 - 89 76 97 37 bu. 101 62 94 - 90 67 95 40 bu. 99 66 92 94 90 84 98 87 1). 101 87 102 - 93 76 97 100 bu. 99 77 102 - 90 69 97 86 bu. 97 56 101 90 94 73 89 89 bu. 99 36 97 93 89 99 100 37 bu. 97 81 98 - 8.....Plus micronutrients 1..... Untreated 17 ....Untreated until 1979 106 27 29 101 44 36 100 45 86 98 20 17 100 29 32 100 27 79 102 24 94 99 16 - values in 1929 were: pH - 5.9, P - 39 lb./acre (M), K - 109 lb./acre (M), Mg - 68 lb./acre (H), Ca 'Average original soil-test 960 lb./acre (H). 2Final soil-test values in 1983 of the 120-60-60 treatment, no. 14, were: pH - 6.2, P - 126 lb./acre (VH), K - 202 (H), Mg - 206 (H), Ca - 1,220 (H). 3N rates for cotton and corn in 1959-67 were 0, 60, and 120 lb./acre. Rates for corn in 1959-78 were 0,60, and 120 lb./acre N, and in 1979-81, 0, 90, 120, and 150 lb./acre N. 4N rates for soybeans in 1968-82 were 0, 30, and 60 lb./acre. 5N rates for grain sorghum in 1982 were 0, 60, 90, and 120 lb./acre. 6N rates for wheat in 1968-78 were 0, 50, and 80 lb./acre. *Indicates significant response at 5 percent level. TABLE 22. AVERAGE RELATIVE YIELDS (PERCENT) OF COTON, SOYBEANS, AT SEVEN LOCATIONS, 1965-82 AND CORN FROM D)IFFERENT RATES OF N, P, AND K Fr SeedcottonSobas Treatment Lb./ acre applied 1965-69 (26)1 1976-80 (23) 1981-82 (8)-8 1no. 90-2(16) residuereiu Nitrogen 1737(1) Corn, 1968023d98-2e8 Soybeans, 0 z 0 -I w N'2 . .. . . . ?*1 . . .. . . . . . . . . . . . . . . 0 30 60 90 120 150 55 79 97 97 100 99 85 92 97 99 100 65 91 95 99 99 2,410 lb. 88 61 82 96 98 100 96 Phosphorus 99 101 98 42 89 98 100 99 85 98 97 101 100 91 98 102 101 101 93 bu. 73 918 97 102 102 idin . . . . . . . . . . . . . . . . P20 5 0 20 40 60 100 7 8 9 10 5 11 12 13 14 15 5 16 6 . . . . . . . . . . . . . . . . 89 96 96 98 100 Potassium 59 80 92 97 99 2,450 lb. 88 85 96 98 10.1 100 88 93 96 101 101 38 bu. 82 99 92 101 99, 101 100 96 97 100 101 106 39 bu. 90 100 z -I N, . . . . . . . . . . . . . . . . K2 0 0 20 40 60 80 100 No lime No sulfur since 1969 iNumbers in parenthesis are location years included in the average. 2Only the 30- and 60-lb./ acre N rates applied on soybeans. TABLE 23. AVERAGE OF SOIL-TEST VALUES FROM DIFFERENT RATES OF P AND K AT SEVEN LOCATIONS, 1954-83 Rates of P2 05 and K 2 O1 1954 1965 1969 Values, pounds/acre 1972 residue residue 29 4 1975 1981 1983 Soil-test phosphorus (P) 0 20 Medium40 60 High 36 33 44 58 70 33 47 6 2 80 29 40 53 66 24 75 28 6 96 100 0 20 Medium 40 60 High 80 100 Soil pH 101 96 98 114 137 160 172 186 5.4 105 110 107 85 97 131 152 166 215 5.3 121 92 118 155 178 188 235 5.3 113 95 122 149 162 175 212 5.1 5.6 Soil-test potassium (K) 93 86 114 102 142 127 164 150 184 168 204 217 5.3 5.3 'Rates of P20 5 and K2 0 applied annually except for 1970 through 1977 when all but the 100-lb/acre rates were discontinued for varying numbers of years among locations. 'Ratings indicated are those used for corn and soybeans on Group 11 soils, which have CEC values from 4.6 to 9.0 mneq/ g. Ratings used for other soils and crops can be found in Alabama Agricultural Experiment Station Circular 251, 1981. Soil Test Fertilizer Recommendations for Alabama. 100 LONG-TERM FERTILITY EXPERIMENTS 47 Responses to P, K, Mg, S, lime, and micronutrients are dependent on the level of these nutrients and of pH in the individual soils. Soil tests for P, K, Mg, and pH should be good indicators of the need for fertilizers or lime to correct deficiencies, if the soil-test values are properly calibrated with responses in field experiments. These average values are therefore of limited use, except that they indicate the magnitude of responses that may be anticipated for the various nutrients. For example, there were few average responses to Mg, S, or micronutrients, but some locations did show need for these nutrients. Averages of soil-test values are also of limited use, but they do show the effects of longterm applications on soil fertility. Responses to Nitrogen Consistent responses to N were found on corn, cotton, wheat, and sorghum at all locations in the TYR. Determination of the rates to which response was found requires careful study of the tables and the footnotes, because the rates varied among the crops. The numbers of location-years (one year at one location) are indicated in parenthesis for each column. Corn, in 90 location-years, responded to 90 pounds per acre N, with a small increase of 5 bushels per acre or less from 120 N over 90 N. During most of these years, corn followed soybeans, which usually reduces the N requirement by about 30 pounds per acre. Corn in the NPK produced 98 percent of 93 bushels per acre from 90 pounds per acre N. The 120- and 150-pound-per-acre rates did not increase yields over 90 N. These data agree with the present recommendation, which is 120 N with a comment that when following soybeans the rate be reduced 20 to 30 pounds per acre. Cotton, in 36 location-years in the TYR from 1959-67, produced 2,420 pounds per acre of seedcotton from 60 pounds per acre of N. The 120 pound-per-acre N rate did not increase the yield over 60 N. Largest increases from N were on the sandy soils and the least increase was found on the Tennessee Valley Substation, which produced 85 percent of top yield without any N. In the NPK experiments, 60 N produced averages of 97 and 96 percent of top yields of 2,410 and 2,450 pounds per acre of seedcotton for the two periods. Increasing the N rate to 150 N slightly reduced yields in both periods. Present recommendations from the soil testing laboratory are 90 pounds per acre N with a comment that where 48 ALABAMA AGRICULTURAL EXPERIMENT STATION excessive growth has caused problems with late maturity, insects, or boll rot, the rate be reduced by 20 to 30 pounds per acre. These data support this comment, especially for the soils similar to those on the Tennessee Valley Substation. Wheat in a three-crop rotation (corn-wheat-soybeans) averaged 37 bushels per acre in 42 location-years. Application of 60 pounds per acre N in the spring after 20 N at planting increased yields 4 bushels per acre over only 30 in the spring and 20 in the fall. Four of the six locations showed response to 80 N over 50 N, with the most southerly locations, which produced the lowest yields, not responding. Sweet sorghum at Sand Mountain and Brewton from 1980 through 1982 in the NPK responded to 60 pounds per acre N at both locations. No further increases were found from 90 and 120 pounds per acre N. Grain sorghum in TYR in 1982 at six locations responded to 90 pounds per acre N and averaged 89 bushels per acre for all locations. Present recommendations are 80 pounds per acre N for both sweet and grain sorghum. This limited amount of data indicates that this rate is correct. Soybeans generally did not show any response to N. Increases of 3 to 4 bushels per acre at Monroeville, Prattville, and Wiregrass are attributed to the soybeans following corn that received no N and produced low yields. This affected the amount of residue returned to the soil by corn and is not a practice that would be recommended. Soybeans that followed well-fertilized corn showed no increase in yield from N. Peanuts at the Wiregrass Substation showed no response to N in either experiment, although peanuts had not been grown on either area for many years, if ever. Responses to Phosphorus The TYR does not include a good rates study on response to P. The no-P plots from 1959-82 had received about 30 pounds per acre P 2 0 5 from 1929 through 1957 and were either Medium or High in P when this comparison was started. They have received no P in 25 years and have dropped to Low or Medium at all locations. The standard treatment has received 60 pounds per acre of P 20 5 annually since 1929 and was High or Very High at all locations throughout this study. Four of the six locations have consistently produced responses to P. The Hartsells soil on the Sand Mountain Substation has LONG-TERM LONG.TERM.FERTILITY FERTILITY EXPERIMENTS EXPERI...MENS 49 " been the most responsive because it has been Low in soil-test P since about 1960. The Dothan soil at the Wiregrass Substation and the Lucedale soil at Prattville have not responded to P on any of these crops because they were Very High in soil-test P in 1959. They dropped back into the Medium range about 1970, but still have not produced response. The other locations have produced moderate response to P at Medium and Low soil-test levels. Average responses to P of the four responding soils have been about 15 bushels per acre of corn and 7 bushels per acre of soybeans. Average increases of all crops from P at all locations in TYR have been about 10 percent. The NPK experiments include an excellent P20 5 rates study and a period when all but the highest rate were in a residual study. As in the TYR, the Hartsells soil, which was Low in P in 1954 and remained at about the same level from 1954 through 1982, produced the greatest response to P. Cotton, corn, and sorghum responded to rates up to 60 pounds per acre of P2 0 5 while soybeans responded to 40 pounds. The Benndale soil at Brewton, which also was Low, produced responses to 20 P2 0O on cotton and to 40 P 20 5 on soybeans, corn, and sorghum. Soils at Monroeville, Tennessee Valley, and Upper Coastal Plain, which were all Low in P, produced responses to 20 P20 5 but not to higher rates. Soils at Prattville and Wiregrass, which were High in P, showed no response to P by any crop from 1954 through 1982. The Dothan soil dropped into the Medium range about 1960 and to Low in 1978, but still did not respond. During the residual period from 1970-72, when P was applied only to the standard treatment, cotton continued to produce at 1965-69 levels at most locations. Some locations were continued in residue through 1977 and all others returned to residue studies in 1976 and 1977 without reduction in yields of soybeans and corn. All rates from 20 P 20 5 through 100 P 2 0)5 annually increased soil-test P in proportion to amounts applied. When applications were discontinued, levels dropped slowly and in proportion to levels attained. Plots receiving 100 pounds per acre P2 O5 annually reached High at all locations in 7 years of application. All except the Dewey soil reached Very High after about 10 more years of application. Data in both experiments show that soil-test P values were reliable indicators of response. Soils that were Low in P generally responded. Soils that were High or Very High did not respond so 5....A............... ALABAMA AGRICULTURAL LA... .... L EXPERIMENT .. . 009%o1 wTA STATION |.l| to P on any crops. Medium soils did not respond in most cases, but rates recommended at Medium increased soil P consistently. The data show that present recommendations based on soil tests are adequate to produce top yields and where levels are not already High to increase soil P reserves. Once these reserves reach High, applications may be discontinued so that growers may benefit from P stored in the soil. Responses to Potassium The TYR includes a good rate of K study. The no-K treatment has had none applied since the experiment was started in 1929. The other treatments had low rates until 1957 when the present 30-, 60-, and 120-pound-per-acre K20 rates were started. Soil-test K in the no-K plots was Low or Medium at all locations in 1957. Cotton responded to 60 pounds per acre K20 at Sand Mountain and Brewton, on soils Low in K. Other sites, which were Medium, responded to 30 pounds per acre K20. The average response to 30 K 20 at the six locations was 1,140 pounds per acre of seedcotton. Greatest response was on the sandy soils with least on the Dewey soil where K increased the yield only 220 pounds per acre. Corn and soybeans generally responded to only 30 pounds per acre K 20, except on the Hartsells where some response was found from the second increment of K 20. Wheat showed little response to K. There were no cases on any crop where 120 K 20 increased the yield over 60 K 20. Application of 30 K 20 annually maintained soil-test K about constant and the higher rates increased soil K on all soils. Response to K in the NPK was similar to that in the TYR. Cotton responded to 60 K 20 at Sand Mountain and Brewton; to 20 to 40 K20 at Monroeville, Prattville, Wiregrass, and Upper Coastal Plain; and to 20 pounds per acre K20 in the first period at Tennessee Valley. Residual K stored in the soils from 1954 through 1969 enabled them to produce at about the same level as before during 3 years without application of K in 1970-72. Corn and soybeans responded to 40 K20 at Sand Mountain and Brewton and to 20 K20 at other locations except Prattville, where soil-test K was High for these crops, and at Tennessee Valley where corn did not respond to K. Sweet sorghum responded to 60 K20 at Sand Mountain and to 40 K20 at Brewton. Soil-test K showed more variation among locations and years than did P. This variation among soil types demonstrates why it LONG-TERM FERTILITY EXPERIMENTS 51 is necessary to classify the soils into groups based on CEC to give reliable recommendations. Highest levels attained on the Benndale would be rated Low on the Dewey. Soil K levels of the no-K treatments dropped little in 30 years, the average drop being from 101 pounds per acre to 95 pounds per acre of soil-test K, table 23. All rates from 20 up to 100 pounds per acre K2 0 applied annually increased soil-test values. Greatest proportional increases were found from the 20- and 40-pound-per-acre K 20 rates except for the 100-pound-per-acre rate, which was applied continuously through the residue periods. Soil-test K ratings based on classification of the soils into three groups served as reliable bases for making fertilizer recommendations. The data show that present recommendations based on this system are reliable and should be adequate for these crops on Alabama soils. This includes the recommendation that no K be applied to soils testing High. Responses to Lime Large responses to lime by cotton and corn were found in the TYR on all except the Dewey soil. Average response to lime by cotton at five locations was 280 pounds per acre of seedcotton. Response by corn increased with time as the pH dropped. Soybeans responded to lime at all locations, averaging 14 bushels per acre response in both periods reported. Grain sorghum was the most sensitive to low pH of all crops grown, averaging a 57-bushel-peracre response to lime for 1982. Wheat averaged 7 bushels per acre response to lime, ranging from 2 to 20 bushels per acre for the six locations. The unlimed treatment in the NPK experiments also produced low yields, and the deficiency grew progressively worse with time as the pH dropped. The sandy soils were affected more than those with higher CEC at Prattville and Tennessee Valley. Soil pH gave good indications of the need for lime. Differences in acidity levels that crops will tolerate are used in making lime recommendations from the soil testing laboratory. Soils of Group III with higher CEC produce satisfactorily at lower pH levels than will the lower CEC soils of groups I and II. Responses to Magnesium and Sulfur The TYR included a comparison between dolomitic and calcitic lime throughout the periods reported. Differences in yield between treatments 5 and 14 are primarily due to Mg, which is 52 ALABAMA AGRICULTURAL EXPERIMENT STATION supplied by dolomitic but not by calcitic lime. Cotton and corn did not respond to Mg on any of these soils, although the Benndale and Hartsells are Low. Soybeans responded to Mg on the Benndale, Hartsells, and Lucedale at Monroeville, averaging about 6 bushels per acre response. The Lucedale soil at Monroeville had 55 pounds per acre of soil-test Mg in 1983, which is just above the adequate level of 50 pounds per acre for Group II and III soils. The Benndale and Dothan are Group I soils where the critical level is 25 pounds per acre of soil-test Mg. Peanuts did not respond on the Wiregrass Substation where the soil-test Mg was 46 pounds per acre in 1983. .Both experiments have included a no-S treatment in recent years after having S supplied to all plots as superphosphate or gypsum for many years. The no-S treatment in the TYR was started in 1978. Corn has produced response at four of the six locations, with the Benndale and Dewey not responding. Soybeans showed responses of 3 to 4 bushels per acre at all locations except Prattville from 1979-82. A no-S treatment was started in the NPK experiment in 1970 on plots that had received 30 pounds per acre of S from gypsum annually from 1954 through 1969. This had no doubt built large reserves of S, especially in the subsoils of these plots. The only case where response to S has been found in these experiments was on corn on the Hartsells soil where there was a 30-bushel-per-acre response in the 4-year period 1976-79. Other crops and locations have not yet shown response to S after 13 years without application. Treatment No. 8 in the TYR included a micronutrient mixture containing Zn, Mn, B, Cu, and Mo applied in alternate years beginning in 1959. The only crop that showed positive responses to this mixture was cotton. Increases of 300, 230, and 180 pounds per acre of seedcotton which were found at Monroeville, Brewton, and Wiregrass are attributed to B, for such sandy soils have been found to respond to this nutrient and 0.3 pound per acre of B is recommended for cotton on all soils in Alabama. Rejuvenating Wornout Soils The TYR included treatments No. fertilizer or lime from 1929 through locations include yield percentages mention has been made of them in 1 and 17 which received no 1978. Yield tables for all for these plots, but little previous discussions. They LONG-TERM FERTILITY EXPERIMENTS 53 produced low yields at all locations, showing that none of these soils will produce acceptable yields without fertilizer and lime. Non-legume crops, suclh as corn, cotton, wheat, and sorghum, produced lowest yields on these plots because N becomes the first limiting nutrient under such conditions. Soybeans produced relatively better because yields were limited by P, K, or lime rather than N. In 1979, treatment 17 was limed and fertilized for the first time in 50 years. Yields of corn and soybeans showed dramatic improvement the first year. For the next 3 or 4 years, 1979-82, corn averaged 79 percent and soybeans 86 percent of the yield of the standard treatment at all locations. Grain sorghum at six locations in 1982 averaged 94 percent while the untreated plots averaged 24 percent of standard yield. These data demonstrate that even when soils were badly mismanaged under continuous cropping for 50 years, they recovered rapidly to near normal production when fertilized and limed according to soil-test recommendations. 54 ALABAMA AGRICULTURAL EXPERIMENT STATION ACKNOWLEDGMENT Contributions of the many soil scientists in the Department of Agronomy and Soils at Auburn University in planning and revising these experiments and of the superintendents of the substations and experiment fields and their assistants are gratefully acknowledged. These Department staff members and station superintendents include: Aublurn: M. J. Funchess, J. T. Williamson, George I). Scarseth, W. Volk, I.. E. Ensminger, J. W. Tidmore, N. 1. Volk, and R. 1). Rouse; G(. Tennessee Valley Substation: Webster; .B.B. Fred Stewart, John Boseck, and Sand Mountain Substation: ner, and J. T. Eason; WXViregrass Substation: Starling; R. C. Christopher, S. E. Gissendan- J. P. Wilson, C. A. Brogden, and J. G. F. E. Bertram, F. T. Glaze, and Prattville Experiment Field: I). P. Moore; Brewton and Monroeville Fields: J. W. Richardson, E. L. Carden, E. Brown, and J. A. Pitts; and Upper Coastal Plain Substation: W. W. Cotney and R. A. Moore. Alabama's Agricultural Experiment Station System AUBURN UNIVERSITY With an agricul: tural research unit in S: every major soil area, - 3 Auburn University serves the needs of field crop, livestock, forestry, and horticultural producers in each region in Alabama. Every citizen of the State has a stake in this research program, since any advantage from new and more economical ways of producing and handling farm products directly benefits the consuming public. Research Unit Identification 2I @ Main Agricultural Experiment Station, Auburn. - E. V. Smith Research Center, Shorter. Tennessee Valley Substation, Belle Mina. Sand Mountain Substation, Crossville. North Alabama Horticulture Substation, Cullman. Upper Coastal Plain Substation, Winfield. Forestry Unit, Fayette County. Chilton Area Horticulture Substation, Clanton. Forestry Unit, Coosa County. Piedmont Substation, Camp Hill. Plant Breeding Unit, Tallassee. Forestry Unit, Autauga County. Prattville Experiment Field, Prattville. Black Belt Substation, Marion Junction. The Turnipseed-Ikenberry Place, Union Springs. Lower Coastal Plain Substation, Camden. Forestry Unit, Barbour County. Monroeville Experiment Field, Monroeville. Wiregrass Substation, Headland. Brewton Experiment Field, Brewton. Solon Dixon Forestry Education Center, Covington and Escambia counties. 20. Ornamental Horticulture Substation, Spring Hill 21. Gulf Coast Substation, Fairhope. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Information contained herein is available to all persons without regard to race, color, sex, or national origin.