i~d~~9- ttr I~i~! II4I >. I 9< EaBIIII%1~BRi~r~s~i~~ ~& ~.* ~L=lia3Jllbi TIJE01QILI.D R,.OT ATIION, 1196 I1996 1 00 YEAIRS 0F SUI STAIINAIBILIE CIRO0PIRIING BY C.C. MITCHELL, F.J. ARRIAGA, J.A. EN TRYJ.L. NOVAK, W.R. GOODMAN, D.W. REEVES, MW. RUNGE, AND G.J.TRAXLER This bulletin is published by the Alabama Agricultural Experiment Station in commemoration of the centennial of the Old ~~ Rtto (1896A- 1996. o th th ceennl evnt ndth~eis bulltin wer spor4ih()ted iiinta rt ycntributins from THE OLID IROTATION, II1896--11996 1100 YEARS OF SUSTAINABL E CROPPING RESEARCH FOREWORD E.TYork W. R. Goodman ALABAMA'S COTTON ECONOMY, 1896-1996 The Cotton Plant Alabama's Cotton Heritage Yields and Acreage INTRODUCTION TOTHE OLD ROTATION History Objectives Leadership and Responsibilities Old Records and Publications Experimental Design Fertilization C.C. Mitchell 6 6 6 6 7 7 SITE CHARACTERISTICS AND SOILS C.C. Mitchell and D.W Reeves Soil Series 8 Soil Test Records 8 Soil Physical Measurements__________________________________10 CROP YIELDS Cultivars Cotton Corn Winter Legumes Small Grain, Cowpea, and Soybean SOIL ORGANIC MATTER MEASURING SUSTAINABILITY Defining Sustainability Measuring Productivity Methods Used TFP Indexes Without External Cost. Total Social Factor Productivity Inter-plot Comparisons CONCLUSIONS REFERENCES APPENDIX C.C. Mitchell and F.J.Arriaga 13 14 15 C.C. Mitchell and J.A. Entry i5 G.J.Traxler, J.L. Novak, and M.W. Runge 16 16 17 17 18 18 19 20 21 POSTSCRIPT Lowell T. Frobish 28 II~Y'CY II IL/ all~3 vll ~wls a / asl \rl \IL~ILR CII~VII 11 III YU II~LYILI \II\LII II VECONOMY, 1896-1 996 H~rit=r~e Id Responsibilities and PublicationsDesi~n ds easurements 1 111 1 I ~ I THE OLD ROTATION, 11896- 11996 IFO IRE W,"OIID By E.TYork, Chancellor Emeritus, State University System-of Florida ne must marvel at the wisdom and foresight of those responsible for initi- ating the Old Rotation experiment a century ago. One must also appreci- ate the wisdom of those responsible for maintaining these historic plots through the years, despite the expense involved, in order that they, today, may reveal the valuable information that only such long-term experi- ments can provide. We should realize that these experiments were started long before the world became concerned with the concept of sustainable agricultural production or the effects of man- agement practices upon sustainability of crop production. Yet we have learned that such practices have significant effects, good and bad, upon the ability to maintain or improve crop productivity. These effects may be due to different soil and crop man- agement practices upon the incidence of diseases, insects, nematodes, and weeds, as well as the effects of chemical and physical properties of soils, water retention, soil losses through erosion, the buildup of harmful chemicals, and other concerns - all of which may impact crop productivity and sus- tainability. The best way to evaluate the impact of such long- term management practices is through experiments such as those of the Old Rotation. Although these experiments have already provided very valuable information, some of which is reflected herein, they become even more valuable every passing year. With global populations increasing at the rate of 90 to 100 million peo- ple annually, the world's agriculture must provide an ever increasing flow of food and other agricultural products to sustain this rapidly growing population, And this must be done without any appreciable increases in arable lands. This means simply that existing agricultural lands must become increasingly productive. Today many fear that our natural resource base for agri- culture is deteriorating rather than improving to facilitate such needed increased productivity. If such deterioration is occurring, and there are many signs that this may be the case, we must fully understand its cause. Much of this under- standing will come about through research such as that con- ducted through the Old Rotation plots. As I was finishing my masters work at Auburn right after World War II, I went in to see Dr. Marion Funchess, who at the time was the dean of the School of Agriculture and director of the Alabama Agricultural Experiment Station, to get his advice on where to go to take my Ph.D. in soil science. He talked about four or five institutions, all of which had strong programs in soils, but then recommended very strongly that I go to Cornell to study with Dr. Richard Bradfield, who at that time was perhaps the preeminent soil scientist in the world. I had great respect for Dean Funchess and took his advice. Dr. Bradfield was a wise man in many ways, and I recall one statement that he made in a seminar that has helped shape my views about agricultural research since. He said "There are many researchable topics in agriculture; some of these are problems" The Old Rotation experiment deals with major problems in agriculture that can best be addressed through such long-term efforts. I salute those who initiated these experiments as well as those who have continued them through the years. 3 896-1996 THE COTTON PLANT Cotton has a long history of use by humankind. The earliest historical records show that cotton was used to manufacture cloth before written records existed, and cotton is mentioned in Indian literature dating from 15 centuries before Christ. Heredotus (484- 402 B.C.) said: There are trees which grow wild there (India) the fruit of which is a wool exceeding in beauty and goodness that of sheep. The Indians make their clothes of this tree wool. Theophrastus (350 B.C.) also mentioned cotton cultivation in India in his letters: The trees from which the Indians make their cloth have a leaflike that of the black mulberry, but the whole plant resembles the dog rose. They set them in plains arranged in rows so as to look like vines at a distance. During the middle ages, Europeans believed that cotton came from a kind of mystical animal-plant, and that each boll contained a little lamb and the lint in each boll was borne by that tiny lamb. Here in the "NewWorld," the history of cotton and the history of the United States have been intertwined from the beginning. Cotton was first grown in what was to become the United States in 1607, selling for about eight cents per pound. However, native cotton plants were found by the first Spanish explorers in the 14th century in the Southwest. Cotton remained a minor crop until the invention the cotton gin in 1793. For much of the history of our nation, cot- ton exports have been very important and, until the end of the 19th century, the United States was the only major exporter of cot- ton in the world. ALABAMA'S COTTON HERITAGE Cotton has been produced in Alabama since before it became a state in 1819, and cotton acreage in Alabama increased steadily until the boll weevil began appearing about 1910 (Figure Ia). Since about 1930, acreage planted to cotton The Old Rotation in Alabama has decreased. By 1982, less than 250,000 acres were planted in the state. Underlying this loss of acreage was a major change in both where cotton was produced in Alabama and the cultural and mechanical practices used to culti- vate and harvest the crop. Mitchell is Professor, Arriaga is Graduate Research Assistant, and Entry is Assistant Professor ofAgronomy and Soils; Novak is Professor and Goodman is Associate Professor of Agricultural Economics and Rural Sociology; Reeves is USDA Research Agronomist and Affiliate Professor of Agronomy and Soils; Runge is Extension Program Associate and Traxler is Associate Professor of Agricultural Economics and Rural Sociology. At the turn of the century, there were about 220,000 farms in Alabama. About 192,000 of those farms grew cotton. Half of Alabama's land resources was under cultivation. Out of a total cropland area of 20 million acres, about eight million acres were in cultivated crops, mostly cotton and corn, and the remain- der was in pasture and hay. Almost all these farms had annual sales of between $ 100 and $ 1,000. In 1900, only 446 Alabama cotton farms (out of 192,000) had sales in excess of $2,500. In 1900, Alabama ranked third in the United States in cotton production, behind Texas and Georgia. Leading cotton counties were, in descending order, Dallas (157,000 acres), Montgomery (134,000 acres), Lowndes (128,000 acres), Marengo (108,000 acres), and Bullock (103,000 acres). There were 3.2 mil- lion acres of cotton in Alabama about the time the Old Rotation was started. At the turn of the century, more than half of the two mil- lion residents of Alabama lived on farms. There were 5 13,000 peo- ple over 10 years of age engaged in agriculture in Alabama. Since most Alabama farms produced cotton, and half the people in the state were farmers, much of the economic activity in the state was linked to cotton production. was listed on the National Register of Historical Places in 1988. In the late 1800s, yields averaged about 120 to 150 pounds of lint per acre in Alabama (Figure la). Yields increased slowly, but in the first decade of this century, farmers in Alabama only averaged between 180 to 200 pounds per acre. From an eco- nomic perspective, it's hard to justify the nearly four million acres planted in Alabama during this period. When all costs were con- sidered, farmers were spending about $75 to $90 on each acre of cotton grown, yielding a cost of about 25 or 30 cents per pound produced. With prices fluctuating between 15 and 25 cents (Figure I b), most years farmers would seem to be losing money. However, L THIE OILID ROI1FATION, i 896 -II 996 cotton was the crop that was generally produced for sale off the farm, most others were mainly grown as feed for draft animals, hay, or for other on- farm purposes. Cotton pro- vided the major source of cash. About half the cost of producing cotton was in labor for planting, cultivating, hoeing, and picking. Much of this was in the form of unpaid family labor. Farmers, their wives, and children provided the bulk of this labor. The "opportunity cost" for this labor was essen- tially zero. So cotton provided a way to turn family labor into cash. YIELDS AND ACREAGE 60P About 1915, the boll wee- Price vil arrived in Alabama. Yields in 140 1988 real heavily infested fields dropped 1 to nearly zero. The state aver- 20 age yield in Alabama fell to 95 00 pounds per acre in 1916. 00 Although acreage decreased 80 initially, the boll weevil did not bring about the instant demise 60 of cotton in the state as is commonly assumed. Acreage 40 fell about 20% at first, but increased to exceed pre-wee- 20 vil numbers by 1930 as prices increased generally into the 0 I 30-cent range. Farmers were 1860 1880 19( able to stay in business in spite of decreased yields. The depression was much Fig. I Trends in Alabama (a) c more devastating to farmers for lint, 1865 to 1995. than the boll weevil. In 1931, the average cotton price fell to 5.6 cents per pound, and didn't recover until the second World War. During this pre-war period, Alabama lost about two million acres of cotton, with each major producing region losing about 40% of total acres between 1930 and 1940'. By the end of the war, Alabama was down to about 1.5 million acres, but acreage actually increased in the 1940s in the Tennessee Valley. Between 1960 and 1970, acreage in the Wire- grass decreased from 124,000 acres to 19,000 acres. Between 1970 and 1980, acreage in the Appalachian region declined by four fifths, to about 22,000 acres. These years were hard on the areas where mechanization, now essential in the economic production of cotton, was difficult. Cotton has migrated out of the Black Belt, once the major cotton producing region, due in part to these problems, with which farm- ers still are struggling. The only region of the state to retain a sig- otton acreage and yield and (b) price received by growers nificant percentage of their pre-war acreage is the Tennessee Valley, which has consistently planted about 200,000 acres since 1960. In recent years, cotton acreage has increased in Alabama, almost tripling since 1980. Alabama may soon plant 750,000 acres of cotton.Although yields have increased four-fold in good years, agriculture in the state is much more diversified than it was early in this century. Cotton has been the most persistent of Alabama's cash crops. In 1859, the state produced 780,000 bales. In 1994 Alabama pro- duced 726,000 bales of cotton. Cotton as a cash crop has had its ups and downs over the years, but the demand for cotton fiber has been steadily increasing for about 100 years, and the trend will like- ly continue. As we move into a new millennium and the second century .of the Old Rotation experiment, sustainable cotton pro- duction will continue to be important to Alabama farmers. 5 TIrlHIIE OILID ROTATIION, II 896 -1996 I1N-IR.ODUCTIION TO THIIE OILID ROTATIION HISTORY The Old Rotation is the oldest, continuous cotton experiment in the world (Steiner and Herdt, 1995) and the third oldest field crop experiment in the United States on its original site. It was placed on the National Register of Historical Places in 1988 (Am. Assoc. State and Local History, 1989). Older experiments include the Morrow Plots at the University of Illinois (c. 1876) and The Sanborn Field at the University of Missouri (c. 1888). The Magruder Plots at Oklahoma State University (c. 1892) are older, but the soil was physically moved to a new location in 1947 (Mitchell et al., 1991). The Old Rotation was begun in 1896 by Professor John F Duggar. In 1896, more than 3.5 million acres of cotton were planted in Alabama, but the average yield was only 130 pounds lint per acre (Figure l a). Alabama cotton farmers planted their crop year after year on the same land. New land for clearing and cropping had almost been exhausted. Very few amendments were applied to the soil, and crop rotations and winter cover crops were a rare practice. Excessive soil erosion, declining yields, and low farm income were common. The economy of the state and the welfare of Alabamians depended upon sustainable cotton production. Some researchers suggested substituting tobacco culture for cotton. However, Professor Duggar undoubt- edly believed that Alabama soils could sustain profitable yields of cotton if a crop rotation system that included winter legumes could be developed. Corn, oat, and cowpea also were familiar crops on 19th cen- tury Alabama cotton farms. These grains and their fodder fed the draft animals that worked the fields. Corn also was a staple in the diet of the people who lived on the land. Almost as many acres of corn as cotton were planted to support the cotton cash crop. Cowpea was one of the few summer annual legumes that grew well in the South. Therefore, corn, oats, and cowpeas were logical crops to include in a crop rotation system. OBJECTIVES A statement of the original objectives of the Old Rotation can- not be found in the historical records. However, the treatments themselves suggest that the objectives of the experiment were to (I) determine the effect of rotating cotton with other crops to improve yields and (2) determine the effect of winter legumes in cotton production systems. These are the same objectives used today. LEADERSHIP AND RESPONSIBILITIES Duggar served as the third director of the Alabama Agricultural Experiment Station from 1903 to 1921. During this time, he continued to oversee the management of the Old Rotation. When the Department of Agronomy and Soils was established in 1919, management of the experiment became the department's responsibility. The area around the Old Rotation became known as the "Agronomy Farm." In 1977,most field crop research was moved from the Auburn University campus to the new farm at EM Smith Research Center near Shorter, Alabama. During the move and for several years thereafter, the logistics of managing the Old Rotation became difficult with no budget and limited equipment and personnel. Some of the yield records were lost during this period. Today, the Department of Agronomy and Soils works with staff of the Alabama Agricultural Experiment Station to maintain the Old Rotation.Agronomy and Soils faculty who have maintained the Old Rotation with the help of numerous technicians include: Project leaders Years J.FE Duggar 1896-1921 E.F. Cauthen and H.B.Tisdale 1922-1 929 E.L. Mayton 929-1944 F.L. Davis 1944-1948 D.G. Sturkie 1948-1959 L.J. Chapman 959-1963 LexWebster 1963-1966 E.M. Evans 1966- 1983 J.TTouchton 984-1985 C.C. Mitchell 1985-present OLD RECORDS AND PUBLICATIONS The original records of the Old Rotation from 1896 to 1919 were destroyed in a fire that razed Comer Agricultural Hall in 1920. However, some hand-written records were later found; average yields for 1896 to 1905 and from 1906 to 1915 had been published and were recovered. There are some gaps in the yield record. The most notable was in the mid-I 970s when records were lost during the move from the old Agronomy Farm to the new E.V. Smith Research Center. The only known publication that summarized all the data to date from the Old Rotation was an article by FL. Davis (1949) in the magazine "Better Crops with Plant Food" published by the American Potash Institute. However, numerous research papers, popular articles, abstracts of professional meetings, and crop rec- ommendations have developed from information gathered from plots in the Old Rotation. No one is certain when it was first called the Old Rotation or who named it. A 1930 Alabama Agricultural Experiment Station bulletin contained a photo taken that year with a caption identifying it as "Old Rotation Experiment" (Bailey et al., 1930). Duggar was a coauthor of this bulletin. The 1949 article by Davis was entitled "The Old Rotation at Auburn, Alabama." The name was obviously associated with these plots by the 1940s. Davis noted in 1949 that it was ". .. probably the oldest field experiment in the United States in which cotton has been grown." A list of known publications from the Old Rotation is given in the Appendix. 6 TIHIE OILID ROTATION, 1189611996 EXPERIMENTAL DESIGN Statistical analysis did not gain widespread acceptance among agricultural researchers until well into the 20th century. Therefore, the Old Rotation, like most 19th century experi- ments, was not replicated. Each plot was a different treatment to be observed. Yield was the only measurement recorded. In the 1950s, routine soil testing allowed quick measurements of soil pH and extractable nutrients, and these measurements were added to the records of the Old Rotation. The Old Rotation consists of 13 plots, each 21.5 feet by 136.1 feet. A three-foot alley separates each of the plots (Figure 2). Plots are identified by numbers. Plots in rotations are essentially replicates as far as soil treatments are concerned. Today, the rotation treatments are often summarized as follows: I. Cotton every year A. No legume/no N fertilizer (plots I and 6) B.Winter legumes (plots 2,3, and 8) C. 120 pounds fertilizer N per acre per year (plot 13) The Old Rotation I. Cotton every year A. No legume/no N (plots #1, #6) B. Winter legumes (plots #2, #3, #8 C. 120 lbs. N/acre/yr. (plot #13) II. Cotton-corn rotation A.Winter legumes (plots #4, #7) B. Winter legumes + 120 lbs. N/acre/yr. (plots #5, #9) III. 3-year rotation (plots #10, #11, #12) Cotton (legumes)-Corn (small grain for grain)-Soybean 3 e . . 4- 5- 6 7- 9- 10 12 S13 < - 136.1 II.Two-year, cotton-corn rotation A.Winter legumes (plots 4 and 7) B.Winter legumes plus 120 pounds N per acre per year (plots 5 and 9) III. Three-year rotation (I) Cotton-winter legumes, (2) corn-small grain for grain (60 pounds N per acre),(3) soybean (plots 10, I I,and 12) FERTILIZATION All plots have received the same annual rate of phosphorus (P) and potassium (K). However, the actual rate applied has gradually increased over the years from a total annual application of 0-22-19 pounds N-P 2 Os 5 -K20 per acre to 0-80-60 since 1956 (Table 1). The changes in the amounts of P and K applied were made to meet obvious fertility needs of the crops (Davis, 1949). In the 1920s, P and K were applied to both the summer crop and the winter legume. Later, treatments were changed so that time of P and K application could be evaluated (e.g., P and K were applied to either the summer crop, the winter legume,pr split). The reason for this change was explained by Davis (1949): Of primary interest is the small but gradual decline in yields of both corn and cotton during the early years of the experi- / ment. This decline was due to the small amount of growth made by the winter Slegumes. (The P and K) applied annually to .. the summer crops did not provide suffcient *o**.**, phosphorus for the winter legumes. When . . 400 pounds per acre of 16% superphos- phate were applied in the fall, the vetch immediately began to make good growth and adequate tonnage of green matter. The subsequent yields of both corn and cot- ton, i.e. after 1923, show the effects of the increased growth of the winter legumes. Since 1956, fertilizer nitrogen (N) as ammonium nitrate has been applied to the cotton and corn rotation in plots 5 and 9 at a rate of 120 pounds N per acre per year, to .oo ~ . .cotton in plot 13 at 120 pounds N per acre .*.*..*.. ~ per year,and to the small grain in plots 10, I I, mor 12 as a topdressing of 60 pounds N per acre. ft. > Fig.. Schematic of the Old Rotation and treatments used since 1956. From 1896 to 193 1,the sources of P and K were acid phosphate (either 14% or 16% P 2 Os) and kainit (12% K 2 0), respectively. In 1932 a change was made from kainit to muri- ate of potash (50% K 2 0). In 1944, 18% super- phosphate and 60% muriate of potash were used. Today, the sources of P and K are con- centrated superphosphate (46% P 2 0 5 ) and muriate of potash (60% K20). Since 1956, all plots have received an annual application of 134 pounds agricultural gypsum (calcium sul- 7 THIE OILID ROTATION, 11 896-- 11i 996 fate) per acre that will provide approximately 20 pounds sulfur (S) been irregular and no records were kept until the 1950s. Since per acre per year. then, soil samples have been taken after harvest about every two Ground, dolomitic agricultural limestone is applied to each years. These have been tested for pH and Mehlich-I (dilute double plot as needed to maintain the soil pH above 5.8. Soil sampling has acid) extractable P, K, Ca, and Mg. TABLE I. CROPS AND FERTILIZER RATES (POUNDS PER ACRE N-P 2 0 5 -K 2 0) USED IN THE OLD ROTATION SINCE 1896 Treatment/plot 1896-1924 1925-1931 1932-1947 1948-1955 1956-Present I corn 0-22-19 corn 0-26-19 cotton 0-72-60 cotton 0-72-60 cotton 0-80-60 cowpeas vetch 0-62-0 vetch 2 corn 0-22-19 corn 0-88-19 cotton 0-72-60 cotton 0-72-60 cotton vetch vetch/clover 0-80-60 3 cotton 0-22-19 cotton 0-26-19 cotton 0-36-30 cotton 0-18-28 cotton 0-40-30 vetch vetch 0-62-0 vetch 0-36-30 vetch 0-18-28 vetch/clover 0-40-30 4,7 cotton 0-22-19 cotton 0-26-19 cotton 0-36-30 cotton 0-36-30 cotton 0-80-60 vetch vetch 0-62-0 vetch 0-36-30 vetch 0-36-30 vetch/clover 0-40-30 corn 0-22-19 corn 0-26-19 corn 0-36-30 corn 0-36-30 corn 0-0-0 cowpeas vetch 0-62-0 vetch 0-36-30 vetch 0-36-30 vetch/clover 0-40-30 5,9 cotton 0-22-19 cotton 8-26-19 cotton 0-36-30 cotton 0-36-30 cotton 120-80-60 vetch vetch 0-62-0 vetch 0-36-30 vetch/clover 0-36-30 vetch/clover 0-40-30 cowpeas 0-22-19 cowpea hay 0-26-19 cowpea hay 0-36-30 cowpea hay 0-36 -30 corn 120-0-0 vetch 0-62-0 vetch 0-36-30 vetch 0-36-30 vetch/clover 0-40-30 6 cotton 0-22-19 cotton 0-88-19 cotton 0-72-60 cotton 0-72-60 cotton 0-80-60 8 (same as #3) (same as #3) cotton 0-72-60 cotton 0-72-60 cotton 0-80-60 vetch vetch vetch/clover 10,1I I,12 cotton 0-22-19 cotton 0-88-19 cotton 0-36-30 cotton 0-36-30 cotton 0-80-60 vetch vetch vetch 0-36-30 vetch 0-36-30 vetch/clover 0-80-60 corn 0-22-19 corn 0-88-19 corn 0-36-30 corn 0-36-30 corn cowpeas/oats oats oats 0-36-30 oats 0-36-30 rye 60-0-0 cowpeas 0-22-19 cowpea hay 0-88-19 cowpea hay 0-36-30 cowpea hay soybeans vetch vetch 0-36-30 13 (same as #5) (same as #5) (same as #5) cotton 0-36-30 cotton 120-80-60 vetch 0-36-30 cowpea hay 0-36-30 vetch 0-36-30 SIITIE CIHIARACTIEfRISTILCS ANID SOilILS The site of the Old Rotation is part of a 90-acre block of land purchased by The Agricultural and Mechanical College of Alabama (now Auburn University) from l.J.B. Gay in 1884 for $1,700. The site straddles the juncture of the southern Piedmont Plateau and the Gulf Coastal Plain physiographic regions in east-central Alabama (32036'N, 85036'W). Average annual precipitation at the site is 53 inches (1,339 mm). Average annual temperature is 640F (1 80C) with 221 days between the last spring freeze and the first fall freeze. SOIL SERIES Soils in this area often have sandy Coastal Plain sediments overlying finer textured, highly weathered Piedmont soils. Although the soil at the Old Rotation site is currently identified as a Pacolet fine sandy loam (clayey, kaolinitic, thermic Typic Hapludults), a char- acteristic Piedmont soil, it has been called a Norfolk fine sandy loam (fine-loamy, siliceous, thermic Typic Kandiudults), a typical Coastal Plain soil. The site appears on the local soil survey as a Marvyn loamy sand (fine-loamy, siliceous, thermic Typic Kanhapludults), another Coastal Plain soil. This confusion arises because the site is on a gradual slope (2-3%) and the surface soil texture changes (Table 2). The upper part of the site (plot I) is more characteristic of the Marvyn soil (Coastal Plain) and the lower part (plot 13) is more characteristic of the Pacolet (Piedmont). SOIL TEST RECORDS Because the Old Rotation experiment is primarily a crop rotation and legume N study, annual rates of P and K applied to each plot have been the same. Although the amounts applied each year have changed periodically, the amount 8 TIH HIE OLID IROTFATION, I 896 - 1i 996 TABLE 2. PARTICLE SIZE ANALYSIS OF THE PLOW LAYER (0-6 INCHES) IN THE 13 PLOTS OF THE OLD ROTATION Particle size Textural Water Plot no. Sand Silt Clay class available % in./in. I 74 16 10 sandy loam 0.06 2 70 18 12 sandy loam 0.07 3 69 19 12 sandy loam 0.07 4 68 22 10 sandy loam 0.08 5 64 24 12 sandy loam 0.09 6 66 19 15 sandy loam 0.08 7 68 20 12 sandy loam 0.08 8 70 18 12 sandy loam 0.07 9 56 21 23 sandy clay loam 0.10 10 59 21 20 sandy loam 0.10 11 64 21 15 sandyloam 0.09 12 61 21 18 sandy loam 0.09 13 58 17 25 sandy clay loam 0.10 Soil pH 7.5 - Cotton every year 7 6.5 5.5 .o . No N, plot #6 5 - --- + legume, plot #8 S...... 120 Ibs. N/acre, plot #13 4.5 I I I I I I 1950 1960 1970 1980 1990 2000 Soil pH 7.5 Rotations 7 - 2-year + legume, plot #4 - -. 2-year + N, plot #5 6.5 - - 3-year,plot#10 5.5 4.5 I I I I I I 1950 1960 1970 1980 1990 2000 Fig. 3. Changes in soil pH since 1956 on selected plots (a) where cotton is planted every year and (b) where cotton is planted in rotation. applied in any one year has always been the same on all plots. Davis (1949) discussed many of the crop growth problems encountered on the Old Rotation during its first 50 years. Most of these, particularly K deficiencies, resulted from low applications of K fertilizers and removal of cowpea hay. Phosphorus deficiencies in the winter legume often led to low dry matter production and low cotton yields as a result of low N in the soil from the legume. This observation led to the split P and K fertilizer applications that continue today in some plots. However, since soil P has accumu- lated to high levels and soil K is in the medium range, deficiencies are no longer observed and there are no cotton yield differences due to split P and K applications. Using data from the Old Rotation, Davis (1949) pointed out that "... cotton as a crop does not deplete the soil or run it down excessively. The cultural practices of leaving the land bare through the winter and of not preventing erosion are responsible for the generally low fertility level of many soils on which cotton is grown." No records of soil measurements before 1950 have been found. Since then, periodic, plow-layer soil samples have been taken for pH and Mehlich-I (dilute double acid) extractable P and K. These are presented in figures 3-5 for selected treatments. There are statistical differences (P<.01) in pH and extractable P among the treatments but no difference in extractable K. In spite of an effort to lime individual plots (using finely ground, dolomitic lime- Extractable P, Ibs./acre 225 - 200 - 175 - 150 - .... 125 - 100 - 75 - Critical P level Cotton every year - No N, plot #6 --- + legume, plot #8 ...... 120 Ibs. N/acre, plot #13 y, I 50 Extractable P, Ibs./acre 225 Rotations - 2-year + legume, plot #4 200 - 2-year + N, plot #5 . 3-year, plot #10 1 7 5 , S150 125 im 75 - . . ", 50 Critical P level"- 25 0 I 1950 1960 1970 1980 1990 2000 Fig. 4. Changes in Mehlich-I extractable plow-player P since 1956 on selected plots (a) where cotton is planted every year and (b) where cotton is planted in rotation. stone) whenever the soil pH drops below 5.8, the critical pH used in Alabama for loamy soils, the treatment receiving 120 pounds N per acre as ammonium nitrate and no winter legume has tended to have a more acid reaction than either the no N treatment or the one receiving winter legumes (vetch and/or clover) as a source of N (Figure 3a). Extractable soil P has been consistently lower over the past 42 years in the treatment receiving only fertilizer N (Figure 4a). 9 IFIHIE OILID ROTATION, 11896 1 996 240 [- Extractable K, Ibs./acre 300 - Cotton every year No N, plot #6 --- + legume, plot #8 ...... 120 lbs. N/acre, plot #13 Critical K level ?? ',so /n P 60- . - 1950 1960 1970 1980 1990 200( Extractable K, Ibs./acre SOIL PHYSICAL MEASUREMENTS Apparent treatment differences TABLE 3. LONG-TERM TREATMENT EFFECTS ON SELECTED SOIL PHYSICAL MEASUREMENTS MADE IN 1994* Cation Plow- Plow- Bulk Cone Water Plots exchange layer layer density penetrometer stable capacity N C 0-30cm resistance aggregates Treatment to 30 cm meq/100g % % g/cm 3 bars % I. Continuous cotton A. No legumes 1,6 3.9 0.1 I 0.39 1.84 29 24 B. +legumes 2,3,8 4.7 0.14 0.75 1.85 28 38 C. 120 lb. N/acre 13 5.4 0.10 0.54 1.73 20 22 II.Two-year rotation A. +legumes 4,7 4.4 0.13 0.82 1.75 19 40 B. +leg./+ 120 lb. N/acre 5,9 5.1 0.1 1 1.00 1.66 20 38 Ill.Three-year rotation 10, I I, 12 4.9 0.14 1.01 1.56 19 39 Analysis of variance P>F <0.01 NS <0.01 <0.01 NS <0.01 *Values presented are an average of samples taken in the winter and spring of 1994. observed by individuals plowing, planting, and cultivating the Old Rotation. Soil on plot 13, which has been planted to cotton every year since 1956 with only commercial N fertilization, has a history of severe crusting after planting. Poor cotton stands frequently result when rains cause crusting prior to seedling emergence. The problem has also been observed on other plots planted to cotton every year with no winter legume (plots I and 6). In spite of the lack of a structured, replicated, experimental design that allows use of traditional statistical analyses, some soil physical measurements suggest that the observed soil tilth problems may be due to long- term treatment effects. In 1994, selected soil physical measurements were made on each plot during the winter and again after planting in the spring. Treatments with observed crusting problems (plots 1,6, and 13) had lower organic C, higher cone penetrometer resistance, higher bulk density, and fewer water stable aggregates (Table 3). This con- firms poor soil structure and soil compaction in these treatments compared to those treatments that use winter legumes and crop rotations. More information on the relationship between soil organic matter and yield is presented in the section on Soil Organic Matter. 10 The plot receiving 120 pounds N per acre per year since 1955 is the last plot in the experi- ment and is at the low- est elevation. The sur- face soil texture tends to be finer than that of the first few plots of the experiment (Table 2) that are more typi- cal of Coastal Plain soils. The apparent lower P may result from the higher P fix- ing capacity of the finer- textured soil in this plot. in soil tilth have been Fig. 5. Changes in Mehlich-I extractable plow-player K since 1956 on selected plots (a) where cotton is planted every year and (b) where cotton is planted in rotation. 0 Records of cultivars planted on the Old Rotation prior to the 1960s do not exist. Part of this omission was probably because the cultivars selected by the pro- ject leaders represented the best that was available based on vari- ety trials by the Alabama Agricultural Experiment Station and those recommended by the Cooperative Extension Service. Specific varieties listed in the records are given in Table 4. Zeasona8 Vaiabthy andc Long-=:e r7Tre:nds. Improving crop yields, primarily cotton yields, has been the principal focus of the Old Rotation since its beginning. Yields were the only consistent records kept throughout the history of the Old Rotation. Seed cotton yield records rom plot 3 (cotton every year with only legume N) are used to illustrate the wide yield vari- ability expected under nonirrigated conditions as used in the Old Rotation and practiced by most Alabama growers (Figure 6). An interesting observation is that yields are rarely high for two consec- utive years. Likewise, two consecutive low yielding years also are rare. Five-year unning average yields seemed to decline slightly during the first 25 years of the Old Rotation. No doubt some of this decline was due to the boil weevil that entered Alabama in 191 1 and became widespread by 19 1 4. Davis (l 949) attributed this decline primarily to a P deficiency ir the winter legume. The 1 924 revision increased P rates from 22 to 88 pounds P2O per acre per year The 193: revision increased K rates from 19 to 60 pounds K20 per acre per year. From the mid-1920s to the mid-1960s, average seed cotton yields on plot 3 crept upward from about 750 to more than 2,500 pounds per acre (approximately 290 to 980 pounds lint per acre). Some of this increase car be attributed to improved soil fer- tily practices, but improved cultivars of cotton and better insect control also contributed. Auburn 56 cotton cultivar was intro- 7ABLE 4. SPECFIC CUUTIVARS OF CO0TO7, CORN, 3MJ2,LL GRA3, ZOCYBEiu\, A N0D \R LEG UW ?iAN- ED ON T- OD Ro 2o0-ot As oumlD NV HANDxtR TSVM REcoRDS Year(s) Co 1968-70 Aubi Corn Fla. 200A Auburn 56 Funks G4949 Small grain wheat (GA I 123) wheat (GA 1123) DPL 26 Funks G4864 rye DPL 26 Funks G4864 rye (Wrens Abruzzi) Coker 56 rye (Wrens Abruzzi) DPL 26 Pioneer 3147 rye (Wrens Abruzzi) DPL 41 Ring-around 1502 DPL 41 Ring-around 1502 1984 DPi 41 1985 DPL 41 sorghum FG 4858 rye (Wrens Abruzzi) rye (Wrens Abruzzi) rye (Wrens Abruzzi) rye (Wrens Abruzzi) 1986 DPL 41 Pioneer 3320 rye (Wrens Abruzzi) 1987-91 DPL 90 1992 DPL 90 1993 DPL 90 1994 DPL 5690 1995 DPL 5690 DK 689 DK 689 DK 689 DK 689 DK 689 rye (Wrens Abruzzi) wheat (Fla. 301) rye (Wrens Abruzzi) rye (Wrens Abruzzi) wheat Winter Soybean" legume Bragg woolypod vetch crimson clover (Autauga) Bragg hairy vetch crimson clover (Dixie) Bragg hairy vetch crimson clover (Dixie) Hutton hairy vetch crimson clover (Dixie) Bragg hairy vetch crimson clover (Dixie) Bragg hairy vetch crimson clover (Dixie) Braxton hairy vetch crimson clover (Dixie) Braxton hairy vetch crimson clover (Dixie) Braxton hairy vetch crimson clover (Dixie) Braxton vetch (Cahaba white) crimson clover (Tibbee) Braxton vetch (Cahaba white) crimson clover (Tibbee) Braxton vetch (Cahaba white) crimson clover (Tibbee) Stonewall vetch (Cahaba white) crimson clover (Tibbee) Stonewall crimson (AU Robin) Hutcheson crimson (AU Robin) Stonewall crimson (AU Robin) duced in 1956. This wilt and nernatode resistant variety became the variety of choice for most producers in Alabama by 1960, and was grown on the Old Rotation longer than any other cultivar. During the late 1950s and 1960s, DDT (dichlorodiphenyl trichloroethane) was a very effective insecti- cide for control of boll weevils and worms. However, its removal from use in the early 1970s may have con- tributed to the tempo- rary decline in yields during this decade. In the 1980s and 1990s, synthetic pyrethroids dominated the market for worm and weevil control in cotton. Efforts to eradicate the boll weevil in East- ttoon urn 56 1971 1978 1979 1980 1981 1982 1983 DPL 26 *Prior to :956, the summer legume was cowpea as a green manure crop or cowpea hay. CiR~B I k iL lDl , r IHIIE OILID RiOTATION, 11896 11996 Seed cotton yield, Ibs./acre 4,000 - 3,500 - -..... 5-yrave. --- -- 3,000 - 2,500 - 2,000 - 1,500 - 500 - 0 I I I lI I I I I 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 Fig. 6.Annual seed cotton yields on plot 3, 1896-1995, where only winter-legume N has been available to the cotton crop. Seed cotton yield, lbs./acre 2,500 - Plot #6 2,000 - ----- Plot #13 1,500 iI I1940 -I ! 1 'I \~. \ 500 - 1,0001 - 1890 -- - I -oto /il 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 Fig. 7. Seed cotton yields on the control plots (plots I and 6) where neither fertilizer N nor winter-legume N has been used on cotton and where 120 pounds fertilizer N per acre per year have been applied on plot 13. Central Alabama also began and may have partially accounted for the upward trend in yield during the past few years. No N and No Legumes. Five-year running average yields on the control plots (plots I and 6) are about the same today as they were when the Old Rotation began (Figure 7). Plot I was in corn during the first 40 years of the Old Rotation. Yield trends on both these plots indicate that with no N fertilization and no legumes, the yield potential gradually declines over a period of 15 to 20 years and then stabilizes at about half of the beginning yields. This may be a reflection of the gradual mineralization of organic N. Soil organic matter in plots I and 6 is less than 1%. Nitrogen removal in the cotton lint and seed (primarily seed) from these plots is estimated to be about 12 pounds per acre per year. This is approximately equivalent to available N from non-symbiotic fixa- tion and rainfall. Relatively higher yields since the late 1980s are found in all treatments and may be a result of favorable growing seasons. However, the 100th growing season, 1995, produced some of the Seed cotton yield, lbs./acre 3,500 - ..... Plot #2 - - r 3,000 Plot #3 SPlot #8 2,500 - -J 2,000 - 1,500 - .- . 500 I 0 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 Fig. 8. Seed cotton yields on plots that have been con- tinuous in cotton with only winter legumes as N sources. Plot 2 was in corn prior to 1932; winter legumes wre added in 1948. Fertilizer P and K are applied to the legume in plot 2, to the cotton in plot 8, and split between cotton and legumes in plot 3. Fig. 9. Seed cotton yields on the two-year rotation with legumes (plots 4 and 7), the two-year rotation with legumes and 120 pounds N/acrelyr. (plots 5 and 9) and the three-year rotation (plots 10, 11, & 12), 1896-1995. lowest yields in more than 50 years. This was attributed to prob- lems throughout the growing season including soil crusting from intensive spring rains following planting, replanting, an extreme summer drought, and increased insect pressure. Winter Legumes Versus Fertilizer N. Including a winter legume as the only source of N for the cotton crop (plots 3 and 8; Table 5; Figure 8) has produced yields as high or higher than those produced from applying 120 pounds N per acre to a cotton monoculture (plot 13;Table 5; Figure 7). Winter legumes were not planted on plot 2 until 1948 (Table I). The N-fertilized plot (plot 13) was not added until 1956. Duggar effectively demonstrated that winter legumes could improve yields of continuous cotton during the first few years of the Old Rotation. Yields since 1956 have been similar using legume N and fertilizer N. Therefore, the choice farmers make obvious- ly depends on costs and management. Planting and growing win- 12 0' I I I I I I I I I I I I I U -1/ I I I IU THIE OILID ROTATFIION, 11896 I996 TABLE 6. ESTIMATED N BUDGET FOR CROPPING SYSTEMS IN THE OLD ROTATION Treatment (plots) N available N Treatment (plots) Legume Fertilizer removed ............. lb./acre/rotation ............. I. Continuous cotton A. No N/no legumes (#1,#6) 0 0 12 B.Winter legumes (#3,#8) 116 0 38 C. 120 lb. N/acre (# 13) 0 120 40 II. Cotton-corn rotation A. +legumes (#4,#7) 232 0 80 B. +legumes/+N (#5,#9) 232 240 94 Ill.Three-year rotation 320 60 275 (#10,# 1,#12) ter legumes in a continuous cotton system requires a higher level of management, and depending upon seed, fertilizer N, and planting costs, growingwin ter legumes can cost more than using fertilizer N. Recent measurements on winter legumes indicate that between 80 and I50 pounds N per acre are fixed in the above- ground portion of the legume. If most of this N is available to cot- ton, it will be adequate for nonirrigated cotton. A N budget for the treatments in the Old Rotation (using yield, fertilization, and crop removal estimates during the past decade) suggests that N use effi- ciency is the same for continuous cotton regardless of the source of N (Table 6). Nitrogen use efficiency appears higher for the three-year rotation because of the high N removal associated with soybeans and because only 60 N use pounds fertilizer N per acre is applied during the efficiency three-year period. % Crop Rotations. There is a definite yield advan- tage to rotating cotton with other crops (Table 5; - Figure 9). However, the two-year cotton-winter 33 legume-corn rotation is as beneficial as the three-year 33 rotation. Low yields for nonirrigated corn in Central Alabama have made a cotton-corn rotation less 35 attractive to growers than continuous cotton. 2 Novak et al. (1990) studied risks and returns for the various "Old Rotation" cropping systems using data for 1980 through 1990. They concluded that "...the optimal farm plan will include a three-year rotation of cotton, winter legumes, corn, small grains, and soybeans. The highest expected return at each target income level will result from planting the entire acreage to (this rotation). As risks are reduced, more and more of the continuous cotton with winter legume rotation will enter the farm plan." CORN In spite of historically low corn grain yields compared to mid- western states, corn has been the principle grain crop produced in Alabama. It was a staple on 19th century Alabama cotton farms because it provided food and fodder for livestock and grain for 13 TABLE 5. TEN-YEAR AVERAGE SEED COTTON AND CORN GRAIN YIELDS, 1896-1995 1896- 1906- 1916- 1926- 1936- 1946- 1956- 1966- 1976- 1986- Treatment (plots) 1905 1915 1925 1935 1945 1955 1965 1975 1985 1995 Seed cotton yields (pounds per acre) I. Continuous cotton A. No N/no legumes (#6) 800 630 340 510 370 510 620 710 610 930 B. + legumes (#3,#8) 860 680 640 I, 160 1,230 1,580 2,360 2,100 1,840 2,230 C. 120 lb. N/acre (#13) 1,960 2,040 1,630 1,860 II. Cotton-corn rotation A. +legumes (#4,#7) 870 750 770 1,260 1,440 1,950 2,640 2,410 1,850 2,290 B. +legumes/+N (#5,#9)* 890 950 1, 150 1, 190 1, 170 1,680 2,500 2,030 2,170 2,560 Ill.Three-year rotation 740 804 704 1,150 1,140 1,690 2,640 2,390 2,210 2,240 (#10,#I 1,#12) Corn grain yields (bushels per acre) I. Continuous corn A. No N/no legumes (#2) 18 II 9 10 B. +legumes (#1) 19 16 18 26 II. Cotton-corn rotation A. +legumes (#4,#7) 18 13 15 29 34 40 69 39 33 73 B. +legumes/+N (#5,#9)* ** 42 96 Ill.Three-year rotation 16 13 15 29 36 47 86 68 33 107 (# 10,# 11,#12) *120 pounds N per acre added as ammonium nitrate since 1956 to cotton and corn. Prior to this, a summer legume (cowpea) was planted in rotation with cotton and winter legumes. **Insufficient data. Note: Corn grain yields are calculated using 56 pounds per bushel at 15.5% moisture. TlHIE OILD ROFOTATION, 11896-- 1996 Seed cotton yield, lbs./acre 3,500 - 3,000- 2,500- 2,000 - 1,500 - 1,000 500 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 Fig. 9. Seed cotton yields on the two-year rotation with legumes (plots 4 and 7), the two-year rotation with legumes and 120 pounds N/acrelyr. (plots 5 and 9) and the three-year rotation (plots 10, I I, and 12), 1896-1995. human consumption. Corn grain yields on the Old Rotation are very similar to Alabama average yields. While grain yields have gradually increased over the 100 years of the Old Rotation, only during the past decade (1986-1995) have they increased dramat- ically (Table 5, Figure 10). The reason for this yield increase is not apparent. It may be a reflection of higher N fixation by the winter Corn yield, bu./acre 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 Fig. 10. Corn grain yields on selected treatments, 1896- 1995. legume (Table 7; Figure I I), improved hybrids, and good weather during the past decade. WINTER LEGUMES Yield records for winter legumes were not kept prior to 1926. Many years since have missing data. In addition, harvest weights were recorded as green weight or fresh weight yield until 1985. Since 1985, all winter legume yields have been reported as dry matter yields. To calculate all yields on a dry 14 ;, C- Plots #4 & 7 M --- Plots#5&9 ..... Plots #10, II &12 i ' I I I TABLE 7. TEN-YEAR AVERAGE WINTER LEGUME (CRIMSON CLOVER ANDIOR VETCH) DRY MATTER, SMALL GRAIN, AND SOYBEAN YIELD, 1896-1995 1896- 1906- 1916- 1926- 1946- 1956- 1966- 1976- 1976- 1986- Treatment (plots) 1905 1915 1925 1935 1945 1955 1965 1975 1985 1995 Winter legume (pounds dry matter per acre)** 1. Continuous cotton A. No N/no legumes (#6) B. +legumes (#3,#8) * * 1090 850 840 730 1470 * 3560 C. 120 lb. N/acre since 1956 *** * i 1050 520 910 (#13) II. Cotton-corn rotation A.+legumes (#4, #7) *** *** 1130 880 720 1120 1250 ** 3560 B. +legumes/+N (#5,#9)* *** 1180 560 890 1100 I 160 3550 Ill.Three-year rotation *** * * 1120 810 790 1040 1530 ** 3960 (#10, I I, 2) Small grain (bushels per acre) Ill.Three-year rotation 16 * * 59 41 50 19 28 27 (#10, II, 12) (oat) (oat) (oat) (oat) (wheat) (rye) (rye) 23 43 (wheat) (wheat) Soybean (bushels per acre) Ill.Three-year rotation 34 33 38 35 (#10, 11 , 12) *120 pounds N per acre added as ammonium nitrate since 1956 to cotton and corn. Prior to this, a summer legume (cowpea) was plant- ed in rotation with cotton and winter legumes. **Prior to 1985, winter legumes (clover and/or vetch) yields were recorded as green, harvest-weight only. Dry matter yields were estimat- ed by assuming 18% dry matter. Since 1985, dry matter yields have been estimated by plot by determining moisture at harvest. ***Insufficient data. Note: Oat, wheat, and rye grain yields are calculated using 32, 60, and 56 pounds per bushel, respectively; soybean yields are calculated using 60 pounds per bushel at 13% moisture. TIHIE OLID IROTATION, I 896-1i996 Winter legumes yield, Ibs./acre 8,000 - - A Plot #3 -T O m Plot #7 S a a O Plot #11 6,000 i- " > > 4,000 2,000- 0 AA A . ? A 0 MO I A 0 I I I 1920 1930 1940 1950 1960 1970 1980 1990 2000 Fig. I I. Estimated dry matter yields of winter legumes in selected plots, 1926-1995. matter basis, earlier data were converted to a dry matter basis assuming 18% dry matter in fresh herbage. This is approximate- ly the average dry matter in herbage harvested since 1985. Apparently, this resulted in low dry matter yield estimates (Table 7; Figure I I). Data in Table 6 would suggest that dry matter yield has more than doubled since the 1960s. Improved varieties and timely fall planting have no doubt contributed to higher dry mat- ter yields of winter legumes. SMALL GRAIN, COWPEA, AND SOYBEAN Small grain (oat, wheat, or rye) and either cowpea or soybean have been planted on the three-year rotation (plots 10, I I,and 12) since 1956. Prior to this time, cowpea was planted as both a sum- mer green manure crop and a hay crop. It was one of the few sum- SOu L OR IGAN C MATT ER Soil organic matter (SOM) is an important indicator of soil quality because it influences soil structure, which affects soil stability and its capacity to provide water. It is also the con- trolling factor in nutrient cycling. Soil organic matter can affect soil productivity. The amount of SOM reflects past balances between rates of humus formation and mineralization. Organic matter loss occurs because the rate of organic matter mineralization is greater than the annual input from plants while increases in SOM are a result of increased plant biomass and/or decreased degradation. Final amounts of soil carbon in agroecosystems are the direct result of the specific farming practices imple- mented on the land. Following a change in land management, SOM changes slowly with time. These changes are difficult to detect until suf- Small grain yield, bu./acre 120 - C1 4 M V n 0 Oat 100 - - -Wheat M M M@oRye Plots #10,11, 12 80 S60 4 140 - 40 20 ** K 00% 0 I I L I I I I I 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 Fig. 12. Grain yields of oat, wheat, and rye, 1896-1995. mer annual legumes that was productive on the soils and climate of the deep South during the late 19th and early 20th century. It could be planted following a spring crop of oat or wheat or fol- lowing corn in the late summer and early autumn. Yields for cowpea when turned under as a green manure crop or as hay are not complete. Data available are in the Appendix. In the early 1960s, soybean became a profitable and widespread crop throughout much of the South. As a cash crop, it replaced cow- pea as the summer annual legume of choice. Oat was produced as grain for animal feed until improved selections of wheat and rye were accepted by southern growers. Although rye is not a profuse grain producer, it is frequently planted as the small grain because it provides rapid fall growth,winter soil protection, early maturity, and high total biomass production. Average yields of small grain and soybeans are given in Table 7 and Figure 12. 15 TABLE 8. SOIL ORGANIC MATTER IN THE OLD ROTATION Year Plot 1988 1992 1994 Average % SOM I. No legume/no N 1.0 1.0 0.6 0.9 2.Winter legumes 1.7 1.7 2.0 1.8 3.Winter legumes 1.6 1.1I 1.7 1.5 4.Two-year rotation/+legumes 2.0 1.8 1.9 1.9 5.Two-year rotation/+ legumes/+N 2.2 1.8 1.9 2.0 6. No legume/no N since 1896 0.9 0.7 0.6 0.7 7.Two-year rotation/+legumes 1.7 1.8 1.5 1.7 8.Winter legumes 2.9 1.7 1.6 2.1 9.Two-year rotation/+legumes/+N 2.4 2.3 .8 2.2 I 0.Three-year rotation 2.8 2.5 1.8 2.4 I I.Three-year rotation 2.6 2.4 1.9 2.3 12.Three-year rotation 2.2 2.1 1.9 2.1 13. 120 lb. N/acre/yr. 1.4 1.8 1.5 1.6 1988-1994 Treatment Treatment average,% I. Cotton every year A. No legume/no N 0.8 B.Winter legumes 1.8 C. 120 lb. N/acre/yr. 1.6 II.Two-year rotation A.Winter legumes 1.8 B.Winter legumes/+ 120 lb. N 2.1 Ill.Three-year rotation 2.3 TIHE OLID ROTATION, I1896 i1 996 Yield relative to Plot 3, % 140 - 120 i @ 80 100 ) [] 60 40 o0 6 01988 40 s 01991 20 - 01993 ? 1994 I 1 1 11 I I I I I 1 I I 0 0.5 I 1.5 2 2.5 3 Soil organic matter, % Fig. 13. Relationship between SOM and cotton yields relative to plot 3 since 1988. ficient time has elapsed for the changes to be larger than the spacial and analytical variability in the soil (Entry et al., 1996). No records were kept of SOM measurements on the Old Rotation before 1988. Measurements in the plow layer have been made in 1988, 1992, and in 1994 using the Walkley-Black procedure and a factor of 1.9 to convert from organic C tc organic matter (Table 8). Results of this investigation show that long-term planting of winter legumes significantly increases SOM. The two-year cotton- corn rotation with winter legumes plus N (treatment IIB; plots 5 and 9) and the three-year rotation (treatment III; plots 10, 1I, and 12) had higher SOM than the other four rotations. Cotton with- out winter legumes (treatment IA; plots I and 6) had lower SOM than all other rotations. These results are not surprising consider- ing the increased biomass returned to the soil from the corn, small grain, and summer legume (soybean) residue. The plots with the highest SOM also are the highest yielding plots. Increased SOM can be viewed as a consequence of improved production. However, Figure 13 suggests that SOM may also be viewed as a predictor of relative crop yield. There is a significant trend toward higher cotton yields in plots with higher SOM. Figure 13 suggests a yield plateau in the Old Rotation above 2% SOM. Cover crops grown on cropland in the southeastern United States are beneficial because they maintain SOM, improve the phys- ical and chemical characteristics, supply the soil with additional N, and reduce erosion of topsoil during the high rainfall winter months. Well adapted winter legume cover crops can replace from 90 to 120 pounds N per acre. After 99 years, data from the Old Rotation indicate that winter legumes increase amounts of both C and N in soil, which ultimately contribute to higher cotton yields. s~( rA U)LA~ MIEASUIIRIING SUISTAIINABIILTY ON TIHIE OILID IOTATIION DEFINING SUSTAINABILITY A sustainable agricultural system should maintain or enhance agricultural production, protect natural resources, be economical- ly viable, and be socially acceptable (Novak and Goodman, 1994; Pfeffer, 1992;Taylor, 1990). Cotton production in the southeastern United States has had a major influence on the economy of this region for almost 200 years. The fact that it continues to be pro- duced as an economically viable crop suggests that cotton is a viable and acceptable component in a sustainable production sys- tem. Yet the historical record of cotton production's destruction of natural resources (soil erosion in the southern Piedmont, stream sedimentation, deforestation, etc.) leaves doubt about its sustainability as a protector of the natural resource base (Trimble, 1974). Pesticide use since the boll weevil entered the Cotton Belt in the early 1900s has added to concerns about sustainability, espe- cially since some of the early insecticides included arsenicals and DDT, which are no longer allowed by EPA because of the health or environmental hazards they pose. MEASURING PRODUCTIVITY Productivity is difficult to assess because it involves more than just yield. Alabama's cotton acreage is 15-20% of what it was during peak production in 1914, yet statewide yields are five times higher (Figure I). The nominal price of cotton has increased over the past 100 years but when adjusted for infla- tion, the real price has not increased. These trends make evalu- ating sustainability difficult. Production indexes have been suggested as appropriate mea- sures of change (Binswanger, 1978; Lu et al., 1979). If all quantifiable inputs and outputs are known or can be reasonably estimated from historical records,then a total factor productivity (TFP) index can be calculated. If externalities such as the cost to society from exposure to pesticides or the negative effects of sedimentation from soil erosion can be factored into TFP, then a total social fac- tor productivity (TSFP) index can be defined and used to evaluate long-term sustainability of a production system. 16 TIHIE OILID ROTATIIION, I 896- 11 996 Records from the Old Rotation where actual inputs and yields are known are extremely valuable for assessing productivity through the use of indexes.The advantage of using indexes is the ease with which they can be developed and compared. The pri- mary interest in addressing the question of the sustainability of cot- ton production is in the effect on the movement of a total factor productivity index over time. Total factor productivity can be a more informative productivity indicator than partial measures, such as output per unit of land or output per unit of labor. The appeal of TFP is that it can be interpreted as "output per unit of input." As constituted here, this index number formula adjusts for the effect of changing input prices, so that changes in TFP can be attributed to a change in production efficiency, rather than chang- ing market prices; a doubling ofTFP implies that twice as much out- put is derived from each "unit" of input. Selected treatments in the Old Rotation were used to calcu- late TFP and TSFP to assess the sustainability of cotton production under different management strategies. METHODS USED Data from three of the original 13 Old Rotation treatments were used to develop TFP indexes to compare the effect of N fer- tilizers and winter legumes on long-term productivity. Treatments were: I. No N: continuous cotton with no nitrogen and no winter legume (plot 6 in Table I); 2.Winter legumes: continuous cotton with a winter legume (crimson clover and/or vetch) used as a green manure and N source (plot 8), and; 3. N fertilizer: continuous cotton with annual application of 120 pounds N per acre (plot 13). The first two treatments have not changed since the experi- ment was initiated in 1896. The third treatment was in a two-year cotton-vetch-cowpea rotation until 1955. The types and levels of all inputs used on the Old Rotation were not recorded. Therefore, management and input practices on the Old Rotation from 1896 to the present (i.e. pest control, culti- vation, labor, etc.) were assumed to be those recommended by the Alabama Cooperative Extension Service (ACES) and the Alabama Agricultural Experiment Station (AAES) for medium to large Alabama farms. Input budgets were constructed by compiling information on "typical" technologies for each era of the experi- ment. Prices received by farmers came from USDA records. Except for shifts from mules and hoes to tractors and herbi- cides, field operations have remained fairly constant since 1896. Cotton production tools were powered by animals and humans from 1896 to 1939. A transition period from animal to tractor power occurred from 1940 to 1955. Since 1955, production has shifted from two-row (1956-70) to four-row (1970-85) to six- row (1986 to present) soil tillage and planting equipment. Marginal improvements were made in plowing, leveling, bedding, and planting equipment. Representative machinery operations for a typical cotton farm consisted of cutting stalks after harvest; flat-breaking the land using a plow; pulverizing the broken land using disk harrows, harrows and/or a drag; bedding; laying off the rows or opening furrows; planting; fertilizing; cultivating; hand thin- ning ("chopping" cotton); pest control; and harvesting and hauling. From 1896 to 1959 harvesting was accomplished almost exclusively by hand picking, requiring approximately 100-I 14 hours per acre. During the 1960s, the cotton picker gradually replaced field hands and family harvesting labor, dramatically decreasing the labor required to produce an acre of cotton. For this analysis, it was assumed (and the historic record supports) that by 1970-100% of Alabama's cotton was mechanically harvested. Variable costs included in the analysis accounted for seed, fer- tilizer, pesticides, defoliation, interest on operating capital, harvest- ing, ginning, and warehousing as well as for machinery operation. Fixed costs of operation included tractor and machinery deprecia- tion, interest, insurance, and taxes. Estimated returns were to land, management, and owner-operator labor. Ginning and labor were the other major inputs into the pro- duction process. Indexes were used to estimate ginning costs from ACES budgets. Budgets also were used to estimate labor rates and wages for the period 1978 to 1991. Alabama wage rates for 1923 to 1976 were taken from Agricultural Statistics bulletins. A United States wage rate multiplied by 57% was used as the Alabama wage rate prior to 1923. The Tornquist approximation to the Divisia index was used in our analysis (Christensen, 1975). Divisia input, output, and total fac- tor productivity (TFP) indexes can be calculated as: (1) 1 (X)t = 7 i (Xit/]X ,t) (Sit + S it-)/2 (2) I (Y)t = j (Xjt/X jt-_) (Wjt + W jt-1) /2 (3) TFP t = I(Y)t/I(X)t where I(X)t and I(Y)t are quantity indexes of input and output use in time t; Xit is the quantity of input i; Yjt is the quantity of output j; sit = rit Xit/ XritXit where rit is the price of input i, and Wt = pjtYjt / IpjtYjt where Pjt is the price of output j. TFP INDEXES WITHOUT EXTERNAL COSTS In 1896, lint yield accounted for 93% of the output shares. This declined to 89% in 1991 because of the higher value assigned to cotton seed (Table 9). Nevertheless, the movement in lint yield and output index over time are very similar. The most dramat- ic shift in input shares since the Old Rotation began has been in decreased labor shares and increased machinery, harvesting, and ginning shares. Surprisingly, seed and fertilizer command small- er shares today than before the turn of the century. Three distinct eras of productivity change are exhibited by the TFP series (Figure 14a). The TFP of the no-N plot eroded steadily over the first 50 years of the experiment, bottoming out in the mid-1940s at less than 40% of the 1900 level. The turn-of-the-century decline on the winter legume plot was shorter and more moderate, reaching its low point in 1921 at 70% of the 1900 level. Productivity on all three plots peaked in the middle 1960s, declined during the 1970s and appear to have leveled off in the 1980s. The TFP of all plots in 1991 is greater than when the experiment began. 17 THIE OLID R OTATION, 11896 11996 TABLE 9. CHANGES IN OUTPUT AND INPUT SHARES FROM 1896 To 199 I 1896 1991 Output shares Seed 7 II Lint 93 89 Total 100 100 Input shares Seed and fertilizer 19 10 Pesticides 2 I Harvesting/ginning 28 40 Labor 34 6 Variable machinery 14 10 Fixed machinery 2 19 Total 100 100 The largest single event affectingTFP was the introduction of the mechanical harvester. This was factored into the TFP calculations in 1959. This technological advancement, more than any other during the history of the Old Rotation, result- ed in a decline in the input index (increase ofTFP) on all three plots. Hand picking of cotton is an extremely labor intensive activity. The appearance of the cotton harvester had the effect of reducing the overall production labor requirement (per acre) by approximately 70%. The productivity decline of the TFP index (1990= 100) 150 - Total factor productivity -NoN ----- Winter legume N fertilizer 00 - -,50 .,,, t I I I I I I I I I I 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 TSFP index (1990= 100) 150 - Total social factor productivity 10 - I 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 Fig. 14. (a) Total factor productivity indexes and (b) total social factor productivity (including externalities) index- es using five-year moving averages or treatments where cotton is planted every year. 1970s is likely due to the effect of poor management when the main research farm was moved from the vicinity of the Old Rotation. Loss of DDT as a cotton insecticide also made insect control difficult during this transition period. TOTAL SOCIAL FACTOR PRODUCTIVITY The TSFP indexes differ from theTFP indexes in that a soci- etal cost is assigned to soil erosion and pesticide use (Figure 14b). The major categories of external costs of pesticide use are regulatory costs, adverse health effects, and damage to the natural environment. The actual net per acre cost (net of exter- nal benefits) of these effects is not known. The approach fol- lowed in this study is to assume external costs equal to 50% of expenditures on herbicides, insecticides, and defoliants. For soil erosion, Ribaudo's (1989) $2.34 per ton of soil loss, for annual off-site damage costs in the Southeast United States, was multi- plied by an estimated average soil loss quantity. Although casual observation indicates that erosion has occurred on the Old Rotation, a historical record of erosion does not exist for the site. Past erosion was modeled using the Erosion Productivity Index Calculator (EPIC) simulation pro- gram (Williams, 1990). EPIC simulated 96 years of soil and organic matter loss for the no N and winter legume experi- mental plots and 35 years of loss for the 120 pounds N per acre plot. Ten years of daily wind and rainfall data for the Auburn area were used to generate historic weather conditions. Average annual soil loss over the simulations were: Treatment No N/no legume Winter legume N fertilizer Soil loss tons/acre/year 9.9 6.1 7.8 Organic N loss lb./acre/year 12 II 16 The inclusion of external costs did not significantly affect productivity trends in any of the treatments (Figure 14b). The no-N plot indexes are not changed at all; the input indexes on the other two plots increased by an average of about 6%. Total Factor Productivity on the legume and N-fertilized plots decreased by 4% and 6%, respectively. The main conclusions from TFP calculations are unaffected. Therefore, total input use remains relatively stable over time when Ribaudo's (1989) external costs estimates are included in TFP. Significant produc- tivity growth has occurred whether measured as conventional TFP or as TSFP. Sustainability of continuous cotton production as measured byTFP was minimally affected by externalities. INTER-PLOT COMPARISONS It is not possible to assess the relative performance of the three plots using either the TFP or TSFP indexes discussed above. However, the input use, output use, and productivity of the winter legume and N-fertilized treatments relative to the no-N treatment can be calculated as indexes (Cooke and Sundquist, 1991). When this was done, the relative output index- es (not shown) merely confirm that the output (yield) of the no-N and winter legume plots were similar during the early 18 ~ \IIA~ I~~ I~ r C ~AII~~ ~~ A THE OILD ROTATION, 11896- II 996 Index (no nitrogen plot= 100) 300 - Relative input index 250 200 150 100 0 - Index (no 300 r 250 - 200 - Winter legume N fertilizer 1 r? ~I '*\ I I I I I I 1900 nitrogen plot = 100) 1920 1940 1960 1980 2000 Relative TFP P 51 ~g /~ VT -'NJ 1900 1920 1940 1960 1980 2000 Fig. 15. Relative input (a) and TFP (b) indexes relative to the no-N treatment. years of the experiment, but then diverged. Yield and output on the winter legume and N-fertilized plots were similar from 1955 to 1991. The input series (Figure 15a), as expected, shows that the winter legume and fertilizer-N plots used more input than the no-N plot. However, much of this apparent input intensity is related to increased harvest costs, which were due to higher yields on the legume and N-fertilized plots. More harvest labor is needed and ginning costs are higher on an acre that yields two bales than on an acre that yields one bale of cotton. The most informative inter-plot comparisons are obtained from the productivity measures of TFP in Figure 15b. The win- ter legume plot relative TFP hovers near 100 until 1920, climbs to 200 in the 1940s, declines in the 1950s through the 1970s, and climbs back to 200 in the 1980s and 1990s. The TFP of the N-fertilized plot follows a similar pattern, but is slightly below the winter legume plot in all years. One of the most interesting findings is that there is a greater difference in relative TSFP than relative TFR The differ- ence is 6-8% in most years on both plots. This implies that accounting for externalities enhances the productivity advantage of the winter legume and chemical nitrogen plots. In other words, the low input system (no N), has less desirable environ- mental consequences than either the winter legume or N-fer- tilized systems. This is due to the decrease in soil erosion brought about through the higher biomass production of the winter legume and N-fertilized plots. CONCILUSIIONS Viewed from the 97-year perspective of the Old Rotation, each of the plots fulfilled at least one criteria required for a system to be sustainable. Output per unit of input is higher in 1991 than in 1896 even when externalities are valued. Other conclusions from this study are: (I)None of the systems show a linear trend in output orTFP over the history of the experiment. Productivity cycles are present in all three systems, despite the positive overall trend. An impor- tant focus of future research will be to attempt to explain whether these cycles are related to weather, technology, or changes in the resource base. (2)The system that has neither an organic or a chemical source of added N is less productive than the other two systems. This system compares even more poorly when externality costs are assigned. (3)Organic and chemical sources of N have similar productiv- ity impacts. (4)Soil erosion and pesticide externalities have a modest effect on measured productivity. (5)The most dramatic single event to affect productivity was the introduction of the mechanical cotton picker. The impact of this technology is powerful enough to offset the effect of many other changes in the system. *Funding for this research was supported, in part, by The Rockefeller Foundation and has been reported by Traxier et al. (1995). 19 Winte40 Iso 1500 THIE OLD ROTATIION, 11896-1996 *Funding for this research was supported, in part, by The Rockefeller Foundation and has been reported by Traxler et al. (1995). RIIE IF E ]RIENC ES Amer.Assoc. for State and Local History. 1989. National Register of Historic Places--cumulative list 1966-1988.AASLH. Nashville,TN. Bailey, R.Y.,J.T.Williamson, and J.F Duggar. 1930. Experiments with Legumes in Alabama. Ala.Agric. Exp. Sta. Bul. 232.Ala. Polytechnic Inst.Auburn,AL. Binswanger, H.P 1978. Measured Biases of Technical Change: The United States. In H.P Christensen, L.R. 1975. Concepts and Measurement of Agricultural Productivity.Amer.J. of Agric. Economics. 75:910-915. Cooke, S., and W.B. Sundquist. 1991. "Measuring and Explaining the Decline in U.S. Cotton Productivity Growth." Sou. J.Agric. Econ. 23:105-120. Davis, FL. 1949.The Old Rotation at Auburn,Alabama. Better Crops with Plant Food. Reprint DD-8-49.Am. Potash Institute, Inc.Washington, D.C.. Entry,J.A., C.C. Mitchell, and C.B. Backman. 1996. Influence of Management Practices on Soil Organic Matter, Microbial Biomass and Cotton Yield in Alabama's Old Rotation. Biol. Fertl. Soils (in press). Lu,Y., P Cline, and L. Quance. 1979. Prospects for Productivity Growth in U.S.Agriculture.Agric. Econ. Report No.435. Economics, Statistics, and Cooperatives Service. USDA. Washington, DC. Mitchell, C.C., R.L.Westerman,J.R. Brown, and TR. Peck. 1991. Overview of Long-term Agronomic Research.Agron. J. 83:24-29. Novak, J., and WR. Goodman. 1994. Profitable and. Environmentally Sound Agriculture:A Sustainable Approach to the Future.J.Agric. Food Information 2:2-4. Pfeffer, M.J. 1992. Sustainable Agriculture in Historical Perspective.Agric. Human Values 9(4). Ribaudo, M.O. 1989.Water Quality Benefits from the Conservation Reserve Program. USDA, Resources and Technology Division, ERS,Agri. Econ. Report No. 606. Steiner, R.A., and R.W. Herdt. 1995.The Directory of Long- term Agricultural Experiments:Vol. I. FAO. Rome. Taylor, D.C. 1990. On-farm Sustainable Agriculture Research: Lessons From the Past, Directions for the Future.J. Sustain.Agric. 1(2). Traxler, G., J. Novak, C.C. Mitchell, Jr., and M. Runge. 1995. Long-term Cotton Productivity Under Organic, Chemical, and No Nitrogen Fertilizer Treatments, 1896-1992. pp. 4 1-61. In V. Barnett, R. Payne, and R. Steiner (editors), Agricultural Sustainability: Economic, Environmental and Statistical Considerations. John Wiley & Sons, Ltd. Trimble, S.W 1974. Man-induced Soil Erosion on the Southern Piedmont 1700-1970. Soil Cons. Soc. of Amer.Ankeny, Iowa. Williams, J.R., et al. 1990. EPIC-Erosion/Productivity Impact Calculator. USDA-ARS Tech. Bul. 1768.Temple,TX. 20 THE QILID IROIFATBON, 11 896--11996 APPIENDRX 10 13 Yields, (lb/aocre) 1,050 965 832 824 1,226 1,032 805 957 1,017 1,016 1,112 794 805 525 452 437 800 656 811 829 800 720 1,039 800 - 593 853 557 1,128 552 848 715 1,464 676 1,01 1 618 626 496 816 570 missing years 1,117 307 578' 227 1,430 546 1,120 152 1, 504 470 missing year 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916-19 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 193 1 1932 1933 1934 1935 1,134 928 1,075 1,192 956 1,045 701 836 800 816 700 768 747 626 628 915 845 590 382 789 512 472 19083 515 1,040 1,222 1,410 602 1,038 1,240 1,030 1,466 1,166 1,096 1,454 1,088 856 1,051I 1,033 874 1,008 717 491 769 821 916 792 784 550 692 979 931 783 480 606 555 392 840 725 1,347 1,077 1,714 537 1,283 1,346 900 1,728 1,262 1,123 1,656 898 733 858 733 887 963 418 772 656 771 568 814 766 633 760 706 685 571 503 571 325 355 965 374 811 986 1,544 633 1,120 1,091 1,015 1,356 1,293 1,061 1,447 Year APPENDixTABLE 1. SEED COTTONYIELDS (POUNDS PERACRE) BYYEAR, 1896-1935 Plot 1, 349 1,483 621 1,144 1,206 840 1,627 1,073 1,183 19337 395 586 254 536 725 492 636 490 466 521 705 763 - 698 953 901 589 481 736 805 1,182 799 804 841 888 754 914 760 378 720 237 499 720 909 1,155 987 1,235 432 1,397 866 924 1,332 1,529 1,358 1,392 701 1,065 896 437 608 959 956 720 1,011 601 562 818 1,346 1,170 967 996 1,363 21 THE QILD ROTATI~ON , 1896--1l996 APPENDIX TABLE 2. SEED COTTON YIELDS (POUNDS PER ACRE) BYYEAR, 1936-1995 Plot Year 1 2 3 4 &7 5 &9 6 8 10,II,1& 12 13 Yields, (lb.! acre) 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972-77 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989Q 1,183 1,061 1,332 994 917 595 627 754 960 l5 662 667 888 886 989 737 696 8 1,121 629 667 626 556 816 389 730 754 612 456 682 778 636 669 991 362 442 634 650 919 514 512 758 512 470 726 290n 530 590 617 559 499 235 235 365 607 67 17 461 2,107 1,735 1,582 1,502 1,022 19793 19202 3,000 1,510 1,824 2707 1,735 2,940 217 2,537 2,246 2,897 2,714 1,872 1,735 29414 2,210 1,915 1,610 1,248 1,308 1,970 2,582 1,234 19252 2,909 945 1132 2,033 3,085 1,888 2,3 2,9 3,9 1,334 970 1,622 1,390 1,068 1,049 1,032 1,459 1,656 811 208 1,603 2,563 1,582 1,529 1,498 998 1,836 1,030 3,026 1,495 1,836 2,788 1,652 2,976 2,039 2,765 2,160 3,026 2,849 1,836 1,668 2,611 2,243 2,011 1,t699 1,212 1,342 1,896 2,616 1,210 1,437 2,792 1,217 1,014 1,597 2,850 1,404 1,145 1,805 1,594 1,327 1,193 1,354 1,594 1,826 1,140 723 1,903 2,597 1,939 1,651 1,798 1,061 2,158 1,262 3,161 1,788 2,196 2,798 2,077 2,947 2,534 3,067 2,602 3,408 1 ,997 missing year 2,038 1,934 2,921 2,441 2,551 2,012 2,234 1,733 missing years 1,634 1,900 1,402 - 1,630 1,577 2,102 - 2,652 2,796 1,685- 1,132 1,241 2,413- 1,329 1,292 1,554- 1,670 2,105 2,722- 2,069 2,178 2,870 2,790 3,410 3,300- 2,940 2,250 1,255 1,051I 1, 630 1,186 1,344 1,097 814 960 1,471 941 833 1,404 2,330 1,678 1,529 1,651 1,022 1,793 1,150 2,534 1,522 2,129 2,776 29102 2,465 2,366 2,693 2,779 3,250 2,995 1,944 442 516 521 502 334 180 264 430 473 74 24 398 482 559 689 694 542 449 461 859 499 586 624 452 756 612 679 684 545 790 768 610 516 650 1,010 439 533 655 734 907 509 631 486 405 484 1,016 817 1,279 1,058 1,488 1,519 1,006 977 1,027 1,435 1,512 838 267 1,637 1,968 1,766 1,466 1,654 1,008 1,838 1,267 2,861 1,582 2,052 2,332 1,897 2,822 2,362 2,443 29412 2,844 29813 2,021 1, 846 2,441 2,087 2,189 19882 19553 1,622 2,134 2,712 1,555 1,143 2,696 972 19124 1,815 3,720 1,193 1, 291 1,478 1,181 1,195 1,111 698 883 1,262 1,130 468 1,639 1,682 1,985 1,651 1,5 10 1,262 1,956 1,358 2,623 2,062 1,968 2,803 2,243 2,686 2,580 2,911 2,885 2,861 3,202 2,647 2,268 2,606 2,541 2,122 1,634 1,730 2,21 0 2,371 2,755 2,030 2530 1,206 1,608 1,670 3,103 I2,14 I2,98 2,87 I2,54 2,36 1135 191 14 833 794 355 626 982 691 785 1,265 1,368 2,054 2,102 1,618 2,405 1,886 1,824 2,006 2,479 19874 2,071 19853 2,086 1,970 2,208 1,361 19522 1,594 1,735 2,333 1,445 860 2,189 1,258 1,960 1,742 22 I~I1J I 1~1 I I 1~~ 1 7 -e APPENDIX TABLE 3A. CORN GRAIN YIELDS (BUSHELS PER ACRE) BYYNFEAR, 1896-1935 Plot Year 1 2 14& 7 10,11, & 12 Yields, (bu./acre) 1896 18 16 I15- 1897 25 20 22 I5 1898 12 12 14 12 1899 18 19 18 1S 1900 30 29 25 24 1901 15 18 IS 14 1902 missing year 1903 21 20 21 22 1904 21 I5 17 14 1905 13 I5 14 10 1906 10 13 12 14 1907 28 16 19 18 1908 I5 9 16 11 1909 19 16 19 17 1910 18 12 14 I5 1911 19 12 16 12 1912 22 10 13 16 1913 17 14 13 15 1914 5 2 3 2 1915 10 6 7 6 19 16-19 missing years 1920 16 9 14 17 1921 19 10 19 12 1922 17 8 12 18 1923 18 7 11 12 1924 17 8 1S 18 1925 missing year 1926 16 13 16 16 1927 27 10 23 32 1928 31 9 42 38 1929 34 9 40 25 1930 23 8 24 27 1931 17 12 22 23 *=missing year of data. APPENDIX TABLE 3C. CORN GRAIN YIELDS (BUSHELS PER ACRE) BYY FEAR, 1978-1994 Plot Year 4 &7 5& 9 10,II1AI2 Yields, (bul1acre) 1978 I5 38 46 1979 41 45 57 1980 13 27 31 1981 40 45 30 1982 79 90 92 1983 27 28 36 1984 -- 102 1985 18 18 SI TiHiiE OLD IllOTA~if ON, 1896:11 996 APPENDIX TABLE 3B.CORN GRAIN YIELDS (BUSHELS PER ACRE) BY YEAR, 1936-1977 Plot Year 4&7 10,II,& 12 Yields, (bul1acre) 1932 ______________ 23 26 1933 _______________21 - -26 1934______________ 27 31 1935 ______________ 44 49 1936 27 2,304 1937______________ _ 16 26 1938______________ _ 46 42 1939______________ _ 25 29 1940 ______________ 41 44 1941 ______________ 46 47 1942______________ _ 34 38 1943__ _ _ _ _ _ _ _ _ _ _ _ _ _ 41 36 1944______________ _ 24 17 1945__ _ _ _ _ _ _ _ _ _ _ _ _ _ 39 51 1946______________ _ 37 47 1947______________ _ 39 52 1948______________ _ 76 69 1949______________ _ 48 54 1950______________ _ 46 54 1951__ _ _ _ _ _ _ _ _ _ _ _ _ _ 18 28 1952______________ _ 26 42 1953_______________ _ 61 55 1954 _ _ _ _ _ _ _ _ _ _ _ _ _ _ 4 14 1955_______________ _ 43 71 1956 ________________ 50 53 1957 __ _ _ _ _ _ _ _ _ _ _ _ _86 111 1958 110 109 1959 _______________ 62 89 1960______________ _ 83 90 1961__ _ _ _ _ _ _ _ _ _ _ _ _ _ 81 100 1962______________ _ 28 47 1963_______________ _ 31 43 1964______________ _ 90 108 1965 _______________ 62 95 1966______________ _ 19 50 1967 _________________ missing year 1968 _______________ 24 63 1969______________ _ 25 38 1970______________ _ 54 72 1971__ _ _ _ _ _ _ _ _ _ _ _ _ 96 118 1972-77_____________ ___ missing years 23 TIHIE OILID ROOTATION, II 896 11996 APPENDIX TABLE 4. ESTIMATED DRY MATTER YIELDS (POUNDS PER ACRE) OF WINTER LEGUMES BY YEAR* Plot Year I 2 3 4 5 7 8 9 10 II 12 13 1896-1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1978 1979 1980 1981 1982 1983 1984-85 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 (Lb./ acre) missing years 2,325 -** 1,078 1,261 1,238 1,260 1,090 974 1,169 1,342 2,268 2,851 2,3 16 1,499 2,846 2,257 2,952 789 1,969 missing year 3,262 - 1,870 3,080 2,771 2,593 1,191 2,342 3,337 1,157 2,182 - 1,443 2,120 1,550 1,918 1,693 1,105 1,021 1,645 2,976 - 1,562 3,124 3,036 2,286 2,842 2,341 2,487 2,551 - - 264 924 569 859 1,060 813 298 993 -- 638 2,049 658 747 856 586 1,047 1,430 - - 1,922 2,259 1,649 1,598 1,175 1,075 2,031 1,384 -- 1,112 1,387 1,084 1,029, 1,025 774 1,252 997 - - 1,444 1,369 1,106 1,088 1,031 904 1,676 975 - - 1,031 280 214 209 207 155 826 588 - - 2,004 3,862 2,332 2,550 3,023 2,195 2,095 2,432 - - 624 530 400 698 564 557 432 393 - - 410 622 346 396 381 383 556 472 - 606 1,111 519 619 487 510 774 838 - 662 1,328 495 837 608 621 994 739 -- 1,766 685 290 1,760 1,670 522 221 947 - - 668 1,576 716 1,008 882 864 1,012 914 - - 1,080 3,330 2, 147 774 1,131 936 2,412 1,391 - - 269 2,178 2,196 320 456 788 1,872 743 - - 2,115 1,850 1,121 1,971 1,949 1,985 1,107 2,327 2,939 1,921 1,094 1,242 1,890 1,755 1,355 2,381 1,818 - 2,066 1,625 1,287 1,584 734 1,413 567 1,328 1,197 - 779 486 945 909 540 373 540 990 464 - 584 375 504 562 322 388 221 162 299 - 3,686 3,501 3,636 4,019 3,501 3,654 3,357 3,209 - 2,408 2,453 3,101 3,209 3,659 3,506 3,582 - 2,484 - 2,685 2,399 3,259 2,862 2,624 3,236 2,603 2,565 2,199 - 1,301 1,247 2,061 1,989 2,053 1,827 2,160 2,097 - 2,052 1,951 2, 165 1,656 2,372 2,363 2,448 - 2,421 - 2,866 2,880 3,101 3,346 2,903 3,167 3,593 3,006 2,934 - 1,978 2,129 1,784 1,851 1,701 1,310 1,980 1,832 - 2,066 2,520 2,903 2,300 2,655 2,264 2,223 - 2,696 - 2,264 3,128 2,777 2,651 2,340 1,958 2,462 2,115 2,075 - 1,944 1,976 1,265 1,089 1,643 1,512 2,016 1,494 - 590 626 977 1,085 720 608 923 - 2,142 - 234 180 207 180 158 135 171 270 185 - 1,998 2,358 2,367 2,133 1,962 2,457 1,197 - - 2,412 2,106 1,602 1,287 2,034 2,079 1,287 - 2,448 - 1,494 1,764 882 774 1,773 1,854 1,539 1,782 1,575 missing year - 2,016 2,426 2,673 2,894 1,301 2,061 1,895 - 1,305 - 1,377 1,188 1,062 1,890 3,330 1,242 2,700 2,322 1,656 - 1,710 2,101 1,885 1,811 1,921 1,699 2,043 1,844 - - 2,769 2,673 2,637 2,931 2,835 2,826 2,871 - 1,143 - 3,285 3,23 I 2,889 3,276 2,664 3,699 3,744 1,458 3,438 - 2,250 2,214 2,187 2,574 2,952 2,187 2,412 2,727 - - 3,447 3,699 2,601 2,430 4,005 3,735 3,600 3,411 - 5,03 I 6,111 6,471 6,057 6,570 6,903 5,085 5,949 7,245 - 3,744 3,339 2,601 2,817 4,095 3,924 3,726 3,339 - - 3,753 4,176 4,059 5,148 4,266 3,906 3,717 - 3,654 missing years - 1,719 2,088 1,760 2,412 3,600 977 1,784 - 2,263 - 3,438 3,006 2,430 2,772 2,610 3,096 2,106 - 2,936 3,402 3,618 3,582 3,618 3,816 4,590 3,618 3,654 - missing year - 3,680 4,090 4,420 3,180 3,330 4,590 3,580 - 3,480 - 3,910 4,680 5,130 5,060 4,300 5,190 4,150 5,140 2,730 - 3,420 4,180 3,970 4,850 3,450 - 3,780 3,620 2,950 - 3,030 2,750 2,610 3,140 3,550 3,490 3,460 2,980 5,790 - 4,513 3,739 3,571 4,118 4,462 4,163 4,489 - - - 3,040 1,910 1,880 2,800 3,520 1,790 4,460 - - 1,459 - 1,558 S 2,123 - 1,155 S 2,758 S 556 - 870 - 774 S 970 - 593 - 449 - 2,277 S 386 - 310 - 410 - 352 - 252 - 941 - 995 - 482 - ,238 S 1,193 - 846 - 347 3,362 2,043 3,753 2,812 - 1,676 2,624 1,517 2,718 - 1,422 - 2,331 1,152 - 900 - 1,999 1,701 - 1,719 - 2,176 - 3,051 - 3,195 - 3,519 - 3,636 2,502 - 24 *NOTE:From 1926 to 1989 the yields were reported in a wet weight basis.The yields here reported are dry, using l 8% dry matter as an estimate. From 1990 to 1995 the yields were recorded dry. **_ = missing year of data. I IIULJ IIIII I ~II I IA I ~- THE OLID I ROTATION, 11896-11996 APPENDIX TABLE 5.YIELDS OF SMALL GRAINS (PLOTS 10, I I, AND 12) IN BUSHELS PER ACRE BY YEAR Year Type Yield I896 I897 1898 I899 I900 I901 I902 I903 1904 I905 1906 1907 I908 1909 1910 9 I1 1912 1913 I 914 1915 1916 1917 1918 1919 1920 1921 I922 1923 1924 1925 1926 I927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat oat Bu./acre 24 25 3 26 9 14 18 32 19 16 13 25 25 25 25 7 7 5 14 45 55 0 20 0 61 71 68 57 62 70 97 39 69 64 66 69 41 * = missing year of data. KNOWN PUBLICATIONs/ABSTRACTS CONTAINING INFORMATION DERIVED FROM THE OLD ROTATION Publications: Bailey, R.Y., J.T Williamson, and J.F. Duggar. 1930. Experiments with Legumes in Alabama.Ala.Agr. Exp. Sta. Bul. 232.Auburn Univ.,AL. Cope, J.T., Jr. 1961. Fertilizing Cotton-corn Rotations. Highlights of Agric. Res. 8(l): I I.Ala.Agr Exp. Sta.Auburn Univ.,AL. Davis, FL. 1949.The Old Rotation atAuburn,Alabama. Reprint DD-8- 49. Better Crops with Plant Food.Am. Potash Inst., Inc.Washington, DC. Entry, J.A., C.C. Mitchell, and C.B. Backman. 1996. Influence of Management Practices on Soil Organic Matter, Microbial Biomass and Cotton Yield in Alabama's Old Rotation. Biol. Fertl. Soils (in press). Evans, E.M.,and D.G. Sturkie. 1974.Winter Legumes Can Help Supply Nitrogen Needs. Highlights of Agr. Res. 21 (3):3.Ala. Agr. Exp. Sta. Auburn Univ.,AL. Donnelly, E.D., 1976. The Role of Vetches in Forage and Cropping Systems. Proc. Miss. Sec.Am. Soc.Agron.Jan. 15, 1976. pp. 69-70. Mississippi State Univ., Starkville, MS. Duggar,J.E 1898. Experiments With Cotton.Ala.Agr Exp. Sta. Bul. 89. Auburn Univ.,AL. 1899. Experiments With Cotton. I 898.Ala. Agr. Exp. Sta. Bul. 101.Auburn Univ.,AL. 25 ___ STHE OILD IROTATION, 11896- 11996 * 1899. Results of Experiments on Cotton in Alabama.Ala.Agr. Exp. Sta. Bul. 107.Auburn Univ.,AL. SI 1899. Winter Pasturage, Hay and Fertility Afforded By Hairy Vetch.Ala.Agr. Exp. Sta. Bul. 105.Auburn Univ.,AL. . 1900. Corn Culture.Ala.Agr. Exp. Sta. Bul. II I.Auburn Univ.,AL. . 1902. Cowpea Culture. Ala. Agr. Exp. Sta. Bul. I 18.Auburn Univ.,AL. 1902.The Cowpea and the Velvetbean As Fertilizers.Ala.Agr. Exp. Sta. Bul. I 20.Auburn CUniv.,AL. . 1905. Corn Culture.Ala.Agr. Exp. Sta. Bul. 134.Auburn Univ.,AL. * 1906. ExperimentsWith Oats.Ala.Agr. Exp. Sta. Bul. 137.Auburn Univ.,AL. ,and E.F. Cauthen. 1913. Oats: Experiments On Culture,Varieties, and Fertilization. Ala.Agr. Exp. Sta. Bul. 173. Auburn Univ.,AL. Ensminger, L.E. 1969. History of Agronomy and Soils at Auburn University. Highlights ofAgr. Res. 16(3).Ala.Agr. Exp. Sta.Auburn Univ.,AL. Funchess. M.J. 1923. Legumes in Relation to Soil Fertility.Ala.Agr. Exp. Sta. Cir. 48.Auburn Univ.,AL. . 1936. Hairy Vetch and Austrian Winter Peas for Soil Improvement--A Progress Report.Ala.Agr Exp. Sta. Cir 74. Auburn Univ.,AL. Kerr, Norwood. 1983. Development of the Alabama Agricultural Experiment Station During Its First IOOYears 1883-1983. Highlights ofAgr Res. 30(1):4-9.Ala.Agr. Exp. Sta.Auburn Univ.,AL. Kessler, K. 1993. Long-term Research Answers Questions About Sustainability. The Furrow magazine. Special Conservation Issue.Jan., 1993. pp. 30-31.John Deere, Moline, IL. Mitchell, C.C., Jr. 1988. New Information from Old Rotation. Highlights ofAgr. Res. 35(4):3.Ala.Agr. Exp. Sta.Auburn Univ.,AL. _ R.L. Westerman, J.R. Brown, and TR. Peck. 199 1. Overview of Long-term Agronomic Research.Agron.J. 83:24-29. . 1992. Long-term Soil Fertility Studies-- What They Tell Us. Proc. 1992 Sou. Soil Fert. Conf. p. 2- 10.The S.R. Noble Foundation, Inc.,Ardmore, OK. G.J. Traxler, and J.L. Novak. 1994. Old Rotation Documents Sustainable Cotton Production. Highlights of Agr. Res. 41 (4).Ala.Agr. Exp. Sta.Auburn Univ.,AL. ,G.J. Traxler, and J.L. Novak. 1994. Alabama's Old Rotation Documents Sustainable Cotton Production. Proc. Sou. Soil Fert. Conf. p. 60-61.The S.R. Noble Foundation, Inc., Ardmore, OK. Novak,J.L., C.C. Mitchell,Jr., and J.R. Crews. 1990. Economic Risk and the 92-year "Old Rotation": Implications for a 250-acre Farm.Ala.Agr. Exp. Sta. Cir. 300.Auburn Univ.,AL. ,C.C. Mitchell, Jr., and J.R. Crews. 1989. AAES "Old Rotation" Results Identify Least Risky Rotations. Highlights of Agr. Res. 36(4): 14.Ala.Agr. Exp. Sta.Auburn Univ.,AL. G.J. Traxler, M.W Runge, and C.C. Mitchell, Jr. 1993. Factor Productivity Indexes of Sustainable Cotton Rotations Under Alternative Nitrogen Fertilizer Treatments, 1896-1992. Proc.Am.Agric. Econ.Asso.,Aug. 1-4, Orlando, FL. G.J. Traxler, M.W Runge, and C.C. Mitchell, Jr. 1995. The Effect of Mechanical Harvesting Technology on Southern Piedmont Cotton Production, 1896-1991. Agr. History 69(2):349-366. C.C. Mitchell, Jr., and G.J. Traxler. 1995. Total Factor Productivity of Continuous Cotton Production in Central Alabama. Proc. 1995 Beltwide Cotton Conf. San Antonio,TX. pp 376-379. Traxler, G.J.,J.L. Novak, C.C. Mitchell,Jr., and M.W Runge. 1995. Long- term Cotton Productivity Under Organic, Chemical, and No Nitrogen Fertilizer Treatments, 1896-1992. pp. 41-61. In V. Barnett, R. Payne and R. Steiner (editors), Agricultural Sustainability: Economic, Environmental and Statistical Considerations.John Wiley & Sons, Ltd. Abstracts: Chapman, L.J., and C.C. Mitchell, Jr. 1989.America's Oldest Cotton Study--Alabama's "Old Rotation." Agron.Abstr. p.2 3 5. McConnell, J.S., R.M. McConnell, C.C. Mitchell, Jr., and D.W Miller 1990. Characterization of Humic and Fulvic Acid Fractions of Soils in Continuous Cotton and Cotton Rotations.Agron.Abstr. p. 234. Mitchell, C.C., Jr. 1988. Long-term Production Systems Research in Alabama. Abst. of MEY Production Systems Conference. Sept. 7-9, 1988. Tifton, GA. R.L.Westerman, J.R. Brown, and TR. Peck. 1989. Overview of Long-term Agronomic Research.Agron.Abstr. p. 247. ,G.J. Traxler, J.L Novak, and M.W Runge. 1993. Measuring Sustainable Cotton Production Using Total Factor Productivity.Agron.Abstr. p. 76. Novak,J.L., G.J.Traxler, M.W. Runge, and C.C. Mitchell,Jr. 1993. Factor Productivity Indexes of Sustainable Cotton Rotations Under Alternative Fertilizer Treatments, 1896-1990. Selected paper,Amer. Agric. Econ. Assn. annual meeting, Orlando, FL.AJAE 75:5. 26 a pL THE YEAR 1996 FINDS AUBURN UNIVERSITY celebrating the centennial of the Old Rotation - 100 years of research condensed into a small plot of land! Much credit is due to our forefathers for their vision and wisdom in initiating long-term experiments and to their successors for continuing these experiments. Today, the public demands immediate solutions to problems, and to establish an experi- ment to last 100 years would likely be deemed nonproductive and impractical. Farming today requires sophisticated management and large investments of money rather than just a way of life. Many external factors, such as government policies, environmental concerns, food safety, and public perception, must be integrated into farm management schemes. The interval from a research idea to implementation of new technology has greatly decreased. These demands have placed an added burden on the research and extension faculty. Despite our fast-moving world, long-term research projects are as important today as the day the Old Rotation was implemented. Much can be learned from these long-term experi- ments about plant nutrient requirements, diseases, fertility, soil texture, compaction, etc. that may not be applicable today, but is necessary for the sustainability of agriculture in the next century. The Old Rotation is not just historical - although there is much history in it. It was intend- ed to establish a base of information and provide a reference point for measuring change over time. Based on the data available, it is successful. The centennial celebration of the Old Rotation provides an opportunity for inventory and rededication. We have traditions and history, and from them facts and precepts for guidance in the future. We have gained knowledge, ability, experience, spirit, and the science and tech- nology to expand our horizon. What will be written of the Old Rotation in the year 2096? We can only hope that it will be said that we were as wise and diligent as our forefathers in maintaining this work and fos- tering long-term projects. Lowell T Frobish Director Alabama Agricultural Experiment Station I U