the
oLO
Rotation
1995
Agronomy and Soils Progress Report No. 128 April 1996
Alabama Agricultural Experiment Station Lowell T. Frobish, Director Auburn University, Alabama
This Progress Report is the first in a planned annual update on the "Old Rotation. "
Crops yields and trends on the Old Rotation seem to reflect experiences that Alabama cotton
producers also experience each year. The Old Rotation is an index of long-term sustainability
of cotton production in Alabama.
Acknowledgments
The Old Rotation exists as the world's oldest, continuous cotton plots, and the third
oldest continuous field crop experiment on the same site in the United States because of the
dedication of cooperation of many individual researchers and administrators at Auburn
University. Recently, the support of the Alabama Agricultural Experiment Station (AAES),
under the leadership ofAAES Director Dr. Lowell Frobish and AAES Associate Director Dr.
David Teem, has been the primary reason the Old Rotation has continued to exist. The help
of the staff at E. V Smith Research Center (EVSRC), Dr. Jim Bannon, EVSRC Director, and
the AAES's Research Operations on the A. U. campus have been necessary to plant, maintain,
and harvest the plots. Most of the day-to-day work and maintenance is conducted by Mr.
Charlie France, Research Technician, who has worked on these plots for more than 35
years.
Information contained herein is available to all persons regardless of race,
color, sex, or national origin.
THE OLD RoTATION--1995
CHARLES C. MITCHELL
1
The "Old Rotation" experiment on the campus of Auburn University contains the oldest, continu-
ous cotton plots in the world. The test was started in 1896 by Professor J.F. Duggar to test and demon-
strate his theories that sustainable cotton production was possible on Alabama soils if growers used crop
rotation and include winter legumes (clovers and/or vetch) to protect the soil from winter erosion and
provide nitrogen (N) for the summer crop. The experiment has continued with only slight modifications in
the treatments used on the 13 original plots. The Old Rotation was placed on the National Register of
Historical Places in 1988. The year 1995 marks the 100th cropping year for the Old Rotation.
OBJECTIVES
The objectives today are very similar to Professor Duggar's original objective, i.e. to determine the
effect of crop rotations and winter legumes on sustainable production of cotton in the southern U.S. In
addition, fertilizer phosphorus (P) and potassium (K) treatments initiated in 1925 allowed early researchers
to evaluate the timing of P and K applications to cotton rotation systems. Today, the site is also used as a
field laboratory for researchers, students, and visitors interested in long-term, sustainable crop production
systems in the southern U.S.
METHODS
The site is at the junction of the Piedmont Plateau and Gulf Coastal Plain soil physiographic re-
gions. The soil is generally identified as a Pacolet sandy loam (clayey, kaolinitic, thermic Typic
Hapluduts). There are 13 plots on one acre of land. Each plot is 136 feet long by 21.5 feet wide with a
three-foot alley between each plot. Originally, each plot was a separate treatment, but today, treatments
may be described as the following cropping systems:
I. Continuous cotton
A. No legume/no fertilizer N (plots 1 & 6)
B. Winter legumes (crimson clover and/or vetch) (plots 2, 3, & 8)
C. 120 pounds N per acre (as ammonium nitrate) (plot 13)
II. Cotton-corn rotation
A. Winter legumes (plots 4 & 7)
B. Winter legumes + 120 pounds N per acre (plots 5 & 9)
III. Three-year rotation
(1) Cotton (winter legumes) - (2) corn (small grain for grain)
(3) soybeans (plots 10, 11, & 12)
I Mitchell is Professor 
of Agronomy and Soils 
at Auburn University.
Of minor interest today is the timing of fertilizer P and K. Originally, the soil was low in both P
and K and the winter legume produced more biomass (and more N) with direct P and K applications. This
provided more N for the following cotton crop resulting in higher cotton yields. Today, all soils test 
high
in P and K and there is no longer a differential response to the time of fertilizer application although the
treatments continue. They are:
(1) P & K applied prior to planting cotton (plot 8)
(2) P & K applied to the winter legume in fall (plot 2)
(3) P & K split, i.e. 1/2 to cotton and 1/2 to legume (plot 3)
(4) P & K split between cotton and winter legumes in a cotton-corn rotation (plots 4 & 7)
All plots have received a total application of 80 pounds P
2
0
5 
and 60 pounds K
2
0 per acre per year since
1956. Fertilizer N or legume N is the only fertility variable. Lime is applied to each plot as determined 
by
a soil test to maintain soil pH between 5.8 and 6.5. Soil samples are taken in even-numbered years.
Crop varieties planted have always been those common varieties recommended and used by area
growers. In 1995, varieties planted and dates harvested were:
Cron Variety/cultivar Date planted Date harvested
Winter legume AU Robin 11/3/94 4/3/95
Cotton DPL 5690 5/10/95 9/20/95
Corn DK 689 4/13/95 8/30/95
Small grain wheat 10/15/94 5/26/95
Soybean Stonewall 6/20/95 11/15/95
Management involves conventional tillage and pest control. Subsoiling under the row just prior to
planting was implemented about eight years ago. The following management sequence is generally used:
early April: clip winter legumes for dry matter yield; disk and turn under winter legume
mid April: subsoil cotton and corn plots; plant corn using conventional tillage
late April: plant cotton using conventional tillage
late May/early June: harvest small grain for grain
mid June: plant soybean following small grain
summer: scout cotton and apply appropriate insecticides
late August: harvest corn for grain
September: plant small grain following corn
early October: harvest cotton; overseed with winter legumes and incorporate with cultivator;
late October: chop cotton stalks when winter legumes are established
early November: harvest soybean
1995 RESULTS
.CROP YIELDS. This year marked the 100th cropping year for the Old Rotation. It also produced one of the
worst cotton crops in the history of the experiment. Long-term yields on plot #3 (cotton with winter
legumes) gives an indication of the variation in yields from year to year and the five-year rnning average
(Fig. 1). Estimated lint yields (assuming 38% lint in seed cotton) was 310 pounds per acre compared to the
2
1985-1994 10-year average of 920 pounds lint per acre (Table 1). Generally, yields from plot 3 are about
20% higher than the state average cotton yields. Cotton was initially planted on April 21. A 1.5-inch 
rain
severely crusted the soil in all plots resulting in a very weak stand. The test was disked and cotton was
replanted on May 10. A very dry May, June, and July devastated the cotton crop in central Alabama.
Fortunately, the Old Rotation escaped some of the insect pressure commercial fields experienced. This
may be related to the relative small size of the test and the large buffer areas which support beneficial
insects.
The corn grain crop was also devastated by the drought. Grain yields on the three corn plots aver-
aged 35 bushels per acre compared to a 10-year average of 90 bushels per acre.
Surprisingly, soybean following small grain did much better in 1995 than in 1994. The 1994 crop
was not harvested due to extremely low yields. Rains in August, September, and especially Hurricane
Opal on October 5, resulted in enough moisture to produce 27 bushels soybean per acre on plot #10.
SoIL ORGANIC MATrER. Interest in sustainable agricultural systems and soil quality dominates agronomic
research today. Surprisingly, little effort has been directed over the past 100 years toward documenting the
effects of the cropping systems in the Old Rotation on soil organic matter and its effect on yields. Soil
organic matter was first measured on plots of the Old Rotation in 1988. Since then, measurements have
been repeated in 1992 and 1994. As expected, the long-term treatments have had a dramatic effect on the
buildup or depletion of soil organic matter. This is reflected in the yields. Yields in 1988, 1992, and 1994
are closely correlated with soil organic matter measurements (Fig. 2).
CONCLUSIONS
In its 100th cropping year, the Old Rotation continues to document the long-term effects of crop
rotation and winter legumes on sustainable cotton production in the Deep South. 1995 was a disaster year
for Alabama's cotton crop and especially for cotton and corn on the Old Rotation. Yields were approxi-
mately one-third of the 10-year average. Nevertheless, the long-term yields suggest that winter legumes
are as effective as fertilizer N in producing optimum cotton yields. Yields are also highly correlated with
soil organic matter which reflects the long-term treatments. While crop rotations benefit long-term cotton
yields, the effect is small and probably not economical considering the crop value.
During fall of 1996, the Centennial of the Old Rotation will be celebrated on the campus of Auburn
University.
TABLE 1. YIELDS OF WINTER LEGUMES AND COTTON LINT FROM THE OLD ROTATION, 1995
Winter legumes Cotton lint
Plot Treatment 1995 10-yr. ave.
Lb. dry matter/acre* Lb. lint/acre
1 no legume -- -- 270 340
2 + legume 3,040 3,350 250 850
3 + legume 1,910 3,340 310 920
4 cotton-corn + legume 1,880 3,260 corn 820
5 cotton-corn + legume
+ 120 N 1,880 3,260 corn 850
6 same as #1 -- -- 0 370
7 same as #4 1,520 3,400 350 980
8 + legume 3,520 3,700 280 920
9 same as #5 1,790 3,490 330 1,190
10, 11 & 12 3-yr. rotation 4,460 3,240 470 840
13 no legume/+120N -- -- 210 770
*There were no long-term, significant differences in winter legume dry matter yield.
TABLE 2. YIELDS OF CORN GRAIN, SMALL GRAIN, AND SOYBEAN FROM THE OLD ROTATION, 1995
Corn grain Small grain* Soybean
Plot Treatment 1995 10-yr. ave. 1995 10-yr. ave. 1995 10-yr. ave.
(Bu./acre) (Bu./acre) (Bu./acre)
4 cotton-corn
+ legume 33 70 -- 
-- --
5 cotton-corn
+legume/+120N 27 97 -- 
-- --
7 same as #4 cotton 58** -- 
-- -- --
9 same as #5 cotton 75** -- -- 
--
10,11 &12 3-yr. rotation 45 104 38 27 27 31
* Small grain is rye or wheat.
**A drought in 1986 devastated corn on these plots reducing the 10-year average yield compared to plots 4
and 5 which were planted in cotton that year.
Seed cotton yield, lb./A
4,0001
1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000
Year
Fig. 1. Yield Trends Over 100 Years.
Seed cotton yields from plot #3 (continuous cotton with winter legumes) are examples of
the yield trends over the 100 years of the Old Rotation. Year-to-year yields vary
tremendously; however, rarely does an exceptionally high yielding year follow another
high yielding year. The same is true for low yielding years. While 1994 was one of the
highest yielding years on record, 1995 was one of the worst. The five-year running average
gives an indication of yield trends. Note that the relative decline in years following the
arrival of the boll weevil in Alabma (1915) was not as dramatic as is commonly thought.
Missing data from the 1970's, poor management of the plots that faced an uncertain future
following relocation of the main station research farm, and the loss of the effective yet
controversial insecticide DDT may all have contributed to the yield decline during this era.
5
Yield relative to plot 3 (%)
140 * 1989
A 1990 O O
120 
E 1988
* 1991
100- A 1992 AO 
o 1993
80- N m1994
60 00" m
40- 0
20- m
20
0 0.5 1 1.5 2 2.5 
3
Soil organic matter (%)
Fig. 2. Soil Organic Matter and Yield.
Long-term treatments have resulted in significant differences in soil 
organic matter. These
differences are reflected in soil structure, water holding capacity of the plow layer,
increased soil buffering capacity (e.g. increased cation exchange 
capacity), total
mineralizable nitrogen, etc. Soil organic matter was measured 
in 1988, 1992, and 1994 and
regressed against plot yield relative to plot #3 (continuous cotton and winter legumes).
6