LJ Te y.c~:'j
Progress Report Series* No. 25 July 1946
AGRICULTURAL 
E XPERIMENT 
STATION
of The Alabama Polytechnic Institute, Auburn, Ala.
* * * * * * M.J. Funchs, Director 
* * * *
FURTHER EXPERIMENTS WITH
COTTONSEED MEAL
DETOXIFICATION OF
FOR HOGS
W. E. SEWELL, Animal Husbandman'
D. Mi TURNEY, Associate Animal Husbandman
IN a previous study (3), free gossypol was
shown to be the primary' cause of the
poisonous effect of cottonseed meal when
fed to hogs in sufficient quantities to con,
stitute the major source of protein in the
ration.
Commercial cottonseed meals produced by
16 mills in Alabama were found to contain
free gossypol in amounts ranging from 0.055
to 0.252 per cent. When fed at a level of
25 per cent of the ration, the meal contain-
ing the least amount of free gossypol proved
toxic. Under the conditions of these experi-
ments, the application of moisure and heat
to cottonseed meats or meal reduced their
free-gossypol content. The reduction was in
proportion to the degree of heat, length of
time heat was applied, and moisture content
of the meal.
The data reported in this publication con-
cerns the application of the principles noted,
as well as other modifications in commercial
milling practices. Possible variations in the
process to reduce cottonseed meal toxicity
were studied in the followirig respects:
1) Denaturization or elimination of free
gossypol through application 
of mois-
ture and. heat.
2) Development 
of a method for 
treating
cottonseed meal with iron to counter-
act its toxicity.
3) Effect of solvent extraction of cot-
tonseed upon free-gossypol content and
toxicity of the resulting meal.
EXPERIMENTAL PROCEDURE
Preparation of test materials.
2 
Cottonseed meal
was prepared by the batch process, using
the maximum amount of moisture that could
be employed without reducing the mill out-
put enough to make the procedure uneconom-
ical. In addition, cottonseed meal pellets,
which receive treatment with moisture and
heat in the pelleting process, were prepared
from the batch-cooked meal and examined
for free-gossypol content and toxicity.
Although iron solutions have been known
to counteract the toxicity of cottonseed meal,
the use of iron salts in the dry meal for hogs
has not always proved favorable (2). In
this investigation, iron-containing pellets
were prepared by adding ferrous sulfate to
the meal before it entered the pelleting ma-
chine, 
in order 
to determine 
if the 
moisture
resulting from the steam spray used in pel-
leting might be sufficient to satisfy the mois-
ture requirement of the iron-detoxification
process.
The cottonseed meal resulting from solvent
extraction of cottonseed meats for oil was
produced with a method formulated by the
cooperating oil company. The meal was made
in a pilot plant by the extraction of unheated
cottonseed kernel flakes, which were then
ground and adjusted to a 41 per cent protein
basis by adding hull bran.
Free gossypoL and toxicity tests. The analytical
method used in determining the free-gossypol
content of the various mill products was
that given by Smith (4).
The toxicity evaluations were arrived at
by feeding hogs on a ration made by mixing
the following materials: Yellow corn meal 73
per cent, alfalfa leaf meal 5 per cent, min-
erals 2 per cent, cottonseed meal or pellets
*Furmerly Mimeographed Series.
1
Resigned.
2The Process Development Staff of the Buckeye
Cotton Oil Company cooperated in the milling
phases of the investigation.
(41 per cent protein) 20 per cent. Minerals
were also supplied ad libitum.
The several tests were begun with pigs
averaging approximately 34 pounds in
weight. The tests were continued until the
animals receiving the less toxic ration aver-
aged above 200 pounds. Each group of hogs
was penned in a dry lot of approximately one-
eighth acre in size and given all of the feed
the animals would eat in two feedings daily.
RESULTS
In the course of the investigation, three
series of products were made, analyzed, and
fed in successive periods. The results are
given in Table 1.
Series I. The free-gossypol content of the
batch-cooked meal, which received maximum
moisture treatment in the cooking process,
was lower than that of the regular process
meal. Likewise, the pellets contained slightly
less free gossypol than the meals from which
they were derived. However, at least one hog
died or became sick with symptoms of cotton-
seed meal poisoning in every lot except that
receiving the pellets fortified with iron.
The mill operators stated that the batch-
cooked meal was prepared with 12.7 per cent
moisture in the second kettle, and that the
time required to cook out additional moisture
slowed production to the point of being un-
economical. It appears, therefore, that the
manufacture of non-toxic meals for hogs by
hydraulic installations will require changes
in the machinery to effect more rapid re-
moval of excess moisture.
Series 2. During the first week of the sec-
ond trial, hogs in the experiment developed
a disease similar to erysipelas. Other pigs
in the breeding herd from which the experi-
mental animals were selected exhibited simi-
lar symptoms, indicating that the trouble
was not due to the test feeds. One or more
animals in each of the lots became ill, al-
though some of the groups were affected
more than others. As a result, the rate and
economy of gains were not normal. However,
the hogs were continued on experiment for
observation of any toxic effects that might
develop. One animal receiving the batch-
cooked meal died. Post-mortem examination
showed both ascites and hydrothorax, which
TABLE 1. FREE GOSSYPOL CONTENT AND TOXICITY OF COTTONSEED MEAL AND PELLETS MILLED 
IN
VARIOUS WAYS
Feed
PRODUCT Free Pis in Toxic Daily gain required
icontent of effects per pig 
100 pounds
Scontent
1  
of grain
Per cent Number Pounds Pounds
Series One
Continuous cooked meal, regular process.................0.074 10 Deaths - -
Pellets from the above containing 2% FeSO
4-
........ 00582 10 None 1.18 450.6
Batch-cooked meal, maximum moisture...................0.066 10 Sickness - -
Pellets from the above ..................... ..................... 0.049 10 Deaths -
Series Two
Batch-cooked meal, regular process, pelleted...........0.072 7 Deaths - -
Undercooked meal, pelleted with 2% FeSO
4-
.......... 0.1122 7 None 0.84 424.5
Batch-cooked meal, pelleted with 2% FeSO
4
......... 0.0722 7 None 0.92 393.1
Solvent extracted meal..........................................0.036 7 None 0.94 403.1
Series Three
Undercooked meal, pelleted with 1% FeSO
4
........... 0.1292 8 None 0.91 448.0
Undercooked meal, pelleted with 2% FeS
4 
......... 0.1292 8 None 0.81 470.9
Batch-cooked meal, pelleted with 2% FeSO
4-
......... 0.0622 8 None 0.82 469.0
Solvent extracted meal pelleted
3
...................... 0.010 8 None 1.31 408.6
Solvent extracted cottonseed meal and
soybean meal, 50% each, pelleted
3
....................... 0.005 7 None 1.37 397.9
'Dry weight basis.
2
Analysis made prior to treatment with 
FeSO4.
BAll of the pelleted materials except these two were ground
[2]
before feeding.
are conditions typical of cottonseed meal
noisoning and not characteristic of erysipe-
las. The iron-treated materials appeared to
be unpalatable. but none .of the hogs receiv-
ing these products or the extracted meal
developed disorders that could be attributed
to the toxic effect of cottonseed meal. The
free-gossypol content of the extracted meal
was considerably lower than any commercial
meal that had been analyzed previously.
Series 3. Despite precautions, the hogs used
for testing the third series of products
showed symptoms similar to those observed
in the previous trial. The symptoms appeared
during the first 10 days of the experiment,
but as before they were also manifest in the
breeding herd. The disease was milder than
in the second test, and the subsequent growth
of the animals was reasonably satisfactory.
However, the rate and economy of gain are
probably questionable as indexes of com-
parison.
None of the products tested in the third
trial gave any evidence of toxicity. The pel-
lets containing iron proved to be unpalatable
when fed whole for several days. There-
after, the pellets were ground and mixed with
the other ingredients of the ration. How-
ever, reaction to the ground material appears
to have contributed to lowered total feed con-
sumption, with consequent reduction in econ-
omy and rate of gain.
The extracted meal contained less free
gossypol than that fed in the second experi-
ment. These pellets were fed whole. The
hogs relished the pellets to the extent of
consuming the pellets before eating the 
re-
mainder of the ration. The daily gain per
pig, 1.31 pounds, and the feed required 
per
one-hundred weight of gain, 407.0 
pounds,
were quite satisfactory for a ration composed
largely of corn and supplemented 
with plant
protein, in the main, from a single 
source.
The soybean-cottonseed meal mixture 
pel-
leted and fed one of the lots in 
the third
series contained too little cottonseed meal to
be toxic, even if hydraulic meal were used.
It would be expected to be superior to cotton-
seed meal in promoting growth, but, due to
the disease troubles, the results probably are
not reliable for comparison with the solvent
meal.
[31
DISCUSSION
The data presented in this report confirm
the principle that the free-gossypol content
of cottonseed meal produced by the hydraulic
process may be reduced through the use 
of
more moisture and heat than is commonly
employed by oil mills. However, it appears
that production of a meal that is 
non-toxic
to, hogs by this procedure will require
changes in the milling machinery to remove
the excess moisture rapidly enough for the
process to be economically feasible.
The rendering of hydraulic meals non-toxic
to hogs by incorporating iron into pellets
made from the meal appears to have possibili-
ties. Such pellets proved non-toxic, 
even
though prepared from undercooked meats
containing relatively large amounts of free
gossypol. While these products were un-
palatable, additional work may show that
poisonless pellets can be manufactured 
with
amounts of iron more acceptable 
to animal
taste.
The solvent-extracted meal studied in
these trials appears to be entirely satisfac-
tory for feeding to hogs at a level of 20 per
cent of the total feed. Such a proportion is
sufficient to supply the protein needed in a
ration composed largely of corn, This pro-
duct contains considerably less free gossypol
than meals being sold at present. The manu-
facturers state that the pilot plant is now
producing meal that consistently contains 
ap-
proximately 0.01 per cent free gossypol, and
that commercial plants will be able to 
dupli-
cate this performance.
Although the two trials with hogs fed
solvent-extracted meal gave favorable re-
sults:, additional tests under a variety of con-
ditions are desirable. It should be noted also
that rigid control of the free-gossypol con-
tent of the meal is necessary if it is to h e
manufactured and sold as a non-toxic prod-
uct. Moreover, cottonseed meals, which may
be produced by other solvent-extraction
methods, will require examination of their
toxic properties before they can be recom-
mended for feeding to hogs in large amounts,
since different adaptations of the solvent
process may alter the nature of the resulting
meals.
] brar~ Q ~~
State Colege 
Mss.
SUMMARY
Production of special cottonseed meals and
cottonseed meal products in mill operations
and feeding of these materials in experiments
with hogs showed the following:
1) A cottonseed meal containing one-third
to one-fifth of the minimum amount of free
gossypol found in hydraulic meals proved
non-toxic to hogs in two experiments, when
fed as 20 per cent of the ration. This meal,
produced by solvent extraction on a pilot
plant scale, contained 0.01 per cent free
gossypol.
2) A non-toxic meal for hogs could not be
produced by the ordinary hydraulic process
due to the failure of the standard equipment
to drive off the excess moisture required for
detoxification rapidly enough to make the
procedure commercially feasible. Changes 
in
mill machinery may correct this 
difficulty.
3) Cottonseed meal pellets were found to
contain less free gossypol than the meal
from which they were made, but the differ-
ence was not large enough 
to be of impor-
tance in the feeding of 
hogs.
4) Iron (ferrous sulfate) incorporated
into cottonseed meal pellets served to reduce
their toxicity, but the amounts of the 
iron
salt used appeared to be 
unpalatable to hogs.
LITERATURE
(1) OLCOTT, H. S. The nutritive value 
of cotton-
seed meal. Jour. Biol. Chem. 133: Proc. 
Am.
Soc. Biol. Chemists. P. LXXII. 1940.
(2) RoBINsoN, W. L. Cottonseed meal for pigs.
Ohio Agr. Expt. Sta. Bul. 534. 1934.
(3) SEWELL, W. E. The detoxification of cotton-
seed meal for hogs. Ala. Agr. Expt. Sta. Bul.
259. 1943.
(4) SMITH, F. H. Revised method for the estima-
tion of gossypol in cottonseed meal. Indus.
and Engin. Chem., Anal. Ed. 9:517-518.
1937.