FINAL REPORT
Midwest Forage Association
THE EFFECT OF MOISTURE AND BALE WRAPPING ON TEMPERATURE AND
FORAGE QUALITY IN HORSE QUALITY GRASS HAY
Krishona Martinson and Craig Sheaffer
University of Minnesota
INTRODUCTION
Since 2000, there have been over 900 livestock and poultry barn fires in Minnesota, resulting in
over 26 million dollars in damages (Minnesota Fire Incident Reporting System - MFIRS).
Although not specifically tracked by MFIRS, a number of these fires have been caused by
spontaneous combustion of hay that was baled too wet. Moisture level at time of baling is a
significant factor leading to the possibility of hay fires. Research has studied the relationship
between forage moisture levels and the resulting bale temperature in small and large square bale
dairy quality hay, but not in horse quality grass hay. Most guidelines currently used in the
equine community are extracted from dairy related research results. However, horses are highly
sensitive to mold. Guidelines for hay moisture at time of bailing established for cattle may be
too high for mold sensitive horses. Ingesting moldy hay can result in both short-term and longterm respiratory problems, specifically heaves and colic in horses. Moldy hay also can cause
human respiratory distress in addition to labor and financial burdens related to replacing and
disposing of the moldy hay.
Previous research suggests that forage needs to be dried to less than 17% moisture for quality
horse hay. In Sweden, researchers have begun wrapping round bales in plastic for horse use at
approximately 35% moisture. Wrapping bales at this moisture level results in minimal
fermentation and has been preliminarily shown to be safe for horse feed in Sweden. Even
though bale wrapping is an accepted means of forage processing for cattle, it has not been
accepted or researched extensively in the horse community in the United States. Bale wrapping
may be a strategy to provide high quality horse hay.
The objectives of this research were to:
1. Determine the relationship between moisture at time of baling and temperature of round
baled horse quality grass hay.
2. Determine the relationship between moisture at time of baling and forage quality,
including the presence and identification of mold.
3. Determine if forage quality of wrapped hay is equivalent to the quality of unwrapped hay
and if wrapped hay can be safely fed to horses.
1
MATERIALS AND METHODS
The experimental design was a randomized complete block with four replications. Treatments
included the following moisture ranges; 17% or less, 20-25% 30-35%, and 30-35% wrapped.
Flowering, first crop orchardgrass hay was cut in the morning of July 8, 2008. Later that same
day, four 4 x 5 net wrapped round bales were baled at 20-25% moisture, and four at 30-35%
moisture (wet basis). An additional four 4 x 5 round bales were baled at 30-35% moisture
and wrapped in plastic immediately after baling. On July 9, 2008, four 4’ x 5’ net wrapped
round bales were baled at 17% moisture or less. Moisture content was estimated by using a
hand-held moisture sensor. Immediately after baling, each bale was cored five times, and the
samples were analyzed for a basic equine forage nutrient composition by a commercial forage
testing lab.
After baling and sampling, three temperature sensors (HOBO onset pendant temp) were placed
in each bale at approximately 15, 30 and 45” from the top of the bale, and approximately 24”
into the bale. The temperature sensors recorded temperature every hour for 50 days or 7 weeks
(experiment ended on August 27, 2008). After 50 days (7 weeks), the sensors were removed and
the temperate data was downloaded. Five additional cores were taken on each bale. The
samples were analyzed for a basic equine forage analysis, and mold count and identification
using the same commercial lab.
The bales were stored on a well-drained surface in a row running north and south with three
bales of each moisture type tightly butted against each other. A check, or non-data round bale,
was placed on both ends of the row and in between each group of bales. The fourth bale of each
moisture type was set apart from the row, and served as a check.
RESULTS AND DISCUSSION
MOISTURE AND FORAGE QUALITY: Table 1 outlines the target moisture, actual moisture, and hay
quality characteristics of baled orchardgrass hay. Actual moisture of the baled hay generally fell
within the target moisture. Most bales were about 18% moisture at the conclusion of the study
with the exception of the wrapped bales, which remained at about 35% moisture. The
unwrapped bales harvested at 20-25 and 30-35% moisture loss water during storage whereas the
moisture content of bales stored at 18% moisture did not change.
The protein content and Equine Digestible Energy (Equine DE) between the bales or at the
initiation and conclusion of the experiment were similar. There were differences between bales
and/or at the initiation and conclusion of the experiment for Acid Detergent Fiber (ADF),
Neutral Detergent Fiber (NDF), Relative Feed Value (RFV), and Non-Fiber Carbohydrates
(NFC). Generally speaking, ADF and NDF increased with time, with the exception of the
wrapped hay where both ADF and NDF decreased. RFV decreased with time, with the
exception of the wrapped hay where the RFV increased. NFC remained similar during the
experiment, except for the 30 to 35% moisture hay, where NFC concentration decreased.
2
Within the different moisture ranges, hay quality remained similar over time for the 30-35%
moisture wrapped bales. Some difference in quality over time where seen in the 17% or less
moisture bales, and the 20-25% moisture bales. The most changes in hay quality over time were
recorded with the 30-35% moisture bales.
Table 1. Target and Actual Moisture, and Hay Quality of Orchardgrass Hay Baled at Three
Different Moisture Types and Wrapped
Actual
Actual
Equine
Target
Moisture Moisture Protein ADF* NDF* RFV* NFC*
Date
DE*
Moisture
(average)
Range
2008
%
%
%
(Mcal/lb)
9-Jul
17
15 to 19
8
41
64
83
17
0.89
17 or less
28-Aug
19
17 to 20
7
42
66
79
17
0.88
8-Jul
23
21 to 24
7
40
64
84
17
0.90
20-25
27-Aug
18
17 to 18
8
42
66
78
16
0.90
8-Jul
27-Aug
8-Jul
30-35
wrapped
27-Aug
LSD (p= 0.05)
30-35
30
18
34
35
1
27 to 31
15 to 20
31 to 36
34 to 35
8
8
8
8
NS
39
42
40
39
1
63
65
64
63
1
87
80
84
86
1
20
16
18
18
1
0.94
0.88
0.92
0.93
NS
*ADF (Acid Detergent Fiber); NDF (Neutral Detergent Fiber); RFV (Relative Feed Value); NFC (Non-Fiber
Carbohydrates); and Equine DE (Equine Digestible Energy).
MOLD AND YEAST POPULATIONS: Table 2 shows mold populations, and yeast counts of
orchardgrass hay baled at the three different moisture types and wrapped. There was a
tremendous amount of variability for mold count and identification, and yeast counts between
and within the different moisture ranges. Due to this variability, we could not detect differences
between the moisture treatments. However, there is a trend of less mold at the higher
temperature ranges in the unwrapped hay. Perhaps the high temperatures discouraged or limited
mold growth.
It is recommended that hay containing more than 10,000,000 cfu/g of mold should not be fed to
livestock (Stearns DHIA). Caution is advised when mold counts are between 10,000 and
10,000,000 cfu/g. Based on these guidelines, only the 30-35% moisture wrapped bales could be
cautiously fed.
Specifically, hay containing Aspergillus flavus and parasiticus can produce aflatoxin, which is
highly toxic to both livestock and humans. All moisture ranges contained some amount of
Aspergillus flavus. Fusarium are toxic to horses and can produce numerous toxic compounds,
including vomitoxin. Aspergillus species, Penicillium species, Cladosporium, and Mucor are
considered common molds and are not likely to produce toxic compounds (Table 2).
3
Yeast can also have adverse affect on horse health and hay containing more than 1,000,000 cfu/g
of yeast should not be fed to livestock (Stearns DHIA). Using these guidelines,, the 20-25% and
30-35% moisture bales could be fed. The 17% moisture or less and 30-35% wrapped hay could
be fed, but yeast populations were near 1,000,000 cuf/g threshold. (Table 2).
Table 2. Mold and Yeast Populations of Orchardgrass Hay Baled at Three Different Moisture
Types and Wrapped
Moisture at Moisture at
Total Mold
Yeast
Target
Initiation* Conclusion*
Identified
Moisture
Populations
Population
(average)
(average)
Molds
%
(cfu/g)
(cfu/g)
Aspergillus flavus
Aspergillus parasiticus
Aspergillus species
17 or less
16.58
18.66
24,800,000
567,750
Fusarium
Penicillium species
Cladosporium
Aspergillus flavus
20 to 25
23.23
18.01
19,700,000
127,500
Cladosporium
Aspergillus flavus
Aspergillus species
30 to 35
27.78
18.16
5,550,000
24,000
Cladosporium
Mucor
Aspergillus flavor
30 to 35
33.75
35.11
85,689
Aspergillus species
669,000
wrapped
Mucor
LSD (p = 0.05)
NS
NS
*Experiment was initiated on July 8th and 9th and concluded on August 27 and 28th
TEMPERATURE: Figure 1 depicts the temperature changes of orchardgrass hay baled at three
different moisture types and wrapped over time (50 days). Moisture range, bale replication,
temperature sensor placement, and time were all significant (p value = 0.05) when evaluating
temperature.
Although all replicate bales within a specific temperate range were statistically different, the
pattern of temperature (or heating) over time was similar. The 17% or less moisture bales
reached the high temperature of 128oF at 7 days after baling. All bales had an almost immediate
reduction in temperature, followed by a steady rise in temperature before leveling off around
80oF at 28 days. For bales 1 through 3, the bottom temperature sensor (45”) appeared to have a
lower temperature, compared to the sensors above. Bale 4 (check) temperatures at the three
different locations within the bale appear to be more consistent compared to the other bales.
The 20-25% moisture bales reached the high temperature of 145oF at 11 days after baling. Most
bales had a steady rise in temperature before leveling off around 120oF at 21 days. For bales 1
through 3, the bottom temperature sensor (45”) appeared to have a lower temperature, compared
4
to the sensors above. Bale 4 (check) temperatures at the three different locations within the bale
appear to be more consistent compared to the other bales.
The 30-35% moisture bales reached the high temperature of 188oF at 11 days after baling. Most
bales had a steady rise in temperature before leveling off around 120oF at 28 days. For bales 1
and 2, the bottom temperature sensor (45”) appeared to have a lower temperature, compared to
the sensors above. Bale 4 (check) had a lower maximum temperature, and temperatures at the
three different locations within the bale appear to be more consistent compared to the other bales.
The 30-35% moisture wrapped bales reached the high temperature of 97oF at 1 day after baling.
Most bales had an initial rise in temperature before leveling off between 70 and 80oF at 14 days.
All bales and temperature sensor locations results in similar temperatures over time. Due to the
wrapping process, it was not possible to ensure the correct placement of the top and bottom
temperature sensor.
According to previous research (Gay et al., 2003), temperature less than 130oF will feel warm to
the touch but not hot (result of the natural curing process) and have a minimal fire risk. The 17%
moisture and less bales, and the 30-35% moisture wrapped bales did not reach 130oF, but still
had a significant amount of mold (Table 2). Temperatures over 150oF are likely to continue to
rise and pose a significant fire threat. The 20-25% moisture bales did not reach 150oF. The 3035% moisture bales did exceed 150oF, but did not combust, likely due to the outside storage
conditions.
CONCLUSION
Moisture at time of baling orchardgrass round bales did have a significant effect on bale
temperature over time. The 17% or less moisture bales reached the high temperature of 128oF at
7 days after baling; the 20-25% moisture bales reached the high temperature of 145oF at 11 days
after baling; the 30-35% moisture bales reached the high temperature of 188oF at 11 days after
baling; and the 30-35% moisture wrapped bales reached the high temperature of 97oF at 1 day
after baling. All bales appeared to reach a more consistent temperature between 14 and 28 days
after baling.
Within the different moisture ranges, hay quality remained similar over time for the 30-35%
moisture wrapped bales. Some difference in quality over time where seen in the 17% or less
moisture bales, and the 20-25% moisture bales. The most changes in hay quality over time were
recorded with the 30-35% moisture bales.
There is a trend of less mold at the higher temperature ranges in the unwrapped hay. Due to
mold and yeast populations, only the 30-35% moisture wrapped bales can be considered horse
quality hay. The wrapped had a positive effect on maintain forage quality and reducing
temperate and mold compared to the unwrapped hay.
Additional data is needed, including establishing a moisture threshold that results in a minimal
amount of mold and yeast.
5
Figure 1. Temperature of Orchardgrass Hay Baled at Three Different Moisture Types and
Wrapped over Time*
17% or Less Moisture Bale Temperature over Time
140
Temperature (F)
130
120
110
100
90
80
70
60
0
7
14
21
28
35
42
49
Days
Bale 1 15"
Bale 2 15"
Bale 3 15"
Bale 4 15"
Bale 1 30"
Bale 2 30"
Bale 3 30"
Bale 4 30"
Bale 1 45"
Bale 2 45"
Bale 3 45"
Bale 4 45"
30‐35% Moisture Wrapped Bale Temperature over Time
190
180
170
160
150
140
130
120
110
100
90
80
70
60
100
Temperature (F)
Temperature (F)
30‐35% Moisture Bale Temperature over Time
90
80
70
60
0
7
14
21
28
35
42
49
0
7
14
21
Days
28
35
Days
Bale 1 15"
Bale 2 15"
Bale 3 15"
Bale 4 15"
Bale 1 15"
Bale 2 15"
Bale 3 15"
Bale 4 15"
Bale 1 30"
Bale 2 30"
Bale 3 30"
Bale 4 30"
Bale 1 30"
Bale 2 30"
Bale 3 30"
Bale 4 30"
Bale 1 45"
Bale 2 45"
Bale 3 45"
Bale 4 45"
Bale 1 45"
Bale 2 45"
Bale 3 45"
Bale 4 45"
*The circle data points represent temperature sensors placed at 15”, the square data points represent temperature
sensors placed at 30” and the diamonds represent temperature sensors placed at 45”. Bale 4 is the check bale, or the
bale that was stored by itself, compared to the other bales that were stored together in a row.
REFERENCES
Gay, S.W., Grisso, R., Smith, R., and Swisher, J.M. 2003. Hay Fire Prevention and Control.
Virginia Cooperative Extension. Publication 442-105.
Stearns DHIA. 825 12th Street South, Sauk Center, MN 56378. http://www.stearnsdhialab.com
6
42
49