Làm giàu protein củ sắn bằng cách lên men với nấm men làm thức ăn cho lợn địa phương ở lào tt tiếng anh

HUE UNIVERSITY
UNIVERSITY OF AGRICULTURE AND FORESTRY

NOUPHONE MANIVANH

NUTRITIVE IMPROVEMENT OF CASSAVA ROOT AND
ITS UTILISATION IN TARO FOLIAGE AND BANANA
STEMS BASAL DIETS FOR LOCAL PIG PRODUCTION
IN SMALLHOLDERS IN LAO PDR

DOCTOR OF PHILOSOPHY IN ANIMAL SCIENCES

HUE, 2019
1


HUE UNIVERSITY
UNIVERSITY OF AGRICULTURE AND FORESTRY
NOUPHONE MANIVANH

NUTRITIVE IMPROVEMENT OF CASSAVA ROOT AND
ITS UTILISATION IN TARO FOLIAGE AND BANANA
STEMS BASAL DIETS FOR LOCAL PIG PRODUCTION
IN SMALLHOLDERS IN LAO PDR

SPECIALIZATION: ANIMAL SCIENCES
CODE: 9620105
DOCTOR OF PHILOSOPHY IN ANIMAL SCIENCES

SUPERVISORS
1: ASSOCIATE PROFESSOR DR. LE VAN AN

2: ASSOCIATE PROFESSOR DR. TRAN THI THU HONG

HUE, 2019
2


1

INTRODUCTION
1. PROBLEM STATEMENT
Pig is one of the most important animals for smallholders in the uplands of Lao
PDR because it can be sold when cash is needed for buying rice and other food, for
paying school fees or if a household member is sick and needs medical attention and
Pork used in traditional ceremonies in households. Pigs can be confined in a small area,
and can covert to meet a variety of crop and kitchen wastes and give a rapid return on
investment (Steinfeld, 1998). About 75% of households in upland areas are raising pig
in the country (FAO, 2017). Overall, native pig around 85.1% under small holder
system (DLF, 2017), they are hardy and able to scavenge at least part of their feed
requirements in free-range condition, Native pigs are mainly raised in extensive lowinput systems that take advantage of naturally occurring feed (Kennard, 1996; FLSP,
2002). In most parts of Laos, agricultural by-products, such as rice bran, and natural
grasses are the main feeds for live stock (ILRI 2002). In Lao villages, where most
farmers are growing paddy rice for sale, the feed for pigs is based on rice bran, which is
fed together with a small amount of green feed. Thus rice bran is available in most farm
households but they cannot support full performance because of their poor nutritive
value. (ILRI, 2002; FLSP, 2002). Since feed accounts for about 50-60% of the variable
costs of production, feed quality is crucial to the success of pig farming operations.
Major problems that may result from low quality feeds are poor appetite, slow growth,
high feed conversion ratio, and low survival. These usually develop as a result of
problems on quality of raw materials, feed formulation, processing technology, storage,
and feed manage. The main problem is the supply of protein as soybean and fish meals

are not available in rural areas and expensive (Phengsavanh and Stür., 2006).
Cassava plantation is mainly for root production. The yields of root are variable
depending on soil fertility, management and irrigation system. Cassava root yields can
be from 10 to15 tonne/ha without inputs on eroded soils (Howeler, 1991). In Laos,
cassava (Manihot esculenta Crantz) known as ‘Man Ton’, it is currently the third most
important crop in Laos, after rice and maize for smallholder farmers in remote upland
areas. Recently, the crop has become an important cash crop for either domestic use or
for export because it can be used for food and feed as well as for industrial processing
into starch (Ministry of Agriculture and Forestry, 2013). Cassava has become a major
crop in Lao PDR mainly because of the export of starch that is extracted from the
cassava root. There are five cassava starch factories with a total planted area of 60,475
ha, giving an average yield of fresh roots of 27 tonnes/ha. Annual production is of the
order of 1.6 million tonnes (Ministry of Agriculture and Forestry, 2013). Cassava farms
are needed not only for a major source of income for rural households but also for use
in pig diets as energy sources because of cassava root content high levels of energy (75
to 85% of soluble carbohydrate) but low crude protein (2 to 3% CP). The root is
composed of highly digestible carbohydrate in the form of starch with little fiber (Kang
et al., 2015; Polyorach et al., 2013). Solid state fermentation of the cassava root is a
promising technology as this has the potential to raise the protein content to levels
required to balance the carbohydrate thus presenting the opportunity to make an almost
1


2

complete feed for pigs (Boonnop et al., 2009). Sengxayalth and Preston, (2017a)
reported an increase in true protein from 2 to 12% in dry matter (DM) of the cassava
pulp. Agreement with Vanhnasin et al., (2016a) true protein increased from 2 to 7% in
dry matter (DM) of the cassava root. Similar findings were reported by Balagopalan et
al., (1988) who developed a solid state fermentation process for the protein enrichment

of
cassava
flour
and cassava
starch
factory
wastes
using the
fungus Trichoderma pseudokonigii rifai. Fermentation with yeast, bacteria has been
studied for reducing non-nutritional components, increasing the nutritive value of agroindustrial by-products (Okpako et al 2008; Aderemi et al 2007; Tran Thi Thu Hong and
Nguyen Van Ca 2013). Additional phosphate results in increased biomass growth of
yeast and bacteria (Papagianni et al 1999). Huu and Khammeng, (2014) reported that
when replacing maize with fermented cassava pulp containing 13% crude protein (DM
basis), digestibility and N retention were similar to the control diet. Protein enriched of
cassava root (PECR) could provide in pig diets up to 25 to 28% of the dietary protein in
a diet based on cassava pulp (or ensiled root), replacing ensiled taro foliage (Vanhnsin
and Preston, 2016b) or soybean meal (Sengxayalth and Preston, 2017b). It similar to
the growth response in pigs reported by Phuong et al., (2013) for cassava pulp enriched
from 3 to 5.5% true protein using the fungus Aspergillus niger.
The local feed used in smallholder systems for pigs include rice by-products,
planted feeds and various green plant materials (ILRI 2002). However, the local feed
contain low nutritive value. Women typically are the key persons in this effort, and,
with traditional practices, they spend 2 to 3 hours each day collecting and preparing
feed for pigs (Australian Center for International Agricultural Research 2010). Farmers
have little knowledge on optimizing use existing feed resources, the growth rate of the
pig only 100 to 120 g/day if depend on local feed staffs. In commercial complete feeds,
the most common protein sources are fish meal and soybean meal. These feedstuffs
provide high quality protein for pigs, but they are imported and are expensive. Due to
their high price, such protein sources cannot be used by smallholder farmers
(Phengsavanh et al., 2010). So, improving nutritive value of local feed that is

abundance in their area especially the application of microorganism fermentation it is
possible to improve the nutritive value of local feed and its utilization as diets for local
pigs in Laos, which helps in reducing feed cost and bringing economic benefits to the
farmers in rural area.

2. OBJECTIVES
The overall aim of this thesis was to improve nutritive value of cassava roots by
fermentation yeast (Saccharomyces cerevisiae), Urea and di-ammonium phosphate
additive and its utilization as protein source in the diets of Moo Lath pigs. Specific
objectives were to:


To study nutritive value of casssava root by fermentation yeast (Saccharomyces
cerevisiae), Urea and Di-ammonium phosphate additive
2


3




To study the limiting factor to the synthesis of true protein from crude protein
in the fermentation of cassava root
To evaluate the use of protein-enriched cassava root as partial replacement of
taro silage in a ensiled banana stem - based diet fed to Moo Lath pigs

CHAPTER 1:
LITERATURE REVIEW
There are main points following (i) Pig production in Laos; (ii) requirement of

protein and amino acid for growing pigs (iii) feed stuffs for pig in Laos; (iv) method to
improve value for feed stuff with low protein content; and (v) utilization of forage
based diet for pigs.

CHAPTER 2
IMPROVING NUTRITIVE VALUE OF CASSAVA ROOTS
(Manihot esculenta Crantz)
INTRODUCTION
The major problems of small-holder pig production in upland areas of Lao PDR
are high piglet mortality and low growth rates. Almost all pigs are of local breed (Mou
Lath), managed in scavenging systems and suffer feed inadequacy in both quality and
quantity. According to the survey by Phonepaseuth et al., (2010) most piglets in upland
areas had a low growth rate (20-50 g/day) and high mortality (30-50%). Weaned pigs
required from 5 to 8 months to reach live weights of 20 to 30 kg.
The cassava root is composed of carbohydrates and is therefore mainly a source
of energy. The starch content varies between 32 to 35% of the mass of fresh root and 80
and 90% of the mass of dried roots (Montagnac et al., 2009). The protein content is
trivial, between 1 and 3% of dry matter (Buitrago, 1990).
One way to improve the protein content of carbohydrate-rich feeds is by solidstate fermentation with fungi and yeasts (Araujo et al., 2008; Hong and Ca, 2013). The
fermentation of cassava meal with S. cerevisiae enhanced the protein level from 4.4%
to 10.9% in DM and decreased the cyanide content (Oboh and Kindahunsi, 2005).
Solid state fermentation of the root with urea and di-ammonium phosphate (DAP) is a
promising technology as this has the potential to raise the protein content to levels
required to balance the carbohydrate, thus presenting the opportunity to make an almost
complete feed for monogastric animals such as pigs and poultry (Boonnop et al., 2009).
The problem, in the studies reported so far, is that not all the added nitrogenous
compounds (urea and DAP) were converted to “true” protein, the levels of which never
exceeded some 50 to 70% of the “crude “ protein in experiments with yeast-fermented
cassava root (Vanhnasin and Preston, 2016a) and cassava root pulp (Sengxayalth et al.,
2017a). Yeast cannot directly use urea which must first be hydrolysed to ammonia by

3


4

urease. However, the activity of urease is inhibited at low pH (Kay and Reid, 1934),
which falls rapidly when the cassava root is fermented.
EXPERIMENT 1:
MATERIALS AND METHODS
Location
The experiment was carried out in the Laboratory of the Animal Science
Department in the Faculty of Agriculture and Forest Resource in Souphanouvong
University. The site is located 7 km from Luang Prabang City, Lao PDR. The mean
daily temperature in this area at the time of the experiment was 27oC (range 22-32°C).
Experimental design
The experiment was arranged as a 2*3*4 factorial in a completely randomized
design (CRD) with 4 replications in each period.
The treatments were:
Root processing
Steamed (ST) and not steamed (NST)
Di-ammonium phosphate DAP:
0, 1 or 2% of root DM
Procedure:
Time (days)
Day 0, 3, 7 and 14
Steaming
Cassava roots were peeled and chopped by hand into small pieces (1-2 cm).
One portion was steamed for 30 minutes in a bamboo basket placed above a pan
containing boiling water. The composition of the substrates were increased the level of of
DAP zero to 2% and balance N in each treatment by urea with constant of Yeast as 3% for

testing the nutritive value of cassava root after fermentation. The steamed cassava root was
removed from the bamboo basket allowed to cool for 15 minutes. The steamed and un-steamed
cassava root were then mixed with urea, yeast (Saccharomyces cerevisiae) and DAP (table 1).
The proportions of urea were varied according to the level of DAP so that the substrates were
iso-nitrogenous. The mixed substrates were then transferred to bamboo baskets covered with
plastic netting to allow free entrance of air (photo 2) and allowed to ferment for 14 days.

Table 1. Composition of the substrates (DM basis)
Treatm
ent
DAP-0
DAP-1

Cassava r
oot, %
95
94.3

Yeast, %

DAP#, %

Urea, %

3
3

0
1

2
1.7

4


5
DAP-2

93.6

3

2

1.4

#Phosphorus 20%

The species of yeast is saccharomyces cerevisiae was used in the experiment. S.
cerevisiae cells are round to ovoid, 5-10 μm in diameter. It reproduces by a division
process known as budding (Feldmann and Horst, 2010). Di-ammonium phosphate
(DAP) contains 16% N and 20% phosphorus (P); Urea has 46% N (DM basis); yeast
(S. cerevisiae) contains 48.6% CP in DM.
Measurements
On days 0, 3, 7 and 14 samples were taken from the treatment. There are four
replicates in each period of the treatments (the sample did not repeated measurements
in each period) and the sample were analyzed for DM, N, OM and true protein. The
fresh weight of the substrates in each treatment were weighed at each time interval to
determine the relative amounts of substrate DM utilized in the fermentation process.

Chemical analysis
DM, N and ash were analyzed according to AOAC, (1990) methods. For
estimation of true protein, 2 g of the fresh sample were put in a 125ml Erlenmeyer flask
with 50 ml of distilled water, allowed to stand for 30 minute, after which 10ml of 10%
TCA (trichloracetic acid ) were added and allowed to stand for a further 20-30 minutes.
The suspension was then filtered through Whatman #4 paper by gravity. The filtrate
was discarded and the remaining filter paper and suspended substrate transferred to a
kjeldahl flask for standard estimation of total N. The measurements of crude and true
protein were done on the fresh sample.
Statistical analysis
The data were analyzed by the General Linear Model (GLM) option in the
ANOVA program of the Minitab, (2010) software (version 16.0). In the model the
sources of variation were treatments, treatment interaction and random error. Turkey’s
pair-wise comparison was used to determine the differences. The statistical models used
were:
Yijk = µ +ci +dj + tk + (c*d*t)ijk + eijk
Yijk are dependent variables; µ is overall mean; ci is effect of cassava root processing
dj is effect level of DAP; tk is effect of time; (c*d*t) ijk is the interaction between the
three factors; eijk is random error.
RESULT AND DISCUSSION
Steaming the cassava root prior to fermentation appeared to have a slightly
beneficial effect (p=0.67) on conversion of crude to true protein (table 2). Increasing
the proportion of DAP from zero to 2% of the substrate DM increased the average level
of true protein from 4.16 to 5.85% in DM (table 2). The level of true protein in the
substrate increased with a curvilinear trend (R 2 = 0.98) from 2.3 to 6.9% in DM as the
5


6

fermentation time increased from zero to 14 days; the crude protein was 10.5 in DM
after mixing the substrate and did not change at the end of the fermentation time. The
ratio of true protein to crude protein increased from 24.6 to 63.7 over the same period
(table 2; figures 1 and 2). However, according to (Vanhnasin and Preston, 2016b)
showed the DM, CP and TP of cassava root fermented (14 days) without ingreedients
were lower such as: (DM 29.5%, CP 3%, TP 1.5% but the value of Ash was hight
(97%).
Table 2. Mean values for DM, OM, crude protein; true protein and ratio of TP/CP at
different stages of the fermentations (% in DM)
DM

OM

CP

TP

TP/CP

Steaming (ST)
ST
28.45
87.25
10.40
5.15
NST
30.34
87.58
10.37
4.90

SEM
0.344
0.929
0.068
0.039
p
0.008
0.812
0.806
0.004
DAP, % in DM
0
29.89
87.14
10.13b
4.16c
1
29.52
87.78
10.53a
5.08b
a
2
28.78
87.32
10.50
5.85a
SEM
0.422
1.138

0.084
0.047
p
0.247
0.921
0.025
<0.001
Times (Days)
0
29.6b
87.41
10.47
2.30d
c
3
24.69
85.71
10.46
4.43c
7
29.34b
86.44
10.47
6.52b
a
14
33.97
90.09
10.14
6.87a

SEM
0.487
1.314
0.097
0.055
p
<0.001
0.199
0.121
<0.001
abc Mean values in rows without common superscript differ at p<0.05

47.10
48.09
1.540
0.665
36.77b
50.05a
55.70a
1.887
0.001
24.59c
42.31b
59.76a
63.72a
2.179
<0.001

DAP: di-ammonium phosphate; DM: dry matter, OM: Organic matter, CP: crude protein,
TP: true protein, ST: steam and NST: not steam

The final values after 14 days of fermentation being increased from 5.63 to 8.33%.
The ratio of true protein to crude protein increased from 46.56 to 81.35 after 14 days of
fermentation (table 3).
Table 3. Effect of level of DAP on concentration of crude protein, true protein and
ratio of TP/CP after 14 days of fermentation (% in DM)
0
CP
TP
TP/CP

9.65
5.63c
46.56b

DAP, % in DM
1
10.51
6.65b
63.25ab
6

2

SEM

P

10.25
8.33a

81.35a

0.168
0.095
3.773

0.279
<0.001
0.049


7
abc Mean values in rows without common superscript differ at p<0.05
DAP: di-ammonium phosphate; DM: dry matter, CP: crude protein, TP: true protein

Changes in the mass of substrate during fermentation
About 30% of the original DM in the substrate had been fermented by the end
of 14 days, the rate of loss showing a curvilinear trend with time, with the major
change taking place in the first 3 days (table 4).
Table 4. Changes in the mass of fresh (FM) and dry (DM) substrate during the
fermentation
Days
FM, kg
%DM
DM, kg
a
b
0
1.00
29.6

0.29a
3
0.94a
24.7c
0.23b
7
0.79b
29.3b
0.23bc
c
a
14
0.62
33.9
0.21c
SEM
0.013
0.708
0.006
p
<0.001
<0.001
<0.001
abc Mean values in rows without common superscript differ at p<0.05

The crude protein and Organic matter g/kg of DM after mixing the substrate did
not change at the end of the fermentation time (Days 0 to 14). The ratio of true protein
was increased from 22.98 to 69.51 from zero to 14 days and there were diference
among time of fermentation (table 5).
Table 5. Chemical composition (g/kg of DM)

Days
OM
CP
TP
0
874.09
104.71
22.98a
3
857.10
104.60
44.32b
7
864.43
104.74
65.15c
14
900.86
101.36
69.51d
SEM
13.137
0.967
0.568
P
0.199
0.121
<0.001
abc Mean values in rows without common superscript differ at p<0.05

EXPERIMENT 2.
MATERIALS AND METHODS
Location: The experiment was carried out in the Laboratory of the Animal Science
Department in the Faculty of Agriculture and Forest Resource in Souphanouvong
University. The site is located 7 km from Luang Prabang City, Lao PDR.
7


8

Experimental design
The treatment was repeated steaming of cassava root with 2% of di-ammonium
phosphate (DAP) the same as for experiment 1
The experimental design was a 2*9 factorial arrangement in a completely
randomized design (CRD) with 2 treatments combination and with 4 replications of
each period. The treatment mixtures (DM basis) of 93.6% cassava root, 3% yeast, 1.4%
urea and 2% di-ammonium phosphate (DAP) were under aerobic and anaerobic
conditions, fermented for 9 periods (0, 3h, 1, 2, 3, 4, 5, 6 and 7 days).
Procedure
Cassava roots were peeled and chopped into small pieces (1-2 cm) and steamed
for 30 minutes in a bamboo basket placed above a pan containing boiling water. It was
then cooled for 15 minutes prior to being mixed with the yeast (S. cerevisiae), urea and
DAP (all these ingredients were use the same kind as experiment 1. One half of the
mixed substrate was transferred to bamboo baskets covered with plastic netting to
allow free entrance of air and allowed to ferment for 7 days (aerobic condition). The
other portion of the substrate was packed tightly into 0.5 liter plastic bags, which were
closed (anaerobic condition), and in which it was stored for 7 days.
Measurements
Samples were taken from each treatment/replicate on day 0 (3 h after mixing the
components of the substrate), and then every 24h until end of day 7 of the fermentation

for measurement of pH, crude protein, true protein and ammonia.
Chemical analysis
The pH of each sample was measured with a digital pH meter, prior to addition
of sodium hydroxide for subsequent analysis for ammonia by steam distillation
(AOAC, 1990). Crude protein was analyzed by kjeldahl digestion with sulphuric acid
followed by distillation according to AOAC, (1990) methods. For estimation of true
protein, 2 g of the fresh sample were put in a 125ml Erlenmeyer flask with 50 ml of
distilled water, allowed to stand for 30 minutes, after which 10ml of 10% TCA
(trichloroacetic acid) were added and allowed to stand for a further 20-30 minutes. The
suspension was then filtered through Whatman #4 paper by gravity. The filtrate was
discarded, and the remaining filter paper and suspended substrate transferred to a
kjeldahl flask for standard estimation of total N. Urea-N was estimated by subtraction
of true protein-N and ammonia-N from the crude protein-N. The measurements of
crude and true protein and ammonia were done on the fresh samples.
Statistical analysis
The data were analyzed by the General Linear Model (GLM) option in the
ANOVA program of the Minitab, (2010) software (version 16.0). In the model the
8


9

sources of variation were treatments, treatment interaction and random error. Turkey’s
pair-wise comparisons was used to determine the differences system condition; time
when the P value of F test P<0.05. The statistical models used were:
Yij = µ +ai +tj + (a*t)ij + eij
Yij are dependent variables; µ is overall mean; a i is effect of system condition
(aerobic and anaerobic); tj is effect of time; (a*t)ij is the interaction between the two
factors; eij is random error.
RESULT AND DISCUSSION

Chemical composition of substrates
The pH decreased with fermentation time, according to an almost linear trend,
from 5.8 immediately after mixing the substrate, to 5.47in 3h and to 3.43 after 7 days
(table 1). The level of crude protein after mixing the substrate and additives was
10.35% in DM and, as expected, did not change over the 7 days of fermentation. The
level of true protein in the substrate increased from 2.37 to 6.97% in DM as the
fermentation time increased from zero to 7 days, such that the ratio of true to crude
protein increased from 22.95 to 66.11 over the same period (table 1). There were no
differences in all these criteria as between the aerobic and anaerobic condition, other
than a tendency for the pH to fall slightly more quickly in the first 4 days in the
anaerobic condition followed by a slower rate of fall to reach almost the same final
value after 7 days, as for the aerobic condition.
Table 1. Changes in pH, crude protein (CP), true protein (TP) and ammonia in cassava
root fermented with yeast, urea and DAP under aerobic or anaerobic conditions
pH
Ammonia
CP
TP
Condition
system
Anaerobic
4.27
0.40
10.40
4.86
Aerobic
4.66
0.42
10.41
4.87

SEM
0.013
0.001
0.045
0.022
p
<0.001
<0.001
0.792
0.773
Time (days)
0
5.83a
0.49a
10.35
2.37i
3h
5.47b
0.48ab
10.26
3.21h
c
b
1
4.87
0.46
10.55
3.74g
2
4.64d

0.45c
10.47
4.23f
e
d
3
4.32
0.41
10.29
5.08e
4
4.11f
0.39e
10.55
5.54d
5
3.90g
0.36f
10.35
6.14c
h
g
6
3.63
0.34
10.29
6.46b
7
3.43i
0.30h

10.55
6.97a
SEM
0.027
0.003
0.096
0.046
p
<0.001
<0.001
0.144
<0.001
abc Mean values in rows without common superscript differ at p<0.05
9

TP/CP

46.66
46.72
0.070
0.537
22.95i
31.30h
35.46g
40.40f
49.39e
52.49d
59.36c
62.75b
66.11a

0.148
<0.001


10
pH: Power of/potential Hydrogen; CP: crude protein, TP: true protein; p: probability; SEM:
standard error of the mean; CP and TP (% in DM)

The proportion of true protein in the substrate after the 7-day fermentation was
doubled from 34 to 62% of the total nitrogen, and appeared to be derived almost
equally from the nitrogen present originally as ammonia (from DAP) and from urea
(figure 3). However, urea was not determined directly but was assumed to be the source
of the N remaining after accounting for the protein-N and ammonia-N at the end of the
fermentation. Two questions to be answered are: i) Why all the ammonia was not used
for yeast growth?; and ii) why was the urea not completely hydrolyzed to ammonia?
The latter question could perhaps be explained as being the consequence of the rapid
fall in the substrate pH inhibiting the action of urease, the action of which is decreased
at low pH (Kay and Reid, 1934).
CP

TP

Power (TP)

12.00

N*6.25, % in DM

10.00
8.00

f(x) = 2.27 x^0.5
R² = 0.99

6.00
4.00
2.00
0.00

0

3h

1

2

3

4

5

6

7

Fermentation, days

Figure 3. Distribution of the nitrogen as urea, ammonia and

true protein at the beginning and after 7 days of
fermentation
DISCUSSION
The increase in the true protein content of the cassava root by fermentation with
yeast, urea and DAP is supported by reports from several researchers. Fermentation of
cassava peels by a pure culture of S. cerevisiae increased the protein content from 2.4%
to 14.1%, according to Antai and Mbongo, (1994). Oboh and Kindahunsi, (2005)
reported that the fermentation of cassava flour with S. cerevisiae increased the protein
level from 4.4% to 10.9% in DM. Krisada et al., (2009) carried out a similar
fermentation with fresh cassava root using urea and yeast. The crude protein was
increased from 3.2 to 21.1% in DM with 90% of the crude protein in the form of true
protein. Phiny et al., (2012) fermented broken rice in an anaerobic system simulating
the “farmer” production of rice wine. The difference in procedure was the addition of
urea (1% of the rice) as well as yeast and no distillation. The crude protein content was
raised from 7% in DM in the broken rice to 23% in DM after fermentation for 3 days.
10


11

The proportion of “crude” to “true” protein was not measured but the 37% increase in
pig growth rate when the protein-enriched rice was included in the diet, compared with
16.5% improvement in growth rate, for supplementation with fish meal, indicated that
much of the increase in “crude” protein was as “true” protein. Manivanh and Preston,
(2016) used cassava root as the carbohydrate source, with incorporation of DAP as a
source of phosphorus to supplement the yeast and urea. True protein content of the
cassava root was raised to 14% in DM (from 2.5% in unfermented root) and growth
rate of Moo Lath pigs was increased by 46% compared with the control diet in which
the protein was from ensiled Taro foliage (leaves and petioles).
Phosphorus is required for the growth of all biological entities, including

yeast, thus the 30% increase in true protein in the fermented cassava pulp by raising
the level of DAP (20% phosphorus) from zero to 2% (in DM) was to be expected.
There appear to be no comparable studies on effects of phosphorus levels in proteinenrichment of carbohydrate with yeast and urea.
There is not all the added nitrogenous compounds (urea and DAP) were
converted to “true” protein, the levels of which never exceeded some 50 to 70% of the
“crude “ protein in experiments with yeast-fermented cassava root (Vanhnasin and
Preston, 2016a) and cassava root pulp (Sengxayalth et al., 2017a). In an attempt to
solve this problem, from experiment 1 increased the level of DAP to 2% of the
substrate, reducing the level of urea to 1.2%. This had the effect of increasing the
phosphorus level in the substrate with a related increase in the proportion of the added
N as ammonia, replacing urea-N. The linear increase in the proportion of true to crude
protein that resulted was attributed to the increased level of phosphorus, but another
explanation could have been the partial change in the origin of the added NPN - from
urea to ammonia. Yeast cannot directly use urea which must first be hydrolysed to
ammonia by urease. However, the activity of urease is inhibited at low pH (Kay and
Reid, 1934), which falls rapidly when the cassava root is fermented.
CONCLUSIONS
 The true protein in cassava root increased with a curvilinear trend (R2 = 0.98) from
2.3 to 6.87% in DM as the fermentation time increased from zero to 14 days, the proportion
of true protein in crude protein increasing from 24.6 to 63.7 over the same period.
 Increasing the proportion of DAP from zero to 2% of the substrate DM
increased the true protein from 5.6 to 7.3%% after 14 days of fermentation.
 30% of the original root DM was fermented in the partial conversion of root
carbohydrate to true protein after 14 days of fermentation.
 Steaming the cassava root prior to fermentation appeared to have a slightly
positive effect on conversion of crude to true protein.
 The pH decreased with fermentation time, according to an almost linear trend,
from 5.8 immediately after mixing the substrate, to 5.4 in 3h and to 3.43 after 7 days.
11

12

 The level of crude protein after mixing the substrate and additives was 10.35%
in DM and did not change over the 7 days of fermentation.
 There were no differences between the aerobic and anaerobic condition, other
than a tendency for the pH to fall slightly more quickly in the first 4 days in the
anaerobic condition followed by a slower rate of fall to reach almost the same final
value after 7 days, as for the aerobic condition.
 It is suggested that the incomplete conversion of urea-N and ammonia-N to
yeast protein is because of incomplete hydrolysis of urea to ammonia due to action of
urease being inhibited by the fall in pH during the fermentation.

CHAPTER 3
REPLACING TARO (Colocasia esculenta) SILAGE BY PROTEIN-ENRICHED
CASSAVA ROOT IMPROVED THE NUTRITIVE VALUE OF A BANANA
STEM (Musa sapientum Linn) BASED DIET AND SUPPORTED BETTER
GROWTH IN MOO LATH PIG
ABSTRACT
A growth trial was conducted with 12 Moo Lath pigs with average 14.8 ±1.89
kg initial live weight in a CRD, with three replications of four treatments. The aim of
the study was to determine the effect of replacing TS with PECR in a basal diet of
ensiled banana stem (BS). Fermentation of fresh cassava root with yeast, urea and diammonium phosphate (DAP) increased the content of true protein in the root from 2.5
to 14.2% in DM. There were positive responses in dry matter (DM) intake, live weight
gain, feed conversion ratio, as the percentage of PECR in the diet was increased (zero
to 15% in DM ). It was concluded that the replacing of taro foliage silage with PECR
improved the quality of the overall diet, which resulted in higher intake, growth rate,
better feed conversion ratio and economical efficiency.
Key words: DAP, yeast, urea, feed conversion, live weight gain
INTRODUCTION

Most pigs in rural areas of Lao PDR are raised in traditional, low input, free and
semi-free scavenging systems, where the pigs are allowed to scavenge freely for feed
all the year round or after the main crops have been harvested (Phengsavanh et al.,
2010). The main feed resources are agricultural by-products and vegetables and weeds
that grow in forests, along the banks of streams and in cropping areas. These feed
resources are vulnerable to seasonal weather changes and in the dry season feed is
always in short supply. Thus, one of the main limitations to pig production in these
smallholder production systems is a shortage of feed. Apart from this feed shortage,
infectious diseases are also a problem that limit productivity (Conlan et al., 2008;
Phengsavanh and Stur, 2006; Thorne, 2005).
In Lao PDR as in most tropical countries the most widely grown crops are
12


13

primarily sources of carbohydrate (eg: rice, sugar cane, cassava). Few crops are grown
specifically as sources of protein. As a result, protein rich feeds such as soybean meal
are imported in order to produce balanced diets for livestock especially pigs and
poultry. An alternative approach that has been studied by several researchers is the solid
state fermentation of carbohydrate-rich byproducts from these crops using
combinations of fungi and yeast (Phiny et al., 2012; Phong et al., 2013; Khempaka et
al., 2011; Hong and Ca, 2015) in order to enrich the content of protein.
The aim of the research reported in this paper was to apply the proteinenrichment technique to raise the protein content of cassava roots and to evaluate the
use of this product as partial replacement of taro silage in a banana stem- based diet fed
to Moo Lath pigs
MATERIALS AND METHODS
Location
The experiment was carried out at the Faculty of Agriculture and Forest
Resource in Souphanouvong University. The site is located 7 km from Luang Prabang

City, Lao PDR.
Experimental design, treatments and management
Protein-enrichment of cassava root
Cassava roots (60 kg) were peeled and chopped by hand into small pieces (1-2
cm), and steamed for 30 minutes. For the steaming a 200 liter steel barrel was used.
This had a false floor of bamboo strips supported by wooden boards 30cm above the
base of the barrel. The space beneath the bamboo strips contained water which was
maintained at boiling point by a wood fire underneath the barrel.
The steamed cassava root was removed from the barrel and cooled for 15
minutes (94.2%) then mixed with urea 0.8%, di-ammonium phosphate (DAP) 3% and
yeast (S. cerevisiae) 2% respectively, of the cassava DM (all ingredients were used the
same kind as experiment 1 and 2). The mixed substrate was then transferred to bamboo
baskets covered with plastic netting to allow free entrance of air (photo 5). On each of 3
consecutive days the contents of the basket were turned so that all the contents were
exposed to the air entering through the top and the sides of the basket. After seven days,
the protein-enriched cassava root was fed to the pigs.
Ensiling taro foliage and banana stems
Taro (Colocasia esculenta) leaves and petioles were collected from areas around the
University and were chopped into small pieces (2-3 cm length). They were wilted for 24h to
reduce the moisture content and ensiled, without additive, in 50 liter polyethylene bags for 14
days. Banana stems were bought from a nearby village. They were chopped by hand into
small pieces and ensiled in 200 liter PVC containers for 14 days.
Experimental design
The experiment was arranged in a completely randomized design (CRD) with 4
treatments and 3 replications.
13


14

Individual treatments were different proportions (DM basis) of protein-enriched
cassava root (PECR) replacing taro silage (TS) with constant proportions of ensiled
banana stem (BS):
• PECR0: Taro silage (TS) 60% + ensiled banana stem (BS) 40%
• PECR5: TS 55% + BS 40% + Protein-enriched cassava root (PECR) 5%
• PECR10: TS 50% + BS 40% + PECR 10%
• PECR15: TS 45% + BS 40% + PECR 15%
Twelve Moo Lath pigs with a mean body weight of 14.8 ±1.89 kg (8 males; 4
females) were bought from a pig farm in Luang prabang Province. They were
vaccinated against swine fever and treated against round worms with ivermectin
(1ml/20kg LW), before starting the experiment. The pigs were housed in individual
pens (width 1m and length 1.2m) made from local materials. The pigs had free access
to water and were adapted to the pens and the feeds for one week before starting the
experiment which lasted 90 days.
The diet ingredients were mixed together and given two times per day at 6:30
am and 5:00 pm, the amount being based on an offer level of 40g DM/kg live weight.
Data collection
The pigs were weighed in the morning before feeding, at the beginning of the
trial and every 15 days. Live weight gain was determined from the linear regression of
live weight on days in the experiment. Samples of feed offered and refused were
collected daily, weighed and sub-samples stored in the refrigerator at 4°C before being
bulked for analysis of DM, N and ash.
Chemical analysis
AOAC (1990) methods were used to analyses the sub-samples of feeds offered
and refused for DM, N and OM. True protein was analyses only PECR, it was
determined by prior treatment of the samples with Trichlor-acetic acid (TCA) before
estimation of N.
Statistical analysis
Data for feed intake, live weight were analysed by the General Linear Model
(GLM) option in the ANOVA program of the Minitab (2010) software (version 16.0).

Sources of variation were treatments and error.
The statistical model was used: Yij =  +  i + eij
Yij are dependent variables; µ is overall mean;  i = treatment effect (i=1-4); eij
is random error

14


15

RESULTS AND DISCUSSION
Chemical composition
After fermentation seven days, the crude and true protein values were 16.7 and
14.2% (in DM). The level of crude protein in the ensiled banana stems was very low
(table1).
Table 1. The chemical composition of feed ingredients (% in DM,
except DM which is on fresh basis)
Taro silage
Ensiled banana stem

DM
26
7.5

N*6.25
15.8
4.6

OM
81.9

92.8

True protein
-

PECR

26.5

16.70

98.4

14.2

Feed intake, growth rate and feed conversion
DM intake was increased linearly with the increase in the level of proteinenriched cassava root (Table 2).
Table 2. Mean values for DM intake (g/day) by pigs fed taro silage (TS) and
ensiled banana stem (BT) supplemented with protein enriched cassava root
(PECR)
PECR0

PECR5

PECR10

PECR15

SEM

p

DM intake, g/day
PECR

0

43

90

127

-

-

TS

431

435

388

383

-

-

BS

301

332

335

345

-

-

Total

732c

810b

813b

854a

4.599

<0.001

g DM/kg LW

39.9b

40.5b

41.1ab

42.4a

0.492

0.002

abc

Means with different letters within the same row differ at p<0.05

The live weight gain of the pigs (Table 3) was increased by 46% with a linear
trend (Figure 2) as the protein-enriched cassava root was increased from zero to 15% of
the diet. The DM feed conversion followed a similar trend with a curvilinear
improvement as the proportion of protein-enriched cassava root in the diet was
increased (Figure 3).
15


16

Table 3. Mean values for live weight changes of growing pigs during the
experiment
Live weight, kg

PECR0 PECR5 PECR10 PECR15

SEM

p

Initial

14.5

15.3

14.8

14.5

1.35

0.966

Final

25.0

27.4

28.1

28.9

1.41

0.302

Daily gain, g/day

125d

150c

167b

183a

2.68

<0.001

DMI, g/day

732c

810b

813b

854a

4.60

<0.001

DM conversion

5.9a

5.4ab

4.9b

4.7b

0.218

0.017

abc

Mean values in rows without common superscript differ at p<0.05

Economic analysis
Feed ingredient costs and an economic analysis of the experimental treatments
are shown in table 4 and 5, respectively. Feed cost/kg DM was lower for PECR0,
although differences between diets were small. However, because feed conversion
ratios were lowest for the PECR15 and PECR10 treatments (4.7 and 4.9 kg feed DM/kg
gain, respectively), feed costs/kg weight gain of the PECR15 and PECR10 treatments
were lowest (15,505 and 15,618 kip/kg live weight gain, respectively) and were highest
for the control diet or PECR0 (16,426 kip/kg gain).
Table 4. Feed ingredient costs (LAK)

Feed stuffs
Taro

Kip/kg as feed
550

Kip/kg DM
2,115

Banana stem
250
3,333
Cassava root
1,000
3,774
DAP
15,000
15,000
Urea
15,000
15,000
Yeast
35,000
35,000
#
PECR
4,111
LAK = Lao Kip (Lao currency) exchange rate: 8,569 = 1 USD;
#
PECR: protein-enriched cassava root was calculate included price of DAP, urea

and yeast from root processing

Table 5. Economic analysis of experimental treatments (LAK)
Parameter

PECR
0

PECR5
16

PECR10

PECR15

SEM

p-value


Days of experiment
LW gain, kg
DMI kg/day
FCR kg DM/kg

90

17
90

90

90

-

-

10.5

c

12.1

b

a

14.4

a

0.255

<0.001

0.733

c

0.81

b

0.854

a

0.005

<0.001

5.9

a

b

0.219

0.017

5.4

ab

13.3
0.813

b

4.9

b

4.7

Feed cost/kg
2,614
2,721
2,838
2,904
Feed cost/kg LWG
16,426
16,391
15,618
15,505
LAK = Lao Kip (Lao currency) exchange rate: 8,569 = 1 USD; LW = live weight; DMI =
Dry matter intake
#
No labour cost included

-

DISCUSSION
The increase in the true protein content of the cassava root by fermentation with
yeast, urea and DAP agrees with the findings of many researchers. Krisada et al.,
(2009) carried out a similar fermentation with fresh cassava root using urea and yeast.
The crude protein was increased from 3.2 to 21.1% in DM with 90% of the crude
protein in the form of true protein. Fermentation of cassava peels by a pure culture of S.

cerevisiae increased the protein content from 2.4% to 14.1% (Antai and Mbongo,
1994). Oboh and Kindahunsi, (2005) reported that the fermentation of cassava flour
(pulp??) with S. cerevisiae increased the protein level from 4.4% to 10.9% in DM.
The increase in growth rate (46%) by replacing Taro silage with proteinenriched cassava root was comparable with that (32%) reported in an earlier experiment
(Manivanh and Preston, 2015). DM feed conversion in the present experiment (4.7)
was also better than Manivanh and Preston, (2016) DM feed conversion (5.7). Hang et
al., (2015) was showed the improvement in pig performance by replacing ensiled Taro
foliage with protein-enriched cassava root could be a reflection of the higher energy
value in the latter (% crude fiber in DM of 3.7 in cassava root compared with 11% in
Taro foliage).
Feed costs were lowest for the PECR15 diet, indicating that replacing taro
(Colocasia esculenta) silage by protein-enriched cassava root can be the most
economical for small-holder farmers.
CONCLUSIONS


There were positive linear responses in DM intake, live weight gain, feed
conversion, when protein-enriched cassava root partially replaced taro silage in
a basal diet of ensiled banana stem fed to Moo Lath pigs.



Utilization of inexpensive, locally available feed resources, such as cassava root
and its can be improving nutritive value by fermented with yeast, urea and diammonium phosphate (DAP) from 2.5% to 14.2%, have the potential to improve
the economical efficiency of smallholder pig production in Laos.

17


18

CHAPTER 4
APPARENT DIGESTIBILITY AND N RETENTION IN GROWING MOO
LATH PIGS FED ENSILED TARO FOLIAGE (Colocasia esculenta) REPLACED
BY PROTEIN-ENRICHED CASSAVA ROOT (Manihot esculenta Crantz)
ABSTRACT
Four castrated male pigs (Moo Lath pig), weighing on average 15 kg were
allotted at random to 4 diets within a 4*4 Latin square design, to study effects on DM
intake, digestibility and N retention of levels of protein-enriched cassava root (PECR)
of 0, 25, 50 and 75% in combination with taro silage 80, 55, 30 and 5% with constant
levels of ensiled banana stem 20% (All on DM basis)
Enriching the protein content of cassava root (PECR) by fermenting it with
urea, diammonium phosphate and yeast and including the PECR at 25% in a diet of
ensiled cassava root, ensiled taro foliage and ensiled banana pseudostem, led to
increases in feed intake, diet digestibility and N retention in native Moo Lath pigs.
These criteria declined linearly when the proportions of PECR were 50% and 75% of
the diet DM. It is suggested that: (i) the benefits from the 25% PECR diet may have
been partially a response to its content of live yeast, the other possibility could be the
increased provision of vitamins of the β-complex from the yeast in the fermented
cassava root; and (ii) the results provide confirmatory evidence that pigs are able to
utilize small quantities of non protein nitrogen NPN recycled to the small intestone for
microbial synthesis to amino acids.
Key words: banana pseudo-stem, diammonium phosphate, probiotic, solid-state
fermentation, urea, yeast
INTRODUCTION
A series of experiments has been carried out recently in Lao PDR (Manivanh
and Preston, 2015; Manivanh and Preston, 2016; Manivanh et al., 2018; Sengxayalth
and Preston, 2017a; Vanhnasin and Preston, 2016; Vanhnasin et al., 2016) with the aim
of defining the potential of the solid-state fermentation process as a means of raising
the nutritive value content of cassava roots and cassava root pulp as the pig diet (Moo

Lath pig in Lao PDR and and Mong Cai pig in Vietnam).
The results of these experiments showed that fermenting the fresh cassava roots
or cassava root pulp with a combination of urea (1-2%), DAP (1-2%) and yeast (3%)
over periods of 5 to10 days resulted in increases of true protein to levels of 7-8% (DM
basis) representing about 60% of the total crude protein derived from the added urea,
DAP and yeast. It has not yet been possible to exceed this true protein fraction by
procedures such as: prior steaming of the carbohydrate source, and/or fermentation
under anaerobic or aerobic conditions. The nature of the approximately 30% of residual
non-protein nitrogen has not been identified. Data presented by Manivanh et al., (2018)
showed that ammonia levels were minimal. These authors hypothesized that the
18


19

residual NPN it could still be in the form of urea, the hydrolysis of which by urease
might be reduced due to the rapid fall in pH with the onset of fermentation.
Despite the apparent limitation of having some 30-40% of the nitrogen in nonprotein form there have been positive results when the protein-enriched pulp or root
was fed to pigs at levels to provide some 25% of the diet protein replacing protein from
ensiled taro foliage (Manivanh and Preston, 2015; Vanhnasin et al., 2016; Sengxayalth
and Preston, (2017b). At levels exceeding 25% of the diet protein, growth rates were
depressed slightly when the PECR replaced a balanced protein source (eg: ensiled taro
foliage) but fell to little over maintenance when the diet was composed of 75% PECR
and 25% of ensiled cassava root, a diet in which all the protein was from PECR
(Sengxayalth and Preston, 2017b).
The following experiment was designed to generate more information on the
effects of increasing dietary levels of PECR replacing a combination of ensiled taro
foliage and ensiled banana pseudo-stem in the diet of growing Moo Lath pigs.
MATERIAL AND METHODS
Location and duration

The experiment was conducted in the experimental area of Souphanouvong
University (SU), in Luang Prabang province, Lao PDR.
Experimental design
Four treatments were compared in a 4*4 Latin Square arrangement with 4 pigs
and 4 periods. The treatments (DM basis) were:
PECR0: Taro silage (TS) 80% + ensiled banana stem (BS) 20%
PECR25: TS 55% + BS 20% + Protein-enriched cassava root 25%
PECR50: TS 30% + BS 20% + Protein-enriched cassava root 50%
PECR75: TS 5% + BS 20% + Protein-enriched cassava root 75
The duration of the experiment was 48 days with 4 periods each of 12 days, the
first 7 days for adaptation then 5 days for data collection (feed residues, feces and urine).
Animals and housing
The experiment was used four male castrated local pigs (Moo Lath pigs), with
average live weight of 15 kg were housed in cages made of bamboo, designed to
separate feces and urine.
Experimental feeds
Taro foliage (leaves and petioles) was harvested from ponds in the Univesity
campus. It was chopped, wilted for 8 hours to reduce the moisture content, and then
ensiled for 14 days before starting the experiment.
19


20

Protein-enrichment of cassava root was processed by following method the
same as experiment 3 (chapter 3) and the ingredient was formulate repeaed from the
best treatment of experiment 1 (chapter 2) the ingreedients were: yeast (3%), urea
(1.4%) and di-ammonium phosphate (DAP) 2%. For urea, yeast and DAP were use the
same as experiment 1, 2 and 3. The mixtures were then allowed to ferment in closed
polyethylene bags for 7 days before feed animal.

Banana stems was processed the same method as experiment 3 (chapter 3)
before feed to animal.
The pigs were fed twice daily (8:00 am and 4:00 pm). The three ingredients
were mixed together before each feed. Feed was offered at ad libitum but with careful
observation of the intake so as to minimize residues. Drinking water was permanently
supplied through drinking nipples.
Measurements and data collection
The pigs were weighed in the morning before the start of each period and one
day after the end of the last period. Feed offered and refused was recorded collected
daily. Samples of feeds offered and refused were taken daily and sored at until the end
of each collection period when they were mixed and sub-samples take for analysis of
DM, ash and N. Feces and urine were collected daily. Each day 20 ml of 15 %
H2SO4 were added to the urine container to maintain the pH of the urine below 4.0. The
feces were stored at 4°C until the end of each collection period when they were mixed
and a sub-sample taken for analysis of DM, ash and N. A sub-sample of urine was
taken daily and stored at 4°C until the end of each collection period when the samples
was mixed and a sub-sample taken for analyses on N.
AOAC, (1990) methods of analysis were followed for analysis of: DM, ash and N in
feeds offered and refused and in feces; N in urine; and true protein after reacting the
samples with trichloro-acetic acid.
Statistical analysis
The data were analysed with the general linear model (GLM) procedure for
repeated measures in the SAS software (SAS, 2010), as a latinsquare split design. The
repeated measures were the data for each of the 5 consecutive days of data collection
within each period.
The statistical model was: Yijk = μ + Ti + Cj+ R (k) + timel + time (pen) + eijk
Yijk = Dependent variables; μ = Overall mean; Ti = Treatment effect (i=1-4), Cj =
period effect (j=1-4); R (k) = pen effect; time effect (l=1-5); pen within time effect
(Error a), and eijk = random error (b).

20


21

RESULTS AND DISCUSSION
Chemical composition
The true protein in the PECR accounted for 53% of the crude protein (table 1), a value
similar to that reported by Manivanh et al., (2016), Sengxayalth and Preston, (2017a);
Vanhnasin and Preston, (2016).
Table 1. The chemical composition of feed ingredients (% in DM, except DM
which is on fresh basis)

Taro silage
Ensiled banana
stem
PECR

DM

N*6.25

25.6

15.3

8.2

4.3

28.4

13.7

OM

True protein

83.
4
93.
1
98.
4

7.3

Feed intake
The daily DM intake followed a curvilinear trend (y = -56.4x 2 + 276x + 366; R²
= 0.70) increasing to a maximum as the proportion of PECR was raised from 0 to 25%
then declining with higher proportions of PECR (table 2). As a function of liveweight
the intakes were high (33 to 44 g DM/kg live weight).
Table 2. Mean values of DM intake by pigs fed protein-enriched cassava root (PECR)
replacing taro silage with constant levels of ensiled banana stem
PECR0
DM intake, g/day
PECR
0
Taro silage
429

Banana
140
stem
Total
569c
g/kg LW
33c
abc

PECR25

PECR50

PECR7
5

SEM

180
381

307
177

422
25

6.13
5.29

180

154

137

3.25

741a
44a

639b
38b

585c
35c

13.5
0.531

p

<0.001
<0.001

Means with different letters within the same row differ at p<0.05

Apparent digestibility
The curvilinear trend for apparent digestibility of DM was similar to that for DM
intake (table 3; figure 2) with increases in the digestibility coefficient as PECR replaced

ensiled taro foliage at the 25% level subsequently declining with increasing degree of
replacement by PECR. However, for protein the depression was a linear negative trend
over the whole range of replacement of ensiled taro foliage by PECR. While organic
matter was led in the diet PECR 25% and different among treatment (p = 0.003).
21


22

Table 3. Apparent digestibility (%) of diets with PECR replacing ensiled taro foliage
with constant levels of ensiled banana stem
Dry matter
Organic matter
Crude protein
abc

PECR0

PECR25

PECR50

PECR75

SEM

p

64.7a
59.7b

75.0a

70.9a
68.4a
74.2a

64.0ab
62.4ab
71.8ab

56.9b
57.0b
68.0b

2.04
2.157
1.63

<0.001
0.003
0.002

Means with different letters within the same row are different at P<0.05

Nitrogen balance
N intake and N retention increased with curvilinear trends reaching maximum
levels with 25% PECR in the diet thereafter decreasing (table 4). Part of the increase in
N retention was apparently due to the increased N intake; however, correction of the
data by covariance for differences in N intake did not change the relative pattern of
response to PECR level in the diet. The improvement in N retention with increasing

levels of PECR replacing mixed silages of taro foliage and banana stem contrasts with
the observed linear decrease in apparent N digestibility (table 4).
Table 4. Mean values for N balance in pigs fed protein enriched cassava root replacing
taro silage with constant levels of ensiled banana stem
PECR0

PECR25

PECR50

PECR75

N balance, g/d
Intake
8.77c
10.6a
9.55b
9.09bc
Feces
2.16b
2.74ab
2.65ab
2.93a
b
b
a
Urine
1.77
1.78
2.56

2.40a
N retention
g/day
4.84b
6.10a
4.34bc
3.77c
a
a
b
% of N digested
73.6
77.3
63.7
60.8b
% of N intake
55.1a
57.5ab
45.6b
41.3b
N retention corrected by covariance for differences in N intake
g/day
4.95b
5.93a
4.33bc
3.83c
abc
Means with different letters within the same row are different at p<0.05

SEM

p

0.188
0.180
0.122

<0.001
<0.001
<0.001

0.16
1.70
1.45

<0.001
<0.001
<0.001

0.168

<0.001

DISCUSSION
It has been shown conclusively (Vanhnasin et al., 2016a; Manivanh et al., 2016;
Sengxayalth and Preston, 2017a) that when cassava pulp (or cassava root) is fermented
with urea, DAP and yeast, not all the NPN is converted to true protein, and that some 30%
of the original urea and DAP remains as some form of NPN possibly ranging from
ammonium salts to peptides and amino acids (AA). There is evidence in humans that NPN
in the form of ammonium chloride was partly converted to amino acids by the action of

microbes in the small intestine (Patterson et al., 1995) and Stein et al., 1996). Colombus et
22


23

al., (2014) infused urea into the cecum of pigs fed a diet deficient in the amino acid valine,
and showed that it was recycled to the small intestine where it was converted by bacteria
into amino acids with the result that N retention was increased. These findings were
corroborated by Mansilla et al., (2015).
Amino acid synthesis from dietary NPN could be the explanation for the
positive effects on growth rate at the lower level of substitution of soybean protein by
protein-enriched cassava pulp, since protein was set at the limiting level of 10% of diet
DM. At this level of dietary protein, additional amino acids resulting from microbial
synthesis of amino acids from NPN could have played a critical role in increasing
growth rate. However, at higher rates of substitution of 50 and 75% of the taro protein
by protein-enriched cassava root, the levels of residual NPN (ammonia and urea) may
have exceeded the capacity of the gut microbes to synthesize it into amino acids with
resultant toxic effects of the ammonia leading to reduced feed intake and therefore
reduced growth rate.
The other factors to be taken into consideration are the possible “prebiotic and
probiotic” effects arising from the live yeast and lactobacilli present in the fermented
cassava pulp. Thiep, (2017) unpublished data) reported concentrations of yeast of 9
million CFU/g and lactobacilli 17 million CFU/g in cassava pulp fermented with yeast,
urea and DAP. Positive effects on N retention in Moo Lath pigs were reported by Sivilai
et al., (2017) when rice distillers’ by-product and brewers’ grains were fed at 4% of the
diet DM. Both these dietary supplements are equally rich in live yeast and lactobacilli
(Thiep, 2017) unpublished data.
The issues that are raised are: (i) the benefits from 25% PECR in the diet were
due to the presence of live yeast (estimated to be about 1% of the DM of the PECR25

diet) acting as a probiotic and: (ii) pigs can apparently utilize small quantities of nonprotein-nitrogen through absorption of the NPN from the large intestine and subsequent
recycling in the blood to the small intestine where microbes use the NPN for synthesis
of amino acids (Colombus et al., 2014).
The other interesting observation from this experiment is that N retention was
only marginally depressed even with as much as 75% of the diet in the form of PECR,
which is in marked contrast with the report of Sengxayalth and Preston, (2017b) that
when 75% of the diet was in the form of Protein-enriched cassava root the growth rates
were reduced almost to zero. The difference between the two experiments was the
nature of the remaining 25% of the diet DM. In the research of Sengxayalth and
Preston, (2017b) this was in the form of ensiled cassava root pulp whereas in the
present experiment it was a combination of ensiled taro foliage (5%) and ensiled
banana pseudo-stem (30%). These latter feeds would have been much richer in
vitamins, minerals and trace elements than the PECR which formed the balance of the
diet in the experiment of Sengxayalth and Preston, (2017b). This cannot be proved in
the absence of relevant biochemical analytical data in both trials, but it is an argument
for the benefits of having at least a part of the dietary protein being provided by
nutrient-ich foliages such as taro.
23