Livestock Research for Rural Development 29 (9) 2017 | Guide for preparation of papers | LRRD Newsletter | Citation of this paper |
Two experiments were carried out to study: (i) the effect of urea levels on conversion of NPN to true protein in a solid-state fermentation of cassava pulp: rice bran (70:30 DM basis) with yeast and diammonium phosphate (DAP); and (ii) how this would affect the performance of growing-fattening pigs fed the yeast enriched pulp (PECP) as partial replacement of fish meal in the diet. In experiment 1, the two factors in a 2*2 factorial design with 3 replications were: urea concentration: 0, 0.5, 1.0, 1.5 and 2% (DM basis). (DAP was constant at 1% of substrate DM); and fermentation time (0, 3, 5 and 7 days). In experiment 2 the treatments in a 4-month growing-fattening trial with 25 pigs were: 0, 25, 50, 75 and 100% replacement of fish meal protein by protein from PECP (equivalent to PECP levels of 0, 3.9, 7.7, 11.7 and 15.6% of diet DM from 15 to 50 kg live weight; and 0, 4.5, 8.9, 13.4 and 17.8% of DM from 50 to 100 kg live weight.
Experiment 1: After 7 days of fermentation, the level of true protein reached a range of 12.3 to 12.5% in DM when urea levels were in the range of 0.5 to 2% of the substrate DM. There were no differences between urea levels of 0.5 and 2%. Comparable values for crude protein were in the range 24 to 27.5% in DM. These increases in concentrations of “true” and “crude” protein were associated with a loss of some 30% of the substrate DM indicating that approximately 3 kg of substrate DM were fermented to produce 1 kg of true (yeast) protein. The overall conversion rate of crude protein (from urea and DAP) to true protein was of the order of 60% when the urea level was between 0.5 and 2% of the substrate DM. The best conversion rate of crude to true protein was with 0.5% urea in substrate DM.
Experiment 2: With 25% fish meal protein replacement (4% PECP in the diet DM), feed intake was reduced, live weight gain did not differ, and DM feed conversion tended to be improved. At higher proportions of PECP in the diet, there was a linear depression in live weight gain (overall reduction of 25%) as protein from PECR replaced that from fish meal in the range of 25 to 100% of the fish meal protein (equal to average PECP levels of from 4 to 16.7% in diet DM). There was no consistent trend in the effect of PECP level on feed conversion. It is suggested that the reduction in live weight gain with more than 4% PECP in the diet DM may be related to sub-acute toxicity caused by residual NPN compounds in the PECP because of incomplete conversion of the NPN to yeast protein.
Key words: fermentation, non-protein-nitrogen, true protein, Pichia kudriavzevii
In Vietnam, cassava is an important food-crop and is commonly planted in upland areas of the country. In Thua Thien Hue province, cassava is cultivated on sloping land due to its tolerance to poor soil conditions. In the past, cassava was often used as feed for livestock, but at present it is used mainly for producing starch. After harvesting, most of the cassava is sold to the cassava starch factory in Phong An commune, Phong Dien district, Thua Thien Hue province. Only one third of the cassava root biomass is extracted as starch leaving two thirds as cassava by-product known as cassava pulp. In the harvesting season, 100-150 tonnes of cassava pulp are discharged per day, which causes environmental problems as it decays rapidly.
At present, the cassava pulp is utilized to a limited extent by farmers, but its use is constrained by the presence of cyanogenic glucosides which are converted to toxic HCN when consumed by animals. The low protein content is another limiting factor.
However, it is possible to reduce the content of HCN precursors and to improve the protein content in carbohydrate-rich industrial byproducts by solid-state microbial fermentation (Brook et al 1969; Obahdina et al 2006; Aro et 2008; Thongkratok et al 2010).
The application of this technology to cassava pulp has been studied recently in Lao PDR by Manavanh et al 2016; Vanhnasin and Preston 2016; Vanhnasin et al et al (2016) and Sengxayalth and Preston (2017a,b). These researchers reported an increase in true protein from 2 to 7-12% in dry matter (DM) of the cassava pulp (cassava root in the case of Vanhnasin et al 2016) and that the protein-enriched product (PECP) could provide up to 28% of the dietary protein in a diet based on cassava pulp (or ensiled root), replacing ensiled taro foliage (Vanhnsin and Preston 2016) or soybean meal (Sengxayalth and Preston 2017b). However, higher proportions of the protein-enriched feed in the diet led to reduced growth performance with almost no growth at 100% replacement of the taro/soybean protein. It was reported that only 50-60% of the added NPN (from urea and DAP) was recovered as true protein which implied that more than one third of the original NPN remained, possibly in the form of ammonium salts or intermediary products formed in the process of yeast growth. It was proposed (Sengxayalth and Preston 2017b) that the improvement in growth rate with 30% PECP protein in the diet was the result of amino acid synthesis by bacteria in the small intestine using the residual dietary ammonia as substrate (Colombus et al 2014); and that at higher levels of substitution of PECP the severe depression in feed intake and in growth was due to the toxicity caused by the residual NPN compounds in the fermented pulp/root.
Two experiments were carried out to determine if the urea levels in the solid-state fermentation of cassava pulp with yeast could be reduced, and how this would affect: (i) the conversion of NPN to true protein; and (ii) the performance of growing-fattening pigs fed the yeast enriched pulp.
Cassava pulp was collected from the cassava starch processing factory in Phong Dien district, Thua Thien Hue province. Rice bran was purchased from a feed shop in Hue city.
The yeast Pichia kudriavzevii was obtained from the laboratory of Animal Science and Veterinary Medicine, Hue University. It was cultured on YPD (Yeast Extract Peptone Dextrose) medium at 37 oC in bottles placed on a reciprocal shaker operated at 150 rpm for 3 days, after which it was stored at 4oC. The Pichia kudriavzevii was harvested by centrifugation at 3025 x g for 30 minutes at 6oC and the pellet resuspended in acetate buffer (0.1 M, pH 5.7) containing 30% glycerol (w/w). This suspension was stored at -70oC. Then 1.0g of the yeast was transferred to a solution containing 20% molasses (w/v), 2% urea (w/v) and 1% diammonium phosphate (DAP) (w/v). The suspension was then incubated at room temperature, with shaking at 150 rpm during 3 days. This suspension, containing approximately 106 CFU/ml of Pichia kudriavzevii, was used to ferment the cassava pulp.
The two factors in a 2*2 factorial design with 3 replications were: urea concentration: 0, 0.5, 1.0, 1.5 and 2% (DM basis); and fermentation time (0, 3, 5 and 7 days).
The substrate was cassava pulp meal and rice bran mixed in a 70:30 (w/w) ratio. The different levels of urea and the DAP were dissolved in 97 liters of water and together with 3 liters of the suspension of Pichia kudriavzevii were mixed with the substrate. The mixture (Table 1) was put in trays at a depth of approximately 5 cm and covered by plastic wrap for fermentation at room temperature during 7 days (Photo 1).
Photo 1. Fermenting the substrate |
Table 1. Composition of the substrates |
|||||
Urea,
|
DAP,
|
Cassava |
Rice bran,
|
PK suspension, |
Water,
|
0 |
1 |
70 |
30 |
3 |
97 |
0.25 |
1 |
70 |
30 |
3 |
97 |
0.5 |
1 |
70 |
30 |
3 |
97 |
0.75 |
1 |
70 |
30 |
3 |
97 |
1.0 |
1 |
70 |
30 |
3 |
97 |
DM, crude and true protein, ash were analyzed in unfermented and fermented cassava pulp according to AOAC (1997). True protein was determined after treatment of the samples with trichloroacetic acid (TCA) to precipitate the protein (AOAC 1997).
The data were analyzed by the GLM option in the ANOVA program of the Minitab (2016) software. Sources of variation were levels of urea, times of fermentation, the interaction levels*times and error.
Figure 1 is based on the analytical data (Table 2) for 0, 3, 5 and 7 days of fermentation (FD 0 to 7). The column ‘Calc CP’ has the values calculated for the theoretical crude protein in the subsrates derived from the addition of urea (N*6.25 = 288% CP in DM) and DAP (N*6.25=128% CP in DM) assuming the mixture of cassava pulp and rice bran contained 6% CP in DM (Table 1). (thus CP in DM of substrate with 2% urea and 1% DAP= 6.0+2*2.88+1*1.28= 13.0 CP in DM)
Table 2. Mean values for true (TP) and crude protein (CP) in the substrates after 0, 3, 5 and 7 days of yeast fermentation (all values on DM basis) with urea levels of 0 to 2% (1% DAP in all treatments) |
||||||||
Urea % |
0 days |
3 days |
5 days |
7 days |
||||
TP |
CP |
TP |
CP |
TP |
CP |
TP |
CP |
|
0 |
3.53 |
6.60 |
4.05 |
7.15 |
5.56 |
8.34 |
6.36 |
8.76 |
0.5 |
5.46 |
13.1 |
7.70 |
17.0 |
11.3 |
21.9 |
11.6 |
24.2 |
1 |
5.63 |
12.5 |
7.53 |
22.9 |
10.1 |
24.5 |
12.2 |
26.6 |
1.5 |
4.73 |
14.1 |
8.24 |
20.5 |
11.5 |
26.3 |
12.7 |
27.0 |
2 |
5.14 |
15.2 |
7.68 |
19.5 |
11.2 |
26.2 |
12.3 |
27.0 |
There were major differences between the theoretical levels of crude protein based on the levels of urea and DAP and those determined by chemical analysis. For all fermentation times (including time zero but after adding the urea and DAP), the values determined by analysis were always higher than the theoretical levels; and the extent of the difference increased with the length of the fermentation (Figure 1). From 3 to 7 days the the levels of CP were twice as high as on day zero. It was especially notable that these differences were only observed when urea was included in the substrate.
Figure 1. Changes in crude protein in the substrate according to level of added urea and duration of ferrmentation |
The analysis of chamges in DM content of the substrate with time of fermentation (Figure 2) shows that when urea was present, there was a reduction in DM content from 30 to 22%, and that this was independent of the level of urea in the range of 0.5 to 2% urea in substrate DM. In a similar solid-state fermentation of cassava root, with yeast, urea and DAP, Manivanh et al (2016) showed that these changes in DM content were the result of a loss of 30% of the substrate (determined by weighing the substrate and analysing for DM% before and sfter fermentation).
A 30% loss of sustrate would have the effect of “concentrating” the yeast protein as a percent of the substrate DM by some 30-40%. However, at the highest urea level of 2% in substrate DM, the CP increased to 27% in DM - - an increase of 100%. Therefore not all the increase in CP is accounted for by the loss of substrate in the conversion of carbohydrate to yeast and carbon dioxide. There is recent evidence that some yeasts can fix atmospheric nitrogen (Nwe Ni Win Htet et al 2013); although earlier reports considered this to be unlikely (Millbank 1969). However, in the absence of urea there was no increase in crude protein with fermentation time. There is no apparent explanation for this difference in N-enrichment of the substrate in the absence and presence of urea.
Figure 2. Change in DM content after 7 days fermentation acccording to level of urea added to the substrate |
After 7 days of fermentation, the proportion of the “crude” protein converted to “true” protein was of the order of 30-40% (Figure 3) when the urea level was between 0.5 and 2.0% in substrate DM. The greatest increase was with the addition of 0.5% urea as compared with zero urea (but with 1% DAP which was present at all urea levels). These rates of conversion of “crude” to “true” protein were lower than the values of 60-65% reported by Vanhnasin et al (2016), who fermented fresh cassava root with 3% urea, 1% DAP and 2% of yeast (Saccharomyces cerevisiae), and Sengxayalth and Preston (2017a) who fermented cassava pulp with similar levels of yeast, urea and DAP. However, it is probably more appropriate to compare actual levels of true protein which reached a maximum of 12.5% in DM in the present experiment, similar to the level of 12.5% in DM reported by Sengxayalth and Preston (2017a) and higher than the values of 7.3 and 7.8% in DM reported by Vanhnasin et al (2016) and Manivanh et al (2016). In the two experiments where levels of 12.5% true protein in DM were achieved, the substrate was either 100% cassava pulp (Sengxayalth and Preston 2017a) or 70% cassava pulp and 30% rice bran as in the present experiment. Where lower levels of true protein were reached the substrate was fresh cassava root (Vanhnasin et al 2016; Manivanh et al 2016).
Figure 3. Changes in true and crude protein in the substrate according to level of added urea and duration of ferrmentation |
There appeared to be no advantage in raising the urea level above 0.5% of the substrate DM, nor in extending the fermentation beyond 5 days, in terms of the production of true protein (Figure 4). On the contrary, the higher levels of urea would appear to have negative consequences for feeding animals as the residual NPN potentially could give rise to ammonia toxicity as was hypothesized by Sengxayalth and Preston (2017b).
Figure 4. Changes in true protein in the substrate according to level of added urea (% in DM) and duration of ferrmentation |
The overall result of the fermentation indicates that, on the positive side, 1 kg of yeast protein is produced from approximately 3 kg of fermentable substrate in the form of a 70:30 mixture of cassava pulp:rice bran (DM basis). The negative aspect is that considerable amounts of non-protein-nitrogen (possibly as ammonium acetate/lactate) remain in the fermented substrate with unknown consequences when fed to pigs.
This experiment aimed to evaluate the feeding of the protein-enriched cassava pulp-rice bran mixture as partial replacement for fish meal in a diet for growing-fattening pigs.
Twenty-five crossbreed pigs with average initial body weight of 10 kg were housed in individual cages at the experimental farm of the Institute of Development Studies of Hue University of Agriculture and Forestry, Vietnam. The experiment was designed with five levels of replacement (0, 25, 50, 75 and 100%) of fish meal protein by protein from protein-enriched cassava pulp/rice bran in a basal diet of maize meal, rice bran and cassava root meal (Table 3). The pigs were allocated to 5 treatments with 3 replicates per treatment, each replicate with equal numbers of male and females. The treatments were:
CP0: All supplementary protein from fish meal (which accounted for 22.6% of dietary N in period 1 and 30.6% in period 2
CP25: 25% of fish meal protein replaced by crude protein from protein-enriched cassava pulp/rice bran (PECP)
CP50: 50% of fish meal protein replaced by crude protein from PECP
CP75: 75% of fish meal protein replaced by crude protein from PECP
CP100: 100% of fish meal protein replaced by crude protein from PECP
The diets were formulated to contain 186 and 157 g crude protein/kg DM in growth periods 1 and 2 (Tables 4 and 5). The PECP was produced following the procedure described un Experiment 1 for the 2% urea treatment, and was assumed to contain 27% of crude protein and 12.5% true protein based on the results of Experiment 1.
Table 3. Content of crude and true protein in diet ingredients (% in DM) |
||
|
Crude protein |
True protein# |
Maize |
10 |
10 |
Cassava pulp |
2.4 |
2.4 |
Rice bran |
12 |
12 |
FM |
60 |
60 |
SBM |
46 |
46 |
PECP |
27 |
12 |
PECP: Protein-enriched cassava pulp/rice bran
|
Table 4. Composition of diet ingredients (on DM basis except for DM which is on air-dry basis) |
|||||||||
|
DM |
CP |
EE |
CF |
Ash |
Ca |
P |
Lysine |
Methionine |
Fish meal |
88 |
60 |
4.8 |
1.5 |
42.9 |
7.34 |
1.67 |
5.25 |
1.77 |
Soybean meal |
90 |
46 |
0 |
0 |
0 |
0 |
0 |
3.39 |
0.66 |
Maize |
88.2 |
9.8 |
3.7 |
2 |
1.7 |
0.22 |
0.3 |
0.27 |
0.26 |
Cassava root meal |
87.4 |
2.87 |
1.68 |
2.95 |
2.18 |
0.23 |
0.15 |
0.07 |
0.04 |
Rice bran |
87.6 |
13 |
12.0 |
7.77 |
8.37 |
0.17 |
1.65 |
0.55 |
0.25 |
PECP |
37 |
27 |
3.33 |
12 |
1.71 |
0.11 |
0.2 |
|
|
Salt |
85 |
||||||||
Premix |
96 |
|
|||||||
Table 5. Ingredients (g/kg DM) in the diets |
|||||
Ingredients |
CP0 |
CP25 |
CP50 |
CP75 |
CP100 |
Period 15 – 50 kg (g/kg DM) |
|||||
Fish meal |
70 |
52.5 |
35 |
17.5 |
0 |
Soybean meal |
160 |
160 |
160 |
160 |
160 |
Maize |
412 |
412 |
412 |
412 |
412 |
Cassava root meal |
200 |
178.5 |
157 |
135.5 |
114 |
Rice bran |
150 |
150 |
150 |
150 |
150 |
PECP |
0 |
39 |
78 |
117 |
156 |
Salt |
3 |
3 |
3 |
3 |
3 |
Premix |
5 |
5 |
5 |
5 |
5 |
CP, g/kg DM |
186 |
186 |
185 |
185 |
184 |
PECP, % of CP |
0.0 |
5.7 |
11.4 |
17.1 |
22.9 |
PECP, % of FM |
0.0 |
25.1 |
50.1 |
75.2 |
100 |
CP, g/kg# |
186 |
175 |
164 |
153 |
142 |
Period 50 – 100 kg (g/kg DM) |
|||||
Fish meal |
80 |
60 |
40 |
20 |
0 |
Soybean meal |
80 |
80 |
80 |
80 |
80 |
Maize |
390 |
390 |
390 |
390 |
390 |
Cassava root meal |
242 |
217 |
193 |
168 |
144 |
Rice bran |
200 |
200 |
200 |
200 |
200 |
PECP |
0 |
45 |
89 |
134 |
178 |
Salt |
3 |
3 |
3 |
3 |
3 |
Premix |
5 |
5 |
5 |
5 |
5 |
CP, g/kg DM |
157 |
157 |
156 |
155 |
155 |
PECP, % of CP |
0.0 |
7.76 |
15.4 |
23.3 |
31.1 |
PECP, % of FM |
0 |
25.3 |
50.1 |
75.4 |
100 |
CP, g/kg# |
157 |
144 |
132 |
119 |
107 |
# Excluding CP (crude protein) from PECP |
Feed offered and refused was recorded daily. The pigs were weighed every month.
Response trends in feed intake, growth rate and feed conversion (Y) were related to percent replacement of fish meal protein by PECR crude protein (X). Comparison of results for the zero and 25% fish meal protein replacement levels were by the GLM option in the ANOVA program of the Minitab (2016) software. Sources of variation were treatments (CPO versus CP25), replicates and error.
The crude protein of the diets was in the range of 15 to 17% in DM in both periods (Table 6). By contrast, the content of "true" protein decreased linearly from 15 to 11% in DM as the fish meal was replaced by PECP. However, this analysis does not take account of the protein that could have been synthesized by bacteria in the pig intestine using recycled or dietary NPN as the source of ammonia (Colombus et al 2014). Thus the apparent reduction in true protein indicated by analysus of the feed may not be a true rrepresentation of the situation at the sites of metabolism.
Table 6. Mean values for crude and true protein the diets used in periods 1 and 2 | ||
Protein in DM, % | ||
Crude | True | |
Period 1 | ||
CP0 | 15.5 | 14.5 |
CP25 | 17.2 | 14.4 |
CP50 | 15.6 | 14.2 |
CP75 | 15.9 | 11.9 |
CP100 | 15.1 | 11.5 |
Period 2 | ||
CP0 | 15.6 | 14.7 |
CP25 | 16.9 | 13.7 |
CP50 | 16.1 | 13.5 |
CP75 | 15.1 | 13.0 |
CP100 | 15.2 | 11.3 |
Overall, there were linear depressions in DM intake and live weight gain as protein from PECP replaced that from fish meal (Table 7; Figures 5 and 6). There was no consistent trend in the data for DM feed conversion as fish meal protein was replaced by PECP protein (Figure 7).
Table 7. Mean values for changes in live weight, DM intake and DM feed conversion of pigs fed diets in which protein from protein-enriched cassava pulp replaced protein from fish |
||||||
PECP protein replacing fish meal protein, % |
||||||
0 |
25 |
50 |
75 |
100 |
SEM |
|
Init wt, kg |
24.2 |
24.1 |
24.1 |
24.1 |
24.1 |
0.563 |
Final wt, kg |
86.3 |
85.1 |
78.2 |
75.8 |
69.6 |
3.87 |
Daily gain, g/d |
702 |
688 |
608 |
584 |
518 |
41.2 |
DM intake, g/d |
1.98 |
1.75 |
1.75 |
1.86 |
1.43 |
0.126 |
DM conversion |
2.82 |
2.53 |
2.93 |
3.19 |
2.81 |
0.182 |
Figure 5. Replacing fish meal with protein-enriched cassava pulp reduces linearly the DM feed intake |
Figure 6. Replacing fish meal with protein-enriched cassava pulp reduces linearly the live weight gain |
Figure 7. There was an indication of better feed conversion by replacing 25% of the fish meal protein with protein from protein-enriched cassava pulp but at higher levels of replacement feed conversion was worse |
When the comparison of treatments was restricted to 0 and 25% replacement of fish meal protein (equivalent to raising the content of PECP from 0 and 4.5% of the diet DM), DM intake was reduced, live weight gain did not differ (p=0.59), and DM feed conversion tended (p=0.08) tended to be improved (Table 8).
Table 8. Effect of low level of PECP on DM intake, growth rate and feed conversion of growing-fattening pigs |
||||
PECP, % in diet DM |
||||
0 |
4.5 |
SEM |
p |
|
LW gain, g/d |
702 |
688 |
16.6 |
0.59 |
DMI, g/d |
1.98 |
1.75 |
0.036 |
0.01 |
DM conv. |
2.82 |
2.53 |
0.09 |
0.08 |
These results are broadly similar to those reported by Vanhnasin and Preston (2016) and Sengxayalth and Preston (2017b), namely that when fed as only a small proportion (4.5%) of the diet DM, PECP had marginal beneficial effects on pig performance but at higher levels there was a marked deterioration in both growth rate and feed conversion.
As only about 50% of PECP nitrogen was present as true protein (Experiment 1), the complete replacement of fish meal by PECP meant that the NPN component of the diet increased to approximately 8 g N/kg of diet DM. If this NPN was in the form of NH4+-N, it poses the question: was this sufficient to cause sub-acute ammonia toxicity? And if this was so, is this the explanation for the deterioration in growth performance as the proportion of PECP in the diet was increased from 4.5 to 15-17% of diet DM?
This research was supported by the Sida-Mekong Basin Animal Research Network (MEKARN II). We thank researchers in the Department of Anatomy and Physiology, and laboratory staff of the Faculty of Animal Sciences, Hue University of Agriculture and Forestry for assistance in carrying out the study and chemical analyses.
AOAC 1997 Association of Official Analytical Chemists International. Official Methods of Analysis. 16th Edition, AOAC, Arlington. USA
Aro S O 2008 Improvement in the nutritive quality of cassava and its by-products through microbial fermentation. African Journal of Biotechnology. Volume 25: 4789-4797.
Brook E J, Stanton W R and Wallbridge A 1969 Fermentation method for protein enrichment of cassava. Biotechnology and Bioengineering. Volume11: 1271-1284.
Colombus D A, Lapierre H, Htoo J K and de Lange C FM 2014 Nonprotein nitrogen is absorbed from the large intestine and increases nitrogen balance in growing pigs fed a valine-limiting diet. The Journal of nutrition 144 (5), 614-620
Millbank J W 1969 Nitrogen Fixation in Moulds and Yeasts—a Reappraisal. Archives of Microbiology. Volume 68, Issue 1, pp 32-39. 27.
Nwe Ni Win Htet, San San Yu and Zaw Ko Lat 2013 Ammonium Accumulation in Culture Broth of Soil Yeast Isolates and Their Effects on Sorghum Plant. Journal of Scientific and Innovative Research. Volumr 2, Issue 2 http://www.jsirjournal.com/Vol2Issue2020.pdf
Nwafor O E and Ejukonemu F E 2004 Bio-conversion of cassava wastes for protein enrichment using amylolytic fungi: a preliminary report. Global Journal of Pure Applied Sciences. Volume 4: 505-507.
Obadina A O, Oyewole O B, Sanni L O and Abiola S 2006 Fungal enrichment of cassava peels proteins. African Journal of Biotechnology. Volume 5. No. 3: 302-304.
Oboh G and Akindahunsi A A 2005 Nutritionnal and toxicological evaluation of Saccharomyces cerevisiae fermented cassava flour. Journal of Food Composition and Analysis. Volume 18: 731-738.
Thongkratok R, Khempaka S and Molee W 2010. Protein enrichment of cassava pulp using microorganisms fermentation techniques for use as an alternative animal feedstuff. Journal of Animal and Veterinary Advances 9(22), 2859-2862.
Manivanh N, Preston T R and Thuy N T 2016 Protein enrichment of cassava (Manihot esculenta Crantz) root by fermentation with yeast, urea and di-ammonium phosphate. Livestock Research for Rural Development. Volume 28, Article #222. http://www.lrrd.org/lrrd28/12/noup28222.html,
Sengxayalth Phoutnapha and Preston T R 2017a Fermentation of cassava pulp with yeast, urea and diammonium phosphate (DAP). Livestock Research for Rural Development. Volume 29, Article #177. http://www.lrrd.org/lrrd29/9/pom29177.html
Sengxayalth Phoutnapha and Preston T R 2017b Effect of protein-enriched cassava pulp on growth and feed conversion in Moo Laat pigs. Livestock Research for Rural Development. Volume 29, Article #178. http://www.lrrd.org/lrrd29/9/pom29178.html
Vanhnasin P, Manivanh N and Preston T R 2016 Effect of fermentation system on protein enrichment of cassava (Manihot esculenta) root. Livestock Research for Rural Development. Volume 28, Article #175. http://www.lrrd.org/lrrd28/10/vanh28175.html
Vanhnasin P and Preston T R 2016 Protein-enriched cassava (Manihot esculenta Crantz) root as replacement for ensiled taro (Colocasia esculenta) foliage as source of protein for growing Moo Lat pigs fed ensiled cassava root as basal diet. Livestock Research for Rural Development. Volume 28, Article #177. http://www.lrrd.org/lrrd28/10/vanh28177.html
Received 25 July 2017; Accepted 21 August 2017; Published 1 September 2017