Back to contents

Workshop-Seminar on Making Better Use of Local Feed Resources. SAREC-UAF, January 2000

Evaluation of water spinach as a protein source for BaXuyen and Large White sows

Le Thi Men, Brian Ogle* and Vo Van Son

Department of Animal Science, College of Agriculture,
 CanTho University, Vietnam
*Swedish University of Agricultural Sciences, Department of Animal Nutrition and Management, PO Box 7024, 750 07 Uppsala, Sweden

 

Abstract

Two on-station trials and a feeding experiment were conducted at the experimental pig farm, CanTho University, to determine the intake and the apparent digestibility of fresh chopped and whole water spinach (WS) by pregnant sows. The effect on reproductive performance of feeding WS to local Ba Xuyen and Large White sows was also evaluated.   The first (intake) trial included 6 multi-parity local Ba Xuyen (6th litter, average live weight 180 kg) and 6 Large White (5th litter, average live weight 202 kg) sows. The sows were mated by AI with the same Large White boar. The trial was designed as 3 multiple Latin Squares with two diets for each breed as follows:

·         1 kg DM concentrate/day plus whole water spinach (WWS) ad libitum

·         1 kg DM concentrate/day plus chopped water spinach (CWS) ad libitum 

For the Large White and Ba Xuyen sows, respectively, results were: fresh WS intake per sow/day, 7.3 and 6.1 kg (604 and 504 g DM) and % of DM intake from WS 37.6 and 33.4 % (P<0.01). Consumption of fresh CWS and WWS by pregnant sows was 6.9 and 6.4 kg/sow/day, respectively (DM intake 575 and 533 g/sow/day, respectively) and  % of DM intake from WS  36.4 and 34.7 % (P<0.01) for CWS and WWS,  respectively.

The second (digestibility) trial included 3 local Ba Xuyen and 3 Large White sows in the first month of pregnancy. The sows were kept in individual cages. The trial was a two-way random complete design with breed of sow as one factor and diet as the other factor. Each breed was given 3 diets in a 3*3 Latin Square design. The experimental diets were as follows: 

Baxuyen sows:

·         1.7 kg DM/day of a control diet (a concentrate based on local feed resources)

·         1.1 kg DM/day of the control diet plus 0.5 kg DM whole water spinach (WWS)

·         1.1 kg DM/day of the control diet plus 0.5 kg DM chopped water spinach (CWS)

 Large White sows:

·         2 kg DM/day of the control diet

·         1.3 kg DM/day of the control diet plus 0.6 kg DM  WWS

·         1.3 kg DM/day of the control diet plus 0.6 kg DM  CWS 

Apparent digestibility  (%) of DM by pregnant Large White and Baxuyen sows of fresh WS, respectively, was 77.7 and 65.1% and CP 77.9 and 73.8 % (P<0.01). The apparent digestibility (%) for CWS and WWS, respectively, was for DM 68.4 and 60.0, OM 76.0 and 71.5, CF 71.6 and 70.8 and for NFE 91.3 and 85.6 (P<0.01)

The reproduction experiment included 6 Large White and 6 Baxuyen sows, and was designed as a 2*2 factorial with two breeds and two diets: 

Gestation for Baxuyen:
            ·         1.7 kg DM daily of a control diet based on local feed resources
            ·         1.1 kg DM/day of the control diet plus 0.5kg DM  of CWS

Gestation for Large White:

·         2 kg DM/day of the control diet

·         1.3 kg DM/day of the control diet plus 0.6kg DM  of CWS 

In lactation the same concentrate was given as in gestation, the daily amounts depending on litter size. Between 4-6 kg (fresh weight ) of WS was given to the sows on the experimental diet.

The total litter weight at birth (12.9 vs 10.6 kg) was higher (P=0.037) for CWS sows. Litter size (9.5 vs 8.5) (P=0.053) and total litter weight at weaning (76.1 vs 68.4 kg)  (P=0.02) were higher for the sows given CWS. Total feed DM/kg piglet at birth (14.4 vs 16.6), at 21 days (5.1 vs 5.8) and 35 days (3.9 vs 4.4) were  lower (P<0.05) for the CWS sows. The gross weight gain of the sows in pregnancy for the control group was 19.3 %, and 21.9 % for the CWS diet (P<0.05). The weight loss during the lactation period was 10.3 % for the control diet and 12.0 % (NS) for the CWS diet.  Mean piglet liveweight at birth was 1.4 vs 1.0 kg, total litter weight 14.0 vs 9.5 kg and total feed DM/kg piglet 13.3 vs 16.8 (P<0.01) for the Large White and Baxuyen sows, respectively. The average gross gain in gestation of the Large White sows was higher (43.5 vs 35.4 kg) (P<0.01) and weight loss in lactation lower (9.5% vs 12.9%) (P<0.01) compared with Baxuyen sows. 

Key words: Large White, Baxuyen,  sows, water spinach,  intake, digestibility, reproductive performance.

Introduction

The Mekong delta is located in the far South of Vietnam, with 21% of the population and 12% of the total area of the country. It is one of the major rice growing areas of Vietnam, with about 50% of the total rice production of the whole country (Vo Hoang Xinh 1998). It is a 3.9 million hectare lowland region of alluvial, acidic and saline soils watered by the Mekong river and its branches, and with an annual rainfall of around 1700mm (Bui Xuan Men 1996). There are large quantities of rice by-products available such as rice bran and broken rice, which have traditionally been important feeds for livestock, particularly pigs. The delta region also has extensive water surfaces, which give opportunities for the widespread, year-round cultivation of water plants, especially water spinach (Ipomoea aquatica).

Water spinach is a variable water and marsh plant with creeping, hollow, water filled stems and shiny green leaves. It is cultivated for food and used as pig and cattle feed throughout Southeast Asia. The fresh leaves and stems of water spinach have a crude protein content of around 28% in DM (variable). Crude fibre and ash concentrations are around 12% and 19% of DM, respectively (Göhl 1981). Water spinach is readily eaten by pigs and is a locally available feed resource throughout the Mekong delta of Vietnam. It has been used successfully to replace part of the protein in a diet of sugar cane juice for breeding sows (Le Thi Men and Bui Hong Van 1993). 

An indigenous pig breed, the Baxuyen, which has traditionally been reared in the remote areas of the Delta, has a number of advantages in these areas as it survives well under the harsh environmental conditions and on poor quality feeds. It is used for breeding, and the sows are usually mated with an exotic boar to produce crossbred fattening pigs. The improved breeds, especially the Large White, have adjusted fairly well to the local conditions, and are raised on family and semi-industrial farms. 

The aims of the study were:

·         To determine the intake and apparent digestibility of chopped and whole water spinach (WS) by Baxuyen and Large White sows.

·         To evaluate the effects on reproductive performance of feeding WS to pregnant Baxuyen  and Large White sows.

 

Materials and methods

Effect of chopping water spinach on intake and total tract digestibility by pregnant Baxuyen and Large white sows

Experimental design and pigs

An intake study with pregnant sows  fed whole and chopped water spinach was carried out at the experimental pig farm of CanTho university in late January 1998. The experimental animals were 6 multi-parity local Baxuyen (6th litter, average live weight 180 kg) and 6 Large White (5th litter, average live weight 202 kg) sows. The sows were mated with the same Large White boar by AI, and were then placed in individual concrete-floor pens. The trial was designed as 3 multiple Latin Squares with two diets given to each breed. The dietary treatments were:

·         1 kg DM concentrate/day plus whole water spinach (WWS) ad libitum

·         1 kg DM concentrate/day plus chopped water spinach (CWS) ad libitum. 

 

The WS was chopped by hand into 4-5cm lengths. The trial was divided into 2 periods, including 2 days for adaptation and 2 days for collection of data per period. The amounts of water spinach offered and refused were recorded to estimate the intake, and the chemical composition (DM, crude protein, crude fibre, ether extract and ash) was determined by standard methods (AOAC 1988). 

Digestibility trial with sows

Based on the results of the intake of water spinach intake, diets were formulated for the digestibility trial, which started in February 1998. The experimental animals were 3 local Baxuyen and 3 Large White sows in the first month of pregnancy. The sows were kept in individual cages. The trial was a two-way random complete design with breed of sow as one factor and diet as the other factor. Each breed was given 3 diets in a 3*3 Latin Square design. The experimental diets were as follows:  

Baxuyen sows:

• 1.7  kg DM/day of a control diet (a concentrate based on local feed resources)
•  1.1 kg DM/day of the control diet plus 0.5 kg DM whole water spinach (WWS) fed separately
• 1.1 kg DM/day of the control diet plus 0.5 kg DM chopped water spinach (CWS) fed separately  

 Large White sows:

• 2 kg DM/day of the control diet
• 1.3 kg DM/day of the control diet plus 0.6 kg DM whole water spinach (WWS) fed separately
• 1.3 kg DM/day of the control diet plus 0.6 kg DM chopped water spinach (CWS) fed separately 

The daily allowances were based on the mean live weights of the two breeds. The concentrate was given first, twice a day, and the fresh chopped or whole WS was available throughout the day. The experimental period was 7 days for adaptation and 5 days for collection of faeces. During the experimental period intakes of the control diet and WS were measured daily by weighing amounts offered and refused. Samples of water spinach, including the refusals, were analysed immediately for DM in a microwave oven. The sub-samples for later analysis were prepared by drying in an oven at 60°C, ground and stored in a fridge at 4°C. The total amount of faeces was collected and weighed every day. An aliquot of 10% faeces/sow was stored in a freezer at –10°C and at the end of each period was mixed thoroughly, and representative samples taken for analysis. DM and N were determined on the fresh samples. A sub-sample of faeces was dried at 60°C (drying oven) for other parameters such as ether extract, crude fibre and ash, determined by AOAC. methods (1988).  

Effect of feeding WS on the reproductive performance of Baxuyen (BX) and Large White (LW) sows

Experimental design and pigs

The experiment included six LW and six BX sows and was designed as a 2*2 factorial with two breeds and two dietary treatments. The sows were allocated at random to one of two treatments three weeks after mating with the same Yorkshire boar by AI.

Diets and feeding

The daily allowances were based on the mean live weights of the two breeds.

Ba Xuyen:

• 1.7 kg DM daily of a conventional diet based on local feed resources (control)
• 1.1 kg DM/day of the control diet plus 0.5kg DM chopped water spinach (CWS) fed separately

Large White:

• 2 kg DM/day of the control diet
• 1.3 kg DM/day of the control diet plus 0.6 kg DM chopped water spinach (CWS) fed separately
  
  The sows were kept in individual concrete-floor pens. Three weeks after mating, the sows were fed the experimental diets, and were given the restricted amount of concentrate twice daily in the morning and afternoon, and then fresh clean chopped water spinach three times per day. The amount of feed and WS consumed was recorded. The lactating sows were fed increasing amounts of concentrate according to the litter size (2 kg/sow plus 0.5 kg/piglet). Sows were given only very small amounts of WS (1-2 kg/d) for the first two weeks after farrowing, as they were reluctant to consume larger amounts. Then they were given 4-6 kg of fresh WS divided between three occasions each day, the amount being restricted because of fears of piglet diarrhoea. Piglets were given the same complete creep diet for each treatment. 

The parameters recorded were:

·         Litter size at birth, 21 and 35 days

·         Litter weight at birth, 21 and 35 days

·         Piglet mortality

·         Total feed DM per kg piglet at birth, 21 and 35 days

·         Sow weight changes          

Statistical analysis

Data were analysed by ANOVA using the General Linear Model (Minitab Statistical Software version 12. 

Results and discussion

Chemical composition of WS and dietary ingredients

Analytical data for the diet ingredients are in Table 1. 

Amino acid analysis showed that most of the important essential amino acid concentrations in WS are higher than those of the Ideal Protein (ARC 1981), and superior to the balance of amino acids in soya bean meal, particularly with respect to methionine, threonine, isoleucine and valine (Table 2).  

Table 2:  Analysed amino acid composition (expressed as percent of  lysine = 100) of the water spinach, soybean meal and concentrate used, and the Ideal Protein for gestating sows (ARC 1981)
Amino acid	Water spinach*	Soybean meal*	Concentrate*	Ideal  Protein
Lysine    Methionine + cystine Methionine Tryptophan Threonine Leucine  Isoleucine              Valine Histidine Phenylalanine+tyrosine	100   42 - 106 178 106 122 46 193	100   23 - 53 84 65 75 37 175	100   42 - 76 150 75 105 50 179	100 67 - 16 84 74 86 107 30 77

*  Analysed values, National Institute of Animal Husbandry, Vietnam (1995)

Water spinach intake and total tract digestibility by pregnant Baxuyen and Large White sows

Intake of WS by pregnant sows

Effect of breed: Fresh WS intake (7.3 vs 6.1 kg per sow/day for the Large White and Baxuyen sows, respectively) was different (P<0.01) between the 2 breeds (Table 3). Even adjusted for live weight, the amounts consumed by the Large White sows were significantly higher than for the Baxuyen breed (3.6 vs 3.4 kg fresh weight/100 kg live weight), which implies that this improved breed has successfully adjusted to the environment of the region. The sows consumed around 6-7 kg (500-600 g DM) of fresh WS/day, plus 1 kg DM of concentrate.  The proportions of the total DM intake from WS were  higher (P<0.01) for the Large White than for the Baxuyen sows (37.6 vs 33.4 %). These results can be largely explained  by the fact that the Large White sows, with a higher live weight (202 kg) than the Baxuyen (180 kg) sows, were fed the same limited level of concentrate (1 kg DM/d). The high intakes of WS showed that it was very palatable for the sows.

Table 3: Fresh water spinach intakes of pregnant Baxuyen and Large White sows
	Baxuyen	Large White	SE	P
Fresh WS, kg/sow/d Fresh WS, kg/100kg LW DM from WS, g/d %of DM intake from WS Total DM intake, g/sow/d	6.1 3.36 504 33.4 1504	7.3 3.61 604 37.6 1604	0.079 0.049 6.537 0.284 6.539	0.001 0.002 0.001 0.01 0.001

Effect of chopping WS:  Consumption of fresh CWS was (P<0.01) higher than of WWS (6.9  vs 6.4 kg/sow/day) and DM intake of CWS  (575 vs 533 g/sow/day) and % of DM intake from WS (36.4 vs 34.7 %) were also higher than for WWS (P<0.01), probably due to the decreased bulk of the CWS (Table 4). This was not unexpected, as it is well known that reducing particle size increases the consumption of high fibre feeds.  Although the relative amount of leaves and stem in the refusals were not weighed, it was observed  that the proportion of  stems in the refusals  was higher than in the whole plant (Table 5). This was confirmed by the chemical analysis, as the crude protein of the CWS refusals (20.5%) was lower than in the whole plant (26.7 %), indicating that the sows had selected the leaves before the stems. However, it was observed that a few individual sows preferred the crisp stems.    

 

 

 

Total tract digestibility of water spinach by pregnant sows

Effect of breed: Apparent digestibility (%) of DM (83.0 vs 80.0) and crude protein  (84.8 vs 81.0) was  higher (P<0.05) for Large White sows than for the Baxuyen breed, but the values for OM, crude fibre and NFE digestibility were not significantly different between the two breeds.  

When the digestibility of the WS was calculated by difference…..  

100*Digcoeff. mixed diet = (100-X)*Digcoeff. concentrate  + (X)*Digcoeff of WS 

where X is percent of WS in the diet (DM basis)  

...the apparent digestibility coefficients of water spinach were DM (77.7 vs 65.1%), and crude protein (77.9 vs 73.8), for Large White and  Baxuyen, respectively,  higher P<0.01) for the Large White breed (Table 7).    

 

 

Effect of diet: The total tract apparent digestibilities of DM (68.4 vs 60.0 %), OM (76.0 vs 71.5 %), CF (71.6 vs 70.8 %) and NFE (91.3 vs 85.6 %) were higher (P<0.01) in CWS than in WWS (Table 8).   Crude protein digestibility was similar for the two physical forms of WS. The generally higher apparent digestibility of the nutrients in CWS was probably a result of the physical disruption of the cell walls of the plant by chopping, and the increased surface area exposed to enzyme action. Bitende (1994) reported that physical preparation of feeds to improve digestibility includes chopping, chaffing, grinding or cooking. Chopping and grinding are common processes of communition for roughage. Coarse fibrous particles may be retained in the caecum longer than more finely ground particles of the same fibre source. Unchopped roughage can also result in increased DM digestibility due to selection of better components of the diet i.e. leaves versus stems. However, in our second trial the sows were offered the same limited amount (30% of dietary DM) of CWS and WWS, so there were no refusals by the sows and no selection of the leaves in the WWS treatment. Consequently most nutrient digestibilities were higher for the chopped water spinach than for the whole plant.   

Effect of feeding water spinach (CWS) on the reproductive performance of Baxuyen and Large White sows

Chemical composition of diets and ingredients

The basal diet included conventional feeds, such as rice-by products (broken rice and rice bran), soybean extract, fishmeal, bone meal and trace minerals, plus a vitamin premix. It had an average DM content of 88 %, and nutrient concentrations as % of DM were: crude protein 15.5; ether extract 8.7; crude fibre 5.1; Ca 0.80; P 0.60 and  ME  13.0 MJ/kg DM (Table 6). The WS had an average DM content of 8.3±1.1 %, and nutrients as % of DM were: crude protein 26.7±2.2; crude fibre 15.5±1.2 and ash 14.1±1.3. Ca and P contents were 0.90 and 0.50 %, respectively.(Table 1). Analysis of the amino acids showed that most of the important essential amino acid concentrations in WS were higher than those of the Ideal Protein (ARC 1981). The balance of essential amino acids in the basal diet was also similar to that of the Ideal Protein, except that threonine and isoleucine levels were lower (Table 2).  

Feed intake

Pregnancy: Most of the pregnant sows consumed CWS without any refusals, except that two Baxuyen sows left around 250 g fresh CWS/day in the week before farrowing. For each breed CWS replaced 30% of the DM of the basal diet. DCP (g/day) from CWS was 42% and 43% of the total DCP intake for the Baxuyen and Large White sows, respectively. DCF (67.0 vs 107.3 g/day), DCP (250 vs 275 g/day) and lysine (15.8 vs 17.4 g/day) intakes for the control diet were lower than for the CWS treatment (Table 9).     

Table 9:  Daily feed and nutrient intake of control and chopped water spinach diets (CWS) by pregnant and lactating  Baxuyen and Large White sows
Parameter	Baxuyen				Large White
	Pregnant		Lactating		Pregnant		Lactating
	Ctrl	CWS	Ctrl	CWS	Ctrl	CWS	Ctrl	CWS
Concentrate, kg/d DM, kg/d WS (fresh), kg/d DM, kg Total DM, kg/d ME, MJ/d CP, g/d CP (WS), g/d DCP, g/d DCP (WS), g/d DCP(WS), % of DCP Lys, g/d Met + Cys, g/d Thre, g/d Ca, g/d P, g/d	1.95 1.74 - - 1.74 22.6 269 - 228 - - 14.8 9.0 11.1 13.6 10.3	1.30 1.16 5.75 0.49 1.65 20.2 312 132 247 105 42.0 16.0 8.6 13.9 13.5 9.3	3.41 3.03 - - 3.03 39.3 469 - 397 - - 25.8 15.8 19.4 23.6 17.9	3.30 2.94 4.30 0.4 3.34 42.4 563 108 446 85.6 19.0 30.0 17.4 24.1 26.5 19.4	2.21 1.97 - - 1.97 25.6 305 - 271 - - 16.8 10.2 12.6 15.4 11.6	1.50 1.34 7.00 0.59 1.93 23.7 367 159 304 132 43.0 18.8 10.0 16.4 15.8 10.9	3.87 3.45 - - 3.45 44.8 534 - 475 - - 29.3 17.9 22.1 26.9 20.4	3.47 3.09 4.91 0.42 3.59 44.5 591 113 490 93.9 19.0 31.5 18.3 25.4 27.9 20.3

Lactation: There were no refusals noted when 15% CWS (DM basis) substituted for the basal diet.   

 

 

 

 Reproductive performance of the sows

Effect of diet: Litter weights at birth, 21 and 35 days were higher (P=0.037) for sows fed the CWS (Table 10). Vestergaard and Danielsen (1998) stated that there are nutritionally induced effects on reproductive performance, especially by vitamin A, which is very important for fertility. Levels of vitamin A are high in green plants, and it is possible that active levels in the control diet were too low, as this vitamin is rapidly destroyed in hot, humid conditions. Comparing the colostrum and milk from ten sows on pasture and ten sows in dry lot, it was reported that mean vitamin A levels (as retinol) tended to be higher for sows in pasture (Bowland et al 1949). 

 Welfare indicators when pregnant sows are fed only concentrate diets are often very low, as the sows are usually hungry and this causes stress. By providing WS, the bulkiness of the diet is increased and the sows should be less hungry and therefore less stressed. In the present study, the control sows were extremely noisy and obviously stressed before feeding time, while the sows given WS were generally calm and contented. These observations agree with Mroz (1986), who concluded that feeding diets with a high fibre content will increase the time needed to consume the daily allowances and decrease competition between pen mates. The total gross weight gain of the sows in pregnancy was lower (P<0.05) on the control  than on the WS diet but the mean net gains were not different. The weight loss during the lactation period was 10.3% for the control diet and 12% for the WS treatment (NS). This agrees with the conclusion of Vestergaard and Danielsen (1998) that high fibre diets increase the gut fill of the sows, leading to a large gross body-weight gain during the pregnancy period, but also to a larger weight loss at farrowing, when the food is changed to a standard lactation diet.  In our study, WS was not provided immediately post-farrowing. The higher total litter weight at birth of the CWS sows would also have contributed to their higher gross pregnancy weight gains.  

Effect of breed: Litter size at birth and weaning was similar for the two breeds. However, the Large White sows produced heavier piglets than the Baxuyen breed (Table 11). The higher mean piglet birth weights would be mainly the result of the higher (by about 10%) mature body weights of the Large White breed. As a result of the early commencement of eating of the crossbred piglets (LW*BX) and their higher creep feed consumption, piglet and total litter weights between two breeds were similar at the end of lactation. Net pregnancy live weight gains were almost identical for the two breeds, but lactation losses were higher for the Large White sows (12.9% of body weight) than for the Baxuyen sows (9.5%) (P<0.01). 

Conclusions

·         Diets including fresh water spinach can be used effectively by breeding sows. 

·         Fresh chopped WS can replace 30% of the DM of concentrate diets for gestating sows and 15% of the diet of lactating sows of both local and exotic breeds, resulting in somewhat improved reproductive performance and welfare.
 

Acknowledgements

The authors are grateful to the Swedish International Development Authority (SIDA/SAREC) for financing this study. We are also grateful to the College of Agriculture, Can Tho University for use of their facilities.Thanks also to my colleagues and my assistants, Khang, Loan, Tam, Hau, Phuong, Duc, Huong and Binh for their help.

References

AOAC 1988 Official methods of analysis, 14th edition Association of Official Analytical Chemists, Washington, DC    

ARC 1981 The Nutrient Requirements of Pigs. Agricultural Research Council, Commonwealth Agricultural Bureaux,  Farnham Royal, UK 

Bitende  S N 1994 The potential of Acacia tortilis and Sesbania sesban leaves as livestock feeds in smallholder farming systems. MSc Thesis. Swedish University of Agricultural Sciences Department of Animal Nutrition and Management  Uppsala 1994  

Bowland J P 1949 In Swine Nutrition. Editors:  Miller E R, Ullrey D E and Lewis A J. Butterworth-Heinemann

Bui Xuan Men 1996 Improvement of local duck production systems in the Mekong Delta, Vietnam. MSc Thesis Swedish University of Agricultural Sciences Department of Animal Nutrition and Management  Uppsala 1996  

Göhl B 1981 Tropical  Feeds. FAO Animal Production and Health Series. No 12 

Le Thi Men and Bui Hong Van 1993 Evaluation of diets based on simulated sugarcane juice and water spinach (Ipomoea aquatica) for breeding sows. In Sustainable Livestock Production on Local Feed Resources.  Ho Chi Minh City Vietnam 1994

Mroz Z 1986  Forages for Growing Pigs. Cited by Andersson C 1997. Doctoral thesis Swedish University of Agricultural Sciences  Department of Animal Nutrition and Management  Uppsala 1997  

National Institute of Animal Husbandry 1995 Composition and nutritive value of animal feeds in Vietnam. Hanoi Vietnam

Vestergaard  E M and Danielsen V 1998 Dietary fibre for sows: effects of large amounts of soluble and insoluble fibres in the pregnancy period on the performance of sows during three cycles. Animal Science 68: 355-362

Vo Hoang Xinh 1998  People´s  Committee  of CanTho Province Vietnam