Livestock Research for Rural Development 6 (1) 1994

Citation of this paper

Raw palm oil as the energy source in pig fattening diets and Azolla filiculoides as a substitute for soya bean meal

Alvaro Ocampo Durán

Universidad de los Llanos, Facultad Medicina Veterinaria y Zootecnia, Instituto de Investigaciones de la Orinoquia Colombiana, AA 2621, Villavicencio, Colombia


An experiment was done to demonstrate: (i) the use of raw palm oil as the basal diet for fattening pigs; and (ii) the potential of the water fern (Azolla filiculoides) to replace part of the soya bean supplement. The raw palm oil was fed at levels of 500 and 800 g/pig/day in the growing (20-60 kg) and finishing (60-90 kg) phases respectively. Fresh azolla was offered at levels of 0, 1.7, 3.48 and 5.21 kg/day for a theoretical replacement rate of the soya bean protein of 0, 10, 20 and 30% (total protein from soya bean meal and azolla was predicted to be 200 g/day). Rice polishing was given at levels of 100 g/day in the growing period and 150 g/day in the finishing period. The most appropriate harvesting interval for azolla was six days leaving 25% of the biomass as seed material. Production rates were in the range 120 to 200 g/m2/day. Dry matter content was 6% and there was 26% N x 6.25 in the dry matter. A mixture (50/50) of cattle manure and poultry excreta was found to be the best fertilizer at rates of 500 g/m2 every 18 days. Production rate was poorest in the dry season and best during the rainy season. Intakes of azolla during the growing phase were only 51, 34 and 26% of projected amounts for the 10, 20 and 30% replacement rates, providing protein intakes of 212, 202, 184 and 167 g/pig/day. During the finishing phase, intakes of azolla increased to reach 100, 97 and 69% of projected levels with total protein intakes of 212, 212, 212 and 194 g/pig/day for the four treatments. The daily liveweight of the pigs was: 0.450, 0.482, 0.457 and 0.407 (SE±0.13; Prob=0.80 for the growing phase; and 0.654, 0.692, 0.666 and 0.528 kg/day (SE±0.18; Prob=0.12) in the finishing phase. Overall the weight gains were: 0.526, 0.561, 0.535 and 0.452 kg/day for the 0, 10, 20 and 30% replacement rates of azolla for soya bean. Feed dry matter conversion (overall) was 2.1, 1.98, 2.0 and 2.2. Optimum replacement rates of azolla were considered to be 10% and 20% in the growing and finishing phases respectively. A demonstration trial was carried out on a commercial farm with 169 pigs, divided arbitrarily into 4 groups, using a ration (g/day) of raw palm oil 500, soybean supplement 500 and rice polishing 500. Liveweight gains were 0.722, 0.628, 0.524 and 0.464 kg/day and dry matter feed conversion 1.8, 2.0, 2.4 and 2.8 for the four groups.

KEY WORDS: Azolla filiculoides, raw palm oil, pigs, fattening


For a livestock production system to be sustainable, the first criteria is that it should be based on the resources available in the country, and preferably on the farm itself. Secondly there should be a balance between the crops and the livestock so that these activities are complementary and, if possible, synergistic (Preston and Murgueitio 1992).

The African Oil palm tree can be considered as a major resource of Colombia with approximately 120,000 ha cultivated with this crop, of which 80% are currently in production. However, the barrier to its use has been the belief that the optimum amount of lipids in a diet for monogastric animals is within the range of 2-6% (air dry basis) (Hutagalung 1987). Earlier work at our University showed that a byproduct from the extraction process -- the oil-rich fibre -- could replace completely sorghum grain as the basis of the diet for fattening pigs (Ocampo et al 1990a; Ocampo et al 1990b; Ocampo 1992). The next stage was the demonstration on commercial farms that the crude palm oil could also be used as the basis of the pig diet for all phases of the production cycle (Ocampo A 1993, unpublished data; Rodriguez Lylian, Cuellar Piedad and Madriñan A, 1993 in preparation).

An advantage from using oil as the energy resource is its high caloric density and the absence of fibre. This creates opportunities for using unconventional sources of protein such as tree leaves (Preston and Murgueitio 1989) and water plants (Van Hove 1986; Lumpkin and Pluckett 1982) whose fibre content would normally be a limitation on conventional cereal grain diets.

The water fern Azolla (Azolla spp) has been used successfully as a partial replacement for soybean in pig fattening diets based on sugar cane juice (Becerra et al 1989). For pregnant pigs fed on cane juice it provided all the protein (Rodriguez Lylian, Cuellar Piedad and Madriñan A 1993, in preparation).

It was hypothesised therefore that Azolla could be equally complementary in diets based on crude palm oil.

Materials and methods

An experiment was carried out in the experimental farm of the Universidad Tecnológica de los Llanos, Villavicencio, departamento Meta. The region is 300 m above sea level with a relative humidity of 70%, average temperature 27 oC and average rainfall 3500 mm/year.

Production of azolla

The species used was Azolla filiculoides which had been introduced into Colombia from Brazil. Three ponds were prepared each with a water surface area of 100 m² and a depth of 0.8 m. Water was continuously added to compensate for evaporation and filtering in the sub-soil. The azolla was harvested every six days, leaving 25% behind as seed material. Three fertilization programmes were used, each one with three treatments.

Programme 1 (6-29 November 1990; rainfall 276 mm):
  Cattle Poultry Calcium Triple
  manure* excreta* phosphate 15N:15P;15K
Pond 1 1 kg/m² 0 10 g/m² 0
Pond 2 0 0.250 kg/m² 10 g/m² 0
Pond 3 1 kg/m² 0.250 kg/m² 10 g/m² 0


* With applications every 6 days.


Programme 2 (30 November 1990 to 6 February 1991, rainfall 183 mm; 7 February to 30 March, rainfall 188 mm):
  Cattle Poultry Calcium Triple
  manure* excreta* phosphate** NPK
Pond 1 1 kg/m² 0 20 g/m² 10 g/m²
Pond 2 0 0.250 kg/m² 20 g/m² 10 g/m²
Pond 3 1 kg/m² 0.250 kg/m² 20 g/m² 10 g/m²


* With applications every 18 days.
** With applications every 6 days.


Programme 3 (April and May, rainfall 2,900 mm)
  Cattle Poultry Calcium Triple
  manure* excreta* phosphate** NPK
Pond 1 1 kg/m² 0 40 g/m² 10 g/m²
Pond 2 0 0.250 kg/m² 40 g/m² 10 g/m²
Pond 3 1 kg/m² 0.250 kg/m² 40 g/m² 10 g/m²


* With applications every 18 days.
** With applications every 6 days.


Once the Azolla was established, it was harvested every six days, taking away 75% of the biomass from the pond's surface; the remaining 25 % served as seed material for the next crop. The protein and dry matter content of the Azolla on each pond was determined.

32 crossbred pigs (Yorkshire and Landrace) pigs (17 females and 15 castrated males) with an average weight of 20 kg were used. At the beginning, the pigs were treated for intestinal parasites.

The trial was divided in two phases:
Growing phase (30 - 60 kg)
Fattening phase (60 - 90 kg).

There were four treatments (Table 1), each one with two replications (pens) and four pigs per pen. In the morning (7 am) the pigs were offered a mixture of the protein supplement, the rice polishing, and half the daily allowance of oil and of azolla. The remaining portion of azolla and oil was given in the afternoon (3 pm). Feed consumption was recorded daily; pigs were weighed every 14 days.

Table 1: The amounts of feed offered according to treatment (kg/animal/day).

Replacement of soya protein by azolla protein, %

  0 10 20 30
Growing phase (20-60 kg):        
Supplement* 0.500 0.450 0.400 0.350
Rice polishing 0.100 0.100 0.100 0.100
Crude palm oil 0.500 0.500 0.500 0.500
Azolla (fresh) 0.0 1.74 3.48 5.21
Finishing phase (60-90 kg):        
Supplement* 0.500 0.450 0.400 0.350
Rice polishing 0.150 0.150 0.150 0.150
Crude palm oil 0.800 0.800 0.800 0.800
Azolla (fresh) 0.0 1.74 3.48 5.21


* Contains (kg): Soybean meal 86; dicalcicum phosphate 10; mixture of vitamins and minerals 2; and salt 2.


Results and discussion

Production of azolla

The biomass production of the Azolla for each one of the programmes and fertilizer treatments is shown in Table 2.

Table 2: Biomass production of Azolla according to each programme and fertilizer treatment (g/m²/day).
  Pond 1 Pond 2 Pond 3
Programme 1 145 146 186
Programme 2 115 118 133
Programme 2* 158 148 200
Programme 3 106 155 185


* Production obtained after taking away the mud from bottom of the ponds.


Programme 1

The greatest production was obtained in pond 3 in which cattle and poultry excreta were combined. The azolla showed rapid recovery and there was an increase in the foliar area of the plant. At the beginning of the summer period (increased temperatures) the azolla changed colour from green to reddish-brown.

Programme 2

The organic fertilization interval was increased to 18 days to reduce the rate of accumulation of organic matter at the bottom of the pond, since this was apparently reducing the oxygen supply and increasing the production of CO2. The ponds were emptied and the organic material was taken away before recommencing the fertilization programme.

Programme 3

This programme was carried out in the rainy season when the plants assumed a green colour.

The protein content of the Azolla (% dry matter basis) was: 27.6 for pond 1, 24 for pond 2 and 27.1 for pond 3. The dry matter percentages were: 5.4, 6.5 and 7.0, respectively. There was no apparent effect of the NPK treatment (Programmes 2 and 3).

The production of Azolla obtained was in the range of 800 to 1,200 g/m² every six days.

The response to the organic fertilizer was greater when the mixture of poultry excreta and cattle manure was used. The application of chemical fertilizer was not justified.

There was a noticeable effect of the winter and summer periods over the production. The production was larger in the rainy period, due to the stimulation that this produces over the reproduction system of Azolla by fragmentation and also, due to the larger water interchange. During the summer period and due to the light effects the production decreases and sometimes the fern takes a reddish colour.

The dry matter content of the azolla was 6.1% and there was 26.2% of protein in the dry matter.

Pig performance

The data for the growing and finishing phases and for the overall period are in Tables 3, 4 and 5.

Table 3: Growth performance of the pigs during the growing phase (70 days).

Replacement of soya protein by azolla protein, %

  0 10 20 30 ±SE/Prob*
No. of animals 7 7 7 8  
Live weight (kg)          
Initial 28.8 29.4 28.8 28.9  
Final 60.3 63.2 60.8 57.4  
Daily gain 0.450 0.482 0.457 0.407 0.107/0.768
Feed intake, kg/day          
Supplement 0.500 0.450 0.400 0.350  
Azolla 0.0 0.88 1.20 1.38  
Oil 0.450 0.440 0.440 0.430  
Rice polish 0.100 0.100 0.100 0.100  
Total DM 0.90 0.97 0.95 0.90  
Conversion DM 2.2 2.0 2.0 2.2 0.57/0.818


* SE of means/Probability


Table 4: Growth performance of the pigs during the finishing phase (42 days).

Replacement of soya protein by azolla protein, %

  0 10 20 30 ±SE/Prob*
Live weight (kg)          
Initial 60.3 63.2 60.8 57.4  
Final 87.8 92.3 88.8 79.6  
Daily gain 0.654 0.692 0.666 0.528 0.143/0.186
Feed intake, kg/day          
Supplement 0.500 0.450 0.400 0.350  
Azolla 0.0 1.75 3.37 3.58  
Oil 0.750 0.710 0.690 0.610  
Rice polish 0.150 0.150 0.150 0.150  
Total DM 1.33 1.33 1.35 1.23  
Conversion DM 2.0 1.92 2.0 2.3 0.42/0.504


* SE of means/Probability


There were no significant difference in any of the performance parameters due to the inclusion of azolla at any of the levels. Only in the finishing phase was there an indication (P=0.21) that growth rate was slightly poorer at the highest level of azolla. Feed conversion at slightly over 2 kg DM/kg liveweight gain reflected the efficient use of the oil for growth.

Carcass yields were extremely high reflecting the high caloric value of the diet and the associated low gut fill. Back fat thickness was normal.

Table 5: Growth performance of the pigs during the overall period (20-90 kg; 112 days)

Replacement of soya protein by azolla protein, %

  0 10 20 30 ±SE/Prob*
Live weight (kg)          
Initial 28.8 29.4 28.8 28.9  
Final 87.8 92.3 88.8 79.6  
Daily gain 0.526 0.561 0.535 0.452 0.09/0.243
Feed intake, kg/day          
Supplement 0.500 0.450 0.400 0.350  
Azolla 0.0 1.17 2.0 2.10  
Oil 0.57 0.55 0.54 0.5  
Rice polish 0.110 0.110 0.110 0.110  
Total DM 1.11 1.10 1.0 1.0  
Conversion DM 2.1 1.98 2.0 2.2 0.34/0.56
Carcass yield, %** 85 84 84 88  
Backfat (cm)* 3.7 3.5 3.6 3.4  


* SE of means/Probability **Means for 3 pigs per treatment obtained 1 hr after slaughter


Table 6: Projected and actual intakes of azolla (kg/day fresh basis)

Replacement of soya protein by azolla protein, %

  0 10 20 30
Growing phase        
Projected 0.0 1.74 3.48 5.21
Observed 0.0 0.88 1.20 1.38
% of projected   51 34 26
Protein intake (g/day)* 212 202 184 167
Finishing phase        
Projected 0.0 1.74 3.48 5.21
Observed 0.0 1.75 3.37 3.58
% of projected   100 97 69
Protein intake (g/day)* 212 212 212 194


* Includes protein from soya bean and 12 g/day from the rice polishing


Projected and observed intakes of azolla are compared in Table 6. During the growing phase the pigs were not able to consume the projected quantities of azolla, the highest amount being 51% on the 10% level of substitution falling to 26% on the 30% level. Relative consumption rates of azolla were much higher during the finishing phase, with 100% of the projected intake being consumed on the 10 and 20% substitution levels falling to 69% on the 30% level. The outcome of the low intakes of protein in the growing phase was that overall protein intake was up to 25% less (167 g/day) on the highest rate of substitution of soya bean by azolla. Despite this, there was no indication of reduced performance, relative to the zero azolla treatment for which the projected protein intake of 212 g/day was achieved.

This is further indirect evidence that protein allowances for growing/fattening pigs fed low protein basal diets can be set at much lower levels than the 262 and 390 g/day, for growing and finishing phases, respectively, recommended by NRC (1988).

It is interesting to compare the results of the present experiment with those reported by Becerra et al (1990) who used similar substitution rates of azolla for soya bean but with a basal diet of sugar cane juice. In that experiment, growth rates were reduced on the azolla treatments during the growing phase but increased during the finishing phase -- the opposite of the trends in the present experiment. It could be hypothesised that this represents a fundamental difference in the interaction between oil and azolla and sugar cane juice and azolla. However, there is no apparent explanation for this.

At intervals during the trial the faeces were analyzed for evidence of parasitic infection, but with negative results. This would appear to be an another important advantage of the feeding strategy.

Table 7: Mean performance values for 4 groups of pigs fed the palm oil based diet
  Group 1 Group 2 Group 3 Group 4
No. of animals 36 40 48 45
Duration, days 79 100 115 123
Live weight, kg        
Initial 32.0 24.2 25.8 26.0
Final 89.0 87.0 86.0 83.0
Daily gain 0.722 0.628 0.524 0.464
Consumption (kg)        
Supplement* 1.0 1.0 1.0 1.0
Oil 0.5 0.5 0.5 0.5
Total DM 1.3 1.3 1.3 1.3
Conversion DM 1.8 2.0 2.4 2.8


* 450 g soya bean meal, 374 g palm kernel meal,
150 g rice polishing, 20 g dicalcium phosphate,
3 g salt and 3 g premix of vitamins and minerals.


Commercial application

A demonstration of the palm oil feeding system was carried out at the commercial piggery "San Nicolas", located in the municipality of Restrepo, Meta Departamento..

The feeding system was based on that used in the control treatment in the Azolla experiment except that the amount of rice polishing was increased to 500 g/pig/day. The data in Table 8 are the mean values for production parameters of a total of 169 pigs in 4 groups, fed the palm oil diet. Economic data are in Table 9.

Table 8: Economic evaluation of the commercial pig fattening demonstration
Item Cost Profit/pig
  (US$) (US$)
Drugs 0.3 Group 1: 26.54
Equipment 0.8 Group 2: 15.04
Salaries 1.8 Group 3: 9.10
Transport 4.0 Group 4: 5.67
Oil 19.0  
Supplement 27.0  
Pig purchase 47.0  


The biological and economic results of the demonstration on the commercial farm confirm those obtained at the experiment station and show that is technically and economically feasible to base pig feeding systems on the crude oil from the African Oil Palm as the basal diet.


(i) It is technically feasible to fatten pigs on a diet in which the principal energy source is the crude oil from the African Oil Palm. On such a feeding system a daily protein allowance of 200 g/pig is adequate provide that the protein is derived from soya bean meal.

(ii) The aquatic plant Azolla filiculoides can replace up to 20% of the soya bean protein with no deterioration in performance (growing and fattening phases).

(iii) The oil-based diet as prepared in these trials results in a higher than normal dressing percentage and no apparent increase in fat deposition.

(iv) The oil-based diet provides some degree of protection to the skin against ectoparasites.


This research was carried out with the support of the International Foundation for Science, IFS Grant B/1629-2


Becerra Maricel, Murgueitio E, Reyes G and Preston T R 1990 Azolla filiculoides as partial replacement for traditional protein supplements in diets for growing-fattening pigs based on sugar cane juice. Livestock Research for Rural Development Volume 2 Number 2 pp15-22.

Becerra Maricel 1991 Azolla anabaena; un recurso valioso para la producción agropecuaria en el trópico. Serie de Manuales Técnicos #1 pp1-53.

Hutagalung R 1987 Palm oil. Animal feed. Wiley: UK

Lumpkin T L and Pluckett D L 1982 Azolla as green manure. Use and management in crop production. Westview Press: Boulder, Colorado, USA

NRC 1988 Nutrient requirements for swine. National Academy Press; Washington, DC

Ocampo A, Lozano E y Reyes E 1990a Utilización de la cachaza de palma africana como fuente de energía en el levante, desarrollo y ceba de cerdos. Livestock Research for Rural Development Volume 2, Number 1: 43-50

Ocampo A, Castro C y Alfonso L 1990b Determinación del nivel óptimo de proteina al utilizar cachaza de palma africana como fuente de energía en raciones para cerdos de engorde. Livestock Research for Rural Development Volume 2, Number 2:67-76

Ocampo A 1992 Oil-rich fibrous residue from african oil palm as basal diet of pigs; effects of supplementation with methionine. Livestock Research for Rural Development, Volume 4, Number 2:55-59

Preston T R and Murgueitio E 1987 Tree and shrub legumes as protein sources for livestock. IN: Forage legumes and other local protein sources as substitutes for imported protein meals (Editor: D Walmsley) CTA:Wageningen and CARDI:Trinidad pp94-104

Preston T R and Murgueitio E 1992 Strategy for sustainable livestock production in the tropics. CONDRIT Ltda: Cali pp89.

Van Hove C 1986 Fisiología de la Azolla. En: Boletín Técnico. Facultad de Ciencias Agropecuarias, Universidad Nacional de Palmira.


(Received 4 November 1993)