1 Contents per kg: 600 IU vitamin A, 160 IU vitamin D₃, 10 mg vitamin E, 2 mg vitamin B₁, 3 mg vitamin B₂, 15 mg vitamin B₆, 0.025 mg vitamin B₁₂, 5 mg panthotenic acid, 300 mg choline chloride, 2 mg menadione sodium bisulphite, 0.5 mg folic acid, 0.4 mg cobalt, 10 mg iron, 0.5 mg iodine

Diets were supplemented with minerals and vitamins as described in Table 1. The control diet contained (g/kg dry matter) ash 107, organic matter 893, N 21.1, crude fibre 24.8, NDF 43.0, gross energy 14.8 MJ.

The macrophytes were either a mixed culture of azolla, where Azolla microphylla was predominant, duckweed or lemna (Lemna minor) or water hyacinths (Eichhornia crassipes Mart.). These plants were collected from a secondary lagoon containing effluents from pig stables of the Swine Research Institute at Havana. A batch of each plant type for the entire experiment was collected and sun-dried on a concrete floor prepared ad hoc, then milled through a hammer mill. The analysed composition of the three resulting meals is listed in Table 2.

Table 2: Composition or the free-floating aquatic macrophytes
	Water hyacinth	Lemna	Azolla
Dry matter, g/kg	934	941	932
Composition, g/kg DM
Ash	190	356	291
Organic matter	810	644	809
N	26.3	45.6	50.9
Crude fibre	370	200	195
NDF	555	382	423
ADF	356	289	266
Energy, MJ	15.4	16.2	16.1

The amounts of macrophytes used varied to some extent the composition of the diets, thus increasing the neutral detergent fibre (NDF) and ash values above the level present in the control diet. As was to be expected, the control diet formulated with sugar cane molasses type "B" and soya bean meal had a relatively low NDF content but a high ash value. Nevertheless gross energy differences among the diets were practically irrelevant.

Three digestibility trials were conducted using nine castrated male pigs averaging 45 kg liveweight at the start of the project in June 1993. The pigs were accustomed to metabolism cages and fed a standard diet of sugar cane molasses type "B" and soya bean meal (Table 1) for one week before being surgically prepared with an end-to-end ileo-rectal anastomosis (Green et al 1987). The animals recovered appetite one week after surgery and then were randomly assigned to one of the three dietary treatment for every macrophyte experiment according to a 3 x 3 Latin square design. The average feed supply was 0.08 kg DM per kg body weight^0.75 per day. The diets were offered twice daily in two equal meals at 9:00 and 15:00 hours. Drinking water was given ad libitum.

Ileal digesta was collected for two consecutive days preceded by five days of adaptation to the experimental diet. The procedure for digesta sampling and preparation for analysis was as outlined by Ly et al (1995). Ileal protein (N x 6.25) digestibility of the studied macrophytes was estimated by difference (Crampton and Harris 1969). Ileal protein digestibility of cane molasses was assumed to be zero therefore the same index was calculated in the content of diets for protein digestibility in soya bean meal from the three in vivo trials (nine animals in total).

The pepsin-pancreatin procedure of Dierick et al (1985) was employed for estimating in vitro protein digestibility in the macrophytes. The sample size was reduced to pass 1 mm mesh by grinding in a cyclone-type mill. In vitro protein digestibility was conducted in quadruplicate and casein (analytical grade) was used as reference protein. The calculation of in vitro undigested dry matter was made as recommended by Boisen and Fernandez (1995). The endogenous protein losses of every feed used in these studies were calculated as outlined by Boisen and Fernandez (1995), but assuming a constant protein intake of 130 g/kg DM.

Dry matter and ash content were estimated on feed and ileal digesta samples by drying at 105 °C and ashing at 550 °C for 24 hr respectively. Energy was determined with an adiabatic calorimeter, and nitrogen was determined on all samples by the macro-Kjeldhal technique. ADF and NDF were measured by the methods of Goering and Van Soest (1970).

Data were subjected to analysis of variance using standard procedures as outlined by Steel and Torrie (1980), with means compared by Duncan's new multiple range test. In addition regression analysis was conducted in the appropriate cases.

The determined composition of diets and ileal digestibility indices in pigs fed water hyacinth are set out in Table 3. The increased ash and fibre content of the diets was noteworthy, whereas gross energy contents were similar in all treatments. Water hyacinth meal was associated with a significant decrease (P<0.05) in ileal DM, ash, organic matter, energy and N digestibility in pigs fed graded levels of this macrophyte. In this connection, ileal N digestibility was markedly depressed from 703 g/kg in the control diet to 507 g/kg in the diet with 200 g/kg of water hyacinth meal. This same effect was absent for ileal crude fibre and NDF digestibility.

Table 3: Composition of diets and ileal digestibility of nutrients and energy in pigs fed water hyacinth
	Water hyacinth, g/kg
	0	100	200	SE/Prob
Composition, g/kg
Ash	126	135	145	-
N	21.1	21.0	21.1	-
Crude fibre	25.0	57.5	88.9	-
NDF	43.0	94.5	146
Energy, MJ	15.0	141	142
Digestibility, g/kg
DM	814a	762ab	685b	±19/0.05
Ash	590a	543ab	386b	±27/0.05
Organic matter	842a	797b	739b	±21/0.05
N	703a	581b	507c	±22/0.05
Crude fibre	175	170	186	±45/0.05
NDF	304	309	300	±53
Energy, J/KJ	858a	795b	752b	±15/0.05

abc Values within a row with the same superscript were not significantly different (P<0.05)

The composition of diets and ileal digestibility of nutrients and energy in pigs fed duckweed are presented in Table 4. The duckweed meal included in the diet increased the ash and fibre fractions. The gross energy concentration was not influenced by the introduction of the macrophyte in the feed but the N content of diets increased concomitantly for this reason, due to the high crude protein concentration found in duckweed (Table 2). On the other hand, DM and energy digestibility up to the terminal ileum were significantly depressed (P<0.05) when graded levels of duckweed meal were included in the diet. However, the digestibilities of organic matter, N, crude fibre and NDF were not influenced by the use of duckweed in the diet of the pigs.

Table 4:. Composition of diets and ileal digestibility of nutrients and energy in pigs fed duckweed
	Duckweed, g/kg
	0	100	200	SE/Prob
Composition, g/kg DM
Ash	121	151	177
N	21.4	23.6	26.0
Crude fibre	25.5	41.4	56.4
NDF	42.9	77.2	111
Energy, MJ/kg	14.6	14.6	14.6
Digestibility, g/kg
DM	800a	782a	734	±4.1/0.05
Ash	662	591	592	±12.2/0.1
Organic matter	819	792	769	±11.0
N	719	662	648	±14.2
Crude fibre	133	203	304	±111
NDF	341	423	488	±80
Energy, J/KJ	809a	798a	754b	±12.0/0.05

ab Values within a row with the same superscript were not significantly different (P<0.05)

Table 5 shows the data on the composition of diets formulated to contain azolla meal, and ileal digestibility indices. In this case, as observed with the other studied macrophytes, as azolla meal was added to the diet, the concentration of dietary ash and fibre increased and no difference in gross energy concentration was found. A sharp increase in N content with increasing levels of azolla meal in the diet was in accordance with the high N content of this macrophyte (Table 2). No evident treatment effect on ileal digestibility of nutrients and energy was found although ileal digestibility of ash and organic matter seemed to decrease (P<0.10) in treatments where azolla meal was included.

Table 5: Composition of diets and ileal digestibility of nutrients and energy in pigs fed azolla
	Azolla, g/kg
	0	100	200	SE/Prob
Composition, g/kg
Ash	72.9	93.2	112
N	20.8	23.7	26.6
Crude fibre	23.9	38.6	53.6
NDF	43.0	81.3	129
Energy, MJ	147	146	145
Digestibility, g/kg
DM	779	762	748	±17.6
Ash	553	407	390	±61.8/0.1
Organic matter	811	800	778	±13.4/0.1
N	696	663	642	±17.5
Crude fibre	120	304	440	±52.4
NDF	316	418	499	±56.7
Energy, J/KJ	779	762	748	±27.6

The regression analysis indicated that energy digestibility (Y, in %) could be calculated from the OM digestibility (X, in %) in the macrophyte-based diets, this relationship being more pronounced in water hyacinth and azolla (P<0.001) than in duckweed (P<0.01). The equations were:

Y = 3.078 + 0.972 X (SE_b = ± 0.14; r = 0.93) for water hyacinth

Y = -15.536 + 1.19 X (SE_b = ± 0.38; r = 0.76) for duckweed

Y = -16.412 + 1.16 X (SE_b = ± 0.053; r = 0.99) for azolla

The estimates of protein digestibility in macrophytes for pigs are in Table 6.

Table 6: Protein digestibility in macrophytes for pigs
	Protein digestibility
	In vitro	In vivo1	EPL2	UDM3
	------g/kg---------		---g/kg DM---
Casein	969±3.24	-	-	-
Soybean meal	914±8.0	899±485	14.9	220
Water hyacinth	412±58	162±88	32.5	706
Duckweed	674±78	560±69	14.8	614
Azolla	701±49	646±79	7.0	434