Supplementing poultry diet with tree leaves or seeds: on-farm research in Nicaragua

The effect of supplementing a traditional sorghum feed for scavenging chicken was studied in a farmer-managed on-farm experiment including 12 farms. The supplements were dried leaves of 3 local trees (Cordia dentata, Gliricidia sepium and Guazuma ulmifolia) or the seeds from the fruit of Crescentia alata. The seeds or leaves were ground together with the sorghum grain in a hand-mill. The participants reported a higher daily egg production, shorter pause between clutches, increased shell thickness and improved colour of the yolk. The new technology was demonstrated to other communities on courses. In 1993, approximately two years after the initial study, 300 families (25% of the farms) were practicing one or more of these methods to supplement the sorghum regularly. This number had increased to 398 by the end of 1994. The range of supplements was extended to include leaves of other tree species, legume seeds, sun-dried fish and whole pulp and seeds of Crescentia alata. Different methods to prepare the feeds were also developed by the farmers.

El efecto de la suplementación de una dieta tradicional de sorgo para gallinas fue estudiado en experimentos manejados por campesinos en sus fincas incluyendo 12 fincas. Los suplementos eran las hojas secadas de arboles forrajeros locales (Cordia dentata, Gliricidia sepium y Guazuma ulmifolia) o las semillas de la fruta de Crescentia alata. Los ingredientes fueron triturados en un molino de mano. Los participantes informaron que la producción diaria de huevos aumentó, el período de cluecez fue acortado, el grosor de la cáscara aumentó, y el color de la yema resultó más amarrillo. Las técnicas fueron demostrados a otras comunidades en talleres. En 1993, aproximadamente dos años despues del estudio inicial, 300 familias estaban utilizando uno o más métodos para suplementar el sorgo regularmente. En 1994 este número se había aumentado a 398 familias. El número de suplementos fue aumentado por los campesinos a incluir por ejemplo hojas de otras especies de árboles, semillas de leguminosas, pescado secado al sol y la pulpa entera secado de C. alata. Diferentes métodos para la preparación de los alimentos fueron desarrollados por las campesinas.

Poultry play an important role in the production system of Nicaraguan small-holders. The peasants are mainly crop-producers, but many keep cattle or pigs, and almost all households keep poultry, mostly chicken. The poultry is scavenging and additionally fed sorghum grown on the farm. Although grain consuming, poultry also serves as an important buffer for the grain production: The flock size is increased in years of good harvest, where low grain prices follow, and is limited to a minimum in years of harvest failure.

Grains are however becoming more scarce due to sub-division of farms by inheritance, lower yields in the traditional slash-and-burn agriculture, and increasing drought problems. Although the use of other feed resources such as the leaves of fodder trees and the fruit of Crescentia alata are well-known practices in cattle feeding, there is no local experience of using these resources for poultry. These feed items are of interest as, compared to sorghum, they have a higher content of protein, some minerals and vitamins (Doggett 1988, Ravidran and Blair 1991, 1992, D'Mello 1992). Green meals are also known to be rich in lysine, the limiting aminoacid of sorghum (Ravidran and Blair 1991, 1992).

On-farm research was therefore started to study the use of these under exploited feed resources in mixtures with the traditional sorghum grain. The methods followed were mainly inspired by the review of Farrington (1988).

Our study was conducted as part of the Rural Development Project "Manuel Lopez" in El Sauce, Nicaragua. The project started in 1990 in 5 pilot areas within El Sauce municipality. The project includes approximately 1200 rural families. The aim is to improve people's welfare by promoting low input agriculture and reforestation.

El Sauce municipality is located at 13ºN. The altitude is between 200 m in the valley and 1000 m in the mountains. Mean monthly temperatures range from 25 to 30ºC in the valley. The mean yearly precipitation is 1750 mm (range: 866 to 2734 mm from 1963 to 1987). There are normally 2 crop cycles per year, the early one from May to July and the late one from August to November. Rainfall is irregular, especially in the early season. The main crops of the area are maize, sorghum (a drought resistant long season variety of Sorghum vulgare), beans (Phaseolus vulgaris) on the slopes and sesame in the valley.

Information about production systems, productivity of the hens, possible feed alternatives and the main problems of poultry production were gathered in two individual and four group interviews before the main study, as well as during the visits to the participants in the on-farm experiments.

Almost all rural families keep poultry, whereas only about 50% keep pigs and/or cattle. Poultry keeping is normally the responsibility of the women. The poultry is mainly chicken of a rather heavy, coloured breed, probably of mixed origin. An egg production of 10 to 15 eggs per clutch and one clutch per month was mentioned as normal. Productivity was reported to vary markedly according to season. Both feed supply and seasonal change of climate were mentioned as reasons.

The chickens are scavenging during the day. They are fed whole grain of sorghum or maize, which is spread on the ground. The daily feed ration is very variable, the highest rate being reported as 100 g per hen per day. Sorghum was preferred to maize for poultry feed as egg production is higher on sorghum and as maize is important for human food. Maize is, however, occasionally used for poultry as two crops can be grown per year whereas sorghum is only harvested once. Maize is known to be good for fattening of poultry.

Poultry meat and eggs are consumed by the family, sold by the farmer in the town, or sold to middlemen from the regional capital. It is normally easy to sell the products and prices are relatively high. The consumers are paying 15 to 25% more for the free-ranged chicken or their eggs than for the corresponding products from the poultry industry. A hen is sold for approximately 15 cordobas, the equivalent of the salary from one day of field work. (The exchange rate in 1993 was: 1 US$ = 6 cordobas.)

When asked about the main limitations to the poultry production, diseases, feeding and predators were mentioned to be important. The high mortality is mainly caused by epidemics, where the symptoms are compatible with Newcastle Disease, and endemic diseases among the young chicken, mostly Fowl Pox. (Parallel to the activities described in the present article, a vaccination program was set up to control Newcastle Disease.) The feed problem consisted both of shortage of grain in years of harvest failure and of seasonal fluctuations in grain supply.

The study was carried out in four communities (two in the valley and two in the mountains), where the women had requested support for their poultry production. Initially, two lectures were given to each community about general poultry nutrition and poultry diseases. During the discussions on these courses the women expressed interest in improving the nutrition of their poultry. The use of commercial concentrates was ruled out mostly due to the cost of transportation. Each of the four groups elected three to five participants to carry out on-farm experiments with local feed resources. A total of 14 women and two men were elected. The number of participants had declined to 10 women and 2 men by the start of the experiments. The participants had small to medium scale farms with from two to 75 chicken. No economic incentives were given.

The basic design of the experiments was: the hens should be fed dried leaves of fodder trees or the seeds of C. alata mixed with sorghum by grinding. The relation between leaves and sorghum was set as one 1kg of dried leaves to 4kg of sorghum. We recommended the feeding of 100 g per hen per day. The feeds should be prepared by grinding in the handmill, which is found in almost all households for the milling of boiled maize for "tortillas". It was agreed that the experiments should be carried out with the farmers' own resources. The productivity before intervention was used for comparison and evaluation of the innovations. Control groups were not included, as it was considered difficult to separate groups of hens on the farms.

The participants were visited individually at least 4 times from June to September 1991 according to the following scheme: during the first visit possible feed alternatives on the farm were identified and it was discussed how to dry the leaves and how to grind the ingredients in the handmill. At the second visit, which was made about one week after the first visit, practical experiences with the preparation were discussed. Demonstrations were included when necessary.

The third visit was made 3 to 4 weeks after the first visit and the fourth about one month later. Problems and solutions to the practical administration of the home-made feeds to the chicken were discussed. Palatability of the different feeds was discussed, and data on productivity were collected. Other changes, like changes in shell thickness, yolk colour or the length of the pause between clutches were registered when mentioned by the participants. (We did not put this question directly, as we wished to use it as a control for what could be called a "please the visitor" bias, which might occur when we asked if her hens had begun to lay more eggs after changing diet).

By the time of the third visit, the participants were gathered for an "innovator workshop" (Chambers and Jiggins 1987) to allow them to exchange experiences and opinions about the different feed alternatives, elaboration of the feeds, and productivity results. Practical demonstrations by the most advanced farmers were included to facilitate the discussion.

As the results were considered positive, the new feed formulations were demonstrated on courses offered to other communities. The acceptance of the new technology was later evaluated by four methods:

Table 1: Chemical composition of available feedstuffs

	DM	CP	EE	CF	Ash	NFE
	%	------------------- % of DM --------------

Sorghum	11.0	3.0	3.0	2.0	81.0
Cordia dentata	20.0	19.6	2.9	12.8	13.7	51.0
G sepium	29.0	22.5	7.9	14.1	9.2	46.0
G ulmifolia	25.0	13.4	9.7	24.8	10.4	41.7
C alata (pulp + seeds)	31.1	12.2	12.8	16.6	6.3	52.1

DM: Dry Matter, CP: Crude Protein, EE: Ether Extracts, CF: Crude fibre, NFE: Non Fatty Extracts. Analyses at Universidad Nacional Agraria. Data for sorghum are from Ravidran and Blair (1991).

During the initial meetings and the first visits a number of trees were identified as possible candidates for further research: the leaves of Cordia dentata (local name Tigüilote), Guazuma ulmifolia (Guácimo) and Gliricidia sepium (Madreado or Madero Negro), and the seeds of Crescentia alata (Jícaro). The later tree is found mainly in the parts of the valley where the soil is of vertisol type. The chemical composition of these feeds is shown in Table 1 together with the corresponding values for sorghum. The data for C. alata is for pulp with seeds, of which the seeds are considered to be the protein source.

Table 2: Individual results and observations from the on-farm experiments

Participant	Number chicken	Supplement	Productivity after suppl	Other observ after suppl

JA San Ramón	70 chicken 8 hens in experiment	C.den	Laying almost daily	Shell becomes thicker, yolk more yellow The sorghum lasts longer
AG San Ramón	5 hen	C.den G.sep G.ulm	Now 5 hens lay daily, before 4-6 eggs from 9 hens
MRV SanNic	10 hens	C.ala	Laying almost daily The chicks grow fast
AB Tololos	15 hens 60 chicks	C.den G.sep	The 15 hens lay daily	Daily ration: 80 g per hen
DT Tololos	25 hens	C.ala G.sep	Laying daily during 22 days, pause for 8 days 18 eggs from 22 laying hens	The mixture attracts humidity
CV Montañita	3 hens	C.den	Laying daily	The sorghum lasts longer
RMVV Montañita	2 hens	C.den G.sep	Now laying daily Before 1 egg per 2 hens or 2 per 3	Almost do not pause between clutches
MPLV Montañita	12 hens	C.den G.sep	The chicks grow fast Now 6-8 eggs from 12 hens, before 4-5	The hens like the feed
MVV Montañ	1 hen 3 chicks	G.sep	The hen laid daily during 2 months
IV Cig	8 hens 8 chicks	G.sep	5 hens lay 5 eggs, before 2 to 3	The shell thicker
BM Cerr	2 hens 6 chicks	G.sep	Laying daily	They don't like the G.sep
SG Montaña	4 hens 4 chicks	C.ala	Laying daily	Pause beween clutches is shorter
Exp. 2	11 hens	C.ala	7-11 eggs daily, before 4-5

C.ala: Seeds of Crescentia alata, C.den: Leaves of Cordia dentata, G.sep: Leaves of Gliricidia sepium, G.ulm: Leaves of Guazuma ulmifolia.

The participants used different methods for preparation of the feeds of which the most accepted was: the branches were cut and the leaves were dried first for a few hours in the sun whereafter the process was finished in the shade. Later, leaves and sorghum were mixed by coarse grinding in the handmill. During this process the veins were separated and discarded. The relation between sorghum and leaves had been set to 4:1 on a weight basis, but as only a few of the participants had scales they often varied the relation, mostly in direction of a lower content of leaves.

The participants normally prepared the feed twice a week. Longer storing was not possible as the ground feed would absorb humidity especially during the rainy season. Some farmers moistened the feed immediately before use to reduce losses by wind or scraping.

Data on the productivity on the different feed mixtures are given in Table 2. There was a good agreement among the participants that the inclusion of the different leaves of fodder trees was beneficial. They reported a higher daily egg production and a shorter pause between clutches. Apart from the improvement in production, some participants observed increased thickness of the shell and a more yellow colour of the yolk. No negative side effects with any of the tree species were reported, although rejection of the mixture with G. sepium was seen in a few cases.

The highest productivity was reported with the seeds of C. alata. On a farm where both these seeds and the leaves of G. sepium were tried, the seeds were superior to the leaves. However, C. alata seeds have an alternative use in a popular soft-drink. Consequently, they have a rather high market price, which limits their incorporation in poultry feed.

The participants considered that the work of preparing the feeds was reasonable. It was observed by one participant that 10 lb (4.5 kg) could be mixed and ground in 30 minutes. The work load was mostly a concern of the women when numbers of birds were high. A frequently expressed problem was the handmill. It is manufactured for the milling of boiled maize for "tortillas" and is worn rapidly by milling dry grains. This problem was overcome by some by using an old mill which was not in use for food. Even an old mill was able to grind the sorghum as it only was necessary to grind the grain coarsely.

The evaluation in November 1991 of the acceptance of the techniques after they had been demonstrated on courses showed that 7 of 15 families interviewed had tried the new feed formulations. As the interviews were carried out just before the sorghum harvest, all 7 had stopped making the concentrates at the time of the interview.

Interviews with key farmers showed that some of the limitations for the acceptance of the new methods were: the wear of the mill, the work load, feeding troughs are necessary if the feed is ground, and it was considered to be easier to distribute a limited quantity of feed between the flock if it was in the form of whole grain.

Table 3: The number of farms which were using different supplements regularly during 1993:

Sorghum with
leaves of Cajanus cajan	30
leaves of Gliricidia sepium	50
leaves of Guazuma ulmifolia	60
leaves of leucaena	20
whole pulp of Crescentia alata	1
seeds of Vigna unguiculata	2
sun-dried fish	4
seeds of Crescentia alata	5
seeds of Cajanus cajan	1
leaves of Cordia dentata	117
Maize with
leaves	10
Total	300

The survey carried out by the end of 1993 showed that 300 persons (of approximately 1200 families) were using the new feed formulations on a regular basis. The combinations they were using are shown in Table 3. The number of items included in the feeds had expanded to include leaves of other tree and bush species, legume seeds, sun-dried fish and dried whole pulp of C. alata.

The prepared feeds were not used following a fixed recipe, but varied according to availability of the ingredients. The main reason given when the women had stopped producing the feeds was shortage of sorghum.

The project monitoring for 1994 showed that 398 persons in 40 communities were using the methods.

New methods for preparation of the feeds were developed by some of the farmers between 1991 and 1994. One method consisted of boiling the ingredients in a small volume of water for 30 minutes, whereafter the mixture was mashed to form a semi-solid mass, which could be fed on the soil or on a simple board. The ingredients were initially: C. dentata leaves, sorghum and cowpea (Vigna unguiculata), but variations were made later on to include other bean species, which had been introduced for soil improvement (eg: Mucuna deeringiana, Canavalia ensiformis). The main limitation of this method was mentioned to be the consumption of firewood.

The main finding of this study was that a substantial improvement in the productivity of scavenging hens was possible, when the traditional sorghum diet was replaced by a mixture made by grinding sorghum with leaves of fodder trees or seeds of C. alata. Increased shell thickness and improved yolk colour were also reported in some cases. These changes are believed to be caused by:

The influence of the two factors - grinding and protein supplementation from the leaves or seeds - could not be separated in the present study. It has been argued that the dominant deficiency in the diet of scavenging chicken is energy. Huchzermeyer (1973 cited by Smith 1990) showed that maize supplementation of scavenging hens alone could improve the productivity. Ravidran and Blair (1991) stated that energy rather than protein was deficient in poultry feed formulations in most Asian countries. In our environment, where grains already are used at a rather high level, we expect that protein supplementation will be beneficial. The total protein content of the leaves is generally high, but their profile of essential amino-acids is also an important factor: Sorghum is deficient in lysine, whereas green meals are good sources of this amino-acid (Ravidran and Blair 1992). Green meals are, however, poor sources of sulphur containing amino-acids, so methionine and cystine probably become the limiting amino acids after supplementation.

Our study did not attempt to find a substitute for the grains as carbohydrate source. Sorghum is grown on almost all farms, and its exploitation can be improved by grinding and supplementation. The use of a well-known feed as a basis for the experiments might also have contributed to the high rate of acceptance by the farmers (Table 3). In the evaluations of acceptance, lack of grain was often mentioned as the reason why the women had stopped using these homemade feeds. Therefore, the introduction of a cheaper carbohydrate source such as sugar cane juice could be the next subject for investigation.

The main biological limitations to the inclusion of dried leaves in poultry diets are their high fibre content and the occurrence of toxic or antinutritional factors especially in leaves of legumes (D'Mello 1987, 1992; Ravidran and Blair 1992). We did not observe any adverse effects in our study, where the participants used leaves of both legumes (G. sepium) and non-legumes (C. dentata and G. ulmifolia). Sun drying of the leaves might have helped in reducing the content of antinutritional factors (D'Mello 1992). However, rejection of the feeds with G. sepium was observed in a few cases. The evaluation of long-term acceptance showed that the most popular species were C. dentata and G. ulmifolia, both of which are non-legumes.

The main technical constraint was identified as the hand-mill, which is not designed for dry grain. Therefore the work load to grind the feeds and the wear of the mill will limit the acceptance by some farmers.

The genetic potential of the rural hens might limit the response to improved feeding. The breeds of our area seem to have sufficient potential to respond to better feeding. Their origin is probably mixed from previous introduction of specialized breeds. Indigenous birds, on the other hand, might not be able to respond to improved feeding. Sazzad (1992) found that the egg production of indigenous birds was almost as high under scavenging conditions as in an intensive system, whereas the exotic breeds performed much better in the intensive system.

The results of the present study have arisen from the use of participatory research methodology (see the review of Farrington 1988). Using the classification employed by Franzel and Houton (1992), our procedure can be classified as jointly designed by farmers and researcher, and as farmer managed. Egg production is well suited for this type of research as the output is easy to quantify and as an improved diet will give an immediate response in productivity.

The conditions on the different farms were quite variable and it was not possible to introduce control groups. Furthermore there might have existed a "please the visitor" bias during the interviews, which could have led the women to exaggerate the benefit of a technology promoted by the visitor. It is therefore important to include long-term acceptance as an indicator of the benefits of a new technology (Table 3).

An important advantage of the participatory research approach in this study was that it initiated a process of farmer innovations, where a large number of farmers began to carry out experiments with different feed alternatives, including other tree species, legume seeds, sun-dried fish and dried C. alata pulp as well as with different ways of preparation. The initiation of this process might in fact be the most important achievement of the work. These observations are in agreement with Sumberg and Okali (1987) who stated that the benefit of a technology should be evaluated by its acceptance and also by the variation that the farmers could give the technology.

In the present study, on-farm research was able to reveal a number of options to the farmers, whereas on-station research is needed to get more precise data on the different ingredients. The inclusion of the legumes, which presently are being widely introduced for soil improvement (Mucuna, Canavalia and Cajanus) should also receive attention.