This paper reports the results of a preliminary experiment to test the hypothesis that a system of underground shelters would be a way of improving the environment of the rabbit, as compared with conventional raised cages, and thus make it possible to increase productivity in a tropical climate. The study was done with the cooperation of 30 small scale farmers in Ha Tay and Hai Hung provinces in North Vietnam.

All the performance parameters were significantly improved in rabbits housed in the underground shelters compared with the cages. The does in underground shelters were 8% heavier after 2 months in the experiment, they gave birth to 39% more offspring and weaned 60% more than those in cages. Survival rate to weaning of the offspring was improved by 16%.

During the growing-fattening phase, the rabbits in underground shelters were 17% heavier after 1 month and 19% heavier after 3 months with average growth rates that were 20% greater.

It is concluded that the underground shelter system of housing is especially suitable for rabbit production for small scale farmers in tropical regions.

Rabbit breeding is practised in almost all developing countries and contributes to family nutrition and economy, through the availability of a source of animal protein, as well as through extra income by sale of animals. Rabbit production has been developed in Vietnam based on the conventional system of cages raised above the ground. The major constraints of this housing system are low reproduction due to high temperatures in summer (30-35 ºC) and high mortality of the offspring due to cold (12-15 ºC) and wet weather in winter. The suitable ambient environment for rabbit production in Vietnam is a temperature range of 18-28 ºC and humidity between 80 and 86% (Nguyen Quang Suc 1985).

The greatest disadvantage of the cage system is that the rabbits are exposed to high temperatures during the hot season. The impact of temperature rabbits has been studied by various authors. Lebas et al (1986) noted that high temperature affects spermatogenesis, reducing the volume and concentration of ejaculates and also affects sperm motility after 8 hours at 36 ºC or 14 days at 30ºC. High temperature also affects growing rabbits negatively due to reduced rates of liveweight gain caused by the reduction in feed intake. When rabbits are raised at 18-20 ºC they can reach a live weight of 3 kg by 112 days, whereas at 30-31 ºC they only reach 2.5 kg (Lebas et al 1986).

It has been proposed that the system of underground shelters, developed by Finzi and Amici (1991), would be a way of modifying the immediate environment of the rabbit and thus make it possible to increase productivity in a tropical climate. This paper reports the results of a preliminary experiment to test this hypothesis carried out on small scale farms in Ha Tay and Hai Hung provinces in North Vietnam.

The most widespread housing system in Vietnam is the one which utilizes hutches or cages. This was introduced during the colonial period and imitates the hutch models which were developed during the first phase of intensification of rabbit production, and were later superceded by the wire net cages presently employed in large scale operations world wide.

The free range system is obviously the most simple since it respects the rabbit's natural living conditions. In a small group (1 buck and 3-5 does) the rabbits will reproduce without leaving the area if feed and water are regularly supplied and they are kept under control by some form of periphery fence.

The underground shelter is based on the premise that rabbits under natural conditions live in underground "burrows" dug out from the soil. This can provide a better environment for the rabbits especially if they are free to move in and out according to the climatic conditions.

The simulation of such "burrows" can be done by making a shelter above or under the ground using stones, bricks, clay or concrete, which can be covered by earth. With this system rabbits can choose the place to make their nest for giving birth. They can also modify the shelter by "digging" to enlarge or lengthen the "burrow".

The integration of the underground shelter and limited free range appears to be the most feasible way for implementing an improved housing system at household level. This was the housing system chosen for the experiment.

Treatments, design and feeding

Thirty farmers, who traditionally raised rabbits, were chosen to implement the experiment which consisted of a comparison of some form of underground shelter versus the conventional system of raised wire cages. Each farmer was asked to keep 3 females and 1 male for the observation of reproduction and lactation and 5 weaned rabbits for fattening on each of the two housing systems.

For the underground shelter, the farmers themselves decided on the exact form in which the shelters were constructed following the principle of a fenced free area of about 0.5-1 m² per adult rabbit and a mound of earth (0.5 x 2 m area and 0.5 m high) in which the rabbits could make their "burrows". The mound area and the immediate points of entry to the burrows were protected with a roof usually made of dry leaves of sugar cane or rice straw to avoid rain entering the shelters.

The conventional housing consisted of cages made of wire or bamboo with 0.6 m² area per adult. Nest boxes were introduced into the cage immediately prior to parturition.

The feeding system was the same for both types of housing and consisted of free access to a molasses block, by-products from the kitchen as available (eg: banana peel, jackfruit, soya bean residue), grasses (mostly guinea grass) and tree leaves (from Trichantera gigantea, mulberry, Leucaena leucocephala and bamboo).

The experiment began in January 1996 and continued until July 1996.

A more detailed study of the effects of the housing system on temperature and humidity was done between 5 and 15 July 1996 at the experimental farm of the Goat and Rabbit Research Centre, where the same treatments as on the farms were applied.

On all farms the following performance data were recorded:

• Liveweight of the does at the beginning and 2 months later
• Litter size at birth and at weaning
• Liveweight of growing rabbits at 1, 2 and 3 months of age

On the experimental farm, the ambient temperature and the humidify were recorded daily at 7:00 , 13:00 and 19:00 hours in the cages and in the underground shelters over a 10-day period from 5 to 15 July.

The results from the on-farm trials are presented in Table 1. All the performance parameters were significantly improved in rabbits housed in the underground shelters compared with the cages. The does in underground shelters were 8% heavier after 2 months in the experiment, they gave birth to 39% more offspring and weaned 60% more than those in cages. Survival rate to weaning of the offspring was improved by 16%.

Table 1: Mean values for parameters of reproduction and growth of rabbits housed in cages or in underground shelters in small scale farms
	Cages	Under-	SE/Prob.
		ground
Liveweight of doe, g
Initial	1652	1659	±20.9/NS
Final	2,137	2,310	±11.2/0.001
Litter size at
Birth	4.88	6.68	±0.13/0.001
Weaning	3.64	5.81	±0.11/0.001
Survival, %	74.1	85.7	±0.68/0.001
Weight of growing rabbit, g
1 month	306	360	±3.9/0.001
2 month	891	1,055	±5.3/0.001
3 month	1,365	1,639	±6.5/0.001
Gain
1-3mth, g/d	17.7	21.3	±0.095/0.001

During the growing-fattening phase, the rabbits in underground shelters were 17% heavier after 1 month and 19% heavier after 3 months with average growth rates that were 20% greater.

It was not possible to obtain detailed records on feed intake but most of the farmers observed that the rabbits in the underground shelters ate more than those in cages. More precise experiments are needed to confirm this aspect.

The results of the study on temperature and humidity during a 10-day period in July in the two housing systems are shown in Tables 2, 3 and 4. Ambient temperature on average was 3.8 ºC lower (P=0.001) and humidity was 4.75% higher (P=0.001) in the underground shelters compared with the cages. It seemed that these mean differences were maintained throughout the day from 7:00 to 19:00 hours (Table 3).

Table 2: Mean values for temperature and humidity in cages and in underground shelters (at 00.70, 13.00 and 19.00 hours during a 10 day period
Day	Temperature °C		Humidity, %
	Grnd	Cage	Grnd	Cage
1	28	32	65	60
2	30	35	65	62
3	28	30	76	72
4	26	28	86	80
5	26	30	83	83
6	26	30	82	72
7	26	30	82	74
8	26	30	83	76
9	27	32	84	83
10	27	31	83	80

Table 3: Effect of housing on temperature and humidity (mean values)
	Cage	Grnd	Diff	SE/Prob
Temperature,ºC	29.4	25.9	3.8	±0.33/0.001
Humidity, %	75.9	80.4	4.75	±0.99/0.001

Table 4: Mean values for temperature and humidity during the day in cages and underground shelters
	Temperature°C		Humidity, %
	Cage	Grnd	Cage	Grnd
07.00	29.6	25.7	77.6	81.5
13.00	30.5	26.9	74.2	78.9
19.00	28.2	25.2	76.0	80.8

• The underground housing system for rabbits appears to be markedly superior to the conventional cage system in North Vietnam at least during the hot summer period from April to August as both reproduction and growth parameters were considerably improved by the new housing system.
• It is recommended that the system be evaluated throughout Vietnam especially at smallholder level.

This study was financed through a "Letter of Agreement" with the Animal Production and Health Division of FAO. The authors are grateful to Dr Manuel Sanchez, FAO, Rome and to Dr A Finzi of the Rabbit Research Centre, Animal Husbandry Institute, University of Viterbo, Italy, for their suggestions and advice.

(Received 31 August 1996)