On-farm introduction of some dry season feeding strategies to cattle farmers on the Accra Plains ofGhana and the response of cattle to these strategies.1. A survey on reasons for non-adoption of strategies

A baseline survey was carried out in the Dangbe East and West Districts of the Accra Plains to ascertain the reasons why cattle farmers had not adopted the dry season feeding strategies which had been known in Ghana for some time. All cattle farmers were also crop farmers producing cash and subsistence crops. Respondents in the Dangbe East District tended to have larger farms than their counterparts in Dangbe West (6.9 ± 0.69 ha vs 4.0 ± 0.93 ha). All cattle owners were involved in the management of their herds in the Dangbe East District. In the Dangbe West District, a lot of absentee owners (53.3 %) left their herds in the care of hired Fulani herdsmen, and only occasionally visited the farm. In the Dangbe East District, the hired herdsman was usually an indigene. Most kraals visited (Dangbe East - 83.3 %; Dangbe West - 63.3 %) had multiple cattle owners numbering at least five. All cattle in the survey area relied exclusively on natural communal grazing lands. Cattle farmers were aware of seasonal production differences, and had also realised that the differences were due to lack of feed and water during the dry season. There were a few instances where calves had been supplemented with dried cassava peels and wheatbran, but generally no supplementation was carried out. Herdsmen grazed cattle for longer periods during the dry season. In cases of very severe drought, they moved the animals and their settlements, temporarily, to a new area with better grazing resources, returning to their old settlement when conditions improved. Farmers had heard about some dry season feeding strategies such as the feedingof crop residues and the use of multi-purpose trees. However, most farmers were not using these technologies because of inadequate knowledge and inadequate contact with extension staff. Maize stover was identified as the crop residue abundant in both districts. Based on an average maize yield of 0.5 t ha-', it is estimated that at least 1.5 t of stover will be produced by each farmer. Cassava leaves and peels as well as leguminous browse plants such as Griffonia simplicifolia and Jasminum dichotomum were identified as possible feed resources.