A Catena in Tropical Moist Semi-Deciduous Forest Near Kade, Ghana

Abstract

Numerous accounts have been written about the West African forests, and forest soil catenas have been described separately by a number of workers (e.g. Greenwood 1938; Vine 1941; Charter 1949; Nye 1954). But there has usually been little or no attempt to relate the soil to the natural vegetation except in the most general way. The present study has been carried out to help to fill this gap in our knowledge. The vegetation, both above and below ground on a catena in Ghanaian forest near Kade, has been examined to see to what extent it can be related to soil differences. Observations have also been made on some of the relevant microclimatic factors. The Agricultural Research Station of the University of Ghana near Kade (0?55' W, 6?09' N) lies in Taylor's (1952, 1960) Celtis-Triplochiton association of the moist semideciduous forest. This association covers the greatest area of the moist semi-deciduous forest of Ghana and is separated from the savanna by a drier type of forest (Taylor's Antiaris-Chlorophora association) to the north and east. In the south-west it is separated from the evergreen forest by a belt of forest intermediate in character between the moist semi-deciduous type and the evergreen type (Taylor's Lophira-Triplochiton association). The distribution of these forest types is shown in Fig. 1. Much of the 2700 ac (1093 ha) of the Research Station has been developed for agricultural purposes but some parts of the forest have probably not been much disturbed for at least 50 years. It is in one such area that the observations described in this paper were made. Teams of staff and students of the Botany Department of the University of Ghana have visited and worked there at intervals since 1964. The bulk of the work described, however, was accomplished during the Easter vacation of 1968. Because of the facilities available at the Agricultural Research Station the forest at Kade has been the subject of several studies by other workers. The area we examined, for instance, stands only a few hundred metres from the area in which Greenland & Kowal (1960) obtained their figures on the mineral content of the soils and the vegetation; as the forest they studied was essentially similar to our sample areas these figures can probably be taken equally to apply also to our area. The main macroclimatic features of the Kade area are summarized in Fig. 2. It will be noted that the rainfall is of the characteristic two-peak type with maxima in July and October, and that January is the only month with less than 5 cm (2 in.) of rain on average. Mean monthly maximum and minimum temperatures, especially the latter, are very steady throughout the year. Relative humidity measured at 09.00 GMT varies only from 86 to 90% but taken at 15.00 GMT it falls to as low as 53Y% during the 'dry' season when the dry Harmattan wind blows from the north-east during January-February.