Abstract 139

Hoyos,-P; Lascano,-C

CIAT, Cali, Colombia.

PasturasTropicales. 1988, 10: 3, 2-7; 9 ref.

In grazing trials on savanna grassland at Carimagua, Colombia from 26 Nov. to 1 Dec. 1985 (end of the rainy season), the frequency of presence (FP) and of consumption (FC) of native species were monitored visually. A preference index (PI = FC/FP X 100) was calculated. A selection index (SI) for each species was determined using oesophageally fistulated animals. The pasture was grazed at 0.33, 0.66, 1.00 or 1.33 animals/ha for 1 year prior to the trials and contained 89 identified species of which Andropogon bicornis, Trachypogon vestitus, Paspalum pectinatum and Andropogon selloanus were dominant. DM availability (measured in concurrent cutting trials) ranged from 10 kg (Axonopus purpusii, Axonopus fissifolius and P. clavuliferum) to 520 kg DM/ha (A. bicornis). Stocking rate affected availability but not preference for native species. The highest FP values were for Schizachyrium hirtiflorum and P. pectinatum (79%), A. leucostachyus (92%) and A. bicornis (24.1%), resp. Corresponding lowest values were 3 and 10% for A. purpusii and P. clavuliferum, and SI was related to species availability. 21% of the diet was composed of prostrate species (av. CP content 8.1% and in vitro DM digestibility 43%) despite only representing 3% of the biomass available. Erect species had av. CP content of 6.2% and in vitro DM digestibility of 29%. The order of preference was (a) prostrate species; (b) erect species of high availability; (c) erect species of low availability.