Animal performance and forage quality in limpograss-nitrogen and limpograss-aeschynomene pasture

'Floralta' limpograss (Hemarthria altissima (Poir.) Stapf et C. E. Hubb.) has been under evaluation since 1974. This perennial grass is very productive and has a relatively high in vitro organic matter digestibility (IVOMD) compared to other tropical grasses. However, some studies have suggested that low N concentrations in Floralta limpograss may limit animal performance. The objective of this study was to compare average daily gain (ADG), carrying capacity (CC), and gain ha$sp{-1}$ of yearling steers grazing limpograss pastures that were (1) overseeded with the legume aeschynomene (Aeschynomene americana L.; LA) or (2) fertilized with N (LN).;Seasonal ADG was higher for LA than for LN pastures in 1985 (P $<$ 0.01, 0.55 vs. 0.27 kg) and 1986 (P $<$ 0.05, 0.85 vs. 0.54 kg). Thus, overseeding aeschynomene in limpograss pastures improved the quality of the diet consumed. Handplucked herbage of aeschynomene had mean crude protein (CP) and IVDOM concentrations of 25 decagram (dag) kg$sp{-1}$ DM and 75 dag kg$sp{-1}$ OM in 1985, and 25 dag kg$sp{-1}$ DM and 72 dag kg$sp{-1}$ OM in 1986, respectively. Handplucked limpograss had CP concentrations that were consistently above 7 dag kg$sp{-1}$ DM. Poor animal performance on LN pastures during the late summer and early fall could not, therefore, be directly related to protein deficiency.;Mean CC over the 126-day grazing season did not differ between systems. From early July to the middle of September, LN pastures had higher CC (6.4 vs 4.4 animals ha$sp{-1}$ in 1985, and 7.0 vs 5.5 animals ha$sp{-1}$ in 1986). These trends toward higher CC in LN pastures suggest that limpograss pastures fertilized with N (in the range used in this study) will generally produce more dry matter than limpograss pastures that are overseeded with legumes (not fertilized with N).;Seasonal gain ha$sp{-1}$ was higher in 1985 (P $<$ 0.01, 296 vs. 157 kg) for LA than for LN pastures and tended to be higher in 1986 (492 vs. 360 kg). There was a period x treatment interaction (P $<$ 0.01) in 1985, and pastures mixture produced higher gains ha$sp{-1}$ only from the middle of August to early October. In 1986 there was no period x system interaction. These data indicate that productivity per unit land area can be increased by associating aeschynomene with limpograss, primarily due to increased diet quality.