Desmanthus virgatus has demonstrated potential as a pasture legume, particularly on tropical and subtropical clay soils that are neutral to alkaline. To increase understanding of its responses to defoliation, a factorial combination of 5, 25 and 100-cm stubble heights and 2, 4 and 12-week harvest intervals was imposed on field plots of D. virgatus accession IRFL 1857 over 2 years. Leaf and stem yield, in vitro digestibility, crude protein and number of growing points following defoliation were determined. Total non-structural carbohydrate (TNC) concentrations in roots and stems were also measured following each of 2 growing seasons of defoliation.Both total dry matter harvested and leaf yield were greatest when plants were harvested every 2 weeks at a stubble height of 100 cm. TNC levels were highest and leaf digestibility was among the lowest in 1990 with this treatment combination. Close and frequent harvesting resulted in low yields, low TNC concentrations, very few shoots, and stand loss following 1 year of defoliation. These results indicate that frequent but light defoliation will maximise productivity. However, leaf digestibility can be lower under such utilisation, perhaps due to high tannin levels. No plants survived into the fourth year.

Evenia aeschynomene, Aeschynomene evenia C. Wright, is a new tropical legume that was introduced to the Agricultural Research and Education Center, Fort Pierce (ARECFP) in 1977. In 1991, seeds of A. evenia plants invading an abandoned field at the ARECFP were collected and designated as IRFL 6945. The object of this report is to present a description and research results of the species in general and of IRFL 6945 in particular. Data included flowering dates, winter liveover, disease and insect incidence, persistence under grazing, productivity, seed characterization, and agronomic characterization. An erect herb (to about 1.5 m), A. evenia is water tolerant and produces seed and green foliage throughout the year, barring frost. When no frost occurs, plants overwinter. Plant populations were greater than 50% at the end of a 5-yr clipping experiment. In September 1992, crude protein (CP) and in vitro organic matter digestibility (IVOMD) concentrations of the top 30 cm of widely-spaced plants were 235 and 665, and 296 and 677 g kg-1 for closely spaced plants, respectively. Seed weight of IRFL 6945 is as much as 42 to 65% greater than that of common aeschynomene, A. americana L. IRFL 470. Seed yield was about 100 kg ha-1 when combine harvested. Cattle graze succulent stems up to 6-mm diam. In the spring and summer, while they select smaller sizes in the autumn. Maturity of stem development rather than stem diameter appears to regulate consumption. In the second year, IRFL 6945 seedlings grew more rapidly than those of IRFL 470, and seedling recruitment occurred in bahiagrass from spring into fall compared with that of IRFL 470.

Low pasture productivity during the cool season limits profitability of beef cattle production in peninsular Florida. An adapted perennial, cool-season grass is not available for much of the region, and annual forages are expensive and vulnerable to possible moisture deficits each autumn. The cultivars Auburn reed canarygrass (Phalaris arundinacea L.), Ioreed reed canarygrass, and AU Oasis phalaris [Phalaris aquatica L. var. stenoptera (Hack.) Burk.] were evaluated for persistence on a Pomona fine sand (sandy, siliceous, hyperthermic Ultic Haplaquod) at Ona, FL. Effects of harvest interval on herbage yield and quality and on stand persistence of Auburn reed canarygrass were subsequently evaluated at clipping intervals of 2, 4, and 8 wk from january through April of 1989 and 1990. A few Auburn reed canarygrass plants in screening plots persisted and grew competitively with warm-season grasses and sedges. Other entries failed to survive the initial summer. Herbage yield of Auburn reed canarygrass increased linearly with extended harvest interval, from 1.2 to 2.8 Mg ha-1 for harvests every 2 and 8 wk, respectively, in 1989. As is typical, forage quality produced opposite results with in vitro digestible organic matter (IVDOM) of 760 to 662 g kg-1 and crude protein (CP) of 231 to 156 g kg-1 at the 2- and 8-wk harvest intervals, respectively, in 1989. Stands at all harvest intervals were sparse by the second year of harvest and were essentially lost following 2 yr of treatment, while adjacent unclipped plants produced vigorous growth. Although cultivars evaluated do not have potential on subtropical Spodosols, responses of individual plants from the broad genetic base of Auburn reed canarygrass indicate that potential may exist for breeding of reed canarygrass germplasm for the subtropics.

Establishment of perennial tropical legumes such as Vigna adenantha (G. Mey.) M.M.&S. and V. parkeri Bak. cv.'Shaw' is hindered by competitive, sod-forming grasses. Openings between bunch grasses could provide an environment where legumes could establish, and tussocks formed by bunch grasses could provide a trellis for these viny legumes. Grasses could also benefit from the association. At Ona, Florida, these legumes were grown individually with five perennial grasses [Panicum maximum Jacq., P. virgatum L. cv. Alamo, Paspalum coryphaeum Trin. PI 303960, Pennisetum purpureum Schum. cv. Mott, and Setaria sphacelata (Shum.) Stapf and C. E. Hubb. cv. Solander]. Yields of grasses grown with legumes were compared to grass yields when grown with 56 kg ha-1 of N fertilizer without a legume. In the second year after planting, ground cover of V. adenantha increased linearly from June through October while V. parkeri cover did not change. Ground cover of both legumes was correlated (P<0.05) with photosynthetic photon flux density (PPFD) penetration in grass canopies at three dates in 1989. In 1990, V. adenantha ( V. parkeri not tested) cover depended on grass entry with greater cover resulting from grasses with weaker stands, i.e. P. maximum and P. virgatum. Legume-herbage mass in October 1989 depended on grass entry and was correlated with grass-herbage mass (r = -0.58) and PPFD penetration (F = 0.57). In October 1990, V. adenantha herbage mass depended on grass entry but was not correlated with grass-herbage mass or PPFD penetration. Grass herbage mass was greater through July 1989 and August 1990 sample dates when N fertilizer (compared to legume) was the N source, but after these dates grass-herbage mass was not different between N sources. Herbage mass of both legumes was reduced by competitive, high-yielding bunch grasses in the second and third years ( V. adenantha) after sowing. Once established, these viny legumes persisted by climbing in the grass canopy, but legume yields were reduced by high-yielding grasses.

The two distinct lowland regions of Costa Rica include important livestock production areas. Pasture improvement has primarily involved introducing new grasses, but potential exists for further pasture improvement through the use of tropical legumes. Fifty tropical legume accessions were selected for screening at a single site in both regions. These accessions were sown in existing stargrass (Cynodon aethiopicus) pastures with plant survival subsequently monitored through periodic visual stand rating. In the Pacific region (Guanacaste Province), Desmanthus virgatus and Centrosema pubescens were the most persistent entries under seasonal rainfall with a distinct dry season. In the Atlantic region (Limon Province), several accessions of Stylosanthes guianensis along with individual entries of C. pubescens and Aeschynomene americana showed the greatest promise under the more uniformly distributed rainfall.

In the subtropics, bahia grass (Paspalum notatum) is a widely adapted but low-quality pasture grass. In peninsular Florida it is the predominant pasture grass, and levels of cattle production supported are typically low. Established pastures of bahia grass with either aeschynomene (Aeschynomene americana), phasey bean (Macroptilium lathyroides), or nitrogen fertiliser were evaluated over 3 years at Ona, Florida, USA for potential to provide sustainable pasture systems of increased productivity. Aeschynomene pastures produced the highest average daily gains (ADG) of yearling steers at 0.57 kg compared with an average of 0.36 kg for the other treatments. Crude protein (CP) concentration of legumes was high (over 20%), however, the legumes comprised only 7.2% (for aeschynomene) and 2.3% (for phasey bean) of the herbage dry matter during the peak period and did not increase CP or in vitro organic matter digestibility (IVOMD) of total herbage samples even during this period. The high nitrogen treatment (224 kg/ha N) produced higher (P < 0.05) total herbage CP and IVOMD than did the zero N control (12.9% CP vs. 8.8% and 48.8% IVOMD vs. 42.3%) but did not produce higher (P > 0.10) ADG. Carrying capacity was lowest on aeschynomene (536 d/ha) and highest at 224 kg/ha N (1322 d/ha). During the late summer period in 1983 and 1984, when aeschynomene comprised 12% and 6% respectively of the available herbage dry weight, ADG of 1.70 and 1.46 kg on the aeschynomene treatment substantially exceeded (P < 0.05) that of 0.79 and 0.84 kg on the control for that period. Stands of the legumes were not sustained, apparently due to dense grass sod and erratic rainfall. N fertilisation of bahia grass increased carrying capacity but failed to enhance individual animal performance despite increases in forage quality.

This research evaluated the effects of source of ruminal fluid inoculum and grass: legume proportion on the kinetics of in vitro neutral detergent fibre (NDF) digestion in grass-legume mixtures. Species used were the grasses, bahiagrass (Paspalum notatum) and limpograss (Hemarthria altissima), and the legumes, aeschynomene (Aeschynomene americana) and hairy indigo (Indigofera hirsuta). Sources of ruminal fluid were from steers fed limpograss hay (6.0 g/kg N) alone, or limpograss hay supplemented with urea or cottenseed meal.In vitro NDF digestion of 100% grass was greater when ruminal fluid from steers fed limpograss hay plus supplemental N was used than ruminal fluid from steers fed limpograss hay only. This suggested that activity of ruminal microorganisms from steers fed limpograss hay only was limited in N and/or other factors necessary for optimal in vitro fibre digestion. When ruminal fluid from steers fed limpograss hay only was used, mixtures of aeschynomene with either grass resulted in in vitro NDF digestion values that were greater than the weighted average of in vitro NDF digestion values of the individual forages. The magnitude of these increases was similar to the improvement in in vitro NDF digestion of pure grass obtained using ruminal fluid from steers fed limpograss hay plus supplemental N, suggesting that aeschynomene provided soluble and/or degradable N to the in vitro fermentation. The likely mode of action of the positive associative effects on in vitro NDF digestion was a reduced lag time for the initiation of fibre digestion.

Newer systems of protein evaluation partition feedstuff N into the proportion degraded in the rumen and that which escapes ruminal degradation. Protein and fibre degradation characteristics were measured in the grasses Pensacola bahiagrass (Paspalum notatum) and Bigalta limpograss (Hemarthria altissima), and the legumes aeschynomene (Aeschynomene americana) and hairy indigo (Indigofera hirsuta). Nitrogen concentration of legumes (30-39 g/kg) was greater than that of grasses (7-13 g/kg). Legumes contained greater absolute amounts of N soluble in a buffer solution, and potentially ruminally degradable N than grasses. Ruminal degradation rate of the potentially degradable N fraction was greater in legumes (24-44%/hr) than in grasses (6-18%/hr), leading to a greater estimated escape N as a percentage of total N in grasses (12.8-25.0%) than in legumes (6.6-11.8%). Major differences in cell wall structure between grasses and legumes occurred in hemicellulose (HC) concentration, with legumes containing much less HC than grasses. Nitrogen in these tropical grasses and legumes appears to undergo rapid and extensive ruminal degradation. Although ruminal degradation of N in these grasses was extensive, low absolute quantities of ruminally soluble and degradable N may limit microbial protein synthesis in ruminants fed tropical grass diets. Legume addition to tropical grass diets may enhance digestion by providing N for rumen function, but a combination of ruminally degradable and escape proteins may be required for optimal animal performance.