The sale of cattle presents a significant opportunity to improve livelihoods for smallholder farmers in eastern Indonesia. An opportunity was identified to grow herbaceous forage legumes either in rotation with, or sown within, staple grain crops (maize, rice) to use surplus soil moisture to produce a feed bank for (mostly penned) cattle and potentially increase nitrogen supply to the grain crops. A series of experiments was conducted on Vertisol, Alfisol and Inceptisol soils in lowland and upland districts of eastern Indonesia to identify legumes from 18 taxa suitable for integrated crop-forage systems. Clitoria ternatea and Centrosema pascuorum were found to have best potential for these systems as they established reliably using local methods, consistently ranked highly for herbage yields 2–4 months after sowing and were relatively easy to harvest. Yields were highest on the Vertisol (greatest number of sites) and Inceptisol soils. Lablab purpureus, Macroptilium bracteatum and Mucuna pruriens also ranked high for yield on Vertisol soils. Clitoria ternatea regrew consistently after cutting and forage removal providing potential for extended forage production as growing conditions allow.

Poster presented at the International Leucaena Conference, 1‒3 November 2018, Brisbane, Queensland, Australia.

There has been a large effort dedicated to the evaluation of a wide variety of sub-tropical and tropical pasture legumes in the past. This large body of information is very valuable for guiding any future legume development activities, yet much of this information was at risk of being lost. This project aimed to collate and store this tropical legume evaluation data and use this and knowledge from past researchers to recommend priority R&D approaches and activities for future pasture legume development. Together with retired pasture researchers, legume evaluation datasets were identified, prioritised, and collated into a database which captured over 180 000 data records collected from 567 sites across northern Australia. Using this large integrated dataset, high power statistical approaches were used to identify legume species which performed well across this large range of evaluation sites. Several species and genera were identified which warrant further investigation and further in-depth analysis of the database in species or genera of interest would be valuable. A gap analysis of commercially proven, underused and prospective legumes was conducted across the key production regions of northern Australia. A range of material was identified which could offer potential improvements in seed production, cold, drought or grazing tolerance compared to the current released varieties.

Perennial herbaceous legumes grown for hay can improve beef and dairy production in north Queensland through providing affordable high-quality (digestible protein) dry season feed. Eleven Arachis ecotypes (A. pintoi (5), A. glabrata (3), A. paraguariensis (2) and A. kretschmeri (1)), two Stylosanthes guianensis varieties and two commercially recommended Medicago sativa varieties were grown for hay under irrigation using standardised populations in replicated small-plots over two wet seasons (summer) and compared for dry matter production and fodder quality using 8 week cutting cycles. All initially grew well but M. sativa plants were damaged by leaf and stem diseases during wet summer periods reducing leaf and stem growth and resulting in open, weedy stands; the Arachis and Stylosanthes were relatively unaffected and exhibited strong summer-dominant growth throughout the study. There were significant species and varietal differences in biomass production and some A. pintoi, M. sativa and S. guianensis produced over 30 T DM (stem plus leaf above 5 cm cut height) over 19 months. Arachis glabrata also yielded well (16-18 T DM) following a prolonged establishment phase. Feed quality was high for all legumes compared, and overall best in the Arachis spp., with crude protein percentages mostly above 16% and high levels of protein and carbohydrate rumen degradability.

A mid-program review of the progress and evolving challenges of a forage legume plant eradication program is presented. The target plants include ecotypes of four perennial and hardseeded legumes well-adapted to various; but extensive areas of Queensland: Acacia angustissima, Aeschynomene brasiliana, Aeschynorriene paniculata and Indigofera schimperi. These were 'assessed and later rejected as potential pasture plants during the 1980s and 1990s during which time seed~producing populations established. A two-phase plant,eradication program was undertaken at 66 locations throughout Queensland between 1999 and 2006. Once detected by regular ground surveys, plants were killed prior to flowering using selective herbicides, manual action, cultivation, controlled grazing and fire. By January 2006 the target plants were contained to original plant locations, emergent plant populations had declined and seeding was prevented at most sites. However, dormant persistent seed populations at most sites prompted extension of the program. At 6 sites larger and more mobile populations of the Aeschynomene ecotypes showed potential for escape from control areas, risking an escalation in resources required for long-term control. Additional activity and funding was necessary at these sites. Knowledge developed during the program was evolutionary, resulting in considerable shifts in strategic approach, scope and funding not perceived at the beginning of the program. Evolving challenges include demonstration of progress to investors, delimitation and predicting long term needs and funding.

Acacia curassavica, Aeschynomene brasiliana, Aeschynomene paniculata, and Indigofera schimperi are exotic tropical legume species to watch out for. All were introduced and evaluated at a number of sites around Queensland as possible forage species but were not released commercially. Their potential as weeds was recognized upon evaluation, with weedy characteristics including: relative unpalatability, hard-seededness, copious seed production, and sticky seeds (one species). In addition, agronomic trials found that they were not significantly better forage than existing pasture cultivars.An eradication program implemented by DPI&F (Queensland) and co-sponsored by Meat Livestock Australia is being undertaken to control and where possible eradicate these species from old trial sites and elsewhere before they escape. A CD-ROM information resource is currently being prepared by DPI&F. It includes full information on identification, distribution, potential weed threat and control methods.

Herbaceous forage legumes can increase agricultural production in Eastern Indonesian crop-livestock farming systems, however, further research into their establishment is required to ensure good forage production and to improve the quality of soils in the region. Studies were conducted testing sowing methods, seed rate, planting time, and weed management to observe their impact on plant population and biomass production during early growth (post establishment). Species studied included butterfly pea (Clitoria ternatea), centro (Centrosema pascuorum), and lablab (Lablab purpureus). Results of the studies indicated that sowing method affected plant population and biomass production. The best planting time was in December. Controlling weeds significantly improved biomass production, however seed rate had no significant effect on plant population or biomass production.

During the 1980s and 1990s large-scale evaluation programs were undertaken in Queensland to identify legumes to improve the productivity of beef pastures. Although the programs were successful, there were concerns that some unpalatable legumes posed a threat to the production and conservation values of grasslands in northern Australia. A six-year MLA-funded program was undertaken by DPI&F, with support from James Cook University, CSIRO, DNRM and EPA, to identify and eradicate legumes of high weed potential before they began to spread from plant sites.