ABSTRACT
Summary Annual species of Medicago (i.e., medics) are the key to successful cereal-pasture rotations in the integrated crop-livestock farming systems on the neutral to alkaline soils of southern Australia. However, many medic stands have deteriorated in recent years. This deterioration has prompted research on the effects of grazing sheep on the seed-seedling dynamics of annual medicbased pastures in the Mediterranean environment of South Australia, aiming to quantify the impact of medic seed reserves in or on the soil upon pasture establishment and productivity.
448Animal-house experiments involving the feeding of purepod and pod-chaff diets to sheep and a field experiment have been used to measure the survival of medic seed following the ingestion of intact pods. In farm survey and sampling studies total seed reserves of annual medic species following cereal crops have been related to medic emergence and establishment, total pasture yield, and botanical composition. In the field experiment on red-brown earth soil at the Waite Institute there was a highly significant linear decline in total pasture, medic pods, and medic herbage (r = - 0.998***, - 0.992*** and - 0.997***, respectively) over a 56-day period of continuous grazing by merino wether sheep. Over the same period there was a highly significant linear decline in pod number/m2 (from 4,167 to 619), pod weight (from 54.5 to 46.9 mg), seeds/pod (from 6.86 to 4.53), and seed weight (from 2.87 to 2.54 mg). Viable seeds/m2 declined dramatically (r = 0.987* * *) from 23,715/m2 on day 0 to 2,652/m2 on day 56. Medic seed extraction from sheep feces showed a highly significant linear decline in seed throughput, from 6,155 to 3,251 viable seeds/sheep/ day, and mean seed weight, from 2.91 to 2.60 mg, during the 56-day period. The survey data from farms on solonized brown soils showed medic seed reserves in April varying from 16 to 282 kg/ha (mean 106.3). The mean emergence count at the end of April was 239/m2, which population declined to 141/m2 by the beginning of June. There was still some emergence in July, giving a mean establishment of 188/m2 (range from 2 to 490/m2), which was only 60% of the mean cumulative emergence of 313/m2. The mean seed reserve in September had declined to 66.2 kg/ha with a negligible change in mean seed weight from 2.805 mg in April to 2.852 mg in September. These studies have shown conclusively that grazing sheep can seriously deplete potential medic seed reserves and that this may result in grossly inadequate seedling density. Medic seed reserves in the top 5 cm of soil are a reliable indicator of potential density and productivity of medic pastures. For this reason core-sampling techniques are being developed to enable farmers to assess their needs for sowing additional medic seed before the autumn rains.
