Karstic features such as swallow holes and dolines are the rule rather than the exception in the Chalk outcrop of southern England, particularly in the region of the Chalk/Tertiary boundary. Previous tracer tests at Bedhampton and Water End have indicated groundwater transport velocities of 2.2 and 5.5km/day. A recent tracer test at Stanford Dingley in Berkshire has indicated comparably high transport velocities in the Chalk: 5.8km/day for peak tracer concentration, and 6.8km/day for tracer breakthrough. In all three cases the swallow hole systems were able to transmit turbid surface run-off contamination over large distances, indicating a short-circuiting of the natural attenuating properties of the unsaturated and saturated zones. In the light of this, the practice of using swallow holes to dispose of agricultural- and road-run-off should be discouraged.

We studied the adaptation of narrow-leafed lupin (Lupinus angustifolius) and yellow lupin (L. luteus) to waterlogging because yellow lupin may have potential as a new legume crop for coarse-textured, acidic, waterlogging-prone areas in Western Australia. In a controlled environment, plants were waterlogged for 14 days at 28 or 56 days after sowing (DAS). Plants were more sensitive when waterlogged from 56 to 70 DAS than from 28 to 42 DAS, root growth was more sensitive than shoot growth, and leaf expansion was more sensitive than leaf dry weight accumulation. Waterlogging reduced the growth of narrow-leafed lupin (60–81%) more than that of yellow lupin (25–56%) and the response was more pronounced 2 weeks after waterlogging ceased than at the end of waterlogging. Waterlogging arrested net root growth in narrow-leafed lupin but not in yellow lupin, so that after 2 weeks of recovery the root dry weight of yellow lupin was the same as that of the control plants but in narrow-leafed lupin it was 62% less than the corresponding control plants. Both species produced equal amounts of hypocotyl root when waterlogged from 28 to 42 DAS but yellow lupin produced much greater amounts than narrow-leafed lupin when waterlogged from 56 to 70 DAS.

Abstract. In the dryland cropping areas of southern Australia, at risk from dryland salinity, tree belts can improve water management by taking up water unused by crops, with the risk that crop yield will be reduced through competition. As there are few direct markets for tree products grown in the medium to low rainfall areas, the design of agroforestry systems becomes important in reducing the trade-off in crop yield.This study examined some factors that influence the trade-off between crop yield and deep-drainage control in order to develop design guidelines for medium to low rainfall agroforestry. Twenty-one sites in the grain-growing region of Western Australia and southern New South Wales were surveyed over 2 years for crop yields, tree leaf area index, and estimated recharge, providing data from 32 tree-crop interfaces on the relative influence of environmental factors and farming system characteristics on the trade-off between water management and crop yield.The factors most strongly correlated with higher yields were water-gaining sites, orientation that provided shelter from southerly to north-westerly (S, SW, W, NW) winds, and tree age (<10 years). The factors most strongly correlated with the area of cropped land protected against deep drainage were tree age (>10 years), lighter soil types, and low rainfall (<400 mm). Economic analysis of the trade-off required to produce a particular deepdrainage reduction target produced 3 groups of sites: (1) those where trees resulted in a gross margin increase of $15/ha and an estimated deep-drainage reduction of 52% (n = 3), (2) those with a gross margin loss of $49/ha and estimated deep-drainage reduction of 47% (n = 11), and (3) those with a gross margin loss of $163/ha and a deep-drainage reduction of 37% (n = 18). None of the 3 sites in the first group were in the most favourable class in both years, highlighting the vulnerability of a relatively fixed farming system to climate variability.

The responses of germination, emergence, and pre-emergent growth of narrow-leafed lupin (Lupinus angustifolius L.) to a range of soil temperatures and matric potentials were determined in 2 replicate experiments with different seedlots and sowing techniques. Although absolute responses to temperature and water availability differed between experiments, the patterns of the responses did not. Optimum temperatures were close to 20�C and base temperatures were 0-3�C. At temperatures >23�C, germination and growth were highly variable and hindered estimation of ceiling (maximum) temperatures; the lowest was about 30�C for germination. Under optimum conditions, germination took 1-1.5 days and emergence from a depth of 4 cm took 4-4.5 days, equating to 75�C. day (base temperature 2�C). The minimum soil matric potential for growth was -1 to -2 MPa, and emergence was more sensitive to low matric potential than was germination. Hypocotyl extension was more sensitive than radicle extension to low matric potential, leading to relatively long roots at the time of emergence.

This research investigated the potential to domesticate an Australian native grass (Microlaena stipoides) to produce a perennial grain crop. Perennial grain crops offer a new solution to the long-standing problems of salinity and soil erosion associated with conventional cropping systems based on annual plants. Seed yield and its components (culm number, spikelet number per culm, seed set, seed weight) were measured in 46 accessions of Microlaena stipoides (microlaena, meadow or weeping rice grass) from Western Australia and New South Wales to quantify potentially useful variation in the species. A high degree of variability was found to exist, with a 20-fold range in seed yield (0.1–2.4 g/plant), 5-fold range in seed weight (129–666 mg per 100 seeds), 2-fold range in spikelet number (14–30 per culm), 8-fold range in seed set (12–98%), and a 5-fold range in culm number (11–59 per plant). Seed yield was positively and significantly (P < 0.05) correlated with culm number, seed set, and seed weight (r > 0.55 for all). No correlation was found between seed yield and spikelet number per culm (r = –0.14). The range in seed yield and its components suggests that there is sufficient variation within microlaena to make selections for higher yielding lines. This variation will enable breeders to exploit genetic diversity more efficiently and identify useful accessions for further work. High priority traits for future work include synchronous maturity and resistance to shattering.