This study investigated the effects of predicted changes in rainfall distribution in marginal (≤325 mm annual rainfall) parts of the south-west Australian wheatbelt and options for management and adaptation of the wheat crop. Field experiments with rain-out shelters and irrigation were conducted in 2008 and 2009 to investigate the interactions of rainfall distribution, row spacing, genotype and timing of nitrogen application on growth, water use and grain yield of spring wheat. Water storage before seeding showed potential to maintain or increase yields despite lower in-season rainfall. Widening row spacing reduced biomass and slowed water use but did not increase grain yield, because of increased soil evaporation and water left in the soil at crop maturity. The Agricultural Production Systems Simulator (APSIM) wheat model was used to investigate the effects of recent and projected climate change on yield in relation to row spacing, phenology and nitrogen. Two climate-change scenarios were applied to historical climatic data to create two plausible future climates (‘optimistic’ and ‘pessimistic’) for the year 2030. None of the strategies tested increased wheat yield under the predicted climate scenarios. Simulated yields at wider row spacings were consistently lower due to insufficient biomass, increased soil evaporation and the inability of the crop to use all of the available water before maturity. Simulated yields of short-season genotypes were always greater than yields of longer season genotypes. Nitrogen regimes had little effect in this study. This study points to several genotypic traits that could improve the performance of wheat grown at wider row spacings. These include early vigour to reduce soil evaporation and increase competition with weeds, greater tillering/biomass to reduce limitation by sink size, and a vigorous root system with appropriate lateral spread and growth to depth to access available soil water.

Borger CPD, Michael PJ, Mandel R, Hashem A, Bowran D & Renton M (2012). Linking field and farmer surveys to determine the most important changes to weed incidence. Weed Research52, 564–574.SummaryAn understanding of weed species incidence and patterns of change in incidence is vital in developing weed management strategies and directing future research endeavours. Weed incidence in fields in the south‐west of Western Australia was surveyed in 1997 and repeated in 2008 to determine any changes. In 2008, farmers were also surveyed to determine their perception of changes to weed incidence and severity. The field survey identified a total of 194 weed species (or groups of species within a genus) in the combined survey data set (i.e. 956 sites from both field surveys). The majority of survey sites were utilised for cropping, and 152 weed species were identified within cropped fields. Between 1997 and 2008, noticeable decreases in incidence (in cropped fields) were observed for Vulpia spp. (−25%), Aira caryophyllea (−21%), Bromus diandrus (−20%), Avena fatua (−18%) and Austrostipa spp. (−13%), with only Raphanus raphanistrum (11%) and Arctotheca calendula (7%) significantly increasing in frequency. Farmer perception of the most severe weed problems did not always coincide with survey results of weed incidence. For example, an exceptionally common weed like A. calendula (with increasing incidence) was of less concern to farmers than the extremely rare Conyza spp. The main conclusion of this research is that the prevalence of a weed species is not always an indication of whether the species is of economic concern to industry. Therefore, it is vital to link field survey results to industry perception of weed species severity, when directing future research efforts into weed management.

The sharp decline in the area of lupin grown in Australia is partly attributed to the failure to control herbicide-resistant weeds in narrow-leaf lupin crops grown with the conventional 25-cm-wide row spacing. Growing lupin with wider row spacing allows for interrow weed control by nonselective herbicides using a sprayshield or physical methods. During 2003 to 2006, two experiments conducted at five sites evaluated the efficacy of interrow weed control techniques in narrow-leaf lupin crops grown in 55- to 65-cm-wide rows within the Western Australia wheatbelt. Interrow herbicides were applied POST using sprayshields, intrarow herbicides were banded on lupin rows at seeding, and interrow weeds were mowed using a garden mower. The main weed species at each site was rigid ryegrass, blue lupin, or wild radish. Paraquat plus diquat applied on the interrow of the lupin crop with sprayshields controlled up to 100% of weeds between rows, leading to increases in lupin grain yield in most of the sites. Glyphosate alone, a mixture of glyphosate plus metribuzin, and glyphosate followed by paraquat plus diquat also controlled interrow weeds, but did not increase lupin grain yield at any site. Thus, paraquat plus diquat is a better choice for interrow weed control in wide row lupin than glyphosate. Mowing did not improve weed control, but mowing followed by paraquat plus diquat increased lupin grain yield at one site. Regression models predicted that there was a strong relationship between weed biomass and lupin grain yield.

Tolerance to metribuzin herbicide is an essential agronomic trait for narrow-leafed lupin (L. angustifolius L.) grown in Western Australia (WA), however, metribuzin causes up to 30% yield loss in cv. Tanjil. Tanjil is widely used as a parent in the WA lupin breeding programme to provide anthracnose resistance. Hence, identification of genotypes tolerant to metribuzin and incorporation of this tolerance into the disease-resistant cultivar is necessary for maintaining lupin production. This study identified tolerance to metribuzin among lupin cultivars and advanced breeding lines under both controlled temperature and natural winter conditions. Differences in dose responses between cultivars revealed that cv. Gungurru was tolerant and cv. Tanjil susceptible to metribuzin. Gungurru seedlings survived metribuzin applications of up to 1600 g/ha, whereas Tanjil seedlings exhibited zero survival at 800 g/ha. The rate of herbicide application that caused a 50% growth reduction (GR50, excluding dead plants) for Gungurru was 2 times greater than that for Tanjil. The level of tolerance in Gungurru is adequate to protect plants against metribuzin damage in the field. Large and consistent differences in tolerance between genotypes were identified among cultivars and advanced breeding lines across controlled temperatures (20°C during the day and 12°C at night) and in natural winter conditions. One breeding line (95L208–13–13) showed marginally better tolerance than Gungurru. A number of advanced breeding lines were as susceptible to metribuzin as Tanjil, indicating that it is very important to select for metribuzin tolerance concurrently with disease resistance in the breeding programme. Of the 6 measures of tolerance used in this study, leaf score proved to be the simplest and most effective measure and could be used for the selection of individual tolerant plants in segregating populations. Tolerance was independent of early vigour, suggesting that it is possible to combine both early vigour and tolerance into a cultivar for better weed management. In conclusion, breeding for metribuzin tolerance in lupin is feasible, and the screening method tested here was simple and consistent, which would assist a breeding programme in making rapid progress towards herbicide-tolerant plants.

Abstract: Of 78 biotypes of wild radish (Raphanus raphanistrum) collected from Western Australia (WA), 42% were resistant and 14% intermediate to acetolactate synthase (ALS)-inhibiting herbicides. Based on the LD50 and GR50 ratios, the resistant biotype K96071 was 81-fold more resistant to chlorsulfuron and 114- to 116-fold more resistant to metosulam than the susceptible biotype K96041. More resistant biotypes were found in northern zones than in southern zones of WA. Resistant biotypes evolved after five applications of chlorsulfuron in a predominantly cereal–lupin rotation. Where resistant biotypes were found, ALS-inhibiting herbicides were not rotated with herbicides with different modes of action as frequently as in fields with susceptible biotypes. Cross-resistance to chlorsulfuron and metosulam was found in the resistant biotypes even though only 15% of the 78 biotypes were exposed to two applications of metosulam over a 10-yr period. All 78 biotypes were effectively controlled by simazine and 2,4-D amine. Nomenclature: Chlorsulfuron; metosulam; simazine; 2,4-D; wild radish, Raphanus raphanistrum L. #3 RAPRA. Additional index words: Weed survey, cross-resistance, crop rotations, herbicide rotations. Abbreviations: ALS, acetolactate synthase; GR50, rate of herbicide required to inhibit growth by 50%; I, intermediate; R, resistant; S, susceptible; SU, sulfonylurea; WAP, weeks after planting.

Of 78 biotypes of wild radish (Raphanus raphanistrum) collected from Western Australia (WA), 42% were resistant and 14% intermediate to acetolactate synthase (ALS)-inhibiting herbicides. Based on the LD50 and GR(50) ratios, the resistant biotype K96071 was 81-fold more resistant to chlorsulfuron and 114- to 116-fold more resistant to metosulam than the susceptible biotype K96041. More resistant biotypes were found in northern zones than in southern zones of WA. Resistant biotypes evolved after five applications of chlorsulfuron in a predominantly cereal-lupin rotation. Where resistant biotypes were found, ALS-inhibiting herbicides were not rotated with herbicides with different modes of action as frequently as in fields with susceptible biotypes. Cross-resistance to chlorsulfuron and metosulam was found in the resistant biotypes even though only 15% of the 78 biotypes were exposed to two applications of metosulam over a 10-yr period. All 78 biotypes were effectively controlled by simazine and 2,4-D amine.

In a trial carried out at the Avondale Research Station, Western Australia, examining the effect of tillage and herbicides on the weed population in a wheat/lupin and a wheat/pasture rotation the following conclusions were drawn: 1. Rotation has the biggest effect on weed population followed by herbicides and then tillage 2. Providing adequate weed control is achieved through rotations and herbicides then reducing tillage from a full cut to a knife point seeding practice saw a decline in annual ryegrass (Lolium rigidum) and wild radish (Raphanus raphanistrum), populations however an increase in medic (Medicago spp.) population was found. 3. Yield of Lupin in the 5th year of the trial was intrinsically related to weed burden.