Six years of survey data taken from 184 paddocks spanning 14 million ha of land used for crop and pasture production in south-west Western Australia were used to assess weed populations, herbicide resistance, integrated weed management (IWM) actions and herbicide use patterns in a dryland agricultural system. Key findings were that weed density within crops was low, with 72% of cropping paddocks containing fewer than 10 grass weeds/m2 at anthesis. Weed density and herbicide resistance were not correlated, despite the most abundant grass weed species (annual ryegrass, Lolium rigidum Gaudin) testing positive for resistance to at least one herbicide chemistry in 92% of monitored paddocks. A wide range of herbicides were used (369 unique combinations) suggesting that the diversity of herbicide modes of action may be beneficial for reducing further development of herbicide resistance. However, there was a heavy reliance on glyphosate, the most commonly applied active ingredient. Of concern, in respect to the evolution of glyphosate resistant weeds, was that 45% of glyphosate applications to canola were applied as a single active ingredient and area sown to canola in Western Australia expanded from 0.4 to 1.4 million hectares from 2005 to 2015. In order to minimise the weed seed bank within crops, pastures were used infrequently in some regions and in 50% of cases pastures were actively managed to reduce weed seed set, by applying a non-selective herbicide in spring. The use of non-selective herbicides in this manner also kills pasture plants, consequently self-regenerating pastures were sparse and contained few legumes where cropping intensity was high. Overall, the study indicated that land use selection and utilisation of associated weed management actions were being used successfully to control weeds within the survey area. However, to successfully manage herbicide resistant weeds land use has become less diverse, with pastures utilised less and crops with efficacious weed control options utilised more. Further consideration needs to be given to the impacts of these changes in land use on other production factors, such as soil nutrient status and plant pathogens to assess sustainability of these weed management practices in a wider context.

On-farm experimentation (OFE) is a farmer-centric process that can enhance the adoption of digital agriculture technologies and improve farm profitability and sustainability. Farmers work with consultants or researchers to design and implement experiments using their own machinery to test management practices at the field or farm scale. Analysis of data from OFE is challenging because of the large spatial variation influenced by spatial autocorrelation that is not due to the treatment being tested and is often much larger than treatment effects. In addition, the relationship between treatment and yield response may also vary spatially. We investigate the use of geographically weighted regression (GWR) for analysis of data from large on-farm experiments. GWR estimates local regressions, where data are weighted by distance from the site using a distance-decay kernel. It is a simple approach that can be easily explained to farmers and their agronomic advisors. We use simulated data to test the ability of GWR to separate yield variation due to treatment from any underlying spatial variation in yield that is not due to treatment; show that GWR kernel bandwidth can be based on experimental design to accurately separate the underlying spatial variability from treatment effects; and demonstrate a step-wise model selection approach to determine when the response to treatment is global across the experiment or locally varying. We demonstrate our recommended approach on two large-scale experiments conducted on farms in Western Australia to investigate grain yield response to potassium fertiliser. We discuss the implications of our results for routine practical application to OFE and conclude that GWR has potential for wide application in a semi-automated manner to analyse OFE data, improve farm decision-making, and enhance the adoption of digital technologies.

Interactions between soil pH and phosphorus (P) for plant growth have been widely reported; however, most studies have been based on pasture species, and the agronomic importance of this interaction for acid-tolerant wheat in soils with near-sufficient levels of fertility is unclear. We conducted field experiments with wheat at two sites with acid soils where lime treatments that had been applied in the 6 years preceding the experiments caused significant changes to soil pH, extractable aluminium (Al), soil nutrients and exchangeable cations. Soil pH(CaCl2) at 0–10 cm was 4.7 without lime and 6.2 with lime at Merredin, and 4.7 without lime and 6.5 with lime at Wongan Hills. A significant lime × P interaction (P < 0.05) for grain yield was observed at both sites. At Merredin, this interaction was negative, i.e. the combined effect of soil pH and P was less than their additive effect; the difference between the dose–response curves without lime and with lime was greatest at 0 kg P ha–1 and the curves converged at 32 kg P ha–1. At Wongan Hills, the interaction was positive (combined effect greater than the additive effect), and lime application reduced grain yield. The lime × P interactions observed are agronomically important because different fertiliser P levels were required to maximise grain yield. A lime-induced reduction in Al phytotoxicity was the dominant mechanism for this interaction at Merredin. The negative grain yield response to lime at Wongan Hills was attributed to a combination of marginal soil potassium (K) supply and lime-induced reduction in soil K availability.