Gaining insight into the risks, returns and value of perfect knowledge for crop sequences by comparing optimal sequences with those proposed by agronomists

Abstract

Crop rotation, in which a legume, pasture, fallow or oilseed ‘break crop’ is grown after a cereal crop to manage soil-borne disease and weeds and, on occasions, to fix nitrogen, is one of the oldest techniques in agriculture. Valuing of crop rotations is complicated because the profitability of particular crop species changes with the prevalence of biotic stresses and varies with seasonal factors such as rainfall. With the Land Use Sequence Optimiser (LUSO) and the Agricultural Production Systems Simulator (APSIM) crop model, we generate an optimum land-use strategy for various biotic stresses and land-use options for a semi-arid grain-growing region in Australia. Over a 10-year time horizon, we compare the performance and variability of an optimal sequence with three sequences recommended by local agronomists. The agronomists recommended strategic sequences to manage weeds and disease and to maximise profit. The optimal crop sequence, with perfect knowledge, selected a mixture of grain legume, oilseed, cereal crops and pastures to manage biotic stresses and generate profit. This sequence precisely timed a period of exploitation, when high-profit crops were repeatedly grown and the biotic stresses increase, with a period of rehabilitation, when low-profit break crops are grown to reduce the biotic stresses. The agronomists’ strategic sequences were either slightly more exploitative, grew more crops and allowed the biotic stresses to increase, or were more conservative and grew fewer profitable crops while managing the biotic stresses. Both strategic approaches were less profitable than the optimal crop sequence. The value of knowledge about a particular stress increases as its rate of accumulation in the farming system increases. With high levels of biotic weed stress, perfect knowledge was worth an additional AU$73 ha–1 year–1. In scenarios with lower levels of biotic weed stress, perfect knowledge was worth just $24 ha–1 year–1. Several measures of risk were defined, but there was no trade-off between profit and risk. Variability at the crop or enterprise scale did not necessarily translate into variability in profit when viewed over 10 years. Tools such as LUSO can help to determine the optimal crop sequence for a given series of enterprise options and a given level of biotic stress and explore the variability and risk associated with different enterprise choices.