Modelling macadamia water use for optimizing orchard irrigation management in periodically water-scarce regions

Macadamia is a high-value tree nut crop experiencing a remarkable global rise in demand.  South Africa is the world’s largest producer and the rapid expansion of macadamia orchards across the country has been driving increased irrigation water use. This, in turn, poses significant challenges to the water-scarce production environment, which is already strained by unsustainable freshwater withdrawals and the growing impacts of climate change. Optimizing orchard irrigation management is therefore essential to minimize unproductive water use. This requires a precise quantification of macadamia trees’ water requirements. To this end, a robust macadamia-specific transpiration model would be needed to provide valuable insights into tree responses to diverse environmental and management factors. Such model, if evaluated properly, would enable upscaling of results from field level across a wide range of cultivation regions and climatic conditions. To date, however, the development of such a model has been constrained by the scarcity of high-quality, long-term transpiration datasets, limitations of existing (overly complex and data-intensive) modelling approaches, and insufficient accuracy.

To address these gaps, we linked the generation of a comprehensive experimental dataset on macadamia transpiration in the sub-humid Levubu region, South Africa, with the adoption and evaluation of a simple, data-efficient modelling approach. Tree sap velocity data were collected from two macadamia cultivars (‘Beaumont’ and ‘HAES849’), alongside continuous monitoring of microclimate and soil water content over two years. These data were analyzed to gain deeper understanding of macadamia water use behavior across seasons and under varying soil water conditions, and were used to calibrate and validate a novel macadamia transpiration model. The model was initially calibrated under non-limiting water conditions using data on tree-intercepted radiation, vapor pressure deficit (VPD), and canopy conductance to simulate potential tree transpiration - representing the upper limit of macadamia water use. It was subsequently refined to simulate transpiration under water deficit conditions, accounting for the seasonally variable and limited water availability typical of southern Africa. Model performance was validated against independent datasets for both cultivars.

Observed macadamia transpiration exhibited pronounced variability, ranging from 0.6 mm d^-1 during the dry season to 1.3 mm d^-1 during the rainy season. This variability was largely driven by microclimatic factors. The trees showed a predominantly water-conserving strategy, with strict stomatal control in response to increasing VPD. Significant differences in water use behavior were observed among cultivars, potentially reflecting variations in productivity and climate resilience. Overall, the observed daily transpiration rates were considerably lower than the industry standard assumption of 2.0 mm d^-1, suggesting that orchards are likely over-irrigated. The model successfully captured the strong stomatal response to increasing VPD and demonstrated satisfactory performance for both cultivars under both non-limiting and water deficit conditions, with lower relative error measures in the latter. This highlights the suitability of this relatively simple and data-efficient model for accurately simulating macadamia tree transpiration across cultivars and under seasonally variable water availability, making it a valuable tool for optimizing irrigation practices and reducing unproductive water use in periodically water-scarce regions.