Quantifying lucerne (Medicago sativa) growth in response to temperature and soil moisture

Abstract

The success of lucerne based farm systems has led to an increased interest in its use in different regions of New Zealand. Annual yields (kg DM/ha-1) and daily growth rates (kg DM/ha/d) have been collated as a reference library for lucerne grown in the Waikato, Taupo, Manawatu, Marlborough, Canterbury and Otago regions. The main aim of this dissertation was to determine whether lucerne yields at Lincoln could be predicted from local weather data. Data from the ‘Maxclover’ long term grazing experiment at Lincoln University was used as a pilot dataset to examine seasonal and annual yield variations of dryland lucerne in response to environmental drivers. Analysis involved quantification of the relationships between dry matter production and weather factors, soil moisture and temperature, that might then be useful for predicting lucerne yield in other areas of New Zealand. The average annual yield in the ‘Maxclover’ experiment was 15 ± 1.1 t DM/ha-1 over the eight years. The analysis of dry matter yield over six harvests per season identified two phases of growth, each influenced by mean air temperature, soil moisture or both. Lucerne dry matter production was ~9.2 ± 0.29 kg DM/°Cd in the spring period until the onset of moisture stress, after which it decreased to ~2.2 ± 0.24 kg DM/°Cd. The breakpoint in dry matter production occurred at ~1410 ± 46°Cd. Dry matter production after the breakpoint was influenced mainly by amount (mm) and frequency of rainfall events. Potential evapotranspiration alone was not an accurate measure of lucerne production therefore potential soil moisture deficit and rainfall were analysed. Potential soil moisture deficit was 250 ± 10.6 mm at the breakpoint x (°Cd) in Years 2-8, which indicates a critical limiting deficit for this soil. Based on this analysis, accumulated thermal time (°Cd) and potential soil moisture deficit (mm) and rainfall (mm) were able to be used to estimate lucerne production. The quantification of these relationships will allow farmers to predict the potential yield of their lucerne stand with the use of specific climatic data for their region and soil type.