Effects of weather conditions on the response of lettuce to applied fertilizers

Abstract

Summary Fertilizer experiments on lettuce were carried out over 5 years on adjacent sites of the same soil. The shapes of the response curves to both N and P varied considerably from experiment to experiment even though the amounts of those nutrients that could be extracted from the soil did not show this variation. Responses to N were of an ‘overturning’ type and the level of N at which maximum yield occurred ranged from 50 to 400 kg/ha. Responses to P were of a ‘diminishing returns’ type and the maximum increases in yield brought about by applying P ranged from 75 to 700% of the yields obtained with no application. A mathematical model was derived which fitted the results from all the experiments satisfactorily. The parameter denning the response to P, and that denning the ‘downward’ component of the response to N were linearly related to the integral of cumulative rainfall with respect to time. Insertion of these relations into the model led to an overall model that predicted lettuce yields for each treatment in each experiment from the values of two weather parameters, the levels of N, P and K and a fitted value for seed weight. This overall model accounted for about 75% of the total variance due to fertilizer applications in the nine experiments. It is argued that most of the year-to-year variation in fertilizer response could be explained in terms of rain influencing the adverse osmotic effects on growth, the leaching of nitrate down the soil profile and dispersion of P from phosphate fertilizer granules especially in the early stages of growth. The model was slightly improved by replacing the parameter ‘time’ with cumulative evaporation from an open water surface, but not with cumulative day degrees, or cumulative evaporation divided by vapour pressure deficit.