Simulating the development of field grown potato (Solanum tuberosum L.)

The simulation of developmental stages is an important part of many crop simulation models because photoassimilates partitioning to different organs varies with plant developmental stage. Developmental models can also help in pest control and fertilizer application programs, in breeding strategies, and in crop harvest programming. There are several models for simulating potato (Solanum tuberosum L.) development. These models have assumptions that are open to criticism, such as the thermal time approach to describe the effect of temperature on potato development, they assume a constant set of cardinal temperatures throughout the crop cycle and they do not take into account the effect of photoperiod on potato development. These assumptions have disadvantages, because they are not completely realistic from a biological viewpoint. The Wang and Engel (WE) model [Wang, E., Engel, T., 1998. Simulation of phenological development of wheat crops. Agric. Syst. 58, 1–24.] seems appropriate to overcome the disadvantages of current potato models, but it has been used to simulate development in annual crops where the reproductive organs are above ground, not in potato. The objective of this study was to adapt the WE model to simulate development of field grown potato. A series of field experiments was carried out at Santa Maria, Rio Grande do Sul State, Brazil, with 15 planting dates: nine in the year of 2003 and six in the year of 2004. The WE model was superior to the thermal time approach for predicting the date of major developmental stages (tuber initiation, beginning of plant senescence, and harvest), with an average root mean square error (RMSE) of 10.4 days, which corresponds to a 28–45% decrease in the RMSE compared with the thermal time methods. Model predictions with the WE model were better for earlier (tuber initiation, RMSE = 3.7 days) than for later developmental stages (harvest, RMSE = 14.0 days).