Surface designs to determine the levels of N and P that maximize yield of tomato

The methodology of response surfaces, originally introduced by Box and Hunter in 1951, has allowed the development of several designs, such as the San Cristóbal Design developed by Rojas in 1962, for research with fertilizers (Chacín, 1998). In order to determine the influence of the application of levels (0 to 360 kg ha) of N and P2O5 on the tomato yield and the similarity of the surface models, estimated by response surface designs, the present research in environments of the National University of San Cristóbal de Huamanga (UNSCH), in Ayacucho, Peru. Tomato was cultivated in a protected environment, using four surface designs (Central Composite Design Rotable: DCCR, Design San Cristóbal: DSC, Design July 3: D3J and Complete Factorial: DFC). The results show that the yields of tomato fruits obey the models: Y = 12292 + 104.77794*N + 87.27649*P2O5 0.12524*N2 0.13156*P2O52 0.05242*N*P2O5 (DCCR); Y = 13121 + 102.98407*N + 51.88371*P2O5 0.14659*N2 0.02645*P2O52 + 0.00512*N*P2O5 (D3J); Y = 13255 + 120.28556*N + 97.79306*P2O5 0.24539*N2 0.16289*P2O52 + 0.07365*N*P2O5 (DSC); Y = 11401 + 103.54156*N + 83.1667*P2O5 0.17911*N2 0.14172*P2O52 + 0.0747*N*P2O5 (DFC). It is concluded that: 1. N (urea) has a greater influence than P2O5 (triple superphosphate) on the yield of tomato fruits; 2. There is greater similarity between the estimated values corresponding to the DCCR and D3J designs, with those of the DFC, showing better results at average levels of N and P2O5 (180 kg ha).