Spatial Variability Of Soil Properties And Its Effects On Drip-irrigated Tomato Yield And Quality Within A Greenhouse

Soil properties, crop yield and quality parameters are naturally variable in both space and time, thus understanding spatial and temporal variability in soil properties, and their influences on spatio-temporal variability of crop yield and quality parameters is the foundation and precondition for precision agricultural management practices. Conventional field management practices, such as intensive and often excessive and uniform fertilizer application, resulted in not only low efficiency in both production and water and fertilizer resources utilization, but also negative impact on the environment. With the development of modern agricultural science and technology, precision agriculture is one of the main effective ways to improve water and fertilizer use efficiency over the entire field. Therefore, comprehensive understanding of the temporal and spatial variability of soil properties and their effects on crop growth, yield and quality parameters is of great practical significance and scientific value for implementing precision agriculture.The experiments were carried out on tomato grown in a solar greenhouse at Shiyanghe Experimental Station for Water-saving in Agriculture and Ecology of China Agricultural University. Tomato was plastic mulched and drip irrigated, and there were total 22 and 48 individual experimental plots in 2012 and 2013, respectively. During the growth period, soil moisture content, nitrogen content and soil physical properties at 0-100 cm soil profile, as well as tomato growth indices, yield and quality parameters were determined. Spatial and temporal variability of soil properties and crop evapotranspiration(ET), crop growth indices, yield and quality parameters were described. The effect of soil properties and ET on crop yield and quality parameters was analyzed. The main results are outlined as follows:(1) Coefficient of variation (CV) of soil moisture content was 6.6%～37.8% through 0-100 cm depth, and CV at the same depth increased gradually during the growth period. CV of soil moisture content decreased with higher spatial mean soil moisture content, and on the relationship between CV and spatial mean soil moisture content was best fitted by the exponential model. Spatial variability in soil moisture content was mainly due to structured spatial variability, and in most cases, range of soil moisture content was found below 10 m. CV of soil nitrate content at different depths varied between 10.6%～62.5%, and great portion in spatial variability of nitrate content was due to random variability. Most of soil nitrate content had the range at about 30 m. CV of soil ammonium content of different times was 6.5%-27.4%, and had the range of about 20 m.(2) During the growth period in two years, CV of plant height and SPAD value ranged between 2.7%～10.8%, and CV of leaf area index (LAI) was 6.1%～18.4%. CV of plant height and LAI decreased with the growth of tomatoes. The growth indices were found mostly spatial structured, but the structure was not the same at different times of the growth period. CV of ET at different growth stages and whole growth period was 7.8%-21.7%, and became larger at later growth stages. Variability in ET was found strongly spatially dependent, and the percentage of structural variation was more than 80%, with the range of 4.2～29.6 m.(3) CV of tomato yield and its associated parameters ranged between 2.8%-22.5%, and CV of quality parameters was 0.5%～33.3% in the two years. Relative to soil properties, most of the yield and quality parameters showed weaker overall variability. Yield and quality parameters showed great extent of structured spatial variability, but with different ranges. Tomato yield and quality parameters were significantly correlated with most of the root zone soil properties and ET.(4) The multiple regression models were developed between tomato yield and fruit quality parameters with soil properties from surface soil and ET at different stages. The selected surface soil properties could explain 35%～80% (with the average of 58%) of total variation in yield and its associated parameters, and 3%～67%(with the average of 38%) of total variation in quality parameters of the two years. When stage daily ET was included into the multiple regression models, the proportions could increase to 38%～92% and 4%～80%, respectively. The PCSR models including stage daily ET performed better.(5) The optimized PLS regression models were more stable than the original PLS models and the selected soil properties from the root zone could explain 35%～92% of total variability in main yield and quality parameters. Soil properties’ effects on tomato yield and quality parameters differed significantly. Overall, soil properties in the shallower soil depth and later growth stages tended to have greater influence on yield and quality parameters. Soil ammonium nitrogen content at the early stage was the most important influence factors of tomato yield and quality parameters.