R. J. Oosterbaan
It is generally found that agricultural crop yields decline at shallow groundwater tables, except hydrophilic (water loving) crops like rice. Most response functions of crop yield versus depth of water table (DWT) are based on a critical depth (CDW) below which the yield reduces while at deeper depth the yield is unaffected. As the water table usually fluctuates during the growing season one normally accepts the seasonal average depth as an indicator for DWT. In some, more complicated, cases one takes the frequency exceedance of DWT over a CDW and one assumes that the crop production is less as this frequency is higher. Such an approach needs a determination of the most indicative frequency (for example 10%, 20%, etc.) and the critical extreme value (for example 10 cm, 20 cm, etc.) need to be established, which complicates the analysis and is not done here, also because the required data are seldom available. In this article the data available in literature on the relation between yield and average DWT are used and concern (1) Banana (data from Surinam), (2) Cotton (data from Egypt), (3) Sugarcane (data from Australia) and (4) Winter Wheat (data from England). The methods of analysis are (a) visual inspection of envelope lines, (b) statistical segmented regression, (c) statistical determination of maximum horizontal stretches in the yield-DWT relationship, and (d) curved (non-linear) regression with the help of (d1) the Power function, (d2) the S-curve functions, and (d3) the quadratic as well as (d4) the cubic regressions. These functions use a generalization by transforming the DWT data by means of an optimized exponent before their application and a back transformation afterwards. Free software for this purpose is available. The generalized quadratic and cubic regressions need the method of matrices and determinants to find their parameters. After the analysis of the four data groups, the conclusions are formulated. The general conclusion is that for each case all types of analysis have to be done while the most appropriate method can be different from one case to the other.
Crop yield, depth of the water table, data from farm lands, statistical analysis, drainage requirements.
Cite this paper
R. J. Oosterbaan. (2021) Crop Yield versus Depth of the Ground Water Table, Statistical Analysis of Data Measured in Farm Lands Aiming at the Formulation of Drainage Needs. International Journal of Agricultural Science, 6, 174-187