| The permanent wilting point(PWP)is a variable factor that determines the available soil water range,plant effectiveness,spatial and temporal evolution of soil moisture and soil water storage.It is often defined as the water content on the moisture characteristic curve corresponding to the soil water potential of-1.5 MPa.Recent studies have shown that the wilting factor is influenced by soil texture and plant drought tolerance mechanisms and that PWP may occur at water contents less than-1.5 MPa.Thus,indirect measurements based on soil properties suffer from the uncertainty of classical values,while machine vision based on plant leaf morphology has the dilemma of image information processing.For this reason,we introduced a trait-based optimisation method to accurately predict PWP from plant responses to soil water content.This study was divided into 2 steps,one to determine the feasibility and accuracy of the optimisation method and the other to explore the specific effects of organic matter and soil texture on the method.The main conclusions were as follows:(1)Under the cubic function fit,the lower threshold of available soil water showed the same pattern across the four soils,four plants and multiple organic matter contents:stomatal conductance g_s responded to soil water before photosynthetic rate P_n and transpiration rate T_r,and had a higher sensitivity to water;the lower threshold of available soil water when g_s was an indicator,in descending order:medium clay loam(red loam)>sandy chalk loam(phaeozem)>light sandy soils(loessial soil)>sandy loams(purplish soil)>heavy sandy soils(sandy loess).In addition,soil texture,plant type and organic matter content significantly affected the PWP(P<0.05).Overall,the PWP of the five soils showed the same pattern as the lower threshold of available soil water,indicating that the higher the sand content,the lower the PWP,and the higher the clay content,the higher PWP;the PWP of the four plant species were,in descending order,soybean,sunflower,sheep grass and alfalfa,indicating that the more drought-tolerant the plant,the lower its PWP;there was a positive relationship between PWP and soil organic matter content.There was a positive relationship between the PWP and the organic matter content,with the PWP of heavy sandy soil increasing more than that of medium clay soil when the organic matter content increased.The water potential of the light sandy soil at PWP was below the classical value(-1.5 MPa),while the medium clay soil was around the classical value.(2)The critical response values of the plant gas exchange parameters(P_n,T_r and g_s)to soil water accurately predicted the PWP,i.e.the lower threshold of available soil water under the cubic function fit was in numerical agreement with the measured PWP.The fitted equations for the lower threshold determined by the dynamic changes in P_n,and T_r,respectively,explained more than 93%of the total variation,with a maximum deviation of the slope from 1 of 0.07,demonstrating that both use as indicators for the determination of the PWP.However,when g_s was used as an indicator,the R~2 was above 0.94 and the slope of the linear relationship with the PWP was closer to 1 than either of the previous,which was in high agreement with the measured PWP,confirming the feasibility of the gas exchange parameter to predict the PWP.In this method,the more representative g_s response to soil moisture content was linked to the measured PWP,which not only ensures plant survival during the measurement process,but was also a simple and easy in situ determination method.Furthermore,when the PWP predicted by the moisture characteristic curve method are compared with those predicted by the gas exchange parameter,the relative error in prediction using the gas exchange parameter was less than that of the moisture characteristic curve on light sandy soils,but both prediction methods were applicable on medium clay soils.(3)The accuracy of predicting PWP based on gas exchange parameters was influenced by organic matter content and soil texture.Organic matter was linearly and negatively correlated with the relative error:the higher the organic matter content,the higher the accuracy of the soil prediction.The relative error using the method on medium clay soils was less than that using it on heavy sandy soils,with sand grains being negatively correlated with the relative error using the new method and silt and clay grains being positively correlated with the relative error.However,in situ measurements do not necessarily mean that soils with a low sand content,high silt and high clay content would have high accuracy when using the method to determine the PWP.The relative error decreases when the sand content decreases or when the silt and clay content increased,and then increased after the critical level.The critical values are 35.63%for sand,39.10%for silt and 25.27%for clay,and the new method was most accurate when the soil was a sandy loam.The new method proposed to measure more than eight key observations(including g_s and corresponding soil water content)to accurately predict the PWP;the field measurements were concentrated after rainfall and after prolonged drought,and the relative error between the predicted values and those measured by the traditional biological method was less than 6%.Therefore,when extended to in situ measurements,the new method was able to predict the PWP correctly,thus reducing the human and material resources required for in situ measurements. |
