|Luuk Fleskens, Coen Ritsema, Zhanguo Bai, Violette Geissen, Jorge Mendes de Jesus, Vera da Silva, Aleid Teeuwen, Xiaomei Yang
|Fleskens, L et al. (2020) Tested and validated final version of SQAPP. iSQAPER Project Deliverable 4.2, 143 pp
Two working papers document the testing of SQAPP by comparing measured and SQAPP-derived soil quality and soil threat parameters carried out. In a report on SQAPP Assessment as a tool to monitor soil quality improvement »Part 1. Correlation results and discussion Teixeira and Basch (2019) discuss the correlation and the agreement between measured physical, chemical and biological soil properties, and the values estimated by the Soil Quality App (SQAPP) for the same location. In »Part 2. Soil threats, Soil Quality Index and recommendations for SQAPP they discuss the accuracy of soil properties and soil threats classification based on soil properties estimates of the Soil Quality App (SQAPP) and the correlation and agreement with the soil properties and soil threats classification using measured physical, chemical and biological soil properties, for the same location. The final goal of these studies was to assess if SQAPP can be used to monitor soil quality improvement, and the adequacy of the recommendation system.
Table 5 presents the correlations between measured and estimated soil parameters. Of the 13 properties analysed, only sand content showed a strong positive correlation between measured and estimated values, 9 presented a moderate correlation and 3 a weak correlation. 6 of the moderate and weak correlations were negative. With the exception of the weak correlated (measured/ estimated) properties (Macrofauna, N and K), and Electrical Conductivity, all other correlations had statistical significance. For all properties, agreement between measured and estimated values was low. The standard error of the estimate was calculated for each SQAPP estimated soil property.
Table 5. Correlation between measured soil properties values and values estimated by SQAPP (dependent variable).
Teixeira and Basch (2019b) consider the soil threat classifications for 8 soil threats: Erosion, Compaction, Salinization, SOM decline, Acidification, Nutrient Depletion, Contamination and Biodiversity Depletion. For each soil, SQAPP estimates accurately classified the level of threat (Low, Moderate, High), on average, in 53% of soil threats. The percentage of soil threats’ classes correctly identified using SQAPP estimates, per soil, varied between 14 and 83%.
Based on these studies, the following qualifications were given to the soil quality and soil threat indicators in SQAPP (Table 6). Of all indicators, 9 are judged to have low accuracy, 8 moderate accuracy, and 2 high accuracy, whereas no verification was made for 7 indicators (mostly soil threats).
Table 6. Accuracy indications of soil quality and soil threat parameters included in SQAPP.
|Soil quality or threat parameter
|Accuracy in iSQAPER field tests
|Depth to bedrock
|Plant-available water storage capacity
|Soil organic carbon content
|Cation Exchange Capacity
|Low (only tested for low EC)
|Available phosphorus using Olsen method
|Soil microbial abundance
|Soil macrofauna groups
|Soil erosion by water
|Soil erosion by wind
|Soil organic matter decline
|Soil nutrient depletion
|Soil contamination by heavy metals
|Soil contamination by pesticides