Precipitation Manipulation Experiments May Be Confounded by Water Source

• Published in: Journal of soil science and plant nutrition 2023-09, Vol.23 (3), p.3775-3779
• Main Authors: Reinhart, Kurt O., Vermeire, Lance T.
• Format: Article
• Language: English
• Subjects: Agriculture; Aluminum; Biomass; Biomedical and Life Sciences; Demineralizing; Dissolved solids; Drinking water; Ecology; Environment; Experiments; Iron; Irrigation; Irrigation water; Life Sciences; Microorganisms; Mineralization; Nitrates; Nitrogen; Phosphorus; Plant biomass; Plant Sciences; Precipitation; Rain; Rain water; Rainfall; Short Communication; Soil chemistry; Soil pH; Soil properties; Soil Science & Conservation; Soil water; Soils; Water quality; Water shortages; Water treatment
• ISSN: 0718-9508; 0718-9516
• DOI: 10.1007/s42729-023-01298-0

This study aims to investigate the consequences of performing precipitation manipulation experiments with mineralized water in place of rainwater (i.e., demineralized water). Limited attention has been paid to the effects of water mineralization on plant and soil properties, even when the experiments are in a rainfed context. We conducted a 6-year experiment with a gradient in spring rainfall (70, 100, and 130% of ambient). We tested effects of rainfall treatments on plant biomass and six soil properties and interpreted the confounding effects of dissolved solids in irrigation water. Rainfall treatments affected all response variables. Sulfate was the most common dissolved solid in irrigation water and was 41 times more abundant in irrigated (i.e., 130% of ambient) than other plots. Soils of irrigated plots also had elevated iron (16.5 μg × 10 cm −2 × 60 −d vs. 8.9) and pH (7.0 vs. 6.8). The rainfall gradient also had a nonlinear (hump-shaped) effect on plant available phosphorus (P). Plant and microbial biomasses are often limited by and positively associated with available P, suggesting that the predicted positive linear relationship between plant biomass and P was confounded by additions of mineralized water. In other words, the unexpected nonlinear relationship was likely driven by components of mineralized irrigation water (i.e., calcium and iron) and/or shifts in soil pH which limit P solubility in soil. Our results suggest that robust precipitation manipulation experiments should either capture rainwater when possible (or use demineralized water) or consider the confounding effects of mineralized water on plant and soil properties.