Spatial variability in soil legacy phosphorus in subtropical grasslands under different management intensities

Jiangxiao Qiu, School of Forest Resources & Conservation, University of Florida, Gainesville, FL, Ran Zhi, School of Natural Resources and Environment, University of Florida, Elizabeth H. Boughton, Buck Island Ranch, Archbold Biological Station, Venus, FL and Jed P. Sparks, Ecology and Evolutionary Biology, Cornell University, Ithaca, NY

Background/Question/Methods
Excess phosphorus (P) fertilizer and manure application in agriculture-dominated landscapes could result in elevated P levels in soils, wetlands, and streams and lake sediments. Such P accumulation over time can serve as a long-term regional non-point source of P to surface waters downstream, even decades after discontinuing P inputs (i.e., ‘soil legacy P’). Long-term P buildup in soils can be further exacerbated by extreme precipitation, thus compromising efforts to improve water quality. However, few studies have examined fine-scale spatial variability of soil legacy P and what factors drive their spatial variations. In this research, we performed an extensive gridded (150-m interval) field sampling of surface soils (0-15 cm) in 2020 in an exemplar subtropical grassland in central Florida managed for cattle production. There are two dominant management intensities typifying of the region: (1) high-intensity where a historical P fertilization lasting 20-30 years was ceased in 1986; (2) low-intensity where no P has been applied. Soil samples were analyzed for total and available P, along with other soil covariates. Empirical Bayesian Kriging was performed to identify hotspots of soil P; linear-mixed effects models were used to analyze factors explaining their spatial variability.
Results/Conclusions
Our preliminary results revealed substantial spatial variability in total soil P and available P (i.e., Mehlich-1, Mehlich-3) across the landscape (based on a total of ~1,400 soil samples). Land-use intensity contributed to variations in total soil P and Mehlich-3 P, where high-intensity grasslands, even after 25 years of ceased P fertilization, still showed overall significantly higher total and Mehlich-3 P levels than low-intensity grasslands (both P<0.001), indicative of persistent legacy effects. However, no significant differences were found in Mehlich-1 P across management intensities. Empirical Bayesian kriging also showed hotspots of soil P levels across the landscape that are most prone to P losses and also where interventions are needed to recover and curb soil P so as to reduce P runoffs. Further analyses will be conducted to determine additional factors (e.g., soil texture, pH, soil hydrologic condition, livestock management, vegetation) in explaining P spatial variations. Our research can help identify factors contributing to soil legacy P, and reveal where and what management practices can be implemented to mitigate soil P levels. Our fine-resolution data can also help inform and parameterize biophysical models to evaluate impacts of best management practices on P loading and predict future losses in soil P under changing climatic conditions.