In‐Field Variability of Soil Test Phosphorus and Implications for Agronomic and Environmental Phosphorus Management

Abstract

Spatial and seasonal variability of P presents a challenge for accurate soil sampling for agronomic and environmental management of farm fields. Our objectives were to: (i) evaluate the accuracy of soil sampling protocols (sampling intensity) during the growing season and after harvest, with and without manure application; and (ii) evaluate the impact of temporal and spatial variability on the accuracy of conversion equations for Mehlich‐3 to Morgan P soil test P. Two 4‐ha corn (Zea mays L.) fields were sampled (150 samples per field; 64 in a regular grid, the remainder in a pattern that optimized distribution of lag distances) in July and after corn harvest. Semivariograms were constructed to investigate spatial dependence. Simulations were done using the regular grid samples to determine the 95% confidence interval (CI) and probability of obtaining a mean within the 95% CI as impacted by sampling intensity. Thirty or more samples per 4‐ha field were needed to obtain an estimate within 10% of the population mean with >90% probability independent of the timing of sampling. Manure application reduced spatial dependence and both summer sampling and sampling within a month after manure application greatly reduced the accuracy of Mehlich‐3 to Morgan conversions used in New York (P < 0.001). Additional research is needed to expand the database with more fields, both tilled and hay and pasture fields, but based on current research, we conclude that current guidelines for the timing of sampling and number of samples per hectare should be adjusted for most accurate P guidelines, especially if Mehlich‐3 to Morgan P conversion equations are used.

Impact of Soil Temperature and Moisture on Mehlich-3 and Morgan Soil Test Phosphorus

“…The results of this study suggest that (i) the accuracy of Mehlich-3 to Morgan soil test conversion equations is predominantly influenced by temperature-driven changes in pH and Ca and moisture-impacted changes in Al and that (ii) moisture and temperature impact the accuracy of the Mehlich-3 to Morgan STP conversion models. These observations are consistent with Grandt et al (2010), who reported that sampling time (summer vs. fall) greatly impacted the accuracy and bias of the Mehlich-3 P to Morgan STP conversion equations, and with the very low ratio of measured Morgan to predicted Morgan STP for Soil 1Y3, reflecting its long history of poultry layer litter and the presence of free calcium carbonate (eggshells). For soils with a long-term history of poultry manure application, current Mehlich-3 to Morgan soil test conversion models reported by Ketterings et al (2002) are unreliable because of the dissolution of free calcium carbonate with the Mehlich-3 extraction method.…”

“…2). Similar results were found in a study with calcareous soils in central New York (Grandt et al, 2010), where the most reliable conversions were obtained when soils were sampled after harvest of corn but before manure spreading. A total of 58% of the variability in ratio (measured/predicted) in our study was explained by moisture content (P = 0.0804), whereas temperature was not a significant factor in the stepwise regression.…”

Erratum

“…Moisture content impacted the change in the ratio of measured Morgan to predicted Morgan STP levels for one of the three soils only (Soil 1Y1) mostly as a result of an increase in Morgan STP without significant changes in Mehlich-3 P, Ca, Al, or pH for this pasture soil. These observations are consistent with Grandt et al (2010), who reported that sampling time (summer vs. fall) greatly impacted the accuracy and bias of the Mehlich-3 P to Morgan STP conversion equations, and with the very low ratio of measured Morgan to predicted Morgan STP for Soil 1Y3, reflecting its long history of poultry layer litter and the presence of free calcium carbonate (eggshells). causing the ratio of measured Morgan to predicted Morgan STP to be soil specific as well (P = 0.091).…”

Guiding soil sampling strategies using classical and spatial statistics: A review

“…In contrast to NO 3 –N, P has much lower mobility in fertility‐limited soils and a high degree of spatial dependence; hence, P is likely to be a better candidate for grid‐based sampling. Several factors may amplify the spatial variability of P. Hot spots from manure application or old homesteads can either increase (Cambardella & Karlen, ) or decrease (Grandt, Ketterings, Lembo, & Vermeylen, ) spatial variability. Surficial or deep‐banding of P fertilizer can increase lateral (Kitchen, Westfall, & Havlin, ) and vertical stratification (Rehm, Scobbie, Randall, & Vetsch, ) of P and K, although the patterns of stratification depend on the tillage system used.…”