Eutrophication of surface waters is a concern from the Northeast and Ohio to Texas and Oklahoma and everywhere in between. Phosphorus (P) losses from agricultural practices are often identified as a significant contributor to the problem. Phosphorus is an essential nutrient in plants for energy storage and transfer (Adenosine di- and triphosphate, ADP and ATP). Soil characteristics such as pH, soil moisture, and content of iron (Fe), aluminum (Al), and calcium (Ca), combined with environmental and management conditions all influence P availability, plant uptake, and loss through runoff and leaching.
Phosphorus losses and concentrations in runoff, drainage discharge, and surface waters are measured as total P (TP), particulate P (PP), and dissolved P (DP). While P losses from agriculture come in all three forms, the DP is the most readily-bioavailable P fraction. Changes in fertilizer source and placement, along with changes in tillage practices, have been listed as a few of the more than 20 possible causes of increased DP loads to surface waters (Smith, et al. 2016).
In the Western Lake Erie Basin, DP loads in three sub-watersheds decreased between 1985 to 2002 and then increased again between 2003 to 2014 (Jarvie et al., 2017). Past changes in field management practices that were designed to reduce PP losses from agriculture may have unintentionally changed the way P moves in the landscape (Jarvie et al., 2017), reflecting the need to study all sources of P when management practices are evaluated. Based on a recent meta-analysis, a systematic quantitative synthesis of available research studies used to provide information to answer larger questions, Ruiz-Diaz and Edwards (2016) reported that the ability to analyze factors affecting P movement are challenged due to the limited number of studies evaluating the long-term effect of tillage and fertilizer placement in specific management systems on yield and P losses to the environment.
Another recent meta-analysis examined the effects of 4R practices on P loss from agricultural fields in Canada and the Eastern United States representing more than 200 site-years (Christianson et al., 2016). A site-year is defined as one site where data is collected for one year; site-years of data can be obtained from either one site monitored for multiple years or multiple sites monitored for at least one year. The greater the number of site-years, the more confidence one can have in the meta-analysis. In this study, on average, data for surface drained site-years resulted in greater TP and DP loads per acre than from subsurface drainage (Christianson et al., 2016). The meta-analysis focused on tile drainage because there were more site-years to compare, 64 for surface drainage versus 423 for subsurface drainage (Christianson et al., 2016). The tile drainage data showed 86 percent of the TP load in discharge was identified as sediment-bound P or PP when both TP and PP were reported (Christianson, et al., 2016). When both TP and DP were reported by a study, 40 percent of the TP load was due to DP, only two studies reported TP, PP, and DP (Christianson et al., 2016). The lack of studies that report all three P loss pathways and the low number of surface runoff studies make it a challenge to determine (thru meta-analysis) what practices will be effective in the field to reduce nutrient losses while meeting crop needs.
Often nutrient loss concerns are focused on runoff and drainage discharge during the growing season when most nutrients are applied. King et al. (2016) reported that the seasonal differences in TP and DP discharge were larger than the differences between discharges from corn and soybeans. Specifically, drainage discharge volume, TP load, and DP load (volume x concentration) were greater during the non-growing season, while TP concentration was higher following a fertilizer application during the growing season (King et al., 2016). This field study found that annual P losses from tile drainage were influenced by drainage flow rate and nutrient application timing (King et al., 2016). Future studies are needed to investigate both non-growing season and growing season tile drainage losses with combinations of the 4Rs (source, rate, time, and place).
Literature based evaluations of P fertilizer are often limited to organic (e.g., manure or litter) versus inorganic sources, usually because the P source is not defined in the cited references. Christianson et al. (2016) reported no difference in SP or TP drainage losses when organic versus inorganic P sources were applied. A recent rain simulation study did compare eight sources of P (monoammonium phosphate (MAP), diammonium phosphate (DAP), triple super phosphate (TSP), polyammonium phosphate, single super phosphate, bone meal, rick phosphate, and poultry litter) to an unfertilized control (Smith et al., 2016). Rainfall was simulated on collection boxes at a rate of 2.5 inches per hour until 30 minutes of runoff was achieved. Samples taken during the runoff period were analyzed for DP. The highest DP loads were recorded for MAP, DAP, and TSP when they were surface applied, 17.4 percent, 16.5 percent, and 19.0 percent relative loss of P (Smith et al., 2016). Surface application of polyammonium phosphate resulted in the lowest soluble P load in the runoff water collected, 0.17 percent relative P loss (Smith et al., 2016). The use of a liquid fertilizer, polyammonium phosphate in this study, may simulate P incorporation as the liquid can infiltrate the soil and have greater soil contact.
In general, P loss tends to be less than 5 percent of the P that is applied in a year, although losses have been reported as high as 20 percent (Christianson et al., 2016). In the Western Lake Erie Basin, P applications in the Sandusky, Maumee, and Raisin watersheds have changed little when compared between 1987 to 2002 and 2007 to 2012 (Jarvie et al., 2017). However, during these same two time periods, crop removal of P has increased, resulting in the annual cropland partial P balances decreasing in all three watersheds (Jarvie et al., 2017).
The results of a meta-analysis conducted on 4R management, conservation practices, and their relationship to P losses indicated that P loss increases with increased P application rate (Qian and Harmel, 2015). Additionally, this meta-analysis reported that the implementation of at least one conservation practice can decrease P loss as P application rates are increased (Qian and Harmel, 2015). However, the lack of detail in reporting the conservation practices implemented in studies included and the low number of studies that evaluated the implementation of each individual practice did not allow for the comparison of specific practices on the reduction of P loss. In the previously noted meta-analysis by Christianson et al. (2016), tile drainage discharge of TP, PP, and DP in relation to the rate of P application was not strongly observed.
Meta-analysis results showed P application timing relative to at planting (41 percent), out-of-season (28 percent) or pre-plant application (24 percent) across data for 142 site-years (Christianson et al., 2016). Among the 200 site-years analyzed, data included 176 site-years with at least one P fertilizer application, 74 site-years with a second P application, and no projects that reported a sidedress P application (Christianson et al., 2016). There were no significant differences in drainage losses of P among application timings (Christianson et al., 2016). Again, this study points out a gap in the research for the management of P application timing, with low site-year numbers for many of the practices. An in-progress study at Kansas State University by Nathan Nelson is helping to fill some the data gaps. The project is comparing the loss of P from a corn-soybean rotation when P is spring injected as ammonium polyphosphate, fall broadcast as DAP, or no P is applied over five cropping years.
More than 90 percent of the site-year applications of P in recent research were reported as banded or injected, limiting the ability to compare P loss in tile drainage among placement practices (Christianson et al., 2016). In a plot study, P loss decreased when MAP, polyammonium phosphate, and poultry litter were banded in the soil prior to simulated rainfall compared to surface applied (Smith et al., 2016). Banded placement of MAP decreased relative P loss from 17.4 percent to 0.13 percent (Smith et al., 2016). A recent meta-analysis of 4R P practices found that placement of P fertilizer can decrease the potential for P fertilizers to enter surface runoff by 59 percent in a corn field (Qian and Harmel, 2015).
Producers and consultants will continue to face challenges to reduce P losses from agricultural production. Recent meta-analyses help determine the effect of some practices like incorporation or banding of P fertilizers and the use of conservation practices to reduce the potential for P loss. While recent research also points to the unintended consequences of practice changes if research is not robust enough. Continued research is needed to help consultants and producers meet the economic, environmental, and social challenges of changes in P management practices.
Sally Flis Ph.D., CCA – Director of Agronomy – The Fertilizer Institute, Washington, D.C.
Christianson, L.E., R.D. Harmel, D. Smith, M.R. Williams, and K. King. 2016. Assessment and synthesis of 50 years of published drainage phosphorus losses. J. Environ. Qual. 45:1467-1477.
Jarvie, H.P., L.T. Johnson, A.N. Sharpley, D.R. Smith, D.B. Baker, T.W. Bruulsema, and R. Confesor. 2017. Increased soluble phosphorus loads to Lake Erie: Unintended consequences of conservation practices? J. Environ. Qual. 46:123-132.
King, K.W., M.R. Williams, and N.R. Fausey. 2016. Effect of crop type and season on nutrient leaching to tile drainage under a corn-soybean rotation. Journal of Soil and Water Conservation. 71(1):56-68.
Ruiz-Diaz, D. and C. Edwards. 2016. Experiences in Looking at Phosphorus Management Data for Meta-Analysis, Challenges and Data Gaps. Great Plains Soil Fertility Conference. Vol. 16 p. 18-21.
Smith, D.R., R.D. Harmel, M. Williams, R. Haney, and K.W. King. 2016. Managing acuter phosphorus loss with fertilizer source and placement: Proof of concept. Agric. Environ. Lett. Feb. 9, 2016.
Qiam, S.S. and R.D. Harmel. 2015. Applying statistical causal analysis to agricultural conservation: A case study examining P loss impacts. Journal of the American Water Resources Association 1-11.