J of Env Quality - 2019 - Hopkins
J of Env Quality - 2019 - Hopkins
Abstract
The discovery and development of phosphorus (P) and P
fertilizers provide context for current management conventions.
Average crop yields were stagnant before the Green Revolution
T he discovery and development of P and P fertiliz-
ers provide context for current and future management
conventions. Understanding historical perspectives as
they flavor current thinking, despite a rapidly changing agricul-
tural world, can illuminate traditional soil P management and
but have steadily increased since. This, along with conventional
P management, has resulted in widely depleting soil P levels.
needed changes.
Improved technology and management are needed to meet
the increasing P demand. Modern hybrids and cultivars have Historical Perspectives
different P demand and uptake patterns that require changes
in conventional P fertilizer placement and timing. Phosphorus Yields and Phosphorus before the Green Revolution
fertilizer recommendations based on soil analysis remains The first agricultural revolution shifted hunter–gather soci-
valid, but evidence suggests a need for recalibrating soil test P
eties to crop cultivation. Subsequent revolutions spread species
(STP) critical levels (the STP concentration at which a response
to P fertilizer would not be expected) and P fertilizer rates to and farming techniques, resulting in widespread cultivation and
accommodate high-yield scenarios. Considering higher P fertilizer distribution of domesticated crops and increasing the land each
rates as a single solution poses environmental challenges, laborer could farm. But crop yields remained low for millennia.
highlighting the need for improved P use efficiency (PUE). Wheat (Triticum spp.) yields were ~0.5 to 1.2 Mg ha-1 in some
Phosphorus fertilization approaches that have the potential
of the earliest, rainfed agricultural communities in the Middle
to improve PUE and enable high yields include crop-specific
precision placement of P, informed timing of P fertilizers, and new East 3000 to 12,000 BP (Araus et al., 2014). Unlike more agri-
enhanced efficiency sources of P fertilizer. This paper examines culturally advanced areas, current yields in the Middle East are
these management approaches from historical, production, and roughly equivalent at 0.7 to 1.2 Mg ha-1 (Araus et al., 2014) and
environmental perspectives in modern cropping systems. were even similar to those in the United States between 1865
and 1940 at 0.7 to 1.1 Mg ha-1 (USDA-NASS, 2019).
Core Ideas Fertilization was undoubtedly initiated by ancient peoples
observing improved harvests where ash, nutrient-rich sediments,
• History of P fertilization illuminates traditional soil P manage- domestic wastes, and green and animal manures were applied.
ment and needed changes.
These observations led to rudimentary fertilization evident in
• Recalibration of STP and P fertilizer recommendations are need-
ed to match increasing yield and rates of P uptake. the archeological record (Araus et al., 2014). But the availability
• Environmental concerns and diminishing P supply necessitate of these fertilizers was minimal and management was not well
improvement in P use efficiency. understood. Other limiting factors, such as pests and weather,
• Placement and timing are improved through understanding of masked the potential benefits of fertilizer. Thus, ancient fertil-
variable rooting patterns. ization maintained the yield baseline but did not dramatically
• Enhanced efficiency P fertilizers can be effective if applied cor- increase yields.
rectly.
In fact, records at Oxford University during the 14th cen-
tury documented that yields declined below the baseline with
repeated cropping of the same soil. Continual cropping on the
same soil was routine in this region due to the limited cultiva-
tion area and large population. Scientists and farmers commonly
observed that release of nutrients through weathering of soil min-
erals and decomposition of organic matter (OM) was adequate
for native ecosystems but not enough in productive agricultural
© 2019 The Author(s). This is an open access article distributed under the CC BY-
NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Brigham Young Univ., 4105 LSB, Provo, UT 84602. Assigned to Associate Editor Jim
Ippolito.
J. Environ. Qual. 48:1265–1280 (2019)
doi:10.2134/jeq2019.03.0130 Abbreviations: APP, ammonium polyphosphate; DAP, diammonium phosphate;
Received 18 Mar. 2019. EEF, enhanced efficiency fertilizer; MAP, monoammonium phosphate; NDVI,
Accepted 9 Aug. 2019. normalized difference vegetative index; OM, organic matter; PUE, phosphorus use
*Corresponding author (hopkins@byu.edu). efficiency; STP, soil test phosphorus; VRF, variable rate fertilization.
1265
15372537, 2019, 5, Downloaded from https://acsess.onlinelibrary.wiley.com/doi/10.2134/jeq2019.03.0130 by INASP/HINARI - AFGHANISTAN, Wiley Online Library on [01/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
systems as the native soil fertility was depleted ( Johnston et al., Breakthroughs in Soil Analysis and Fertilization
2014). They learned that to prevent soil degradation and asso-
The Green Revolution is indicated by the sharp and steady
ciated starvation, mineral nutrients removed through harvest
increase in yields in many nations, including the United States
needed to be replenished to sustain an agriculturally dependent
(Fig. 1). This seminal event in history was spawned by advance-
population.
ments in agricultural science and related disciplines, including
Hopkins (2015) reviewed the history of formal fertilization
pest management, irrigation, crop breeding and genetics, mecha-
science, which began in the late Renaissance with develop-
nization, communications, transportation, computerization,
ments in the core sciences, including the discovery of chemi-
and the widespread availability of inexpensive fertilizers and
cal elements. Elemental P was discovered in 1669 by Hennig
scientific knowledge related to their use, including soil analysis.
Brandt (Emsley, 2000; Sharpley et al., 2018). Emsley (2000)
Soil scientists began to successfully predict P fertilizer response
masterfully discussed the many positive and negative impacts
with soil analysis, which was an important part of this revolution
of P in society, including its role in developing agriculture and
(Anderson, 1960; Peck, 1990; Johnston et al., 2014).
as an essential nutrient for all biological life. In 1800, Erasmus
Several chemical soil extractants have been successfully
Darwin (1731–1802), grandfather of Charles Darwin, wrote
developed to measure a fraction of soil P with reasonable cor-
in The Philosophy of Agriculture and Gardening that N and P are
relation to yield and/or crop quality response to P fertiliza-
plant nutrients taken up by roots and that compost, bone ash,
tion (Mallarino, 2003; Hopkins, 2015). Correlations of soil P
and manures should be applied as fertilizers. He even suggested
extraction with crop response were reasonable and proved to be
exploration of P-bearing minerals (Darwin, 1800). Although
a valuable advance in soil fertility (Peck, 1990) and continuing
largely ignored for decades, this work eventually led to discov-
today (Mallarino, 2003; Poulton et al., 2013). These extractants
eries of P as an essential nutrient and P fertilization science.
are not perfect in their prediction of this complex relationship
Early scientists understood P was essential and commonly
and their performance varies with soil types (Heckman et al.,
deficient in soil. However, sources of P, such as manure, were
2006). For example, the Bray P1 extractant should not be used
limited. This was a large societal problem as food was some-
on calcareous soils because the carbonates neutralize the poorly
what scarce. But with the increasing scientific understanding
buffered acid extractant; rather, the Olsen bicarbonate and the
of the day, effective P fertilizers were discovered and the fertil-
Mehlich 3 extractants are much more effective for calcareous
izer industry was born, with development of widely available
soils (Mallarino, 2003; Hopkins, 2015). Appropriate soil test
effective, affordable products. As such, P fertilizer use multi-
P (STP)–based fertilizer recommendations continue to be
plied exponentially, especially during the decades immediately
important for crop production and water quality (Fulford and
before the Green Revolution in the mid–20th century.
Culman, 2018), including in high-yield environments (Boring et
However, informal observation and scientific studies revealed
al., 2018).
that fields receiving frequent P fertilizer applications in past years
The use of STP as a decision tool is widely proven and
would often no longer exhibit yield increases with continued
accepted. As with Mallarino (2003) and Sucunza et al. (2018),
P application. Hall (1909) discussed the history of P fertiliza-
the typical approach is to evaluate the P response in the year of
tion and stated that at that time, it was impossible to predict if
application. This is highly valuable information, but as society
P fertilizer application would result in a
response or not and, if it was responsive,
the proper rate to apply for each cropping
system and environment. A predictive
tool for P fertilizer recommendations was
needed, which was especially challenging
given the known complexity in soils and
cropping systems that prevented accurate
fertilizer recommendations.
Scientists worked to solve these dilem-
mas beginning in the latter half of the
19th century and into the next (Anderson,
1960; Peck, 1990). The first efforts inves-
tigated plant tissue analysis, but these were
often not adequate because P deficiency
would occur in irreversible fashion before
the test results were available and fertil-
izer could be applied (Hopkins, 2015).
Despite widespread availability of P fertil-
izer prior to the Green Revolution, grow-
ers continued to lack a reliable, predictive
tool that would inform them, before
planting, whether they should apply P
Fig. 1. Average annual yield in the United States relative to the pre–Green Revolution averages
and, if so, at what rate. (set to equal 0) of 6.4 (potato), 1.6 (maize), 1.9 (rice), 0.9 (wheat), 1.0 (soybean), and 24.8 (sugar
beet) Mg ha-1 (adapted from USDA-NASS, 2019).
1266 Journal of Environmental Quality
15372537, 2019, 5, Downloaded from https://acsess.onlinelibrary.wiley.com/doi/10.2134/jeq2019.03.0130 by INASP/HINARI - AFGHANISTAN, Wiley Online Library on [01/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
moves toward higher yields there is a need to expand the view
to include the long-term approach to promote sustainable soils,
yields, and society. Most of the P taken up by a crop comes from
residual in the soil (Fixen and Bruulsema, 2014; Johnston et al.,
2014). In research trials, it is common for control plots (no P
fertilizer added) to yield reasonably well in the short term, even
with low to moderate STP (Schlegel and Havlin, 2017). For
example, Hopkins et al. (2010a, b, c) measured a significant
increase in potato (Solanum tuberosum L.) yields when P fertil-
izer was applied, but the treatments with no added P had yields
within ~90% of the fertilized ones. Note that this effect is only
with 1 yr without P (these fields received P in prior years).
Schlegel and Havlin (2017) demonstrated the important role
of residual and fertilizer P in a long-term, irrigated, continu-
ous maize (Zea mays L.) study (Fig. 2). The control treatment, Fig. 2. Annual maize grain yields for a long-term fertilizer study near
with no N and P fertilizer (0N–0P), had average yields that Tribune, KS, from 1961 to 2010. Untreated control (0N–0P; X), N
remained at background levels–despite other advances in agri- fertilizer only (179N–0P; C) at 179 kg N ha−1, and N + P (179N–20P; @)
at the same rate of N plus 20 kg P ha−1 (Schlegel and Havlin, 2017).
culture (Schlegel and Havlin, 2017). Fertilization with N and P
(179N–20P) resulted in an immediate, significant yield increase
over the control, with the average yields trending higher over High-Yield Agriculture
time. Initially, yields were similar when N was applied without Yield Trajectory Requires Phosphorus Innovations
P (179N–0P), as the maize thrived on the existing soil P. After
The upward yield trajectory for the last eight decades shows
this, P gradually depleted and yields declined significantly to the
no signs of reaching a plateau (Fig. 1). This continued pattern in
point where P became the most limiting factor in the system. The
high-yield agriculture requires understanding the highly variable
gap between adequate and absent P fertilizer widened over time
differences in total P demand and uptake efficiency across species
as residual P depleted. The STP concentration gradually reduced
and cropping systems (Hopkins, 2019).
from 17 to 6 mg P kg-1 (Bray P1) for plots receiving no P fertil-
izer over the six decades of the study. It is noteworthy that the Key Species for Examining the Yield–Phosphorus Relationship
decline in STP was not dramatic between the first year and the Given limited space herein, we selected eight representative
years when yield declines were first measured. In the absence of species for the following discussion. We based our selection on
applied P fertilizer, these plants obtained P from the natural, low- crops with global importance as well as crops that have unique
level cycling that occurs in soil. This residual P supplied enough plant–soil P relations. Grain crops are the most important glob-
to grow a base yield, but fertilizer P was necessary for elevated ally in value and acreage (USDA-NASS, 2019) and as such utilize
yields. This and similar studies demonstrate the importance of the majority of fertilizer P sold. The top four grain crops—maize,
viewing fertilization from a long-term viewpoint and gives clues rice (Oryza spp.), soybean [Glycine max (L.) Merr.], and wheat—
on how we should move forward in fertilization science. were selected for discussion (Table 1). In addition, potato and
Understanding the role of soil and fertilizer P was vital in apple (Malus pumila Mill.) are the most globally important veg-
obtaining the steadily increasing yields through the Green etable and fruit crops, respectively. Finally, onion (Allium cepa
Revolution. One important concept going forward is under- L.) and sugar beet (Beta vulgaris L.) are globally significant and
standing the relationship between P soil supply and plant are included as key species as they represent unique root systems
demand. The transformation of solid phase soil P to plant-avail- and P relations. Each of these that have reliable long-term yield
able forms can be described in one of two processes: decompo- histories show an increase in relative yield (average for the last
sition of organic materials and solubilization of P containing decade compared with the baseline in Fig. 1 at 638, 522, 327,
minerals. The rate of microbial decomposition of crop residues 241, 215, 172% for potato, maize, rice, wheat, soybean, and
or soil OM is largely independent of the rate of P uptake by sugar beet, respectively). Historical yield records for onion and
plants—with both similarly dependent on soil temperature apple are not as reliable (and therefore not shown in Fig. 1), but
and other factors. This is similar for the rate of dissolution of evidence shows upward yield trajectories (Lazicki et al., 2016;
P-containing minerals, although equilibrium chemistry will USDA-NASS, 2019). The upward yield trajectories and unique
result in a slightly elevated dissolution rate as soil solution P soil–plant P relations in these eight species, which are largely rep-
concentration drops. However, the bottom line is that the rate resentative of the diversity of crops found in the world, warrant
at which P is converted to plant-available forms in the soil is further examination of the role of P in high-yield agriculture.
not proportionally linked to uptake rate. In other words, supply Cropping Systems Outpace Fertilizer Recommendations
will not completely keep pace with demand, and therefore, new
and innovative ways to enhance supply need to be established as Initially, crops were fertilized very similarly regardless of spe-
demand for P increases due to increasing yields. Although adding cies. With time, individual nuances with species, and even within
high rates and building the labile and nonlabile P fractions will species, were realized and custom fertilization resulted in ben-
improve supply, it is often not enough to meet the high demands efits. The past several decades have seen significant advances in
of high yield agriculture (Hopkins, 2015). soil, nutrient, and fertilizer sciences. However, the approach to
manage P fertilizers has also remained largely similar. The primary
method of fertilization for most growers involves soil analysis to other aspects of modern cropping systems changed dramatically.
determine the P rate with application based on one of the follow- Changes (such as new crop cultivars and hybrids, new fertilizer
ing approaches: (i) applying a sufficient amount of P to maximize materials and application methods, and even changes in soil and
the economic return of the fertilizer in the year of application, climate conditions) may require adaptation of fertilizer strategies.
(ii) maintaining P by applying the amount of P removed by the Fulford and Culman (2018) called into question the suitability
crop, or (iii) applying an excess of P to build up the STP and then of current fertilizer P recommendations aimed at maintaining
applying a maintenance amount once STP reaches a high level STP due to higher yielding systems. They stated that the recom-
(Olson et al., 1982). Dry fertilizer (most commonly diammo- mendations require updating to better reflect fertilizer needs
nium phosphate [DAP] or monoammonium phosphate [MAP] with modern maize and soybean productions systems. Heckman
is broadcast applied and incorporated into the soil with tillage. et al. (2006) took a similar view, pointing out a need to at least
Growers in high-yield environments will often apply a liquid fer- apply a removal rate of P to soils that are not excessively high
tilizer in a concentrated band near and/or on the seed or seed in STP. Hopkins et al. (2018) conducted a meta-analysis with a
piece (Borges and Mallarino, 2001; Rosen et al., 2014; Preston et variety of crops, with results suggesting P responses at STP con-
al., 2019). In reduced tillage systems, ammonium polyphosphate centrations higher than typical critical levels.
[APP] is commonly dribbled in a concentrated band on the soil Yields Are Largely Proportional with Phosphorus Demand
surface or, more effectively, injected into the soil near the seed
(Preston et al., 2019). These practices remain the standard means Increases in overall biomass clearly result in increases in the
of P management despite dramatic changes in yields. amount of P in that biomass. Although concentrations are vari-
These traditional P management approaches are insufficient able by species, plant part, and development stage (Table 2), every
as standalone techniques in high-yield systems (Hopkins, 2019). plant cell contains P. Despite these differences in plant tissue P
While the principles and practices are still relevant, many gen- concentration, that variation is less important than variation in
erally accepted soil critical levels and corresponding fertilizer crop yield when determining total P requirements (Wortmann
recommendations remained unchanged over decades while et al., 2018). Table 2 shows that a majority, although not all, of
Table 2. Phosphorus uptake (International Plant Nutrition Institute, 2014), average crop yields (Lazicki et al., 2016; USDA-NASS, 2019, averaged
across 2007–2016), P removal at average yields (USDA-NASS, 2019), leaf P concentrations (Bryson et al., 2014), and P concentrations in the harvested
portion (Lardy and Schafer, 2014; USDA-ARS, 2019; USDA-NRCS, 2019) for eight key species.
Uptake Yield Removal Leaves Harvested portion
kg P Mg -1
Mg ha -1
kg P Mg -1
kg P ha
-1 % %
Potato 0.92 46.9 0.56 26 0.20–0.50 0.04
Maize 4.19 10.1 2.70 27 0.25–0.50 0.21
Rice 3.67 8.2 2.54 21 0.10–0.18 0.10
Wheat 5.24 3.1 3.83 12 0.20–0.50 0.42
Soybean 7.86 3.1 6.48 20 0.25–0.50 0.70
Sugar beet 0.61 64.6 0.45 29 0.45–1.10 0.24
Onion 0.21 56.6 0.14 8 0.35–0.50 0.03
Apple 0.25 34.8 0.04 2 0.10–0.40 0.01
Fig. 5. Relative shoot and root sizes at early stages of growth (~60 d) and near maturity for (a) onion, sugar beet, potato, soybean, spring wheat,
rice, and maize and (b) apple. Adapted from Weaver (1926), Weaver and Bruner (1927), Atkinson and Wilson (1980), Atkinson (1983), Yamaguchi
and Tanaka (1990), Thorup-Kristensen (2006), Iwama (2008), Andresen et al. (2016), Fan et al. (2016), and Ordóñez et al. (2018).
Table 3. Comparison of root characteristics throughout the soil profile at the stage of maximum shoot growth for six species (adapted from
Yamaguchi and Tanaka, 1990).
Dry weight Length Surface area Volume Average diameter
gm -2
km m -2
m m
2 -2
×10 cm m
3 3 -2 mm
Maize 316 50 88 12.9 0.56
Wheat 247 89 98 9.1 0.36
Rice (flooded) 203 78 86 7.6 0.35
Soybean 152 39 48 4.8 0.39
Sugar beet (excluding tap root) 114 52 64 6.5 0.39
Potato 61 21 22 1.8 0.33