Just what the grass requires: Using minimum levels for sustainable nutrition
Good turf performance can be achieved at lower nutrient levels.
Micah Woods, Ph.D.; Larry Stowell, Ph.D.; and Wendy Gelernter, Ph.D.
Read this story in GCM's digital edition
In experiments with creeping bentgrass at Cornell University, a wide range of soil potassium levels were established in these research plots, but no benefit to applied potassium was observed even when the soil potassium levels were well below the conventional guidelines.
Photos by Micah Woods
In 2012, we introduced the minimum levels for sustainable nutrition (MLSN) as an
alternative to conventional soil nutrient guidelines (7). Conventional
guidelines are epitomized by the low, medium, high and very high classification
scheme described in the third part of the “Clarifying Soil Testing” series
published in GCM
years ago (1).
light of recent trends in reduced inputs and increased sustainability, and
taking newly published data into account, the conventional approach requires
scrutiny and significant revision. Conventional guidelines are not only complex,
they are also relatively static, without regular or systematic updates.
regular updates seem like a good idea, because many research projects suggest that
high-quality turf can be produced at levels below the conventional guidelines
(2,3,4,6,8). As an alternative to the conventional guidelines, the MLSN
guidelines are an attempt to identify not the optimum levels for soil
nutrients, but rather the minimum levels of soil nutrients at which we can be
confident of good turf performance.
This creeping bentgrass green at Takarazuka GC near Osaka, Japan, has calcium, magnesium and potassium levels not only below the conventional guidelines, but also below the MLSN guidelines, yet still has produced excellent turfgrass conditions since the soil was first tested in 2009.
may have seen the same thing yourself: high-quality turf with no problems, growing
in a soil classified as low in one or more essential elements. The question
then arises, if the soil is lacking in these elements, why is the grass
performing so well? Adding a nutrient may change the soil test result to move
the level up to a desired range, but if the nutrient addition has no effect on
the grass performance, is it necessary?
How to use the MLSN guidelines at
MLSN guidelines (Table 1) take a new approach to soil test guidelines.
Turfgrass managers will have two questions about fertilizer application, and
the MLSN guidelines answer both of them. The first question is, “Does this element
need to be applied as fertilizer?” As a follow-up to the first question, one
also needs to ask, “If this element is required, how much should be applied?”
Penn A-1 creeping bentgrass grown in soils with decreasing levels of potassium from left to right; keeping soils at or above the MLSN guideline provides a level of safety that such deficiency symptoms will not occur.
answer the first question, simply compare the MLSN guideline value for an
element to the soil test level for that element. If the element is below the
MLSN guideline, or if the estimated use of that element will drop the soil to
the MLSN guideline during the course of the growing season, then that element
should be applied. If the element, as measured by the soil test, is above the
MLSN guideline, and if estimated use of that element during the growing season
will keep the soil above the MLSN guideline, then that element is not required as
answer the second question, regarding how much of an element to apply, simply
add enough of that element to keep the soil at or above the MLSN guideline at
the end of the growing season. To calculate that, compare the soil test result
to the MLSN guideline and to an estimate of how much of that element the grass
try the MLSN approach, you will need some recent soil test results from tests
done using the Mehlich 3 extractant. You will also need an estimate of how much
nitrogen will be applied to your turf in the upcoming year. Because nitrogen
controls the uptake of other nutrients (5), we can use the nitrogen estimate to
predict the grass’s use of other elements. Before making the calculations, we
will make some assumptions about grass growth and the relationship between
fertilizer applied to the two-dimensional soil surface and the soil test levels
within the three-dimensional root zone. These include:
- The grass cannot use more of an element than it harvests.
- The growth and consequently the nutrient uptake are determined by the amount of
- The concentrations of macronutrients and secondary nutrients in the leaves will
be estimated as in Table 2.
gram of an element spread over 1 square meter on the surface is equivalent to 4.4
ppm of that element in the root zone of 1 square meter to a 15-centimeter
depth, and vice versa.
- One pound of an element spread over 1,000 square feet on the surface is
equivalent to 22 ppm of that element in the root zone of 1,000 square feet to a
6-inch depth, and vice versa.
Example 1: Potassium
say the potassium soil test level is 52 ppm, and we plan to apply 3 pounds of
nitrogen/ 1,000 square feet in the upcoming year. How do we determine the potassium
requirement to ensure we stay above the MLSN guideline for potassium of 35 ppm?
As shown in Table 2, the grass is expected to use half (0.5) as much potassium
as it does nitrogen. That is, we predict the grass will use 1.5 pounds of potassium/1,000
square feet, which is equivalent to a depletion of 1.5 * 22 = 33 ppm from the
soil. Because we want to keep the soil at or above the MLSN guideline, the
total amount of potassium required is the plant use (33 ppm or 1.5 pounds)
added to the amount we want to ensure remains in the soil (35 ppm or 1.6 pounds).
In our example, this is 68 ppm or 3.1 pounds. The amount of potassium in the
soil test is 52 ppm (2.4 pounds). The amount required as fertilizer is the
difference between the amount required (68 ppm or 3.1 pounds) and the amount
actually present (52 ppm or 2.4 pounds), which comes to 16 ppm or 0.7 pound. Thus,
the fertilizer requirement for potassium in this situation using the MLSN
guidelines is 0.7 pound potassium/1,000 square feet.
Example 2: Magnesium
the soil test level for magnesium is 75 ppm and we plan to apply 3 pounds of
nitrogen/ 1,000 square feet in the upcoming year, how do we determine the
magnesium requirement to ensure we stay above the MLSN guideline for magnesium
of 54 ppm? As shown in Table 2, we expect the grass to use 20 times more
nitrogen than magnesium. That is, we predict the grass will use 0.15 pound of
magnesium/ 1,000 square feet (3 pounds nitrogen * 0.05 = 0.15), which is
equivalent to a depletion of 0.15 * 22 = 3.3 ppm from the soil. We want to keep
the soil at or above the MLSN guideline, so the total amount of magnesium
required is the plant use (3.3 ppm or 0.15 pound) added to the amount we want
to ensure remains in the soil (54 ppm or 2.5 pounds). In our example, this is
57.3 ppm or 2.6 pounds. The amount on the soil test is 75 ppm (3.4 pounds). The
amount required as fertilizer is the difference between the amount required
(57.3 ppm or 2.6 pounds) and the amount actually present (75 ppm or 3.4
pounds). Because the amount present is more than the amount required, we do not
need to apply any magnesium to keep the soil above the MLSN guideline.
How the guidelines were developed
From 2006 to 2009, more than 50 varieties of warm-season grasses were grown
at the Asian Turfgrass Center research facility north of Bangkok in soils
with nutrient levels below the conventional soil guidelines, yet the turf
still met all performance goals.
started with soil test data from the PACE Turf database. This consisted of data
from more than 17,000 individual soil samples, each drawn from a stand of turf
that was performing well. Because the data in those samples were from sites
where turf performance was good, we could expect that whatever the nutrient levels
were at those sites, those levels would be sufficient to produce turf that
performed well. Then we filtered the data, selecting only the data from sites
with a cation exchange capacity (CEC) less than 6 cmolc/kilogram.
filter removed all the soils with high nutrient-holding capacity from the
working data set. We wanted to look at only the soils that had a relatively low
nutrient-holding capacity, yet still produced good turf conditions, to
investigate and identify the individual nutrient levels in those soils. For the
MLSN guidelines, we assume that if there is enough of an element to produce
good turfgrass in a low-nutrient-holding soil (such as a sand root zone from a
golf course putting green), then the same amount of that element will be
sufficient to produce good turfgrass conditions in a more nutrient-rich soil
that has a higher CEC. We think that if there is enough of an element to
produce good turfgrass in a sand root zone on a golf course putting green, then
the same level of that element in a soil-based green or on a golf course
fairway will produce good turfgrass as well.
added one more filter to the data. This was for pH. We selected only those
samples with soil pH from 5.5 to 7.5. The purpose of this was to develop
guidelines that would be accurate for a range of elements using the widely used
Mehlich 3 soil test extractant. When soil pH is less than 5.5, we recommend
application of liming materials to reduce soluble aluminum, to increase soil
microbial activity and to reduce the risk of toxic soil-soluble ammonium levels.
Because of that, there was no reason to include soils with a pH of less than
5.5 in the data set.
soils with a pH above 7.5, there is a high probability that the Mehlich 3
extractant may dissolve some soil minerals that contain calcium or magnesium.
Such dissolution would have introduced error into the guidelines, which we
avoided by selecting for a pH range at which mineral dissolution is minimal,
and above which magnesium and calcium would not be deficient.
the two filters were applied, we were left with a working data set of more than
1,500 soil samples. These were from turf that performed well, had a relatively
low CEC typical of golf course putting greens or relatively sandy soil, and a
pH of 5.5 to 7.5. Because all of these soils were producing good turf, one
could conclude that all the soils had sufficient nutrients, so anything at or
above those nutrient levels would be fine. Rather than divide the data from
these soils into low, medium and high classifications, we took a different
approach, in which we modeled the distribution of the data for each element
concentrations in the soil are a continuous random variable with a minimum possible
value of zero and a virtually unlimited maximum possible value. We analyzed the
filtered data set using EasyFit distribution-fitting software from Mathwave
(www.mathwave. com) and found a good fit for each element in these soil test
results with a three-parameter loglogistic distribution. From this modeled
distribution, based on the actual data from turfgrass sites that had good
performance, we identified the MLSN guidelines. A visual representation of the
cumulative distribution function is shown for the potassium data in Figure 1
and for the phosphorus data in Figure 2.
High-performance turf at Keya GC near Fukuoka, Japan, is maintained in soils with sulfur and magnesium near the MLSN guideline and potassium below conventional guidelines. Adding data from sites like this helps to improve the accuracy of the guidelines as they are updated.
we look at the data for potassium, for example, we see the cumulative
proportion of the samples at any particular level as we go from 0 to 280 ppm.
The conventional guidelines would seem to be taking a number of sites with good
performance and then choosing to target the higher end of that range as a
MLSN, we take a different approach, taking the data from thousands of sites
with good performance, assuming that there must be enough nutrients available
to produce good turf because the sites are already performing well, and then
selecting a conservative value at the 0.1 level at the lower end. Because we
have already omitted the sites with bad performance from our data set, we can
have some confidence that these apparently low levels are sufficient to meet
the requirements of the grass.
Four advantages of this approach
The guidelines are based on real data from actual turfgrass sites. We worked
only with a data set from sites with good performance, omitting soil test
results from problem areas and nutrient-deficient soils. The modeled distribution
is a mathematical representation of the soil nutrient levels as they are
distributed on actual turfgrass sites. Because the data are carefully selected
from soils that are already producing good turf, there is a layer of safety in
the model. That is, any clearly deficient soils were not included in the model,
so the results are not skewed lower by nutrient-deficient soils.
Once the model has been fit to the actual data, we can select a base level we
wish to stay above. Again, this model and the associated level are based on the
actual nutrient levels in the soil at sites where turfgrass performs well. We
chose the nutrient level coinciding with the 10th percentile to define the MLSN
guideline for each element. At this level, 10% of the samples in the data set
would have a lower soil nutrient level than the selected MLSN guideline.
We can calculate a sustainability index for each element, based on a comparison
of the concentration of that element on a soil test with the modeled MLSN
distribution for that element. The sustainability index is the proportion of the modeled distribution that reports values lower than the sample soil test value. This is a metric that assists turf managers in the evaluation of soil
nutrient levels over time. It also provides a guide for the development of nutrient
management programs. Perhaps most important, the sustainability index
identifies and rewards the restriction of nutrient inputs when they are not
necessary to meet turf performance goals.
The MLSN guidelines are easily updated as we add new data from turfgrass sites with
good performance (see the sidebar on page 138).These guidelines are
self-correcting. Using this method and continuously adding to the reference
data set with soil test data from turfgrass sites that perform well, we will
see the guidelines move up if they are too low or down if they are too high. In
short, these guidelines are designed to be updated as the core data set grows,
and the MLSN guidelines will adjust based on samples added to the data set from
turfgrass that performs well on various soils and across a wide geographic
more about these guidelines, videos explaining the guidelines, and a link to
the most current version of the guidelines, see: www.paceturf.org/journal/minimum_level_for_sustainable_nutrition. To join other turfgrass managers from around the world
in a discussion of these guidelines or to pose questions about the guidelines,
go to the MLSN page on Facebook: www.facebook.com/mlsnturf. For even more
examples of how the MLSN guidelines fit into turfgrass nutrient requirements and
how these requirements can be calculated, download Understanding Turfgrass
Nutrient Requirements at: http://calendar.asianturfgrass.com/understanding_
gratefully acknowledge the hundreds of golf course superintendents who have
submitted soil samples over the past 20 years, and whose soil test results were
the basis for development of the MLSN guidelines.
- Carrow, R.N., L. Stowell, W. Gelernter et al. 2004. Clarifying soil testing:
III. SLAN sufficiency ranges and recommendations. Golf
Course Management 72(1):194-198.
- Dest, W.M., and K. Guillard. 2001. Bentgrass response to K fertilization and K
release rates from eight sand rootzone sources used in putting green construction.
Turfgrass Society Research Journal 9:375-381.
- Fulton, M. 2002. Creeping bentgrass responses to long-term applications of
Course Management 70(2):62-65.
- Kreuser, W.C., P.H. Pagliari and D.J. Soldat. 2012. Creeping bentgrass putting
green Mehlich-3 soil test phosphorus requirements. Crop
- Kussow, W.R., D.J. Soldat, W.C. Kreuser and Steven M. Houlihan. 2012. Evidence,
regulation, and consequences of nitrogen-driven nutrient demand by turfgrass.
2012, Article ID 359284. doi:10.5402/2012/359284 (www.hindawi.com/isrn/agronomy/2012/359284/).
Verified Dec. 9, 2013.
- Raley, R.B., P.J. Landschoot and J.T. Brosnan. 2013. Influence of phosphorus
and nitrogen on annual bluegrass encroachment in a creeping bentgrass putting green.
Turfgrass Society Research Journal 12:649-655.
- Stowell, L., and M. Woods. 2013. Minimum levels for sustainable nutrition.
Proceedings: Constructed Rootzones 2012. Online. Applied
Turfgrass Science (www.plantmanagementnetwork.org/pub/ats/proceedings/2013/rootzones/8.htm).
Verified Dec. 9, 2013.
- Woods, M.S., Q.M. Ketterings, F.S. Rossi and A.M. Petrovic. 2006. Potassium
availability indices and turfgrass performance in a calcareous sand putting green.
MicahWoods is chief scientist at the Asian Turfgrass
Center and an adjunct assistant professor in the department of plant sciences
at the University of Tennessee, Knoxville. Larry Stowell and Wendy Gelernter are
the principals of Pace Turf LLC, San Diego, Calif.