High expectations


I've rarely been so excited to read an article. Last week when I saw Energy use and greenhouse gas emissions from turf management of two Swedish golf courses, by Tidåker et al., I immediately dropped what I was doing and read it.

If you've talked with me about turfgrass management sometime in the past 18 months, our conversation may have touched on differences in energy use, and the difference in carbon emissions, caused by differences in grass selection and maintenance practices. In fact, this is one of the topics Dave Wilber and I discussed as part of our wide-ranging conversation during episode 14 of the Turfgrass Zealot Project. I don't know how to make these calculations yet, but finally with this article I've read something that provides the calculations, and that I can study so I can figure out how to do this myself.

Gelernter et al. wrote in 2014 about quantifying sustainability on golf courses. We suggested measuring and tracking the annual:

  • quantity of fertilizers applied
  • quantity and toxicity of pesticides applied
  • quantity of water used
  • fuel volume
  • labor hours
  • electricity used

One can keep track of those quantities, together with the associated costs, and from that one can check the efficiency of the operation. These quantities also serve as some of the basic data requirements for the GEO OnCourse program.

But the quantities we wrote about in the GCM article are all different: kg of N, kg of fungicide, L of water, L of diesel, kWh of electricity. By expressing all the turf maintenance activities in units of greenhouse gas emissions (expressed as CO2 equivalents) or energy use, one then has a single number for the entire course, or for an area of the course, or per square meter, that can be used to compare to other courses next door or around the world. And the use extends well beyond comparisons to other golf courses; one can use these units to compare the maintenance of a golf course to anything that has greenhouse gas emissions or energy use.

I had high expectations for the article, and I wasn't disappointed. The authors described the fertilizer rates, topdressing rates, water use, mowing frequencies, and much more, for the two courses, and then expressed those units in GHG or energy use. N rates were up to 22 g/m2, as were K rates (I think the rates for golf course turf in Sweden should usually be less than reported in the article -- using precision fertilization, or temperature-based growth potential and MLSN, will lead to lower recommended amounts of fertilizer). Sand topdressing on greens was about 10 mm/year. Irrigation of greens was about 300 mm/year. Mowing of fairways was about 85 times/year, and greens were mown about 180 times/year.


I think this is fascinating because one can consider Sweden to have relatively low inputs. If you're familiar with golf course maintenance in a tropical environment, let's say in Phuket, you might expect fairways to be mown more than 150 times a year, greens more than 300 times, about double the fertilizer, and more than twice the water use. Now imagine what happens when comparing irrigated vs non-irrigated rough? Seashore paspalum wall-to-wall vs. manilagrass? A 60 ha sandcapped golf courses vs. one with drainage and 2 cm of sand topdressing? Overseeded vs. not? The differences in energy use and greenhouse gas emissions will be huge.

What did Tidåker et al. find in their analysis? The entire paper is worth a careful study, but in summary they found mowing was the most energy-consuming activity, and mowing together with the production and application of fertilizers (especially N) contributed the most to greenhouse gas emissions. They suggest:

Appropriate measures for reducing energy use and carbon footprint from lawn management are thus: i) reduced mowing frequency when applicable, ii) investment in electrified machinery, iii) lowering the mineral N fertiliser rate (especially on fairways) and iv) reducing the amount and transport of sand for dressing. Lowering the mineral fertiliser rate is of particular importance, since GHG emissions originate from both the manufacturing phase and from N turnover after application.

Jason Haines has some interesting reads about how turf condition can be improved while at the same time reducing inputs:

New paper on variability of hybrid bermudagrass used on putting greens

If you work with warm-season grasses, you will want to have a look at this new paper by Reasor et al. on the variability of hybrid bermudagrasses used on putting greens.


Ever see anything like this? Off-types growing in a green? Wondered if the off-types are contamination by a completely different grass, or if the grass has mutated?


This paper explains what can happen, what has happened, and why. Plus it has a historical review of these hybrid bermudagrasses used on greens. Find out where they came from and how the grasses are related.

ReasorSometimes I write about papers that are behind a paywall and most people can't read (or at least don't want to pay the high fees to purchase). I'm glad there won't be that problem with this article, as Reasor et al. have published this open access so everyone can read it.

I've just spent a couple weeks with the lead author Eric Reasor (pictured at right in Japan) collecting data from bermudagrass putting greens in Asia.

He's been doing a lot of interesting research about ultradwarf bermudagrass, off-types within those grasses, and the management of putting greens to minimize problems with off-types. Watch out for more interesting research from him on this topic.


The MLSN guidelines, data, and reproducible research

Our preprint on the MLSN guidelines is now available. It was published today at PeerJ Preprints, as Minimum soil nutrient guidelines for turfgrass developed from Mehlich 3 soil test results. We wanted to share what we have done so far, make this paper available for citation in case anyone needs to cite something more technical than our 2014 GCM paper, and also solicit feedback about this paper before we submit it for peer review.

If you are interested in this, you probably care just about the article. Maybe just the abstract of the article. Maybe the abstract and a glance at the introduction and then a skip to the discussion and conclusions. That's fine. We'll be glad if you read any of it.

Beyond the article itself, I want to share what I'm most interested in with this project. That's the reproducibility of it. And the openness of it. We are sharing the results, and also the data and the code to generate the results and the code to generate the paper itself.



We  want to make sure that anyone who wants to read it can do so, so we share it as a preprint, and will make sure if a later version is published, that it is open access. You won't have to worry about clicking to read the article and hitting a paywall. I hit a couple paywalls this afternoon in my own research, and snapped these screenshots.



Those type of paywalls won't happen with this project.

Beyond that, however, the paper is reproducible*. That is, we are sharing all the data, all the code, and all the text; you can run the files and generate the exact same results -- in fact, the exact same pdf. You probably don't have all the software on your computer to do that, but you could. It is all open source and free. R, LaTeX, and some R libraries. We used knitr, VGAM, xtable, and dplyr in this project. You can check our files and see which libraries we used. You can check the code to see how we made the figures. How the values in the tables were calculated. You can see what functions we wrote to calculate the MLSN guidelines.

With this type of work, you can see what we did, and you can also see how we did it.

Furthermore, we've made the data, as we did with the Global Soil Survey data, freely available with no copyright. You want to study soil test results and have a need for more than 16,000 samples, or a subset of them? Have at it!

*reproducible research -- if you are interested in this, I suggest reading this post at Simply Statistics:

The Real Reason Reproducible Research is Important

High quality turfgrass is often produced in soils that don't have enough nutrients to produce high quality turfgrass.

That's the first sentence of our article about the development of the MLSN guidelines, published today as a preprint at PeerJ Preprints. You can read the article there and find out how (and why) we developed the guidelines.

We have also shared all the data used to develop the guidelines, and you can find the code used to calculate the guidelines in the 2016_mlsn_paper folder on GitHub.



That's not the way it is supposed to work

Of the many interesting things in the report by Gelernter et al. on the GCSAA's second nutrient use survey, I was especially intrigued by the part about soil testing.

First, a little background. If one has no idea how much of any mineral element is in the soil, then the logical amount to apply as fertilizer is just a little bit more than the grass can use. This guarantees that the grass will be supplied with all of each element that it can use. That's like an estimate of the maximum amount of fertilizer to apply.

Why soil test? Because soil testing allows for more efficient application of fertilizer. After finding out how much is in the soil, one can often reduce the quantity of fertilizer applied, because one knows that the soil can supply some portion of the plant's requirements.

With no soil testing, it makes sense to apply all that the grass could use. With soil testing, it makes sense to apply only the amount that the grass could use that can't be supplied by the soil. It's evident that the maximum amount of fertilizer should be applied when one doesn't know the nutrient content of the soil, and that in the most infertile soils, the quantity of nutrients required as fertilizer will be close to the maximum, with the quanity required as fertilizer decreasing as soon as the soils have some quantity of nutrients.

In the last chapter of A Short Grammar of Greenkeeping, I wrote that "I'd recommend soil testing, because in most soils the correct interpretation of soil tests can reduce the quantity of fertilizer that is applied."

You may have heard me say that soil testing is broken. For more background:

Now back to the GCSAA nutrient use survey results. Here's what the survey says:

"Despite the fact that respondents said that they used soil tests to reduce reliance on fertilizers, higher use rates were observed for respondents who conducted soil tests (Table 7). This apparent contradiction may be due to some of the turf fertility guidelines currently in use, which target higher nutrient levels than are required for acceptable turf growth ...

As a result, those who conduct soil tests with the belief that it will help them to reduce fertilizer inputs may end up unintentionally increasing fertilizer instead, likely because the guidelines used to evaluate their results may be higher than necessary."

That's not the way it is supposed to work. For more, check out the fertilizer and soil categories on the blog.

Every spring when the snow melts ...

I look forward to some photos from Doug Soldat. For the past three years, he's had some fascinating photos to share of snow mold on creeping bentgrass. And each year, there was more snow mold where potassium fertilizer was applied, and less snow mold where potassium wasn't applied.

Spring of 2014

In the spring of 2014, there was more snow mold where K was applied.

Spring of 2015

Last year, there was also more snow mold where K was applied.

Spring of 2016

This year, it happened again. There was more snow mold where K was applied.

Doug will be talking about K in a TurfNet webinar in April: Is Your Potassium Program Hurting or Helping Your Turf?

On those creeping bentgrass plots in Wisconsin, adding K increases snow mold. No K had less snow mold.

At Rutgers, annual bluegrass plots deficient in K have had more anthracnose in summer and more winter injury. Eliminating the deficiency reduced those problems.

Then there is the MLSN guideline for K of 37 ppm. I recommend keeping the soil K above 37 ppm (Mehlich 3 extractant).

And there are hundreds of other studies about K. Some show a benefit from adding K, and some don't. I haven't read all of them, but I have read a lot of them. This sounds like it could be pretty complicated.

Actually, I don't think it is. Here's what seems to be the case, for both warm-season and cool-season grasses:

Ensuring the grass is supplied with all the K it can use will provide all the benefits associated with K. Adding more than that usually has no effect, other than wasting time and money, but sometimes has a negative effect.

As a turfgrass manager, all one has to do is ensure the grass is supplied with all the K it can use. This can be accomplished in 2 ways. One is by keeping the soil K above the MLSN guideline. A second is by applying N:K in a 2:1 ratio for cool-season grasses, a 1:1 ratio for seashore paspalum, and a 3:2 ratio for other warm-season grasses. I wrote about that in the final chapter of A Short Grammar of Greenkeeping and in The (New) Fundamentals of Turfgrass Nutrition.

Note that I do not recommend tissue testing for K (or any other element).

If you want to read more about K specifically, and about how the benefits of K come from correcting a deficiency, I recommend:

"Knowing which soil test results are important can simplify turf management"


Bill Kreuser's guide to soil test interpretation; read it! Here's how he describes it:

"While soil tests can be useful, their results are frequently overanalyzed and overinterpreted. Sometimes soil test results can be more confusing than helpful. It doesn't have to be so difficult. The goal of this publication is to explain which soil test values are important and which values can be ignored."

After reading the publication, I think it achieves that goal.

This is what PAR looks like

I downloaded NOAA quality-controlled data for Corvallis on 5 minute intervals. An analysis of these data, and a comparison to Ithaca, are in this report.

These charts show the photosynthetically active radiation (PAR) by time of day, using the 2014 data.

This is the photosynthetic photon flux density (PPFD) every 5 minutes on 13 June 2014.


It was cloudy for most of 13 June, with only a few of the 5 minute intervals having maximum potential PAR.

On a sunny day, like 10 August 2014, shown here, the PPFD increases from sunrise until a peak at solar noon, then decreases until sunset.

10 August 2014 PPFD at Corvallis

In these charts, the times shown are standard time, and the morning and afternoon separation is made based on the time being before (morning) or after (afternoon) solar noon.

That same week that contains 13 June is the 24th week of 2014, and all the measurements from that week are shown in the next chart.

PPFD for a week in June

For the entire month of June, the PPFD looks like this.

PPFD for June 2014

Then for an entire year, the PAR looks like this.

PPFD at Corvallis in 2014

Soil test phosphorus in turf soils: 2 datasets

Selection_090Landschoot et al. wrote about a large dataset of Mehlich 3 phosphorus data in Summary of Mehlich-3 P Data from Home-Lawn Soil Tests in Pennsylvania.

How are the data distributed? The median of the Pennsylvania data is 57 ppm, and 40% of the samples are less than or equal to 45 ppm, which is the cutoff level for P fertilizer recommendations to home lawns by the Penn State Agricultural Analytical Services Laboratory (AASL). Home lawn samples greater than or equal to 45 ppm P will not receive a fertilizer recommendation from AASL. For golf course putting greens the recommendation is different; P fertilizer is recommended by the AASL for putting greens when the soil test P is less than 90 ppm.

I looked at the Global Soil Survey (GSS) Mehlich 3 P data through August 2014. For these GSS data, the median value is 76 ppm, compared to 57 ppm for the Pennsylvania data, and only 35% of the GSS samples are less than or equal to 45 ppm, compared to 40% falling below that value in Pennsylvania lawns. Here is a histogram of the GSS data collected through August 2014:

Histogram_gss_m3pIn the GSS data, 11% of the samples were below the MLSN guideline of 21 ppm for P. The average expected P use by grass on putting greens is about 22 ppm, based on the average nitrogen application rate to putting greens in the US. Using the MLSN interpretation of the GSS data, and assuming P use at an average level, 33% of the GSS samples would receive a P fertilizer recommendation, compared to 40% of the Pennsylvania samples.

2 similar approaches to fertilization, with 1 notable difference

At first appearance, the demand-driven fertilization of STERF seems almost the same as the growth potential (GP) and MLSN approach. If you are not familiar with this approach from STERF, you can download their Precision Fertilisation -- from theory to practice, written by Tom Ericsson, Karin Blombäck and Agnar Kvalbein.

And I recommend you do download it. It is a great explanation of turfgrass nutrient use and requirements in only 20 pages.

Selection_089First, the similarities. These quotations will sound familiar, but they are not from me. These are quotes from the Precision Fertilisation handbook -- text in [ ] brackets is mine:

"fertilisation can be adapted based on the nitrogen requirement of the grass"

"light and heat control the growth potential of grass"

"warm summer days lower the nutrient requirement" [of cool season grasses adapted to a Nordic climate]

"when photosynthesis is slower, there is a decrease in the growth capacity of grass and thus also in its nutrient requirement"

"nitrogen is the nutrient that grass plants require most"

"By controlling the nitrogen concentration in the leaves through fertilisation, the growth rate is also controlled. A growth rate corresponding to 60% of maximum growth is often sufficient to produce a surface with good playing qualities. However, if the turf needs to repair some form of damage, the growth rate needs to lie around the maximum capacity for a period and therefore the nitrogen concentration also needs to be higher.

An experienced greenkeeper can judge from the colour of the grass whether the fertilisation level is right or wrong. The amount of clippings produced also sends a clear signal about the nitrogen of the grass."

"Since the potential growth of grass is controlled by the availability of light, heat, and water, the fertilisation level must be adapted to the growing conditions."

[When the growth capacity is reduced:] "The same argument applies when the cutting height of the turf is lowered before competitions. When the leaf area is reduced, the capacity of the grass to capture solar energy is also reduced. This decreases the growth capacity and the nutrient requirement. In order to avoid changes in the growth pattern of the grass above and below the ground and to maintain leaf structure and carbohydrate levels in the tissues, the fertilisation intensity must be decreased." [see also this on the lower nutrient requirement of stressed turf]

"Adding extra potassium in late summer/autumn to turf which already contains a surplus of this compound therefore has no additional effect on the ability of the grass to survive low temperatures. Addition of extra phosphorus in the spring when the soil temperature is low is also superfluous when a well-balanced fertiliser is supplied in small, frequent doses. As discussed, a moderate lack of nitrogen poses no serious problems for the health of turfgrass and in fact actually increases the quality of the turf."

So what is the difference between demand-driven fertilization of STERF and the MLSN approach I use? It is accounting for nutrient supply from the soil.

I say, follow the approach exactly as described by Ericsson et al. in their Precision Fertilisation, but if the soil is above the MLSN guideline for a particular element, then that element does not need to be applied as fertilizer, because the soil can supply enough to meet the grass requirements.

Ericsson et al. write that "greens built according to USGA norms have a low capacity to bind and supply nutrients to the grass. On this type of green, it is very important that the fertiliser compound used contains all essential nutrients." I would say that the fertilizer does not need to contain all essential nutrients, if the soil can supply them. And if the soil has nutrient levels above the MLSN guideline, then I am confident the soil can supply them. If the soil will drop below the MLSN guideline, then I agree, it is important that the fertilizer compound used contain that element.