Skip to Content

Iowa is Hemorrhaging Nitrogen

Posted on November 7th, 2019

Someone once told me “for god’s sake start at the end” when giving a presentation, and since there is a lot to unload with this one, I’ll give you a couple of nuggets if you don’t have time to read the rest:

  • For the 2019 Water Year (10/1/18 to 9/30/19) Iowa stream nitrate load was 980 million pounds.
  • Since 2003, statewide stream nitrate loads have increased 100.4% (i.e. doubled), as measured by the 5-year Running Annual Average.

Ok, here goes:

A “water year” (WY) is a convention used in hydrology to describe the period from October 1 to September 30. Scientists use this when evaluating hydrologic data because autumn and early winter precipitation accumulates in the soil as liquid water or on the surface as snow, and thus doesn’t reach rivers until spring.

Today I am going to present some nitrate-nitrogen (NO3-N) data for the 2019 WY along with similar data accumulated since 1999. This analysis didn’t all happen at once; much of this was aggregated and analyzed a couple of years ago and published in this paper and this paper. This blog post is just an update. Notes on data sources are at the end.

One note now: I do not presume that this accounting is my personal plaything; if somebody else wants to take a stab at this, hey, dive in, the water’s fine (not really).

Herein I look at the state in three ways: areas draining to the Missouri River Basin (31% of Iowa, abbreviated as IAMoRB), areas draining to the Upper Mississippi River (69% of Iowa-IAUMRB), and Iowa as a whole (IATot). There are objective reasons for dividing Iowa in this way—soils, climate, hydrology, and some other things.

Iowa areas draining to the Missouri River, Upper Mississippi River, and the Mississippi River watershed draining to the Gulf of Mexico.

Water from all Iowa streams eventually makes it to the Gulf of Mexico, where a “Dead Zone” develops every year as a result of Midwest nitrogen and phosphorus loss from farmed fields.

The amount (“load”) of NO3-N leaving the state during WY 2019 was 980 million pounds. This was the second-highest load since 1999 (2016 was the biggest—1.25 billion pounds). Of this total (2019), 39% left the state in the Missouri and 61% in the Upper Mississippi. One caveat here: the statewide load includes 2300 square miles of Minnesota that drain into Iowa, about the size of four Iowa counties.

When comparing watersheds, we often look at load per watershed area: NO3-N yield. In 2019 this was 34.2 pounds per acre (lbs/ac) for areas draining to the Missouri; 24.0 lbs/ac for areas draining to the Mississippi; and 27.2 pounds statewide.

Since greater than 90% of stream nitrate is from agricultural sources in Iowa, we also look at load per cropped area. These values are shown in the map below. Using a current price of $0.31/pound for anhydrous ammonia and with corn at $3.79/bushel, recommended nitrogen application rates range from 145-197 lbs/ac, so you can see that the losses are substantial relative to corn nitrogen needs.

I don’t have precipitation data aggregated, but 2019 was a wet year and river discharge was well above normal. When we calculate water discharge, we often use a convention called water yield. This is water volume per watershed area. Water yield “normalizes” discharge data and allows us to compare apples to apples when looking at watersheds of varying size. Average water yield for Iowa since 1999 is 11.1 inches; in WY 2019 it was 20.7 inches. These values for Iowa areas draining west to the Missouri are 8.7 (average) and 18.8 (2019) inches and east to the Mississippi, 12.0 (average) and 21.4 (2019) inches. Take home: eastern Iowa is wetter than western Iowa.

2019 Nitrate losses in pounds per crop acre.

It’s a good idea not to put too much stock into one individual year when looking at stream pollutant loading. Weather does a play a role in year-to-year variations. It’s not the driver of the problem, but it is important because we’ve removed all the natural resilience from the landscape. To account for anomalous years, we often look at the “5-year Running Annual Average” (5YRAA), which as the term implies, is the average of the current year and the previous four.

The 5YRAA is at its highest level since 2003 for Iowa as a whole (my dataset begins in 1999; the first year I can calculate the 5YRAA is 2003), and for the two sub-regions draining to the Mississippi and Missouri.

Nitrate loading to the Missouri is increasing far faster than areas draining east toward the Mississippi. The 2019 5YRAA is about three times larger for western Iowa than in 2003. For areas draining east to the Mississippi, the increase is 74%, and statewide, 100.4% (i.e. doubled).

5-Year Running Annual Average Nitrate Loads for Iowa areas draining to the Missouri River (brown), Upper Mississippi River (blue) and the Gulf of Mexico (green).

Assuming 2013 as a starting point for the Iowa Nutrient Reduction Strategy, the 5YRAA for IAMoRB, IAUMRB, and IATot has increased 77%, 34%, and 46% since then, respectively. That 45% reduction goal? It’s 65% now, if we are to use actual water monitoring data as the metric. This illustrates the folly of trying to measure success by cataloging implemented practices.

I also look at where a raindrop has its biggest impact—i.e. where does it dissolve the most nitrogen. Here I use the metric of Flow Weighted Average Concentration (FWA). This is the total N load divided by the total water discharge. Units are mg/L (ppm).  This is shown the graph below. The NW Iowa watersheds of the Floyd and Rock usually are at the top of the heap, and 2019 was no exception. I tend to think this is because livestock concentration.

Flow weighted average nitrate concentrations in WY 2019. Streams draining to the Upper Mississippi River are denoted by blue bars; streams draining to the Missouri River by brown bars.


The WY 2019 stream nitrate load was just short of a billion pounds. Is that a lot? A billion to us seems like a big number but I can increase or decrease it by changing the units, which are a human construction. So we must put these things in perspective:

  • A human being excretes about 10 pounds of nitrogen per year. Thus the fecal and urinary waste of Iowa’s 3.2 million people in 2019 included about 32 million pounds of nitrogen, which is about 3.3% of the total stream load.
  • A billion pounds of nitrogen spread out among 85,000 Iowa farmers is ~12,000 pounds per farmer. 12,000 pounds is enough to fertilize about 80 acres of corn (if grown after a previous year’s soybean crop).
  • You would need 236,000 thousand-gallon anhydrous nitrogen nurse tanks (shown below) to hold a billion pounds of nitrogen.
  • A billion pounds of anhydrous nitrogen is worth more than $300 million at current prices (note: anhydrous is the usually the cheapest N form of many that farmers buy). Think about what that number might be extrapolated across the several cornbelt states. This is an objective fact: companies are making money selling this lost nitrogen, and taxpayers are asked to shoulder the burden for the resultant pollution.

The contents of 236,000 of these left Iowa in its rivers over the past year.

Final thoughts

I will be surprised if we ever have enough cost share (public) money to address the problem of water quality degradation from nitrate loss. Farmers can buy it for $0.31 per pound but it can cost $2 per pound (or more) to keep it out of our streams. I wrote about this more than two years ago in this space, before anybody read this blog. In that piece, so far read by only 41 people, I calculated that reaching water quality goals for nitrogen using cover crops would cost at least $175 per Iowan per year. Walk down your street and ask a household of four to fork over $700 every year to mitigate nitrate pollution. Something tells me people aren’t going to run to get their checkbook.

One last thought. Why have things degraded so much more in western Iowa in the last 20 years? This is an interesting question. I have some ideas but I don’t really have the answer.

Notes About Data

Stream nitrate data from 1999-2016 is from the Iowa DNR database. Since then, that data has been supplemented with IIHR and USGS real time water quality sensor data. I use linear interpolation to estimate nitrate levels on days without data. This method has been shown to be reasonable.

Stream discharge data is from USGS.

Individual site loads are calculated by multiplying actual or estimated daily nitrate concentrations by stream discharge data.

Region and statewide loads are calculated by looking at overall N yields for the measured areas, and then multiplying that yield by the actual areas of IAMoRB, IAUMRB, and IATot.

Sites do change some from year to year based on data availability. One important watershed—Maquoketa River—dropped out of the monitored areas in 2019.

In total, 79.5% of Iowa is included in the monitored areas; 62.1% of IAMoRB and 89.2% of IAUMRB.

Crop area data is from USDA (NASS).

Tags: , , , , , , , ,

17 Responses

  1. Brenda Brink says:

    Thank you for your and Larry Weber’s work. I attended the Drake Ag Law SOIL conference last week, mainly because you two were going to be there. Keep up the good work.

  2. Madeline Voss says:


    At the end of the article you say,
    “Why have things degraded so much more in western Iowa in the last 20 years? This is an interesting question. I have some ideas but I don’t really have the answer.”

    Can you share your thoughts? I am very interested in this. My family and I are debating moving back to Western Iowa and are looking for any advice/information.

    • cjones says:

      Less pasture, more row crops, more terraces (which increase N loss), more tiles.

      • Cindy Hildebrand says:

        I appreciate this information about terraces. I’m guessing the N loss is because most terraces are hooked into tile lines. Is that correct?

        I tried looking up information about terraces and water quality. I found terrace information on the Iowa Agriculture Water Alliance website under “Water Quality, Soil Health Solutions at Work: Terraces.” That entry talked a lot about how terraces reduce soil and P loss, but didn’t mention nitrogen. At all.

        A second entry was “Why do they do that? – Terraces and Tile Lines” on the Iowa Agriculture Literacy Foundation website. That entry also talked about soil and P, and not at all about nitrogen.

        • cjones says:

          Thanks for reading Cindy. Yes terraces can increase both flow and nitrate. See line from this paper below: file:///C:/Users/csjns/Documents/Journal%20of%20Soil%20and%20Water%20Conservation-2010-Gassman-381-92.pdf

          “The increased NOx-N levels were also unexpected and may have resulted from increased transport due to the large influx of tile terraces during 1988 to 1999 in the SMCW.”

          • Cindy Hildebrand says:

            Thank you, Chris. I was not able to access the paper itself, but that line from it is very interesting.

            It seems to me now that ignoring the terrace/nitrate issue when terraces are discussed on websites intended to inform the general public about farm conservation is disingenuous at best.

  3. Karin McKeone says:

    Is there any research about nitrate runoff from homeowners’use of chemicals on lawns? I see those granules on sidewalks and streets, waiting to be flushed in streams. Is this a part of the dead zone problem, or is farm run off the bigger problem?

    • cjones says:

      Thanks for reading Karin. Turf fertilizer is lost to streams primarily when the material is applied to impervious surfaces like concrete. There is not any evidence that i am aware of that shows nitrogen applied to the turf itself substantially affects nitrate levels in Iowa.

  4. John Norwood says:

    Fabulous discussion, rich in detail — data and interpretation.
    Thank you.

    So, here are some observations. Sorry for the length of my responses but your work inspires much thought!

    1) I think your distinctions between what’s happening in the western and eastern part of the state is informative and relevant for farmers, policy makers and scientists. 3x increase vs. 0.75 increase in loading. Let’s drill down more into the why is this happening and what do we do about it? The difference is considerable. I agree that our animal livestock holding capacity may play a big role as well as poor definition as to the value of manure being land applied vs. commercial N fertilizer. Manure may be viewed more as a waste product to be gotten rid of to enable contract payments for hogs to continue.

    2) If we are sending a 1 billion pounds of nitrogen downstream at $0.31/lb market value or $300 million, and part of the reason is nitrogen is cheap to apply (as you note) and farmers use it as a form of insurance to drive yield, we ought to align our crop insurance policy to good agronomy and water quality goals.

    3) What’s the total amount of nitrogen we use in this state? If we don’t have a dedicated fund to help our 3,700, mostly county controlled, drainage districts modernize, we should consider expanding the existing fee we collect on the production and distribution of commercial nitrogen sales to help create a fund that would generate $60-$100 million annually for design and engineering improvements that allow tile water to be filtered in constructed wetlands. This fund would also help acquire ag lands for this use and we could come up with a CRP program for continuing payments that wetlands provide, including aquifer recharge, wildlife habitat, and in some cases irrigation. In Floyd County, the County parks district owns some of this underlying green infrastructure. That’s an idea worth exploring.

    4) While the cost of nitrogen is cheap on a per pound basis, ISU extension information shows it’s the second most costly input after seed. What is further down on the cost of production is crop insurance and it is not priced to reflect some of these externalities we are concerned about. We might ask policy makers to look at crop insurance not only with respect to desired yields but also the stress certain practices put on our system, corn on corn, for example, because it drives more nitrogen. We could add a small fee to crop insurance to go into the fund, for example, for those who choose corn on corn.

    5) Going back to the idea of a dedicated infrastructure fund to address our tiling systems (12 million of our 23 million acres is tiled, 6 million of the 12 million is tiled). The cost to tile is not cheap. $800-$1,200 per acre or more. The cost of a CREP style wetland, I’m told is about $600,000. Divide that by 2,000 treated acres (for example) and we have an all-in cost of about $300/acre for a 100 to 150 year solution. If we constructed the 1,400 or so of the initial CREP style wetlands in this state, we could filter about 2.8 million acres of tile ground. That would be significant. In order to expand the fairness of this “user fee” on nitrogen, I think everyone, in my vision, who uses or touches nitrogen would get to help pay to address its negative effects. This means urban and rural users of nitrogen would contribute. Users of tile pay a fee on new tile and there might be county connection fees for new tile or expanded capacity like we see in cities. And largely exempt users of nitrogen via manure, which is largely unregulated and non-commercial material since the content of the manure is not consistent, would get to pay into the fund based on the volume of manure generated each year. I’m confident if we made the fund broad enough, we’d begin to generate the kind of dollars we need to invest in our drainage system that we have spent decades building with largely private (landowner) money generated from the sales of corn and beans over the past century. What is needed now is a public finance system harnessed to a statewide program that can address this infrastructure at scale and with scope economies.

    6) Last, I think your mention of cover crops is vital as well. We have just over 1 million acres which is not nearly enough — the nutrient reduction strategy calls for north of 10 million. But our failure rates of producers continuing once subsidies run or because of bad experience or weather is not where it needs to be. I’ve heard failure rates as high as 50%, defined by those continuing the practice after state or federal funding runs out. I think we need to examine how we create targeted programs at certain critical areas of watersheds and we create financial incentives that build and reward teams of producers for achieving watershed by watershed developed goals and objectives.

    Thank you again for the discussion and your great work.

    John Norwood
    Polk County, Soil and Water Commissioner

    • cjones says:

      John: thank you for your detailed and thoughtful comments. And i appreciate you reading the blog.

      3) About 2.5 billion pounds of commercial fertilizer nitrogen (as N) is sold in Iowa every year. Nitrogen generated from manure is probably in the 500 million to 1 billion pound range.
      5) i agree that taxation or fees could be a pathway toward funding improvements.
      6) i agree watershed approaches toward solving this are imperative. There is no sense spending money where it will not have an impact.

  5. gene veltkamp says:


    Thanks for the informative article. It is eyeopening to say the very least. I have a couple of questions on the report.
    1. Is the amount of nitrogen-nitrate leaving the state through the waters the total amount or is there additional nitrates/phosphates that are captured by the water borne grasses that are present in the rivers, streams and lakes?
    2. Does anybody have any idea of how much nitrates/phosphates are given off by the deciduous trees and grasses each fall after the killing frosts that occur every year? I am anticipating that there is a substantial amount of nitrates and phosphates that are released after the leaves/grasses fall down and are degraded by the breakdown of the said leaves and grasses. How much of this occurs after the ground is frozen, and the nitrates/phosphates are then unable to be absorbed into the ground. With late fall rains/snow falls, where do these nitrates/phosphates end up and can this be accounted for as well?

    I do not discount the effects of nitrates leaving the lands, as I think that there is an excess of fertilizers added to the land. If the plant cannot uptake it, the nitrate leaves the growth zone, and is eventually entered into the groundwater. This ground water eventually is entered into the rivers and streams, where it becomes a problem with increased levels of nitrates and phosphates that enter the rivers.

    It is also interesting that talking to some of the waste water treatment facilities, the general consensus was that the treated water (In Iowa) probably did not contribute to the overall problem of the dead zone down at the Mouth of the Mississippi. If every waste treatment operator holds that mentality, no wonder that we have a problem with excess nitrates/phosphates in the gulf of Mexico.

    We have a tendency to want to blame the farmers for the problem, and if we only get rid of the extra nitrate/phosphates, then our problem will magically disappear. What about the excess amounts of fertilizer that are put onto lawns within the city limits, and how much does that contribute to the overall problem. Same issues with the farmland also occurs in the city limits. If the grass fertilizer is not taken up by the grass lawn, that nitrate will eventually end up down in the aquifer, and will eventually show up as elevated nitrate levels in the waters leaving the state.

    This is a complex issue, much more than assigning the responsibility to the farming community. I am not a farmer, but in addressing the issue, one needs to look at the pressures put upon farmers to produce additional supplies each and every year, due to pressures from financial institutions, as they want to show a profit, but then when the farmers do produce more, then they are compensated less for yields from the crops. All to keep cheap food for the masses.

    We need to develop solutions for the extra amounts of nitrates and phosphates in the surface waters. That amount of nitrates/phosphates should be looked at as our “crude oil”, and should be recycled and reused, made into organic nitrates/phosphates that can be more easily taken up by the plants. Until there is a change in the political climate, I do not believe that there will be a solution. With the change in the amounts of moisture that the state receives and subsequent floods happening, that will be a very difficult task to manage.

    I appreciate the article, very interesting. I am looking for solutions that make economic sense to solve the issues.

    Gene Veltkamp

    • cjones says:

      Thank you for reading Gene, and for the thoughtful and insightful comments. My recollection of the science assessment that was conducted for the nutrient reduction strategy is that about 80-85% of the phosphorus and 90-95% of the nitrogen is coming from non-point sources, with agriculture being the dominant driver there. Certainly decomposing vegetation is a factor, but we know from historical water quality data that our streams likely had < 2 and probably < 1 ppm of nitrate. Statewide average now is around 6-8 ppm.

      Thanks again Gene, and i hope you keep reading.

  6. Cindy Hildebrand says:

    There is a strong and very understandable desire in Iowa to “not just blame farmers” for nutrient pollution. Two observations.

    First, Iowa is not like many other states which have rowcrop fields located within larger matrices of woodlands, wetlands, and grasslands. In Iowa, the rowcrop fields are the large matrices, and the remaining grassland, wetland, and woodland areas are small. Iowa has less of its original landscape left than any other state.

    With such a high percentage of Iowa’s land in agriculture, including very large livestock numbers, it is not surprising at all that agriculture is by far the largest source of non-point nutrient pollution, as determined by research and acknowledged in Iowa’s official Nutrient Reduction Strategy. It’s a predictable result of Iowa land use patterns.

    Second, a farm-conservation-knowledgeable friend keeps saying, with reason, that Iowa soils don’t have a nitrogen problem as much as they have a carbon problem. Our soils have lost a sizable percentage of their original SOM (soil organic matter). That means our soils are much less able to hold nutrients in place than they were before the original Iowa landscape, mostly tallgrass prairie, was plowed up and repeatedly tilled.

    The kinds of agricultural changes that would improve soil health would also help solve a variety of environmental problems, including nutrient pollution, flooding, drought, climate change, air pollution, and biodiversity loss. All of us in Iowa would benefit from those changes, and it makes sense that all of us would help pay for them.

    Some reasonable ways to do that have been proposed. And some of us, myself included, would rather help subsidize the kinds of agriculture that help solve environmental problems than continue to subsidize the kinds of agriculture that make the problems worse.

  7. Rick says:

    I don’t recall what the effect was on nitrogen, but the South Fork Watershed Alliance (S Fork of the Iowa) had some success in reducing soil and P loss through surface tile inlets. Seems that some combination of practices might also reduce Nitrates.

    • cjones says:

      Thanks for reading, and thanks for this comment. Yes, most (but not nearly all) our stream phosphorus enters from the surface. Unfortunately nitrate is dissolved from the soil profile as water percolates/infiltrates down to the tile depth. Tile also captures high-nitrate groundwater as the water table rises during wet periods.

  8. Vishal Raul says:

    Thank you for the detailed article. I have been reading your papers and interested in finding a relation between FWA No3-N with respect to Nitrogen input. In your paper (Livestock manure driving stream nitrate) you have shown a relation of FWA No3-N vs N-surplus for Iowa areas draining to the Missouri River Basin. Do you have such a relation for Iowa areas draining to the Mississippi River Basin?

    • cjones says:

      thanks for reading. I have not quantified this information (yet) for the areas draining east toward the Mississippi.

Leave a Reply