Skip to main content
Overhead irrigation in a field

A Clean Water Odyssey through the Chesapeake Bay

Dr. Amy Shober, University of DelawareDr. Amy Shober, a professor and extension specialist in nutrient management at the University of Delaware, was the third speaker of the 2020 Virtual Ag BMP Summit. Dr. Shober discussed the current state of the Chesapeake Bay Watershed and the actions that are being taken to protect the Bay.

The Chesapeake Bay Watershed spans 64,000 miles across six states and DC (i.e., jurisdictions). It provides $750 million to local economies and serves as a habitat for 3,600 species of plants and animals. However, pollution problems have led to water quality issues like nutrient fueled algal blooms result in dead zone conditions and a decline in habitat quality and quantity.

While land use in the watershed is primarily forest (59%), nutrient and sediment losses from agricultural (pasture/hay – 18% and cropland – 10%) and urban land uses (8%) contribute to water quality issues. Delaware makes up a relatively small proportion of the Chesapeake Bay watershed. Approximately 33% of Delaware’s land drains into the Bay, contributing 1% of the total flow and 2% of the total nitrogen and phosphorus loads entering the Bay. Delaware’s land use is dominated by agriculture (44%) and, therefore, agriculture is estimated to contribute a majority of the nutrient and sediment loads when compared to other sectors (e.g., wastewater, urban stormwater, forest). Since agriculture has such an impact on water quality in the state, farmers are required by state law to complete nutrient management certification training and follow a nutrient management plan.

Keeping the watershed healthy is achieved using a partnership approach, where regional leaders and experts unite to share resources across jurisdictions (6 states and DC). Chesapeake Program partners include federal and state agencies, local governments, academic institutions, and a variety of non-governmental organizations. The program is structured so that the executive council (governors, agencies, etc.) set the vision, and then the goal implementation teams (and their associated work groups) implement that vision.

A total maximum daily load was established for the Chesapeake Bay in 2010, which required Bay jurisdictions to have all actions to meet water quality goals in place by 2025. Watershed Implementation Plans were developed to outline what actions each jurisdiction will take to achieve the nutrient load reduction goals. In addition, each Bay jurisdiction must meet 2-year milestone goals. Progress toward these goals is assessed using load estimates that are developed using a suite of models (currently on Phase 6). Inputs to the models include land use, climate data, air quality, point source data, etc. The models are calibrated and validated using real-world water quality data.

There are over 45 best management practices (BMPs) included in the Bay models, with 275 variations. Examples of approved agriculture sector BMPs include cover cropping, grazing, manure transport, runoff control, and animal mortality management. Each of these BMPs has a load reduction tied to it, which was determined by Bay Program experts based on analysis of the scientific literature and other data sources. Jurisdictions can use an online version of the modelling suite to test different BMP options and observe how they would affect the nutrient load estimates. These scenarios help determine which actions jurisdictions will include in their Watershed Implementation Plans. When prioritizing BMPs, each jurisdiction must balance the potential load reduction of a practice with the willingness of landowners to adopt that BMP. Delaware’s experience showed that land managers are more willing to adopt practices that do not take land out of production.

Due to concerns about how jurisdictions were reporting BMP data, there are now specific BMP verification requirements before adoption data can be used in the modeling suite. These verification methods include farm inventory, farm records, transect surveys, agency-sponsored surveys, and remote sensing.

In Delaware, nutrient management is a regulatory BMP that has been successful, with approximately 85% compliance with core practices. Delaware has prioritized “supplemental” nutrient management practices (e.g., targeted placement, rate, and timing) in their current Watershed Implementation Plan. Verified supplemental nutrient management practices can reduce estimated loads of nitrogen by up to 15% and phosphorus loads by up to 20%, depending on the crop being grown. Cover cropping is a voluntary practice that Delaware has also prioritized in the Watershed Implementation Plan. Depending on the cover crop and planting/termination timing and method, cover crops are estimated to reduce nitrogen loads by 11-45%, phosphorus loads by 0-15%, and sediment loads by 0-20%. Agriculture BMPs are often prioritized because they are more efficient in reducing nutrient and sediment loads than BMPs implemented in urban or forest areas.

Jurisdictions report BMP data to the EPA through the National Environmental Information Exchange Network (NEIEN). If a jurisdiction has not met their load reduction targets, the EPA will take action to ensure the goal is met. Jurisdictions have not met their water quality goals yet, but they are taking action to meet the 2025 goal. For example, there is a downward trend in both phosphorus and nitrogen loads from Delaware, but the state is much closer to reaching the phosphorus load reduction goal than nitrogen. The health of Chesapeake Bay is also being assessed by watching for qualitative indicators such as more female crabs, restoration of oyster reefs, and increased submerged aquatic vegetation.

Water quality improvement is likely to lag behind implementation because nutrient concentrations in the water often reflect past practices. Future challenges for the Chesapeake Bay Program include climate change, land use conversion due to population growth and urbanization, and changing BMPs to reflect needs and new technologies. For more information on the Chesapeake Bay and the progress on this project, please visit: and

Watch the recording of her webinar.

Written by Natasha Roberts, CLUE Communications Intern

Question and Answer Session

What’s the difference between developed and wastewater (loading)? Here in Florida, a lot of our wastewater comes from developed land uses, such as commercial and residential.

It depends on what data you’re looking at as to whether those things are lumped together, but in a lot of program materials, “developed” is related to stormwater. The wastewater treatment plants are considered the point sources and they actually report their load separately than the urban stormwater. And just to give you an idea, in Delaware all of our municipalities are considered MS4s, so there are significant stormwater efforts at the state level here.

Do economic considerations play a role in defining what is a BMP and what is not?

I don’t know how much they use the economics to determine whether the BMPs go into the model. I can say for sure that jurisdictions are always looking for things that they think are economical to promote as new BMPs or to get those numbers updated, so I think in that way it does. But once it gets to that expert panel phase, they’re just looking at the data to see if that particular BMP would be effective at reducing loads in our region.

How much data do you gather from agricultural producers regarding BMP implementation?

That depends on how they’re doing the verification. For something like nutrient management, the Department of Agriculture will audit about 20% of the farms every year, where they will go out, look at farm records, and talk to the crop consultants to determine whether or not everything is in place. For things like cover crops you can use the cost-share data, for manure transport – we use the cost-share data. But for a lot of things there must be a visual or remote sensing assessment. For some practices like conservation tillage or cover crops, we have that down pretty well; but for other practices we might not have it down very well. I think for a lot of the urban stormwater its visual assessment and people actually have to visit the sites.

How are climatic conditions, such as rainfall and rainfall rate entered and considered in the model?

The precipitation data is entered into the modeling process for the watershed model. As far as modeling ahead of time, there are models that forecast but I’m not really familiar with that level of detail in the modeling suite.

Does the University provide on farm demo plots for the high priority BMP’s?

The Universities in the region do a lot of BMP research, but we do not all have the resources to provide demos for all of those practices. A lot of the numbers that come for the cover crops come out of the University of Maryland research for nitrogen load reduction. They’ve done a ton of research looking at the efficacy of different crop types and planting dates under a variety of conditions. So, where we can, where we have the resources to, we would implement those things but a lot of the time we do not have resources to do that.

For manures that are being removed from the basin via manure transport, how is the material being handled – where is it going?

On the Delmarva Peninsula in Delaware we do distribute poultry litter outside the watershed. A lot of it still goes to land application, but there are periods of the year where they’re doing cleanouts and we can’t land apply any manure. So, it typically is going to the mushroom industry up in Pennsylvania and they’re using it as a substrate for their mushroom soil. As a side note, there was a very large effort to pelletize poultry litter in our region and move it out of the watershed, but there were a lot of logistic issues with that – they were never allowed to move that material on the same train cars that brought the grain to feed our chickens. Purdue lost a lot of money on that enterprise, so they abandoned the process about 2-3 years ago. Manure to energy hasn’t really kicked off here. A lot of that is related to the fact that the companies want the farmers to sign contracts with them to provide a specific amount of manure, but the grower does not have a contract from their poultry integrator saying they’re going to get chickens. There is a possibility they could not get chickens and they would not have the manure to give to these companies; that’s been one of the biggest failures of those endeavors.

What about load estimates? How accurate are those estimates in relation to your comments about water quality improvement of cover crops?

As far as the data they have on cover crops, from the studies where they collected that data, I would say it’s very good, but there’s definitely a range there where you see more benefit for load reductions – specifically for nitrogen leaching. In years where it’s very dry and we didn’t achieve our yield goals. We are not allowed to fertilize cover crops, not allowed to put manure out on any cover crops, so there’s actually been a lot of anecdotal evidence that we’ve seen where farmers, after really good years of corn (upper 200-300 bushels), can’t even establish a cover crop because there’s just no nitrogen there. So, in those cases, if you can’t establish the cover crop, it can’t even grow root systems to scavenge any nitrogen that would have leached below the root zone during the growing season anyway.

How is the cost of BMP’s determined? Who does that?

I got all of that information from the Chesapeake Bay Program. Somewhere in that structure there is somebody whose job it is to look at those things. I think Olivia Devereaux (a consultant for the Chesapeake Bay Program) put together the spreadsheet I used. I imagine they pull it from expert panel reports and a variety of other program resources.

On slide 35, how does the atmospheric nitrogen loading decrease from 2009 to 2018?

Those estimates come from Clean Air Act efforts and there is an air-shed model that determines the atmospheric contributions. I’m not exactly sure how they got the numbers, but there’s obviously some kind of process for inputs and such into that model that are trying to capture the different implementation of air quality standards among things like fossil fuel burning power plants and such.

Regarding nitrogen reduction from cover crops, do the land models include agricultural yield data and cover crop biomass data?

Part of the inputs into that watershed model include data we pull from the ag census about average yields and acres planted, etc. With regard to the cover crops though, they just get a specific load reduction for every acre. There is modelled load for each acre depending on the crop system. For example, with corn, they would come up with some loading for that land use based on the average yield. And then if you put a cover crop on top of that land use, it would reduce that number by 15-45% depending on what variation of cover crop you have.

I wonder how the land use change is accounted for in the model – the load reduction for all land uses seem to drop even though there is probably a land conversion (land use change) going on?

There is a land use model that they use to estimate the different changes in land uses over time. I am not an expert on the model, but I can certainly send anybody to the experts. I’m sure they’d be happy to explain how they get all that information.

On slide 38, SAV (submerged aquatic vegetation) is increasing. Is all of that natural regrowth or is vegetative replanting taking place?

I’m pretty sure there is some planting that is also taking place. But determination of SAV is based on visual or using remote sensing. But I think, just like the oysters, there is some effort to restore the sea grasses and the other submerged aquatic vegetation.