Here are some questions about the Chesapeake Bay restoration effort that might have surprising answers:
- Is the Bay region successfully curbing nutrient runoff from farms?
- Is nitrogen runoff from developed lands really increasing?
- Is the region actually on track to meet its phosphorus reduction goals?
A recent modeling exercise by U.S. Geological Survey scientists suggests the answers to all of those questions might be “no” — conclusions that run counter to conventional notions within the Bay cleanup effort.
The results published in a paper last year are based on a computer modeling exercise that relied heavily on water quality monitoring data within the Bay watershed over a 20-year period.
The exercise found fewer nitrogen reductions from agriculture than estimated by the state-federal Bay Program for the same period and sharply different phosphorus trends. It also concluded that nitrogen runoff from cities and suburbs is decreasing.
To be sure, there are many caveats to those conclusions, and they do not necessarily mean that Bay Program estimates are wrong.
But the analysis highlights the longstanding question about whether on-the-ground nutrient reduction efforts are producing the expected water quality improvements. It’s an issue scientists have highlighted for years, and it’s drawing increased attention from the scientific community as the region approaches its 2025 Bay cleanup deadline.
The implications are huge. The region has been working since the mid-1980s to reduce the amount of two nutrients, nitrogen and phosphorus, reaching the Bay. In the Chesapeake, they spur water-staining algae blooms that draw oxygen from the water when they die, creating “dead zones” that are off-limits to most aquatic life.
Both the USGS analysis and the Bay Program agree that most nutrient reductions achieved to date are from technology upgrades at wastewater treatment plants. With those largely completed, states are counting on greatly ramped-up efforts on agricultural lands, which generate the majority of the nutrient runoff, to meet Bay goals
But the USGS analysis indicates that it’s uncertain when and if those farm-based practices — such as nutrient-absorbing cover crops, vegetated buffers along waterways, or plans to guide manure management — will achieve clean water goals.
A second USGS paper published this summer laid out a number of reasons for that uncertainty. For one, it can take a long time for some on-the-ground actions to benefit the water. But other factors could be at play, too. For instance, runoff control practices may not be as effective as thought.
“People want to understand as they’re putting in the practices, are we getting the reductions we anticipate?” said Scott Phillips, USGS Chesapeake Bay coordinator. “We’re seeing that the nutrient reductions in monitoring data vary greatly across the watershed, making the comparison to reduction efforts more difficult. We’re continuing to use multiple tools to explain trends.”
Different approaches, different results
The Bay Program tracks cleanup efforts using its Watershed Model, which has been refined and peer-reviewed over three decades. It predicts how cleanup actions and other factors, such as land-use changes and population growth, will affect the number of nutrients entering the Bay. It suggests slow but steady progress in reducing nitrogen and phosphorus.
But its results have not always aligned with the data collected at scores of water monitoring sites throughout the watershed that paint a more nuanced picture.
From 2009–18, for instance, monitoring showed that for nitrogen, 41% of monitored sites had reductions, 40% had increased and 19% had no trends.
For phosphorus, 44% of locations had reductions, 33% had increases and 23% had no trend.
By the time the water that passes through those monitoring stations reaches the Bay, the results — at best — are mixed. Most of the Bay’s nine major nontidal rivers have no trends or worsening trends for nitrogen and/or phosphorus over the past decade. Only the James and Patuxent rivers show improving trends for both nutrients.
The USGS analysis examined the implications of those trends over time using data from 1992, 2002, and 2012 in its highly regarded SPARROW model (that stands for SPAtially Referenced Regressions On Watershed attributes). Using those results, and other data — such as nutrient inputs, land use, and geographic settings — it analyzed what those various monitoring results meant for the Bay.
The SPARROW model indicated that the amount of nitrogen entering the Bay declined, but about 25% less than what the Bay Program estimates for the same time.
More than 80% of the reductions stemmed from wastewater plant upgrades. Most of the rest came from reductions in air pollution: a decrease in nitrogen oxide, emitted by power plants and vehicles, which enters waterways after it falls to the ground.
The analysis also indicates that nitrogen runoff from developed lands — often called the only major source of nutrients still on the rise — also declined during the study period, though phosphorus was largely unchanged.
Most worrisome, though, it found no overall nitrogen or phosphorus reductions from the region’s vast agricultural lands. The Bay Program model estimates a 17 percent nitrogen reduction from farms and an even larger phosphorus reduction during that 20-year period.
And while the Bay Program considers phosphorus reductions to be largely on track to meet cleanup goals, the USGS analysis showed that overall phosphorus loads actually increased 9% during the study period. The Bay Program estimates phosphorus declined by nearly a third during that time.
Why the difference?
The USGS analysis is part of an effort to explain those differences, a task that is becoming more urgent as the 2025 cleanup deadline approaches. The Bay Program has a workgroup exploring the issue, and its Scientific and Technical Advisory Committee is identifying Bay-related science needs for a report expected next year; a better understanding of factors behind nutrient trends is one of the areas being examined.
In one of its recent papers, the USGS cited four likely reasons for the differences.
- Lag time. Unlike wastewater discharges that go directly into rivers, most other nutrients are applied to the land, usually as fertilizer or manure. A portion of those nutrients may wash directly into streams when it rains. Most of the nitrogen, though, soaks into the soil and flows into waterways through groundwater, a journey that may take years or decades. Phosphorus, which tends to bind with soil particles, flows slowly downstream, typically moving relatively short distances during large storms. Not only is the movement of nutrients to streams slow, but some practices, such as a newly planted forest buffer, can take years to reach its maximum effectiveness.
- Unrealistic expectations. The Bay Program’s computer model uses assumptions about the expected nutrient reductions from a wide variety of on-the-ground pollution reduction practices. But the number of studies about the measurable impact of those practices is often limited. And estimates are complicated because differences in soil, topography and other factors may result in different effectiveness from place to place.
- Insufficient monitoring. Much of the monitoring in the watershed is conducted at scales too large to detect small changes. More monitoring in small watersheds where a large number of runoff control practices are installed could provide better insight about their effectiveness.
- Competing factors. The benefits of runoff controls may be offset by other issues. Intensification of farm activities, such as converting low-runoff hay fields to high-runoff crop lands, increasing numbers of farm animals, or changes in fertilizer or manure applications, can offset the impact of nutrient reduction efforts. Over time, the Bay Program has made efforts to address some of those issues, such using more up-to-date data to account for the intensification of farm activities.
Scientists generally agree that lag times almost certainly play a role in explaining some of the differences between the monitoring data and Bay Program estimates.
The Bay Program’s Watershed Model is essentially trying to predict whether management actions being taken now will meet cleanup goals when they are fully effective in the future. That predictive capability is needed for states to estimate how many buffers must be installed, cover crops planted, or wetlands restored, to meet Bay goals. Each year, states provide data about actions they took in the past 12 months so the model can estimate the impact they will have at some point in the future — if they perform as expected.
Because the USGS model relies on water quality monitoring, it would not quickly reflect the impact of those actions because of the lag time in nutrients reaching streams.
The Bay Program model did not include a mechanism to account for lag times until three years ago, and that function is not used for management. But the first attempts to factor lag times into model estimates did show that nitrogen trends in many areas became more similar to monitoring results.
“When we make an estimate of what happens when you incorporate lag time, it makes a big difference,” said Gary Shenk, a USGS hydrologist who coordinates the Bay Program’s watershed model.
While lag times likely cause real delays in water quality responses, many scientists say it’s unclear how much of the differences they account for.
Particularly disturbing, some say, is the failure of the SPARROW model to detect any major changes in nutrient pollution from farm lands, even though it covered a 20-year period when the use of runoff control measures was accelerating.
Bill Dennison, vice president of the University of Maryland Center for Environmental Sciences, and co-chair of a Bay Program workgroup that coordinates monitoring and modeling analysis, said attributing all of the differences to time lags is “a common fallback for dealing with uncertainty and dealing with impatient people. Officials and resource managers really want to see a response.”
The good news, he said, is that the Bay has shown improvements from reduced wastewater discharges. Underwater grass beds have expanded and smaller oxygen dead zones have shrunk. “The Bay is very responsive,” Dennison said. “You can turn off the sewage and get a response and in a year or two.”
But trying to relate actions in the watershed to impacts on the Bay, he said, is “an order of magnitude” more difficult.
Studies show little change in the total amount of nitrogen applied to farmland over the years. That means it’s critical to find out whether nutrient control actions are succeeding in keeping them out of the water.
“We make some assumptions about how effective these practices are, but they may not be performing as we expect.” — Zach Easton, Virginia Tech
“We make some assumptions about how effective these practices are, but they may not be performing as we expect,” said Zach Easton, a professor at Virginia Tech and member of the Bay Program’s Scientific and Technical Advisory Committee.
One way to help reduce uncertainty, Easton said, would be to increase small-scale monitoring to understand how well, or whether, nutrient concentrations in streams respond to on-the-ground actions. “The level of monitoring just is not sufficient to detect the signals,” Easton said.
More local monitoring could also shed light on the USGS study’s conclusion that nitrogen runoff from developed lands is decreasing.
The exact driver for that change is uncertain, but the papers suggested a variety of factors, including impacts of runoff control efforts, reduced sewer line leaks, or better efforts to pick up pet waste. “I don’t know if we know the answer to that at this point,” acknowledged Scott Ator, a USGS hydrologist, and lead author of the recent papers.
But his papers were not the only ones to detect that trend. USGS scientists say several limited monitoring efforts have shown a nitrogen decrease from developed lands, and a separate USGS modeling effort reached a similar conclusion.
The different monitoring and modeling results for phosphorus also are not fully explained by lag times. Some factors are known: More phosphorus bound to sediment is passing through Conowingo Dam on the Susquehanna River because its reservoir is filled and no longer trapping it. And some agricultural areas where soils are saturated with phosphorus are leaking it into waterways at increasing rates. But scientists say the full reason for increasing trends is unknown.
Understanding these and other uncertainties has major ramifications for 2025. If all cleanup actions are implemented, would it be acceptable to wait — potentially for decades — to see whether the Bay responds as anticipated? Or, should more work be done as a hedge against the possibility that some actions are not as effective as thought?
The Scientific and Technical Advisory Committee will produce recommendations next year.
But, cautioned Kurt Stephenson, a Virginia Tech professor overseeing the effort, “whatever we recommend is not going to all of a sudden solve the mystery. The uncertainty is inherent. We’re never going to eliminate it, so we have to manage in the face of it.”
This article originally appeared on BayJournal.com on October 14, 2020, and is republished with permission from Bay Journal Media.