Arthur Gold

1. Advancing the adoption of artificial nitrogen sinks
Kelly Addy, Art Gold1*, Louis Schipper2, Mark David3, Nicole Sacha1 and Brian Needelman4
1University of Rhode Island, 2University of Waikato, 3University of Illinois, 4 University of Maryland *agold@uri.edu
This material is based upon work supported by
the National Institute of Food and Agriculture,
USDA, under agreement No. 2011-51130-31120.
Any opinions, findings, conclusions, or
recommendations expressed in this poster are
those of the authors and do not necessarily reflect
the view of the USDA. Contribution #5352 of the
RI Agricultural Experiment Station.
Citations from this poster and meta-analysis :
Cameron and Schipper. 2010. Ecol. Eng.
Christiansen et al. 2011. Agr. Water Manage.
Christiansen et al. 2011. Journal of Environ. Eng.
Christiansen et al. 2012. Tran. Am. Soc. Ag. Eng.
Christiansen et al. 2013. Ecol. Eng.
Chun et al. 2009. Biosystems Eng.
Crumpton et al. 2008. Final Report: Potential of
restored and constructed wetlands to reduce
nutrient export.
David et al. in review. J. Environ. Qual.
Elgood et al. 2010. Ecol. Eng.
Gibert et al. 2008. Bioresource Technol.
Greenan et al. 2006J. Environ. Qual.
Healy et al. 2006. J. Environ. Sci. Heal.
Healy et al. 2012. Ecol. Eng.
Healy et al. 2014. Ecol. Eng.
Lepine et al. in review. J. Environ Qual.
Kadlec. 2012. Critical Reviews in Environmental
Science and Technology.
Long et al. 2011. Agr. Ecosyst. and Environ.
Mitsch and Day. 2006. Ecol. Eng.
Moorman at al. 2010. Ecol. Eng.
Pluer et al. in review. J. Environ. Qual.
Robertson et al. 2009. J. Environ. Qual.
Robertson. 2010. Ecol. Eng.
Robertson and Merkley. 2009. J. Environ. Qual..
Rosenburg et al. 2000. Metawin statistical
software for meta-analysis.
Schipper et al. 2010ab. Ecol. Eng.
Schipper and Vojvodić-Vuković. 2000. J. Environ.
Qual.
Schmidt and Clark. 2012. J. Environ. Qual.
Warneke et al. 2011. Ecol. Engin.
Warneke et al. 2011bc. Water Research.
Woli et al. 2010. Ecol. Eng.
IntroductionIntroduction
Acknowledgements & References
Constructed Wetlands
Constructed wetlands (Fig.1) are artificial systems
that provide ecological services, such as flood
water storage, nutrient (nitrogen or phosphorus)
storage and cycling, and erosion control. They are
modeled after natural wetland systems. These
systems are placed alongside ditches or streams
where they retain water for hours or days to allow
nitrate-N removal through denitrification.
Figure 1: (a) Constructed wetland in Iowa and
(b) conceptual diagram of constructed wetland
treating tile drainage (Mitsch and Day 2006).
The two main types of denitrifying bioreactors
are beds (Fig. 3a) and walls (Fig. 3b). Beds are
closed systems that receive tile drain inflow.
Walls are placed in the natural flowpath of
groundwater leaving from agricultural fields.
Constructed wetlands and denitrifying bioreactors are artificial “sinks” (hotspots of nitrogen (N) removal)
that are created to resemble natural systems that promote denitrification. When installed at the edge of
an agricultural field, they can intercept groundwater or tile drainage water and reduce the nitrate-N
content of agricultural discharge. The goal of our project is to advance the adoption and proper
placement of denitrifying bioreactors and constructed wetlands in agricultural settings.
Meta-analysis: N removal in denitrifying bioreactors
Outreach approachOutreach approach
The International Atlas identifies artificial N sink research
and demonstration sites from across the globe. The map not
only gives information about where the systems are but also
provides project contact information, links to case studies,
papers, and websites with additional information. We welcome
any additions to our atlas or website (email kaddy@uri.edu).
Another important feature of the website is the Frequently Asked
Questions (FAQs) which provides basic information on the function
and processes of artificial N sinks. Basic information about point source
pollution, dead zones and other related details are also present on this
page. Most questions have links to EPA definitions and various videos
which give a fuller understanding of the subject.
Resources include fact sheets, case
studies, videos, workshop presentations
and research summaries about constructed
wetlands and denitrifying bioreactors.
Research papers written by scientists are
further studied and condensed into
summaries in our case study section which
highlights the situation, actions, and take-
home messages. We also have videos and
fact sheets that highlight similar subject
matter.
www.artificialnsinks.org
This website, primarily targeting land managers and farm
advisors, offers guidance, an International Atlas of
constructed wetlands and denitrification bioreactors, fact
sheets, case studies, presentations, videos and other
resources.
We also have a listserv to share news on artificial N sinks
and share project successes. To join our listserv, email
kaddy@uri.edu.
Figure 2: Denitrifying beds under construction.
Meta-analysis results for denitrifying bioreactors
pLimitations & Next steps
Next Steps
• Two papers (one on this meta-analysis) have been submitted to a special issue of the Journal of
Environmental Quality on “Moving Denitrifying Bioreactors beyond Proof of Concept.” Our project was
the impetus for this special issue and Dr. Schipper is a Guest Editor at JEQ for it.
• Develop bed design recommendations to maximize nitrate removal and minimize any
potentially adverse byproducts.
Denitrifying Bioreactors
A denitrifying bioreactor (Fig.2) is an artificially
constructed system that mimics selected functions
of riparian wetlands. These systems are
composed of an added source of carbon (often
woodchips) that intercept groundwater or tile
drainage. The wood chips create an anaerobic
environment in which bacteria transform the
nitrate-N in the water into nitrogen gas.
Meta-analysis: Methods
Bioreactor design: Nitrate removal rates in
bed and lab column studies were not
significantly different, but both were higher
than wall designs (p<0.05; note: small n
value of walls).
Bed influent nitrate concentration: Beds with
influent N > 30 mg l-1 had higher nitrate removal
rates than beds with intermediate (10-30 mg l-1;
p<0.01) and low influent N (<10 mg l-1; p<0.05).
Fig 3a: Denitrifying bed to treat tile
drainage (Schipper et al. 2010)
Figure 3b: Denitrifying wall to
intercept groundwater flow
(Schipper et al. 2010)
Artificial N sinks are not suitable for all locations and conditions. It is important to
consider site conditions and hydrology at all sites to maximize N removal function.
Flowpaths: If a denitrifying bioreactor is
intended to intercept natural groundwater flow,
site hydrology must be understood. For
instance, deep groundwater flow can bypass
treatment. (Schipper et al. 2010)
Residence time and hydraulic loading: For high N removal
in both constructed wetlands and denitrifying bioreactors,
longer residence times need to be accommodated when
hydraulic loading is high.
(Crumpton, ISU)
Cost and Social Barriers to adoption are
other important limitations for artificial N sinks.
A meta-analysis was conducted to analyze the response of
nitrate removal rate to study type (wall, bed, laboratory
column) and various parameters using MetaWin version 2.1
software (Rosenberg et al. 2000). We used a random
effects model of meta-analysis as the data were collected
from bioreactor units that differed in a number of design
features. Effect size calculations used in meta-analysis
approaches are intended to weight data based on the
variance and number of replicates for each bioreactor unit
(Rosenberg et al. 2000).
After determining the effect sizes per category, we tested for
heterogeneity between categories within each parameter.
Significant heterogeneity suggests that the individual studies
likely come from different statistical populations, potentially
justifying a categorical representation of the full dataset for
that parameter (Rosenberg et al., 2000). A bootstrapping
(sampling with replacement) procedure with 999 iterations,
corrected for bias was then conducted, which generated
95% bias-corrected confidence intervals. We used level of
p<0.1 to indicate trends.
Bed Hydraulic Residence Time (HRT):
Cumulative nitrate removal (the nitrate removal
rate normalized for time) in beds with HRT < 6
h was significantly lower than in beds with HRT
from 6-20 h and > 20 h (p<0.05).
Bed Temperature: Beds with temperatures <
6°C had lower nitrate removal response than
those at intermediate temperatures of 6-16.9
°C (p<0.1) and temperatures >16.9 °C
(p<0.05).
M. David
(2b)
(1b)
D. Jaynes
(2a)
W. Crumpton
(1a)
0
5
10
15
20
25
Nitrate-N(mgL-1)
0
5000
10000
15000
20000
25000
Inflow(m3d-1)
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
5
10
15
20
25
Nitrate-N(mgL-1)
0
10000
20000
30000
40000
50000
Inflow(m3d-1)
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Residence
time
Longer
Shorter
Hydraulic
Loading
Rate
Lower
Greater
Inflow concentration Outflow concentration Flow rate
Each of these plots depict mean nitrate removal rate effect size and 95% bias-corrected confidence interval by
category. Bars with different letters were significantly different; n values represent number of bioreactor units in that
category used in the analysis. Inset table indicates the back-calculated mean nitrate removal rate with 95% bias-
corrected confidence interval.