Neuse Estuary Monitoring and Research Program
Center for Applied Aquatic Ecology, NCSU
www.waterquality.ncsu.edu
“Anthropogenic impacts on water quality are strong
and the need for characterizing ambient conditions
and temporal trends in these conditions is
correspondingly urgent.” - Jassby et al. (1997)
More than 2/3 of coastal rivers and bays in the U.S.
are moderately to severely degraded from cultural
eutrophication. – Howarth et al. (2000)
Program objective (ongoing)
Using a decadal+ dataset, characterize physical,
chemical, biological trends in the Neuse Estuary
Thanks to: CAAE staff and students; statisticians D. Dickey, C.
Brownie; collaborators in 16 agencies and academic institutions;
funding from the NC General Assembly, EPA, private foundations
Inventoried and groundtruthed major sources of
mammalian wastes (GIS); working to quantify sources of
non-point pollutant loadings
NPDES sites
Swine CAFOs
Cattle
Poultry CAFOs 0 100 200 km
10% Water
Urban
8%
Cropland 23%
Pasture 3%
23%
Forest (+silvaculture) 44%
Neuse Watershed
Nutrient loading to the Neuse - past decade
Not well quantified – non-point pollution from runoff
(row crops, CAFOs), stormwater (impervious
surfaces), industrial / residential lawn care;
groundwater; atmospheric deposition (wet, dry).
~ 2 million people, ~ 2 million pigs
Human population 16%
Swine population 285%
Urban land cover 110%
Synthetic fertilizers 150%
Watershed profile
P banned from domestic detergents (1985-).
Biological nutrient removal (BNR) at some major
WWTPs (also reduced bypassing of raw sewage during
moderate-major storm events) (2000-).
Buffers (~17 m [50 ft.] along certain streams, 2000-).
Cropland farmers signed agreement to reduce their
use of N fertilizer by 40% (2003).
Improvements – Neuse watershed
Methods
Collected water quality data (10 yr, ~40 variables, weekly –
biweekly April-Oct., monthly Nov.-March, more frequently
during blooms, other events); also, ADCP flow data,
augmented (past 4 yr) with RTRM platforms.
Calibrated time series models (ARIMA; log-transformed
data – then transformed back to the original units
for graphs) to improve prediction of long-term trends
in nutrient loadings.
Developed an ADCP-calibrated flow model using a
delivery gate at entrance of estuary, Mills Branch
(previous models estimated flow from ~70 km upstream).
Developed a univariate model for dissolved oxygen
to evaluate long-term trends over time.
Neuse Estuary study area - CAAE
Pop.
Center
N
Most detailed data set on the Neuse Estuary for the
past decade.
cm / sec
Flow Out (%)
ADCP Std. Deviation (cm / sec)
ADCP Current Velocity (cm / sec) cm / seccm / sec
Flow Out (%)
ADCP Std. Deviation (cm / sec)
ADCP Current Velocity (cm / sec)
Cross-transect ADCP
detailed data set –included
bottom-tracking capability
for accurate bathymetry
Mass Water Transport
Reed et al. (2004), Estuarine and
Coastal Shelf Science
0
400
800
1200
Transect
Dis
ch
arg
e (m
3 s
-1)
ADCP (Mills Branch, x-transect)
4 8 12 16 20 24 0 28
USGS fixed point (Kinston)
Model (Mills Branch)
Modeled mass water transport, vs. measurements
Historical approach underestimates mass water transport (Mills Branch
= better gate).
Burkholder et al. (2004), Proceedings of the National Academy of Sciences
CAAE’s Remote Data Acquisition Platforms (1999-)
Meteorological measurements –
wind speed / direction, air temp.,
rel. humidity, barometric pressure,
incident solar radiation, precip.
Hydrological measurements – water
level, water temp., salinity, DO,
redox, pH, turbidity.
Discrete water samples (ISCO
sampler; short / long intervals); e.g,
> 204,000 nutrient samples analyzed.
www.waterquality.ncsu.edu Glasgow et al. 2004 –
J. Exp. Mar. Biol. Ecol.
Dissolved oxygen (example)
www.waterquality.ncsu.edu
DO 10-15 Aug. 2001
Profiler Record for DO (Carolina Pines, ~2,400 casts)
De
pth
(m
)
DO
13 June – major fish kill
Winter Spring Fall Fall Summer
DO DO DO DO DO DO 0
16 DO
16
12
8
4
0
Glasgow et al. (2004), J. Exp. Mar. Biol. Ecol.)
Color infrared photograph
Powerboat
Neuse dinoflagellate bloom (Prorocentrum minimum)
near south shore at Mills Branch, Dec. 2000
Prorocentrum minimum
More intensive bloom sampling
Springer et al. 2004, Harmful Algae
Chlorophyll a concentrations
Springer et al. (2004), Harmful Algae
www.waterquality.ncsu.edu
Phytoplankton chlorophyll a
April – 28, 2001
250
200
150
100
50
µg / LFeb. 6
TP
280
240
200
160
120
80
40
0
NO3-
µg / LApril 9
NH4+
Oct. 19
100
80
60
40
20
0
µg / L250
200
150
100
50
µg / LFeb. 6
TP 250
200
150
100
50
250
200
150
100
50
µg / LFeb. 6
TP
280
240
200
160
120
80
40
0
NO3-
µg / LApril 9
280
240
200
160
120
80
40
0
NO3-
µg / LApril 9
NH4+
Oct. 19
100
80
60
40
20
0
µg / LNH4+
Oct. 19
100
80
60
40
20
0
µg / L
Related to nutrient pulses
Springer et al. (2004),
Harmful Algae
Annual Water Delivery Volumes at Mills Branch
6
7
8
9
10
11
Flo
w (
m3 x
10
9)
1991 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 1992
Emily* and above-
normal winter
Danielle*
Bob*
Allison*
Fran, Bertha,
Arthur*, Josephine*
Bonnie*, and
wet spring
El Niño
mid’97-
early’98
Dennis*,
Floyd, Irene*
100-yr drought
Burkholder et al. (2004), Proceedings of the National Academy of Sciences
TN loading – Neuse Estuary
TN
0
5
10
15
20
25
28%, p = 0.017
TN
(kg
x 1
05)
94 95 96 97 98 99 00 01 02 03 93
TN
0
5
10
15
20
25
15%, p = 0.313
TN
(kg
x 1
05)
Drought-
94 95 96 97 98 99 00 01 02 03 93
Parametric PROC ARIMA, non-parametric Seasonal
Kendall Tau models agreed: Significant decrease
(~28%) in TN loading. However, drought years drove
the trend.
TN loading estimates (Neuse Estuary)
PROC ARIMA (parametric), Seasonal Kendall Tau (non-
parametric) trend analyses: trends (+ drought) were not
affected by the hurricanes.
No significant change in TN loading over the past decade –
“running to stand still”.
1996
1999
Dennis
Floyd
Irene
Precip. contours Fran
No change in trend T
N (
kg
x 1
05
)
0
5
10
15
20 TN loading
95 96 97 98 99 00 01 02 03 93 94
~200 – 500% increase (significant, with or without the
sustained drought). Potential drivers: changes in land
use and waste inputs, regeneration etc. [NO3- concs.
increased 120%, but not significant ].
NH4+ concentrations - Neuse Estuary
Year
94 95 96 97 98 99 00 01 02 03
50
100
150
200
250 1994 - 2003
503%, p < 0.0001
NH
4+ (
µg
L-1
)
Year
1994 – 1999 (excluding drought)
190%, p = 0.015
Drought
94 95 96 97 98 99 00 01 02 03 0
No significant trend for all stations considered
collectively, but significant increases (up 60%) at some
stations.
Chlorophyll a – Neuse Estuary (1994 – 2003)
Chla
(µ
g L
-1 )
94 95 96 97 98 99 00 01 02 03 0
100
200
300
Summary (Neuse, 1993 – 2003)
TN loading 28% (significant; but drought-related
[2000-2002]. Loading is again increasing). TN trend
highly sensitive to Tinitial (no change from 1994-). No
change in TP loading + drought.
NH4
+ concentrations 200 – 500%; significant + drought.
Chla in localized areas, but no change overall.
From analysis of this decadal database
Drought-related TN loading decrease (loading again
increasing.
Highly significant increase in NH4
+ concentrations,
related to non-point sources. Improved treatment of some major point sources (sewage);
Overall, these analyses point to the need to control
additional non-point sources, as in many estuaries
throughout the world.
Ongoing and Future Plans – Neuse Monitoring
and Research Program (CAAE and partners)
Assess factors controlling harmful algal blooms
(meteorology, nutrient fluxes, grazing);
Improve quantification of nutrient loading, and examine
the role of non-point sources on algal blooms (RTRM);
Finish validating the U.S. EPA EFDC/WASP models
to improve prediction of phytoplankton chlorophyll,
harmful algal blooms.
Ongoing and Future Plans (CAAE and partners)
Environmental education outreach (various secondary schools, high schools, colleges)
Floating Classroom Program – RV Humphries