*>EPA
United States
Environmental
Protection Agency
EPA/600/R-22/215.
Background Specific Conductivity
and Associated Five Percent
Extirpation Estimates in Arkansas
Prepared by
United States Environmental Protection Agency,
Office of Research and Development for USEPA Region 6
The views expressed in this report are those of the authors and do not necessarily represent the views or policies of the U.S. Environmental Protection Agency.
-------
Bae,"™""el Specific Conductivity and
Associated 5% ation Estimates in Arkansas
ABSTRACT
This report describes the analyses and models used in a weight of evidence for characterizing
background specific conductivity (SC) in Arkansas. The ionic composition of waters in the state
are described. Formulae for converting total dissolved solids (TDS) to SC are provided for each
ecoregion. Stream background was estimated using observed and empirically modeled data by
choosing among three options, the objective being to identify the best available estimate of
minimally affected background for estimating 5% extirpation. The lowest of three values is
recommended for estimating background: the median observed SC at the station, the median of
stations within 5 km, or the estimated default background for the ecoregion. In the Mississippi
Alluvial Plain, Mississippi Valley Loess Plains, Ouachita Mountains, Ozark Highlands, or South
Central Plains, the default background is the stream segment empirically modeled background.
In the Boston Mountains and Arkansas Valley, the default background is the station background
or the ecoregional 75th centile. The ecoregional estimate is recommended rather than the
predicted stream segment estimate because the empirical model consistently over-predicts
background in these two ecoregions. These methods, data, and models may be used to assess the
protectiveness of site-specific water quality criteria proposed by third parties.
Preferred citation: Cormier, S. M, Wharton, C, and Wang, Y. 2022. Background Specific
Conductivity and Associated Five Percent Extirpation Estimates in Arkansas. United States
Environmental Protection Agency, Office of Research and Development. EPA/600/R-22/215.
Cover Photo: Used by permission from Marysia Jastrzebski.
-------
Bae,"™""el Specific Conductivity and
Associated 5% ation Estimates in Arkansas
TABLE OF CONTENTS
1 Executive Summary 6
2 Background Estimates 9
2.1 Data Sets 10
2.2 Ionic Composition and SC Analytics 11
2.3 Observed Least Disturbed Level III Ecoregional Background 12
2.4 Relationship between TDS and SC 13
2.5 Predicted Least Disturbed Background for Stream Segments 17
2.6 Development of Flow Chart to Select Background SC 23
3 Calculation and Assessment of Extirpation Estimates 29
3.1 Extirpation Based on National B-C Model and Arkansas Background Estimates 30
4 General Conclusions 33
5 Quality Assurance and Supplementary Data 35
6 Acknowl edgem ents 35
7 Citations 35
8 Appendices 39
TABLES
Table 1. Summary of regression models for converting TDS (x) mg/1 to Conductivity (y)
(|iS/cm) by Ecoregion 6
Table 2. Estimated specific conductivity (|iS/cm) background and value above which 5%
extirpation is expected 8
Table 3. Summary median statistics for SC (|iS/cm) by Ecoregion, least disturbed stations 12
Table 4. Summary of regression models and r2 values for converting TDS (x) mg/1 to
Conductivity (y) (|iS/cm) by Ecoregion based on the Arkansas data set from 1990-2021.
17
Table 5. Descriptive statistics of the predicted background SC and centiles estimated from an
empirical random forest model 18
Table 6. Weight of evidence used to select scale, data set, statistic, and method for estimating
background SC 25
in
-------
Bae,"™""el Specific Conductivity and
Associated 5% ation Estimates in Arkansas
Table 7. Background estimates obtained from different data sets 28
Table 8. Ecoregional XCD05 values for SC based on the 5% extirpation model and 25th centile
all observed stations and 75th centile of least disturbed Arkansas stations 33
FIGURES
Figure 1. Arkansas data sampling stations sampled (circles) within the seven USEPA
Ecoregions 10
Figure 2. Relationship between TDS and SC for Arkansas data set by ecoregion. Ouachita
Mountain has a markedly different slope 14
Figure 3. Relationship between TDS and SC for Arkansas 15
Figure 4. Relationship between TDS and SC for Arkansas data set by ecoregion 16
Figure 5. Comparison of absolute difference between median predicted background SC and
median observed SC in Arkansas least disturbed stations (N=135) 20
Figure 6. Comparison of absolute difference between predicted median background SC and
median observed SC with Arkansas data set for all stations 21
Figure 7. Scatter plot of Arkansas least disturbed stations 22
Figure 8. Flow chart depicting considerations for selecting background SC in Arkansas streams.
29
Figure 9. Background-to-criterion model (Cormier et al., 2018a) 31
iv
-------
Bae,"™""el Specific Conductivity and
Associated 5% ation Estimates in Arkansas
ABBREVIATIONS
Abbreviation
Definition
|iS/cm
Microsiemens per Centimeter
ADEQ
Arkansas Department of Environmental Quality
ANCOVA
Analysis of Co-Variance
B-C model
Background-to-Criterion Model
Ca2+
Calcium Cation
CFD
Cumulative Frequency Distribution
cr
Chloride Anion
e. g-
Exempli Gratia, For Example
GAM
Generalized Additive Models
HCO3-
Bicarbonate Anion
K+
Potassium Ions
MAE
Mean Absolute Error
Mg2+
Magnesium Ions
N
Number of Stations
n
Number of Samples
NA
Not Available or Not Applicable
Na+
Sodium Ions
NSE
Nash-Sutcliffe Error
SC
Specific Conductivity
S042"
Sulfate Anion
TDS
Total Dissolved Solids
UAA
Use Attainability Analysis
USEPA
United States Environmental Protection Agency
USGS
United States Geological Survey
XC95
Extirpation Concentration
XCD
Extirpation Concentration Distribution
XCD05
5th Centile of an Extirpation Concentration Distribution
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
1 Executive Summary
This report describes the methods to process and curate stream data to estimate background
specific conductivity (SC) in Arkansas. The report is organized into two major sections:
background SC estimates and 5% extirpation estimates. USEPA ORD conducted analyses as
technical support to USEPA Region 6 and the state of Arkansas. The report includes:
1. Characterization of the data set used in the report:
2. Conversion formulae for total dissolved solids (TDS) to SC (linear regression models),
3. Estimation of ecoregional least disturbed background from observed stream
measurements (method: Cormier et al., 2018c),
4. Estimation of stream segment least disturbed background using an empirical random
forest regression model based on geological, climate, and soils and other data (method:
Olson and Cormier, 2019), and
5. A flowchart for selecting recommended background values to estimate 5% extirpation of
benthic invertebrates (Cormier et al., 2018c; USEPA, 2017).
In this report, minimally affected conditions are the physical, chemical, and biological habitats
found in the absence of recognizable human disturbance. Least disturbed conditions are the best
available physical, chemical, and biological habitat conditions given the present degree of
disturbance of the landscape or habitat type (Stoddard et al., 2006). In this report, the term least
disturbed is used when referring to stations designated as "high quality" in the Arkansas data set
because these stations appear to be a mix of minimally affected and least disturbed conditions.
Relationship between TDS and SC: TDS and SC are highly correlated but differ slightly
among ecoregions; therefore, different regression equations were generated to convert TDS to
SC for each ecoregion (Table 1).
Table 1. Summary of regression models for converting TDS (jc) mg/1 to Conductivity (y)
(jiS/cm) by Ecoregion.
Ecoregion
Equation
Arkansas Valley
y = -26.46 + 1.82x
Boston Mountains
y = -14.89 + 1.8x
Mississippi Alluvial Plain
y = -42.62 + 1.86x
Mississippi Valley Loess Plains
y = -69.92 + 1.8x
Ouachita Mountains
y = -13.11 + 1.67x, breakpoint 141.75 mg/1,
y= 157.6 + 0.94x
Ozark Highlands
y = 12.63 + 1.69x
South Central Plains
y = -38.49 + 1.7x
6
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Observed Least Disturbed Ecoregional Background: The background levels of dissolved ions
in streams in Arkansas are among the lowest in the country (Cormier et al., 2018c; Cormier et
al., 2021; Griffith, 2014). Streams often have dissolved ion levels of less than 50 |iS/cm. On
average, ecoregional background estimates based on the 75th centile of least disturbed station
medians are: 63 |iS/cm in the Arkansas Valley, 92 |iS/cm in the Boston Mountains, 105 |iS/cm
in the Ouachita Mountains, and 134 |iS/cm in the South Central Plains. The 75th centile of least
disturbed background SC was higher in the Ozark Highlands (381 |iS/cm) and Mississippi
Alluvial Plain (329 |iS/cm) and background based on the 25th centile of all stations oddly is
much less, 248 and 116 |iS/cm, respectively. The difference between all stations and least
disturbed estimated background suggests that the stations identified as least disturbed include
disturbed stations or that the range of least disturbed conditions is broader than half the sampled
locations. A provisional background estimate for the Mississippi Valley Loess Plains (69 |iS/cm)
is based on the 25th centile of the ecoregion in and outside of Arkansas from an EPA probability
data set because there was only one sample in Arkansas.
Predicted Least Disturbed Background for Stream Segments: Comparison of Arkansas's
least disturbed stations with predicted least disturbed background levels suggests that estimates
were over-predicted in the Boston Mountains and Arkansas Valley. SC levels were higher than
predicted and may be due to an anthropogenically shifting baseline in the Spring River in the
Ozark Highlands.
Flowchart for Selecting Recommended Background Values Weight of Evidence for Least
Disturbed Background Selection: A weight-of-evidence approach was used to select the scale,
data set, and assessment statistic to estimate background for each ecoregion and for site-specific
estimates (Cormier et al., 2018c; USEPA, 2017). Minimally affected background is the preferred
estimate. Where such data or conditions are unavailable, the next most relevant estimated
background is one from one or more nearby, minimally affected locations. Where neither of
these options is available, values recommended for use are shown in Table 2. In the Mississippi
Alluvial Plain, Mississippi Valley Loess Plains, Ouachita Mountains, Ozark Highlands, or South
Central Plains, the default recommended background is the stream segment background
predicted from the empirical model. In the unlikely case that minimally affected background
cannot be estimated for the Boston Mountains and Arkansas Valley, the default recommended
background is the observed ecoregional 75th centile because the model consistently over-
predicts background in these two ecoregions. In all ecoregions, where station SC is less than the
values recommended in Table 2, the station SC takes precedence. These methods, data, and
models may be used to assess the protectiveness of proposed site-specific water quality criteria
proposed by third parties.
5% Extirpation Levels Based on National Extirpation Model: The SC levels expected to
extirpate 5% of benthic invertebrates (XCD05) were estimated from a linear regression
extirpation model (B-C model). The log-log linear model was developed from 24 ecoregions
within the contiguous United States for which XCD05 values had been regressed against the 25th
centile of observed SC values in each ecoregion (Cormier et al., 2018a). Where available,
minimally disturbed background SC estimates are recommended as the independent variable to
7
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
estimate XCD05 values at the smallest scale deemed reliable. When site-specific minimally
disturbed background is not known and cannot be ascertained, XCD05 values may be estimated,
as described above, using the B-C model and the empirically predicted least disturbed
background or observed SC as recommended in Table 2 and the flowchart (Figure 8).
In sum, for any station where minimally affected stations are available for comparison or where
the observed SC is less than an estimated recommended background SC, the observed
background is most relevant and likely more accurate. Therefore, wherever it is possible to
directly measure minimally affected SC background with confidence, that observed background
is more reliable, relevant, and defensible. Anthropogenic background (i.e., least disturbed
background) is a useful metric, but it suffers from the effect of a shifting baseline and can
contribute to worsening water quality (Pauly, 1995; Campbell et al., 2009; Gillon et al., 2016;
Kaushal et al., 2018).
Table 2. Estimated specific conductivity (jiS/cm) background and value above which 5%
extirpation is expected.
Ecoregion
Geomean of predicted stream
segments [range]
75th centile of observed
ecoregional least disturbed
stations frangel
Background
5% extirpation
Background
5% extirpation
Arkansas Valleya
63
ri8.5-423cl
180
[80 - 629cl
Boston
Mountains51
92
T18 - 2591
231
\19 - 4561
Mississippi
Alluvial Plain
125
T55 - 3501
283
T164 - 5561
Mississippi Valley
Loess Plains'3
270
T55 - 3631
469
T164 - 5691
Ouachita
Mountains
100
r54 - 3501
244
T162 - 5561
Ozark Highlands
301
T86 - 4091
503
T221 -6151
South Central
Plains
90
[48 - 375]
228
[150- 582]
'Predicted values were overestimated, therefore XCD05 calculated from 75th centile of ecoregional least disturbed
stations is likely to be more accurate.
background and 5% extirpation levels are provisional.
0 Although identified in data set as least disturbed, these maxima may represent an anthropogenically altered
background.
8
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
2 Background Estimates
This section describes factors influencing natural background SC and then describes analyses
used to develop and recommend least disturbed background estimates.
Stress from elevated ionic concentration, measured as SC, causes adverse effects on a range of
freshwater ecosystems (e.g., Canedo-Argiielles, et al., 2016; Kaushal et al., 2018; USEPA,
201 la). The sources of ions in surface waters may be anthropogenic or natural, reflecting the
level of alteration of soils and geology. Nationally, the two most common background anionic
mixtures in streams are those dominated by either chloride anions (CI") or bicarbonate (HCO3")
plus sulfate (SO42") anions based on mass (Hem, 1985; Griffith, 2014). Calcium (Ca2+) is the
most common cation nationally.
Different mixtures of ions that increase SC are associated with multiple anthropogenic sources,
including discharges from wastewater treatment facilities, ground water recharge, surface and
underground mining, oil and gas exploration, runoff from urban areas, and discharges of
agricultural irrigation return waters, among others (USEPA, 2016). Different ionic mixtures
have been shown to have different toxicities in laboratory tests (Mount et al., 2016; Erickson et
al., 2017; Erickson et al., 2022a, b; Hills et al., 2022) and field studies (Mooney et al., 2020).
However, at low SC-effect levels, SC or a sum of all ions demonstrate only marginally different
toxicities.
To our knowledge, map layers based on stream SC or ionic concentrations have not been
delineated. However, background stream SC is known to be "rock dominant" and affected by
soils, air deposition, precipitation, evapotranspiration, relative ground and surface water
contribution, and other factors (Hem, 1985; Olson and Cormier, 2019). And given that geology
and soil parameters were used to develop national ecoregions (Omernik, 1987, 1995), groups of
streams were sorted by USEPA ecoregions or two groups of combined ecoregions for some
analyses (USEPA, 2013). We found that the different ecoregions of Arkansas have different
natural background SC partly owing to their unique natural geological characteristics. In
Arkansas, the most abundant sediments are sand, clay, silt, gravel, and marl. The most abundant
sedimentary rocks are shale, sandstone, dolostone, limestone, and chert (AGS, 2022). Ionic
sources arise from the dissolution of minerals in rock. For example, sandstone grains of quartz
are resistant but the cement that holds the particles together is not and weathering of exposed
surfaces releases calcium and carbonates.
Arkansas can be divided into a highland area in the northwest and a lowland region in the south
and east. Sedimentary rocks in the highlands are marine. In the southern and eastern parts of the
state, the sedimentary deposits are predominantly fluvial, arising from fresh-water processes.
Rocks in the Ozark Highlands, Boston Mountains, and Ouachita Mountains are dominated by
well-lithified sandstones, shales, limestones, and dolostones. The Arkansas Valley and other
valley floors are associated with the streams and rivers having alluvium deposits of
unconsolidated clays, sands, and gravel formed by freshwater erosional processes and also have
strata of coal. The sedimentary deposits of the lowlands are varied including clay, sand, gravel
9
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
silt, limestone, and lignite, marl, and chalk (Haley B.R. and Arkansas Geological Commission
staff, 1993). Additional details for each ecoregion are included in Appendix A-1.
2.1 Data Sets
2.1.1 Ecoregional Groups
Most analyses were performed by ecoregion (Figure 1). However, for one analysis, ecoregions
were combined to form two groups to allow comparison with a previous EPA report (USEPA,
2016). Ecoregion Group 1 consists of the Arkansas Valley, Boston Mountains, Ouachita
Mountains, and South Central Plains. Ecoregion Group 2 consists of the Ozark Highlands,
Mississippi Alluvial Plain, and Mississippi Valley Loess Plains (Figure 1).
&o°§>o% °
o
JSP ° J 0° ^ oft/n • ° °$S> ° ° 74
f '*•***/, •• °o °°
*•37 * #o i• VV »»••' °°o_o °
. * *.* . o. V .
• **• * ^ «• jl*' • %ot>° o *
/ . fc . o o
;'Y.3.6' . */ *: °o°
«P * • o
• • o
<*\f\v
• O o O ® •
o## 35 • • •
• O Cu©0 O • •
o ® ® °
© _
' oo„oo « ® • ®°°
#o o ®° # O • o
Figure 1. Arkansas data sampling stations sampled (circles) within the seven USEPA
Ecoregions.
TDS records were converted to SC and combined with SC to produce this map. Cooler colored
circles indicate lower SC (e.g., violet and blue); warmer colors indicate higher SC (e.g., yellow
to red); gray circles are <10 or > 5000 nS/cm. Group 1: (35) South Central Plains, (36) Ouachita
10
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Mountains, (37) Arkansas Valley, (38) Boston Mountains. Group 2: (39) Ozark Highlands, (73)
Mississippi Alluvial Plain, (74) Mississippi Valley Loess Plains.
2.1.2 Observational Data Sets
Descriptive statistics (e.g., centiles, minima, maxima, means) were estimated for Arkansas
stream observations to characterize chloride, sulfate, TDS, SC, and relative ionic composition
(Appendix A.2.1-14). Data were obtained directly from the Arkansas Department of
Environmental Quality (ADEQ) for the years 1990-2021 but are also available from the EPA
Water Quality Portal (https://www.waterqiialitydata.iis). Data processing and links to R-scripts
are available from the authors. Of the 16,745 records, 21 records designated as not available
(NA) were removed as were 6 with values < 1 |iS/cm which were assumed to be data
management errors or measures of rainwater. TDS values below detection limits were removed
(TDS < 5 mg/1). For maps, TDS records were converted to SC and combined with SC. For
characterizing SC, only SC measurements were used.
For analyses of ecoregional predicted SC, a stream segment that crossed an ecoregional
boundary was included in both ecoregional estimates. Therefore, the sum of ecoregional stream
segments exceeds the total number of stream segments in Arkansas.
2.1.2.1 Arkansas Data set (1990-2021)
The frequency and period of sampling varies greatly among the stream stations. Some stations
(ambient stations) were sampled monthly from the early 1990s to the present, while other sites
were sampled intermittently (roving stations) or to address a particular concern (e.g., special
survey, toxicological survey, or use attainability analysis (UAA)). For most of the analysis in
this report, concentrations are based on station medians for a given water quality parameter so
that all stations are given equivalent weight regardless of sampling intensity. Parameters
included SC, individual ions, TDS, pH, and genus level occurrences.
2.1.2.2 Arkansas Least Disturbed Stations (1990-2021)
The Arkansas least disturbed stations data set is a subset of the Arkansas data set. For this
analysis, rivers and streams classified by ADEQ as high-quality resource waterbodies were
treated as least disturbed stations. ADEQ identifies high-quality stations based on best
professional judgment including confirmation of no (or very few) point source discharges and no
substantial areas of nonpoint source disturbances; land use data and other factors (Green, 2014;
ADPCE, 1987a, b). A complete list of station names, station descriptions, and where applicable,
the corresponding least disturbed stream classifications is available from the authors.
2.2 Ionic Composition and SC Analytics
In the Boston Mountains, Mississippi Alluvial Plain, and Ozark Highlands ecoregions, the
relative proportion of divalent cations, [Ca2+] + [Mg2+] is greater than monovalent cations, [Na+]
+ [K+] based on mass. In the Ouachita Mountains, Arkansas Valley, and South Central Plains
ecoregions, [Na+] + [K+] is dominant at 5%, 33% and 58% of least disturbed stations,
respectively (Table A.2.5). Although the possible cause was not investigated, oil and gas wells
11
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
predominately occur in these two ecoregions (Map A.6.1). This suggests that the higher Na+
levels may be influenced by anthropogenic causes rather than minimally affected ion
composition. In single species tests, Na+ + K+ are more toxic than Ca2+ + Mg2+ on a mass-basis
(e.g., Mount et al., 2016; Erickson et al., 2022a, 2022b; Hills et al., 2022).
The anionic composition was characterized using measured alkalinity, pH, sulfate, and chloride
from each station (Tables A.2.1 to A.2.4). For all ecoregions except the South Central Plains
(96%), 100% of the least disturbed stations were dominated by conditions of bicarbonate
(HCO3") plus (SO42") being greater than chloride (CI"), often by factors >10 (Table A.2.6). Of the
28 least disturbed stations identified by ADEQ in the South Central Plains, only one station
appeared to be chloride-dominated. It is unclear if the station was misclassified as least disturbed
conditions or is an anomalous station. HCO3" and SO42" dominance is consistent with Hem
(1985) and Griffith (2014). Therefore, where chloride is the dominant ion, then the natural ionic
mixture is likely due to anthropogenic inputs exhibiting altered relative amounts and
concentrations of ions. Exceptions may occur at natural salt springs, but they were not among
the least disturbed stations included in this analysis and not typical of water in Arkansas streams.
Sodium and bicarbonate are the dominant ions in Arkansas hot springs in the Ouachita
Mountains. (Kresse and Hays, 2009).
2.3 Observed Least Disturbed Level III Ecoregional Background
The background levels of dissolved ions in streams in Arkansas are among the lowest in the
country (Cormier et al., 2018c; Cormier et al., 2021; Griffith, 2014), a testament to the natural
characteristics of the region and little or moderate anthropogenic alteration in many areas.
Streams in Arkansas are often below 50 |iS/cm (Tables A.2.1 to A.2.4). Based on the 75th centile
of least disturbed station medians, ecoregional background values are: 63 |iS/cm (Arkansas
Valley), 92 |iS/cm (Boston Mountains), 105 |iS/cm (Ouachita Mountains), and 134 |iS/cm
(South Central Plains). Background SC was higher in the Ozark Highlands (381 |iS/cm) and
Mississippi Alluvial Plain (328 |iS/cm) (Table 3).
Table 3. Summary median statistics for SC (jiS/cm) by Ecoregion, least disturbed stations.
Individual values represent the median of all measurements for a given parameter at that site.
Arkansas
Valley
Boston
Mountains
Mississippi
Alluvial Plain
Ouachita
Mountains
Ozark
Highlands
South
Central
Plains
Minimum
18.5
18
57
3
5
6
10th centile
30
31
110
29
148
38
25th centile
35
44
152
38
201
57
50th centile
45
67.5
239
58
279
92
75th centile
63
92
329
105
381
134
12
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Geo Mean
49
64
224
61
263
91
Maximum
423
259
891
472
568
1150
n (samples)
13
26
5
32
55
22
N (stations)
593
400
253
1203
1847
841
St. Dev.
31.39
43.08
146.05
46.95
105.22
114.07
2.4 Relationship between TDS and SC
TDS and SC are two common measurement endpoints for characterizing the total ionic strength
or salinity of a water body. In freshwater systems, TDS, usually expressed as mg/1, is a measure
of the total mineral content of water typically determined by the weight of the evaporates. SC is
expressed as [j,S/cm. It measures dissolved mineral concentration from the relationship in which
the conductance of the flow of electrical current increases as the concentration of dissolved ions
increases. In the past, total ionic concentration was often measured by the ADEQ as TDS but
more recently as SC. In this section, the development of ecoregional models is described for
converting TDS (mg/1) to SC (|iS/cm).
Linear regression models of TDS and SC were calculated for each ecoregion using the Arkansas
data set. Prior to analysis, values of <5 mg/1 TDS were removed and assumed to be below the
detection limit. Values of <1 |iS/cm conductivity were also removed as likely below detection
limit.
Visual inspection of linear regression models suggested that the slope of the Ouachita Mountains
was different from the other ecoregions and appeared to have at least two distinct slopes. The
cause of the distinct slopes was not investigated. However, anthropogenic inputs different from
freshwater ionic composition may be reflected in the less steep slope, which occurs at higher SC.
A broken-stick regression analysis revealed an inflection at 141.75 mg/1 TDS (Figure 3). Note
that there are two regression models with different range limits for Ecoregion 36 that are used to
convert TDS to SC, but the single r2-value relates to the combined performance of both
regression models.
Overall, TDS and SC are highly correlated with r2 values ranging from 0.86-0.99. An analysis of
co-variance (ANCOVA) showed that the slopes and intercepts of each ecoregional regression
model (Ecoregions 35, 37, 38, 39, 73, and 74, excluding Ouachita Mountains, Ecoregion 36) are
significantly different from one another. Also, the revalues of individual regression models are
stronger than the combined data of these six ecoregions (Figure 3). Therefore, conversions from
TDS to SC are expected to be more accurate using regression models tailored to each ecoregion
(Figure 4, Table 4). These models were used for maps to convert TDS to SC.
13
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Arkansas Valley Mississippi Alluvial Plain Ouachita Mountains South Central Plains
Ecoregion
Boston Mountains Mississippi Valley Loess Plains Ozark Highlands
Figure 2. Relationship between TDS and SC for Arkansas data set by ecoregion. Ouachita
Mountain has a markedly different slope.
Values less than or equal to 5 mg/1 TDS and 1 jiS/cm were removed prior to analysis.
14
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
TDS (mg/L)
• Arkansas Valley Mississippi Alluvial Plain Ouachita Mountains South Central Plains
Ecoregion
Boston Mountains Mississippi Valley Loess Plains Ozark Highlands
Figure 3. Relationship between TDS and SC for Arkansas
(a) All ecoregions except Ouachita Mountains and (b) Ouachita Mountains. Broken stick
regression at 141,75 mg/1 TDS is characterized by steeper slope at low SC. Values less than or
equal to 5 mg/1 TDS and 1 |iS/ctn were removed prior to analysis.
15
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
1250-
1000-
750-
500-
250-
0-
1000-
x—«.
E
750-
o
cn
500-
=L
250-
-4—'
>
o-
o
13
"O
c
o
O
2000-
o
M—
1500-
o
ID
n
1000-
CO
500-
0-
Arkansas Valley
Ouachita Mountains
R2 = 0.96
Eq1: y = -24.26 + 1,9x
% *•
Eq2: y = 104.06 + lx
Ozark Highlands
400
500
1000
1500
250
500
750
TDS (mg/L)
Figure 4. Relationship between TDS and SC for Arkansas data set by ecoregion.
Ouachita Mountain has two equations with a breakpoint at 141.75 mg/1 TDS. Values less than or
equal to 5 mg/1 TDS and 1 |iS/cm were removed prior to analysis.
16
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Table 4. Summary of regression models and r2 values for converting TDS (jc) mg/1 to
conductivity (y) (jiS/cm) by ecoregion based on the Arkansas data set from 1990-2021.
Ecoregion
N
r2
Eql
Eq2
36
Ouachita Mountains
3046
0.96
y = ~24.26 + 1.9x
y = 104.06 + lx
(breakpoint: 141.75
mg/1)
39
Ozark Highlands
3085
89
y = 6.46 + 1.78x
NA
38
Boston Mountains
600
0.97
^ = -14.89+ 1.8x
NA
37
Arkansas Valley
2066
0.99
y = -26.44 + 1.82x
NA
74
Mississippi Alluvial Plain
1743
0.90
y = -44.68 + 1.87x
NA
35
South Central Plains
4298
0.96
y = -38.83 + 1.7x
NA
73
Mississippi Valley Loess
Plains
73
0.86
y = -69.92 + 1.8x
NA
2.5 Predicted Least Disturbed Background for Stream Segments
The mean least-disturbed predicted background SC was empirically estimated for stream
segments throughout Arkansas using an empirical random forest regression model of
geophysical attributes and ecoregional background (Olson and Cormier, 2019). Descriptive
statistics are shown in Table 5. Because the predicted background is based on least disturbed
stations, these predicted values may be useful for evaluating whether observed SC has been
altered by anthropogenic activity (Cormier et al., 2018c). Stations identified as least disturbed
may be the best in an area, but that area may be anthropogenically altered and quite different
from a stream that is minimally affected by people. For example, whereas elsewhere in the
Ozark Highlands the predicted background SC values are within 100 |iS/cm of observed values,
the observed SC near Spring River is greater than 100 |iS/cm of the predicted SC and may
indicate an anthropogenically shifting baseline. Least disturbed background is a useful metric,
but it suffers from the effect of a shifting baseline and can contribute to worsening water quality
overtime (Pauly, 1995; Campbell et al., 2009; Gillon et al., 2016; Kaushal et al., 2018).
In contrast, in the Boston Mountains, observed SC is consistently less than the empirically
modeled least disturbed background. The lower observed SC is a better estimate of minimally
affected background than a model of least disturbed stations. Therefore, where observed SC is
less than the predicted background SC, then the observed SC is a more relevant and accurate
measure of minimally affected conditions than modeled estimates.
17
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Table 5. Descriptive statistics of the predicted background SC and centiles estimated from
an empirical random forest model.
Centiles
Ecoregion
Min
10th
25th
50th
75th
Max
Mean
GeoMean
N
Arkansas Valley
57
96
108
122
144
449
142
133
11975
Boston Mountains
69
134
176
237
295
464
235
222
6382
Mississippi
Alluvial Plain
55
103
115
125
230
350
165
150
20718
Mississippi Valley
Loess Plains
55
97
117
270
319
363
233
208
1576
Ouachita
Mountains
54
69
89
100
114
350
110
104
9605
Ozark Highlands
86
242
267
301
315
409
290
288
12620
South Central
Plains
48
69
73
90
126
375
116
104
26265
Arkansas
48
74
97
124
249
464
166
145
82679
The total number of stream segments from all ecoregions (summing all the values in the N column from row 1 to row 7) will be larger than the
total number of the entire Arkansas data set. This is because a segment crossing the boundary of two ecoregions is included in each ecoregion,
but only once for the state estimate.
2.5.1 Validation of the predicted background model for Arkansas ecoregions
The performance of the predicted background model for Arkansas ecoregions was evaluated by
comparing predicted background SC with observed SC at least disturbed stations from the
Arkansas data set. A map of the absolute difference between observed SC and predicted
background SC shows that the predictive background model overestimates SC in the Boston
Mountains and Arkansas Valley (Figure 5) and underestimates in the northeastern Ozark
Highlands in parts of the Spring River drainage. The same patterns were observed with values in
the non-reference Arkansas data set (Figure 6). In both analyses, the model underestimated SC in
parts of the Ozark Highlands and Mississippi Alluvial Plain, suggesting a greater level of altered
SC in these two ecoregions (green circles) (Figures 5 and 6).
A scatter plot of the absolute difference between observed SC and predicted background SC
shows that the observed values at Boston Mountain and Arkansas Valley stations consistently
fall below the 1/1 line, indicating that the predicted values are overestimated (Figure 7). For the
least disturbed Arkansas data set excluding the Boston Mountains and Arkansas Valley, 62.5% of
observed SC values were within 61 |iS/cm and 81.2% were within 100 |iS/cm of predicted
background SC values. Stations in the Ozark Highlands varied greatly and observed values in
the Spring River drainage were often slightly more than 100 |iS/cm than predicted by the
random forest model, suggesting that baseline SC for the river may be shifting toward more
mineralized conditions or that there are unusual natural factors not considered by the empirical
model at these stations. A site-specific assessment would be needed to identify probable causes.
18
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Statistical metrics for predictive performance were calculated after removing the Boston
Mountain and Arkansas Valley. To compensate for bias from repeat sampling, stations with the
same geographical location (samples with the same catchment identified for the ComID, where
ComID is a unique stream segment identifier in NHDPlus) were pooled and the medians of
unique stations were calculated. For the least disturbed Arkansas data set, the model explained
most of the variation in SC and produced fairly accurate predictions for the least disturbed
Arkansas data (Mean Absolute Error (MAE) = 61 |iS/cm, Nash-Sutcliffe Error (NSE) = 0.62,
and r2 = 0.64, percent bias 6.3%).
Due to the over-estimation of background SC in the Arkansas Valley and Boston Mountains, we
attempted to calibrate the predicted values using long term SC observations collected at gaging
stations by the United States Geological Survey (USGS) (Appendix A-2). Although the temporal
patterns were qualitatively similar, rising and falling at the same time for observed and predicted
background SC, a consistent difference in average magnitude was not discernable and so
calibration was not done.
The percentage of sites greater than the predicted background was estimated for all stations in
the Arkansas data set (Figure 6 and Figure 7). For the Arkansas data set excluding the Boston
Mountains and Arkansas Valley, 53.5% of observed SC values are within 61 |iS/cm and 75% are
within 100 |iS/cm of predicted background SC values. This indicates that more than half of the
monitored streams in the state are within the MAE of the predicted background model's
estimate.
Because the predictive model consistently overestimated background SC for the Boston
Mountains and Arkansas Valley compared to observed SC, that poses a risk for under-protecting
the aquatic life in those two ecoregions. Because observed background in freshwater is
substantially lower than the modeled predicted background, the ecoregional 75th centile of least
disturbed stations is recommended in lieu of the predicted background for the Boston Mountains
and the Arkansas Valley, where an ecoregional estimate is needed.
19
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
O o
5
\
i
Specific
Conductivity,
US/cm
|r
0 to 100
100 to 200
200 to 300
300 to 400
400 to 500
Missing
Predicted Results
O Overpredicted
O Within +/- 100 pS/cm
O Underpredicted
Figure 5. Comparison of absolute difference between median predicted background SC
and median observed SC in Arkansas least disturbed stations (A/=135).
Predicted background levels are overestimated in the Boston Mountains and Arkansas Valley. SC
is underestimated in the Ozark Highlands near Spring River and may be due to anthropogenic
alteration or unusual natural sources. No least disturbed stations were identified for the
Mississippi Valley Loess Plains.
20
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Specific
Conductivity,
[iS/cm
Oto 100
100 to 200
200 to 300
300 to 400
400 to 500
Missing
Predicted Results
O Overpredicted
O Within +/-100 |jS/cm
O Underpredicted
Figure 6. Comparison of absolute difference between predicted median background SC
and median observed SC with Arkansas data set for all stations.
Observed values in the Arkansas data set (A; 811) confirm the bias for over-estimation in the
Boston Mountains and Arkansas Valley (red circles). The model underpredicted in parts of the
Mississippi Alluvial Plain and Western portion of the South Central Plain, indicative of altered
SC regimes (green circles).
21
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
100 150 200 250 300
Predicted background SC (pS/cm)
Figure 7. Scatter plot of Arkansas least disturbed stations.
Oblique line is 1:1 line. Background SC values in the Boston Mountains (blue circles) and
Arkansas Valley (red circles) are overestimated and indicate that the model is unreliable for this
ecoregion. Three Boston Mountain stations are questionably low with observed SC near zero
and are either below detection limit or indicative of data management errors. No least disturbed
stations were identified for the Mississippi Valley Loess Plain.
JOO
£ 300
O
CO
o
CO
"O
•D
>
o
c
<5
~a
o
200
Ecoregion
• Arkansas Vaiey
• Boston Mountains
• Mississippi ADuvial Plain
• Ouachita Mountains
• Ozark Highlands
O Soulh Central Rains
22
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
2.6 Development of Flow Chart to Select Background SC
2.6.1 Weight of Evidence Used to Select Background SC Estimates
A weight-of-evidence approach was used to select the scale, data set, and assessment statistic to
estimate background for each ecoregion and site-specific estimate (Cormier et al., 2018c; USEPA,
2017). Comparing available and relevant evidence provides greater confidence than one line of
evidence and increases transparency in the decision process. Relevance and reliability were scored
using symbols indicating support (+) or weakening (—) of the option for selecting a data set, the
statistic for background, and appropriate spatial scale (Table 6). Based on the weight of evidence,
least disturbed Arkansas ecoregional background estimates were calculated (Table 7). A flow chart
was developed that depicts considerations for selecting background SC in Arkansas streams based
on location, comparisons, and available data (Figure 8).
Minimally affected background is more relevant and therefore the preferred estimate. Where such
data or conditions are unavailable, the next most relevant estimated background is one from one or
more nearby, minimally affected location. Where neither of these options is available, values
recommended for use are shown in Table 7. In the Mississippi Alluvial Plain, Mississippi Valley
Loess Plains, Ouachita Mountains, Ozark Highlands, or South Central Plains, the recommended
background is the stream segment empirically modeled background. In the unlikely situation that
minimally affected background cannot be estimated from observed SC for the Boston Mountains
and Arkansas Valley, the recommended background is the ecoregional 75th centile because the
model consistently over-predicts background in these two ecoregions. These methods, data, and
models may be used to assess the protectiveness of site-specific water quality criteria proposed by
third parties.
Ouachita Mountains, Ozark Highlands, and South Central Plains: The predictive model
appears to reliably and realistically estimate background for stream segments in the Ouachita
Mountains, Ozark Highlands, and South Central Plains with an error less than 100|iS/cm.
Therefore, the stream segment predicted background is a reasonable default for estimating
background SC where observed measures are greater than the predicted background, or the
background is known to be altered from minimally affected conditions.
Mississippi Alluvial Plain: The predicted background of the Mississippi Alluvial Plain (125
|iS/cm) was within 20 |iS/cm of the 25th centile of all stations in the Arkansas data set (116 |iS/cm)
and of the stations in the ecoregion wide probability data set (132 |iS/cm). Although the background
for least disturbed stations should be lower than an estimate of all stations, the estimated
background at the 75th centile is about double the 25th centile of all Arkansas stations (N= 36) and
ecoregion-wide stations (N= 27), and the 75th centile was greater than half of all stations in the
ecoregion in Arkansas. This may be attributed to the paucity of least disturbed stations (N= 5) in the
Mississippi Alluvial Plain Arkansas data set. Owing to the few least disturbed stations and the
disparity between the resulting 75th centile of least disturbed stations and other metrics, the
predicted background was judged to be a better estimate than one based on five least disturbed
23
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
stations in the Mississippi Alluvial Plain. Therefore, the stream segment predicted background is
recommended for estimating background SC where observed measures are greater than the
predicted background.
Mississippi Valley Loess Plains: There is insufficient data to estimate background for the
Mississippi Valley Loess Plains based on the Arkansas data set, and there is a large discrepancy
between the observed EPA ecoregion-wide probability stations from the entire ecoregion (69
|iS/cm, N= 26) compared to the predicted background median (270 |iS/cm, range 55-363 |iS/cm)—
which may be reflecting altered vegetative cover that is almost entirely agricultural rather than
native forest. The background for this ecoregion is uncertain and no one approach is strongly
recommended at this time. As a practical provisional default estimate, the empirically modeled
background is recommended in the absence of least disturbed stations for comparison.
Boston Mountains and Arkansas Valley: The predictive model consistently overestimates
background for the Boston Mountains and Arkansas Valley. This may be due to the high proportion
of intact ecosystems in this ecoregion in Arkansas and the challenge of modeling in a left censored
data set bounded by zero. For this reason, the predicted background is not recommended. Instead,
either observed SC at the station less than the 75th centile or the 75th centile for ecoregional least
disturbed stations is recommended to estimate 5% extirpation levels, whichever is the lower
estimate.
2.6.2 Recommended process for estimating background from predicted and observed
data
For any station where nearby minimally affected stations are available for comparison, the
observed background is most relevant and likely more accurate. Minimally affected conditions
are the physical, chemical, and biological habitat found in the absence of significant human
disturbance (Stoddard et al., 2006). Therefore, wherever it is possible to directly measure
minimally affected SC background, not least disturbed background, with confidence, that
observed background is the more reliable, relevant, protective, and scientifically defensible
estimate.
Anthropogenic background or least disturbed conditions are the best available physical,
chemical, and biological habitat conditions given the present degree of disturbance of the
landscape or habitat type (Stoddard et al., 2006). Least disturbed background is a useful metric,
but it suffers from the effect of a shifting baseline and can contribute to worsening water quality
(Pauly, 1995; Campbell et al., 2009; Gillon et al., 2016; Kaushal et al., 2018).
Where it is necessary to estimate background SC in the absence of minimally affected stations,
predicted background SC is recommended for the Mississippi Alluvial Plain, Mississippi Valley
Loess Plains, Ouachita Mountains, Ozark Highlands, and South Central Plains or the observed
SC, whichever is less. The Arkansas Level 3 ecoregional estimates are recommended as a
default estimate for the Boston Mountains and Arkansas Valley or the observed SC, whichever
is less. A flow diagram depicts the process (Figure 8).
24
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Table 6. Weight of evidence used to select scale, data set, statistic, and method for
estimating background SC.
Relevance
Reliability and Justification
Scale
Scores: supports (+), weakens (—), not applicable (NA)
Level 3
ecoregions
The objective is to estimate site-
specific background, so this scale
is likely to be less relevant than at
state scale.
+
Reliability is less at greater scales
because natural variations are expected
to increase with scale. However, sample
size is larger than when constrained to
Arkansas which increases confidence,
especially for Ecoregion 74.
+
Level 3
ecoregions
within
Arkansas
More relevant because scale is
smaller than entire level 3
ecoregion and is within Arkansas.
++
This is an intermediate scale and
reliability is also intermediate for
estimating at a stream reach.
+
Ecoregion 74 is very small within
Arkansas and number of samples are
also few in number, so an estimate
would be unreliable.
-
Stream
segment
(e.g.,
predicted,
comparison
stations)
Stream segment is very relevant to
the application of site-specific
benchmarks.
+++
This is among the smallest scales and
can be reliable depending on available
data and predictive performance of a
model.
+
Stream
station (e.g.,
observed)
Stream station is most relevant to
the application of site-specific
benchmarks.
+++
This is the smallest scale and can be
reliable depending on availability of
comparison stations and rigor of ground-
truthing.
+
Summary
Stream segment or station is most relevant to the application of site-specific benchmarks.
However, depending on data sets, larger scales may be more reliable estimates when there is
a paucity of localized data or model predictions are weak.
Data set: Source of Estimate
Predicted
The data used in the model screens
for anthropogenic disturbance,
which makes it more relevant, but
excludes local natural sources.
+
At the regional scale, validation
indicated predicted background was
estimated with reasonable accuracy. For
Arkansas least disturbed stations
excluding the Boston Mountains and
Arkansas Valley, 81.2% were within 100
+
25
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Relevance
Reliability and Justification
(iS/cm of predicted background SC
values.
The Boston Mountains and Arkansas
Valley were consistently overestimated.
-
Observed
Arkansas
stations
Ecoregion metric: Measurements
are inherently relevant to current
conditions but include various
levels of anthropogenically
disturbed waters.
-
Ecoregion metric: Data set was quality
assured, but records of <1 (iS/cm suggest
not all values are valid and may affect
the 25th centile estimate.
+
Stream station or segment: Many
streams in Arkansas have low
background SC even though they
have not been designated as least
disturbed and therefore may be
relevant for comparison.
+
Stream station or segment: Data quality
must be reviewed on a case-by-case
basis. Not reliable on its own, but where
it is less than other metrics it is
justifiably the best estimate of
background.
+
Observed
Arkansas
least
disturbed
Measures of least disturbed
stations are most relevant, because
they represent the type of stream
condition of interest, and the
samples are from Arkansas.
However, stations may represent
best available (least disturbed)
rather than natural background
condition (minimally affected).
++
Data set was quality assured, but records
of <1 (iS/cm suggest not all values are
valid but influence on the 75th centile is
likely to be negligible. Ground truthing
by state agencies strengthens the
designation of least disturbed status for
this data set.
+
Summary
Predicted estimates are relevant, accurate, and minimally influenced by anthropogenic
alteration but less reliable for the Boston Mountains and Arkansas Valley. Observed data
from least disturbed stations are more relevant but may include stations with anthropogenic
influences. The Arkansas least disturbed station data set is more relevant than other data sets
but is not available for all stream segments and there are none in the Mississippi Valley
Loess Plains.
Background statistic
10th
The small difference between 10th
and 25th centiles suggest that the
10th centile may be too
conservative for all ecoregions,
especially for Ecoregions 39 and
74.
Low centile sites are more likely to
characterize background but may be
conservative because there were many
sites with low SC across the state. Also,
the 10th centile is less reliable than a
central tendency or larger centile
especially with a small number of
available of stations.
+
26
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Relevance
Reliability and Justification
25th
The 25th centile of a mixed
probability data set may be
comparable to the 75th centile of
best available, least disturbed
stations (USEPA, 2011), but
others have disputed
appropriateness of using a fixed
centile (Herlihy and Sifneos,
2008).
+
The small SC increase from 10th to 25th
centile suggests that the 25th is a
reasonable background estimate in this
non-randomized Arkansas data set and is
less influenced by potential errors among
low values.
+
75th
The 75th centile of best available,
least disturbed stations may be
comparable to the 25th centile of a
mixed probability data set
(USEPA, 2011), but others have
disputed the practice of using a
fixed centile (Herlihy and Sifneos,
2008).
+
The 75th centile of least disturbed
stations tended to be greater than other
estimates, but this may be due to the
overall higher percentage of high-quality
streams in the state. Alternatively, the
higher value may be due to the inclusion
of non-reference stations in the data set.
The 50th centile of all stations was often
similar to 75th centile of least disturbed
stations (Table 7)
++
Central
tendency
For the model prediction, geomean
or median is relevant because the
model predicts an annual estimate
of least disturbed stations.
+
Central tendencies are the most robust
statistic because they represent all the
data.
++
Summary
Depending on the data set, any statistic may be appropriate to estimate background in
Arkansas except the 10th centile.
Weight-of-
Evidence
Summary
For any station where nearby minimally affected stations are available for comparison or
where the observed SC is less than an estimated background SC, the observed background
is most relevant and likely more accurate.
Where it is necessary to estimate background SC:
• Mississippi Alluvial Plain, Mississippi Valley Loess Plains, Ouachita Mountains, Ozark
Highlands, South Central Plains—the mean predicted background SC is recommended
because the model is reliable and the need is for local predictions.
• Arkansas Valley and Boston Mountains—the 75th centile of Arkansas ecoregional least
disturbed stations is recommended because the ecoregion is the smallest reliable scale
and because the predictive model overestimates background.
Note: Scores: supports (+), weakens (—), not applicable (NA)
27
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Table 7. Conductivity background estimates (jiS/cm) obtained from different data sets.
Number of stations in the data set in parentheses, range in brackets. Cells in green were
identified by weight of evidence as the more reliable estimate.
Predicted
Observed
Arkansas least
disturbed, 75th
centile0
Observed
Observed
ecoregion
Arkansas stream
Arkansas
station
Arkansas
station
Observed Ecoregion-
wide EPA probability
segments
geomeanb
medians, 25th
centile
medians, 50th
centile
samples, 25th centile3
Arkansas
122 (11975)
(t i
38
61
32
Valley
[56.82 - 449.451
(l'i
(43)
(43)
(47)
Boston
237
53
(39)
91
(39)
23
(26)
Mountains
(6382)
[68.74 - 464.361
(2(>i
Mississippi
Alluvial
Plain
125(2u"IS)
329
116
215
132
154 (>2 - U'J 5X|
(5)
(36)
(36)
(27)
Mississippi
Valley Loess
Plains3
2"( i
( 15"<>)
154 (.2 - '<>2 5X|
NA
89
(1)
164
(1)
69
(26)
Ouachita
|oo (lJ<>o5)
105
38
56
22
Mountains
154 2<> - '5o 4~|
(32)
(85)
(85)
(50)
Ozark
'ill ( I2(.2(H
381
248
318
362
Highlands
|S(. 24 - 4o<) iKi|
(55)
(128)
(128)
(54)
South Central
•>o <2<>2<>5)
134
60
94
51
Plains
|4X IS - '"5 '4|
(22)
(129)
(129)
(60)
"Disparity between Ecoregion-wide and only one observed station in Arkansas suggests a conservative provisional
estimate.
b Sources: Cormier, S.M., Zheng, L., Hill, R.A., Novak, R.M. and Flaherty, C.M. 2018c. A flow-chart for developing
water quality criteria from two field-based methods. Science of The Total Environment, 633: 1647-1656.
c Cormier, S., Wharton, C., and Olson, J. 2021. USEPA Freshwater Explorer. V: 0.1. U.S. Environmental
Protection Agency. https://arcg. is/KHb 9S
28
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Recommended
Location Comparison background
Figure 8. Flow chart depicting considerations for selecting background SC in Arkansas
streams.
3 Calculation and Assessment of Extirpation Estimates
Field observational data have been successfully used for developing exposure-effect models and
risk levels associated with salts (USEPA, 2011; Cormier et al., 2020; Humphrey and Chandler,
2018; MPCA, 2020). Although EPA released a method for deriving benchmarks for SC based on
the extirpation of benthic invertebrates using large regional data sets with paired biology and
chemistry data, these types of data are not always available or of sufficient size. ORD developed
an approach requiring no biological data. The approach uses local background SC to predict the
5th centile from distribution of extirpation concentration (XCD05) using a national SC benthic
invertebrate extirpation model (B-C model) (Cormier et al., 2018a).
The method using the B-C model was also selected because it measures the diversity of
vulnerable species' responses, species interactions, autecology, and routes and dynamics of
exposure (Gerritsen et al., 2015) rather than using SC alone to set an environmental threshold.
The results are ecoregional estimates that reflect localized conditions, allowing XCD05 values to
be estimated at the stream reach spatial scale.
29
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
3.1 Extirpation Based on National B-C Model and Arkansas Background Estimates
The underlying basis for the model is the SC range that is occupied by different benthic
macroinvertebrate species. Natural conditions limit where species can thrive. The lowest SC
niches are not necessarily filled by the same species at each location. Where a niche is absent
due to natural factors affecting background, species specialized for that absent niche are also
absent. In other words, species specialized for niches less than natural background in a region
are unable to compete and survive because conditions are not suitable for their survival. As a
result, biological communities differ from place to place, but the lowest tolerance limit cannot be
lower than the lowest SC niche. So, as background SC increases, the tolerance values of the
most sensitive species are greater and likewise, the 5th centile of those tolerance values also
increases. This translates into a positive mathematical relationship between increasing
background and increasing minima of niches.
This basic ecological relationship was mathematically modeled using species sensitivity
distributions from many data sets with different background SC regimes and therefore different
ionic-niche structures. The model was constructed using 24 data sets with XCD05 paired with
the 25th centile SC for the data set. The 25th centile does not necessarily reflect minimally
affected or least disturbed conditions; rather, it is an estimate of the background of the dataset
experienced by the biota in that dataset. The resulting model is a linear logio-logio least square
regression model that can be used to estimate the SC likely to cause 5% extirpation with just the
input of background SC (Figure 9) (Cormier et al., 2018a). The formula for the mean model
prediction (Eq. 1) is shown below with an example calculation.
30
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
1000
E
O
c/5
=L
J 500
>
o
"O
c
o
o
o
"o
CD
Q_
en
s
O 100
x
50
Figure 9. Background-to-criterion model (Cormier et al., 2018a).
Circles represent 24 ecoregions. In the formulae, y- and x-values are expressed as logio. The
formula is logioy = 0.658* logiox + 1.071. The solid oblique line is the least squares regression
model, with 90% confidence limits shown as dashed lines. An example background value was
inserted into the model as the independent x variable to yield the SC value likely to cause
extirpation of 5% of benthic invertebrates (XCD05). The example calculation shows the vertical
dashed line at 100 |iS/cm intercepts the mean regression line at 244 |iS/cm.
Using the background SC estimates in Table 7, example XCD05 values were calculated for each
ecoregion in Arkansas using the B-C extirpation model. For each stream segment, the SC
predicted from the empirical model was used as the independent variable to estimate an XCD05
(Table 8) except for the Boston Mountains and Arkansas Valley Ecoregions. For these two
ecoregions, the observed least disturbed background for each ecoregion was used as the
example.
An example calculation using background SC to estimate an XCD05 from a stream segment or
ecoregion is shown below (Eq 1.). SC least disturbed background is the independent variable
used in the 5% extirpation model to estimate dissolved ion levels likely to extirpate 5% of
aquatic life. As an example, an estimated background SC of 100 |iS/cm was used as the
independent variable (x) in the B-C model to generate the SC level expected to extirpate 5% of
aquatic life (XCD05)(y) in Eq 1.
25th centile specific conductivity {uS/cm)
31
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
25th centile Specific Conductivity (|iS/cm)
logioy = 0.658* logiox + 1.071 Eq. 1
logioy = 0.658* logio(100 |iS/cm) + 1.071
y = XCD05 =244 |iS/cm
As an example, the SC 5% extirpation level from Eq. 1 (244 |iS/cm) was converted to TDS
using the SC-TDS regression generated from the Arkansas data set for Ecoregion 35, South
Central Plains (Table 4, Figure. 4). The XCD05 from Eq. 1 was used as the independent variable
(x) to predict TDS XCD05 (y). An example is shown in Eq. 2.
y (244 |iS/cm) = -38.83 + 1.7*x mg/1 Eq. 2
244 |iS/cm + 39 =1.7 x mg/1
282.83/1.7 =x mg/1 = 166.37 mg/1
TDS XCD05 = 166 mg/1
32
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Table 8. Ecoregional XCD05 (jiS/cm) values for SC (jiS/cm) based on the 5% extirpation
model.
Green cells are recommended XCD05 values.
Median geomeans of predicted
stream segments and
[range]
75th centile of median ecoregional
least disturbed stations and
[range]
Ecoregion
Background
5% extirpation
Background
5% extirpation
Arkansas Valleya
122
[57 - 449]
278
63
[18.5 - 423c]
ISO
| Si) _ o2^|
Boston Mountains51
237
[69 - 464]
430
92
[18 -259]
231
17^ - 4501
Mississippi Alluvial
Plain
125
[55 - 350]
2X3
| IM - 55o|
328
[57 - 891°]
533
Mississippi Valley
Loess Plains'3
270
[55 - 363]
| IM - 5(^|
NA
NA
Ouachita
Mountains
100
[54 - 350]
244
| 102 - 55o |
105
[3 - 472°]
252
Ozark
Highlands
301
[86 - 409]
5<)3
1221 - M 51
279
[5 - 568°]
479
South Central
Plains
90
[48 - 375]
22S
| I5i) - 5S21
134
[6-1150]
296
aPredicted values were overestimated, therefore XCD05 calculated from 75th centile of ecoregional least
disturbed stations is more accurate than predicted.
hObservation at only one station in Mississippi Valley Loess Plains.
cAlthough identified in data set as least disturbed, these maxima may represent an anthropogenically
altered background.
4 General Conclusions
These analyses demonstrate that many streams in Arkansas have SC at nearly natural
background levels and thus are a national treasure. The ionic compositions of these natural
waters are primarily due to rock dominance and are thus well suited for estimating their SC
levels from a random forest model that relies heavily on geological parameters (Olson and
Cormier, 2019). However, the model does not perform well in the Boston Mountains and
Arkansas Valley ecoregions and estimates based on observational data are recommended. These
stream segment and ecoregional estimates of background SC may be useful for estimating site
specific benchmarks or criteria that take into account not only the water chemistry but also how
freshwater organisms are expected to respond to changes in SC (Cormier et al., 2018a).
33
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
A weight of evidence was used to assess the choice of scale, data sets that would be used for the
background estimate, and the statistic used to characterize background (Table 6). Considerations
were weighted based on relevance and reliability with justifications for each score.
A key objective is to provide information to inform assessments of third-party site-specific
aquatic life criteria. Therefore, the stream segment or station scale is the most relevant
background scale. However, depending on data sets, larger scales may provide more reliable
estimates when there is a paucity of localized data or model predictions are weak within a
particular area. Consequently, a flow chart was developed that recommends scale based on
information about the location (Figure 8).
Similarly, the relevance and reliability of the source of SC values is influenced by location.
In most of the state, SC estimated from an empirical model was reliable and offered the
advantage of being scaled to the stream segment. However, based on a comparison of SC of
high-quality stations identified by ADEQ and SC estimated from an empirical model, modeled
least disturbed SC is less reliable in the Boston Mountains and Arkansas Valley (Figure 7).
Therefore, observed SC is recommended in combination with a comparison with other possible
background estimates in these two ecoregions (Figure 8).
In minimally affected fresh water, the lowest observed SC is the background for a defined
temporal period, typically 1-year, but seasonal backgrounds can also be characterized and can be
important in arid regions where deep groundwater has a greater seasonal influence (Clark and
Davidson, 2009; Olson and Cormier, 2019). In general, where observed SC is less than
predicted, the observed SC is a more relevant and reliable estimate of background conditions.
Where observed SC is greater than predicted SC, the observed SC is likely affected by
anthropogenic sources. The unknown true minimally affected background SC may be less than
the predicted SC, but the predicted SC is the more protective estimate than an altered
background.
The 10th, 25th, 75th, and central tendencies were evaluated as possible choices for estimating
background: (USEPA, 2011; Herlihy and Sifneos, 2008; Stoddard et al., 2006; Cormier et al.,
2018c). Depending on the data set, any of these statistics may be appropriate to estimate
background in Arkansas except the 10th centile. Low centile sites are more likely to characterize
background but may be too conservative because there are many sites with low SC across the
state. Also, the 10th centile is less reliable than a central tendency or larger centile, especially
with the small number available of least disturbed stations.
In summary, for any station where nearby minimally affected stations are available for
comparison or where the observed SC is less than predicted background SC, the observed
background is most relevant and likely more accurate. Where it is necessary to estimate
background SC in the absence of verified minimally affected SC background, the selection
process differs by ecoregion. In the Mississippi Alluvial Plain, Mississippi Valley Loess Plains,
Ouachita Mountains, Ozark Highlands, and South Central Plains, the mean predicted
background SC is recommended because the model is reliable in these ecoregions and there is
34
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
need for local predictions. In the Arkansas Valley and Boston Mountains, the 75th centile of
Arkansas ecoregional least disturbed stations is recommended because the predictive model
overestimates background and because the ecoregion is the smallest reliable scale. However,
because there are many low SC observations in these two ecoregions, nearby stations may
inform background estimates at a finer spatial resolution than the ecoregion scale.
The development of a robust data set has enabled these analyses and may provide a valuable
resource for future research. For example, the data set may enable characterization of ionic
signatures for commonly encountered sources that may be helpful for stressor and source
identification. As new data are added to the database, the analyses can be revisited. One
opportunity for improving confidence in the findings of this report and the database is the
development of an explicit process for selection of high-quality stations. Currently,
documentation relies on technology and guidance from more than 30 years ago (ADPCE, 1987a,
b). There is also a research opportunity to perform similar analyses using data augmented with
data from outside Arkansas but from within an ecoregion. Also, these data may be analyzed in
time series to monitor salinization of fresh water in the state, a condition that that has become an
international problem (Canedo-Argiielles et al., 2016; Kaushal et al., 2018; USEPA, 2011a).
5 Quality Assurance and Supplementary Data
Data, metadata, R-codes, and quality assurance procedures are contained in Wang, Y-C.,
Wharton, C., Cormier, S. M. 2022. Data sets: Background Specific Conductivity and Associated
5% Extirpation Estimates in Arkansas. United States Environmental Protection Agency, Office
of Research or can be obtained from the authors at cormier.susan@epa.gov.
6 Acknowledgements
Susan Cormier designed and led the project and analyses and wrote the report. The maps in this
document were prepared by Christopher Wharton, TetraTech, Inc. The statistical analyses,
figures and tables were prepared by Yu-Chen Wang, U. of Michigan. The document was
reviewed by USEPA scientists Drs. Mark Bagley, Michael Elovitz, Marc Weber, and Jonathan
Jimenez, and independently reviewed by two non-EPA experts by a contract managed letter
review. Josh Criss edited and Molly Windsor formatted the report. The views expressed in this
report are those of the authors and do not necessarily represent the views or policies of the U.S.
Environmental Protection Agency.
7 Citations
ADPCE (Arkansas Department of Pollution Control and Ecology). 1987a. "Physical, chemical, and
biological characteristics of least disturbed streams in Arkansas' ecoregions." Volume I: Data
compilation. 685 pp.
35
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
ADPCE (Arkansas Department of Pollution Control and Ecology). 1987b. "Physical, chemical, and
biological characteristics of least disturbed streams in Arkansas' ecoregions." Volume II: Data analysis.
148 pp.
AGS (Arkansas Geological Survey) 2022. "General Geology."
https://www.geologv.arkansas.gov/geologv/general-geologv.html.
Campbell, L.M., Gray, N.J., Hazen, E.L. and Shackeroff, J.M. 2009. "Beyond baselines: rethinking
priorities for ocean conservation." Ecology and Society, 14(1).
Canedo-Argiielles, M., Hawkins, C. P., Kefford, B. J., Schafer, R.B., Dyack, B. J., Brucet, S.,
Buchwalter, D., Dunlop, J., Fror, O., Lazorchak, J., Coring, E., Fernandez, H. R., Goodfellow, W.,
Achem, A. L. G., Hatfield-Dodds, S., Karimov, B. K., Mensah, P., Olson, J. R., Piscart, C., Prat, N.,
Ponsa, S., Schulz, C.-J., and Timpano, A. J. 2016. "Saving freshwater from salts." Science, 351(6276):
914-916.
Clark, M.L., and Davidson, S.L. 2009. "Specific conductance and dissolved solids characteristics for the
Green River and Muddy Creek, Wyoming, water years 1999-2008." U.S. Geological Survey Scientific
Investigations Report, 5168.
Cormier, S.M., and Suter II, G.W. 2013. "A method for deriving water-quality benchmarks using field
data "Environmental toxicology and chemistry, 32(2): 255-262.
Cormier, S.M., Suter, G.W., Fernandez, M.B., and Zheng, L. 2020. "Adequacy of sample size for
estimating a value from field observational data." Ecotoxicology and Environmental Safety,
203:110992.
Cormier, S.M., Suter, G.W., and Zheng, L. 2013. "Derivation of a benchmark for freshwater ionic
strength." Environmental Toxicology and Chemistry, 32(2): 263-271.
Cormier, S., Wharton, C., and Olson, J. 2021. USEPA Freshwater Explorer. V: 0.1. U.S. Environmental
Protection Agency. https://arcg.is/KHb9S
Cormier, S.M., Zheng, L., and Flaherty, C.M. 2018a. "A field-based model of the relationship between
extirpation of salt-intolerant benthic invertebrates and background conductivity." Science of the Total
Environment, 633: 1629-1636.
Cormier, S.M., Zheng, L., Hill, R.A., Novak, R.M., and Flaherty, C.M. 2018b. "A flowchart for
developing water quality criteria from two field-based methods." Science of The Total Environment,
633: 1647-1656.
Cormier, S.M., Zheng, L., Suter II, G.W., and Flaherty, C.M. 2018c. "Assessing background levels of
specific conductivity using weight of evidence." Science of The Total Environment, 628: 1637-1649.
Erickson, R.J., Mount, D.R., Highland, T.L., Hockett, J.R., Hoff, D.J., Jenson, C.T., Norberg-King, T.J.,
and Peterson, K.N. 2017. "The acute toxicity of major ion salts to Ceriodaphnia dubia. II. Empirical
relationships in binary salt mixtures." Environ. Toxicol. Chem, 36(6): 1525-1537.
Erickson, R.J., Mount, D.R., Highland, T.L., Hockett, J.R., Hoff, D.J., Jenson, C.T., Norberg-King, T.J.
and Forsman, B. 2022s. "Acute Toxicity of Major Geochemical Ions to Fathead Minnows (Pimephales
promelas). Part A: Observed Relationships for Individual Salts and Salt Mixtures." Environmental
Toxicology and Chemistry, 41(9): 2078-2094. https://doi.org/1.0.1002/etc.5390
Erickson, R.J., Mount, D.R., Highland, T.L., Hockett, J.R., Hoff, D.J., Jenson, C.T., Norberg-King, T.J.,
and Forsman BB. 2022b. "Acute Toxicity of Major Geochemical Ions to Fathead Minnows
(Pimephales promelas). Part B: Modeling Ion Toxicity." Environ Toxicol Chem, 41(9): 2095-2106.
https: //do i .o rg/10.1002/etc .5389
Farrar, D., Alexander, L.C., Yuan, L.L., and Gerritsen, J. 2015. "Regional observational studies
addressing confounding." Ecological Causal Assessment, S.B. Norton, S.M. Cormier, G.W. Suter II
(Eds.). CRC Press, Taylor and Francis Group, Boca Raton, FL. 203-212.
36
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Gerritsen, J., L.L. Yuan, P. Shaw-Allen, and S.M. Cormier. 2015. "Regional observational studies,
deriving evidence." Ecological Causal Assessment, S.B. Norton, S.M. Cormier and G.W. Suter II
(Eds.). CRC Press, Taylor and Francis Group Boca Raton, FL. 203-212.
Gillon, S., Booth, E.G., and Rissman, A.R. 2016. "Shifting drivers and static baselines in environmental
governance: challenges for improving and proving water quality outcomes." Regional Environmental
Change, 16(3): 759-775.
Green, J. J. 2014. "Re-evaluating Least disturbed Reference Streams in Arkansas' Ecoregions." Arkansas
American Fisheries Society (AFS), Rogers, AR.
https://www.adeq.state.ar.us/water/planning/pdfs/presentations/20140226-re-evaluating-least-
disturbed-reference-streams-in-arkansas-ecoregions.pdf
Griffith, MB. 2014. "Natural variation and recent reference for specific SC and major ions in wadeable
streams of the coterminous U.S." Freshw Sci, 33(1): 1-17.
Haley B.R., and Arkansas Geological Commission staff. 1993. Geologic Map of Arkansas.
https://www.geologv.arkansas.gov/maps-and-data/geologic maps/geologic-map-of-arkansas-.1.993-
revised-from - .1.976-edition .htm 1
Hem, J. 1985. "Study and Interpretation of the Chemical Characteristics of Natural Waters." USGS
Water Supply Paper 2254. Department of the Interior, U.S. Geological Survey, Washington, DC.
https://pubs.usgs.gov/wsp/wsp2254/pdf/wsp2254a.pdf.
Herlihy, A.T., and Sifneos, J.C. 2008. "Developing nutrient criteria and classification schemes for
wadeable streams in the conterminous US." Journal of the North American Benthological Society,
27(4): 932-948.
Hills, K.A., Hyne, R.V., and Kefford, B.J. 2022. "Bicarbonate alone does not totally explain the toxicity
from major ions of coal bed derived waters to freshwater invertebrates." Ecotoxicology, 1-9.
Humphrey, C.L., and Chandler, L. 2018. "Use of field-effects information to derive a surface water
guideline value for magnesium in Magela Creek, NT Australia." Supervising Scientist Report 212.
Supervising Scientist, Darwin, NT Australia.
Kaushal, S. S., Likens, G. E., Pace, M. L., Utz, R. M., Haq, S., Gorman, J., and Grese, M. 2018.
"Freshwater salinization syndrome on a continental scale." Proc. Natl. Acad. Sci., 115(14): E574-E583.
Kresse, T.M., and Hay, P.D. 2009. "Geochemistry, comparative analysis, and physical and chemical
characteristics of the thermal waters east of Hot Springs National Park, Arkansas, 2006-09." U.S.
Geological Survey Scientific Investigations Report 2009-5263, 48 p. Revised February 2011.
https://pubs.usgs.gov/sir/2009/5263/downloads/SIR2009-5263.pdf
MPCA (Minnesota Pollution Control Agency). 2020. "Implementing the Aquatic Life Narrative
Standard." Class 3 & 4 Water Quality Standards Revision: Technical Support Document, December
2020. . https://www.pca.state.mn.us/sites/default/files/wq-wwprml-36.pdf
Mooney, T.J., McCullough, C.D., Jansen, A., Chandler, L., Douglas, M., Harford, A.J., van
Dam, R. and Humphrey, C. 2020. "Elevated magnesium concentrations altered freshwater
assemblage structures in a mesocosm experiment." Environmental Toxicology and Chemistry,
39(10): 1973-1987.
Mount, D.R., Erickson, R.J., Highland, T.L., Hockett, J.R., Hoff, D.J., Jenson, C.T., Norberg-King, T.J.,
Peterson, K.N., Polaske, Z.M.,and Wisniewskiz, S. 2016. "The acute toxicity of major ion salts to
Ceriodaphnia dubia: I. Influence of background water chemistry." Environ. Toxicol. Chem, 35(12):
3039-3057.
Olson, J.R., and Cormier, S.M. 2019. "Modeling spatial and temporal variation in natural background
specific conductivity." Environmental science & technology, 53(8): 4316-4325.
Omernik, J.M. 1987. "Ecoregions of the conterminous United States." Ann Assoc Am Geograph, 77: 118-
125.
37
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Omernik, J.M. 1995. "Ecoregions: a spatial framework for environmental management." W. S. Davis and
T. P. Simon (Eds.). Biological assessment and criteria: tools for water resource planning and decision
making. Lewis Publishers, Boca Raton, Florida. 49-62.
Pauly, D., 1995. "Anecdotes and the shifting baseline syndrome of fisheries." Trends in ecology &
evolution, 10(10): 430.
USEPA. 2011. "A Field-based Aquatic Life Benchmark for Conductivity in Central Appalachian
Streams." National Center for Environmental Assessment, Office of Research and Development.
EPA/600/R-10/023F. http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=233809
USEPA. 2013, "Level III ecoregions of the continental United States" National Health and
Environmental Effects Research Laboratory, https://www.epa.gov/eco-research/level-iii-and-iv-
ecoregions-continental-united-states
USEPA. 2016. "Final Report: EPA Technical Support: Evaluation of Several Approaches to Develop
Mineral Criteria in Arkansas." 85.
USEPA. 2017. "Weight of Evidence in Ecological Assessment." EPA/100/ R16/001.
https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P100SFXR.TXT
38
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
8 Appendices
Appendix A-l. Ecoregion Level 3 Descriptions
The South Central Plains region (Ecoregion 35) is at the western edge of the southern coniferous
forest belt, much of which is now used for pine plantations. Two-thirds of Ecoregion 35 is forest
and woodland and a sixth is cropland. The dominant economic activities are timber or oil and
gas related. The surface geology of this area is characterized by unconsolidated deposits of sand,
gravel, silt, and clay from the ocean bottom, beaches, and estuaries that have eroded (USEPA,
2013).
The Ouachita Mountains region (Ecoregion 36) is defined by east-west trending ridges of tilted
strata of eroding sedimentary rock formations. Sandstone and shale were formed from deep
marine sediments. Natural vegetation is oak-hickory-pine forest but most of the region is now
dominated by pine forest, and the major land use activity is commercial logging (USEPA, 2013).
The Arkansas Valley (Ecoregion 37) is primarily forested valleys and ridges. Located north of
the Ouachita Mountains, one-fourth of the region is grazed and one-tenth is in cropland. The
area of the Arkansas River Valley warped downward into a trough that repeatedly filled and
eroded as the Ouachita Mountains formed by folding upwards. Swamps of river deltas
accumulated clay, and plant remains were buried under later sediments and transformed into coal
and natural gas. Additional alluvia became layered above these strata (USEPA, 2013).
The Boston Mountain (Ecoregion 38) is defined by a deeply dissected sandstone, shale, and
limestone plateau, covered originally by oak-hickory forests, which continue to dominate the
sparsely populated ecoregion. The principal land use is recreation (USEPA, 2013).
The Ozark Highlands (Ecoregion 39) has irregular physiography and soils derived from cherty
carbonate rocks. The dominant bedrock in the interior region is dolomite and sandstone while
the western outer region bedrock is dominated by limestone. Karst features are common.
Ecoregion 39 is forested by oak and mixed oak-pine stands. Less than one-fourth of the interior
of the ecoregion has been cleared for agricultural uses. The outer half of the periphery is pasture
and cropland (USEPA, 2013).
The Mississippi Alluvial Plain (Ecoregion 73) is a broad, flat alluvial plain of sand, silt, and clay
with fine textured soils though some areas have coarser, better drained soils. Bottomland
deciduous forest covered the region before most of it was cleared for agricultural purposes in the
north and central sections of the ecoregion (USEPA, 2013).
The Mississippi Valley Loess Plains (Ecoregion 74) are characterized by thick loess deposits
with deep steep, silty and erosive soils (USEPA 2013).
39
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Appendix A.2. Data set descriptive statistics
Table A.2.1. Summary station median statistics for least disturbed stations.
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Arkansas
Valley
Alkalinity, total
mg/l
CaCC>3
0.050
3.00
7.20
11.80
18.30
117.18
14.70
10.67
12.41
1017
18
Arkansas
Valley
Bicarbonate
mg/l
0.063
3.77
8.92
14.81
22.79
147.17
18.40
13.30
15.69
982
18
Arkansas
Valley
Chloride
mg/l
0.190
1.49
2.19
3.33
4.54
112.00
3.76
3.13
4.18
1473
18
Arkansas
Valley
Hardness, Ca, Mg
mg/l
4.000
9.00
11.00
13.30
19.00
95.00
16.52
14.72
9.46
751
18
Arkansas
Valley
Sulfate
mg/l
0.020
2.14
2.72
3.63
5.81
71.60
4.89
3.93
4.77
1471
18
Arkansas
Valley
Calcium
mg/l
1.200
1.90
2.27
2.81
3.72
26.80
3.33
2.99
2.09
755
18
Arkansas
Valley
Total Recoverable
Calcium
mg/l
1.330
2.00
2.33
2.93
3.88
28.30
3.41
3.10
2.01
348
12
Arkansas
Valley
Sodium
mg/l
0.020
1.30
1.80
2.70
4.30
167.00
3.63
2.71
6.42
755
18
Arkansas
Valley
Potassium
mg/l
0.010
0.63
0.95
1.45
2.20
11.00
1.77
1.43
1.20
755
18
Arkansas
Valley
Magnesium
mg/l
0.212
0.90
1.12
1.59
2.41
7.30
1.99
1.70
1.24
755
18
Arkansas
Valley
Specific
Conductivity
US/cm
18.500
29.92
34.70
45.30
63.10
423.00
54.13
48.67
31.39
593
13
Arkansas
Valley
Total dissolved
solids
mg/l
8.000
27.50
33.00
44.00
58.63
387.00
48.04
44.35
22.75
1472
18
Arkansas
Valley
Aluminum
mg/l
0.008
0.01
0.02
0.05
0.06
0.93
0.07
0.04
0.09
755
18
Arkansas
Valley
Total Recoverable
Aluminum
mg/l
0.020
0.05
0.08
0.19
0.43
2.14
0.30
0.19
0.31
348
12
Arkansas
Valley
Iron
mg/l
0.008
0.05
0.09
0.18
0.30
2.16
0.25
0.17
0.27
755
18
40
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Arkansas
Valley
Total Recoverable
Iron
mg/l
0.103
0.22
0.31
0.52
0.86
3.27
0.67
0.52
0.51
348
12
Arkansas
Valley
Ammonia-
nitrogen
mg/l
0.002
0.02
0.02
0.02
0.04
3.44
0.04
0.02
0.11
1466
18
Arkansas
Valley
Nitrite + Nitrate
as Nitrogen
mg/l
0.005
0.03
0.06
0.15
0.28
3.01
0.21
0.12
0.23
1473
18
Arkansas
Valley
Total Kjeldahl
nitrogen
mg/l
0.025
0.10
0.18
0.31
0.53
3.88
0.39
0.29
0.32
1238
18
Arkansas
Valley
Total Phosphorus
mg/l
0.002
0.01
0.02
0.04
0.07
1.17
0.06
0.04
0.07
1450
18
Arkansas
Valley
Orthophosphate
mg/l
0.003
0.01
0.01
0.01
0.02
0.45
0.02
0.01
0.03
1469
18
Arkansas
Valley
Total suspended
solids
mg/l
0.500
0.50
1.50
4.00
9.89
392.50
10.13
4.09
22.90
1472
18
Arkansas
Valley
PH
NA
4.000
6.22
6.50
6.75
7.06
8.82
6.78
6.76
0.51
1451
18
Arkansas
Valley
Temperature
°C
1.000
6.81
10.50
17.50
25.00
36.30
17.75
15.53
8.07
1491
18
Boston
Mountains
Alkalinity, total
mg/l
CaCC>3
0.299
8.30
13.80
24.25
35.58
164.00
27.93
21.25
20.89
854
35
Boston
Mountains
Bicarbonate
mg/l
0.376
10.43
17.43
30.78
44.63
204.47
35.23
26.79
26.29
829
35
Boston
Mountains
Chloride
mg/l
0.250
1.19
1.50
1.95
2.46
239.00
2.58
1.98
7.88
1310
36
Boston
Mountains
Hardness, Ca, Mg
mg/l
1.000
9.07
13.00
23.00
35.38
231.00
27.25
21.83
21.31
762
34
Boston
Mountains
Sulfate
mg/l
0.500
2.38
3.03
4.03
5.60
82.60
4.79
4.16
3.65
1313
36
Boston
Mountains
Calcium
mg/l
0.100
2.30
3.40
6.87
11.50
83.20
8.52
6.31
7.42
763
34
Boston
Mountains
Total Recoverable
Calcium
mg/l
1.110
2.19
3.13
6.47
11.50
42.00
7.92
5.99
6.11
293
28
Boston
Mountains
Sodium
mg/l
0.020
0.97
1.25
1.62
2.20
161.00
2.52
1.69
7.08
764
34
41
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Boston
Mountains
Potassium
mg/l
0.200
0.50
0.70
0.92
1.22
6.12
1.04
0.92
0.58
764
34
Boston
Mountains
Magnesium
mg/l
0.200
0.84
1.04
1.30
1.70
24.70
1.46
1.33
1.04
764
34
Boston
Mountains
Specific
Conductivity
US/cm
18.000
31.00
44.00
67.50
92.00
259.00
74.46
64.15
43.08
400
26
Boston
Mountains
Total dissolved
solids
mg/l
14.500
28.00
35.00
46.00
57.50
519.00
49.70
45.75
26.73
1099
35
Boston
Mountains
Aluminum
mg/l
0.008
0.01
0.01
0.02
0.06
1.28
0.05
0.03
0.07
764
34
Boston
Mountains
Total Recoverable
Aluminum
mg/l
0.010
0.04
0.06
0.11
0.26
3.57
0.25
0.13
0.41
293
28
Boston
Mountains
Iron
mg/l
0.006
0.01
0.02
0.04
0.06
0.94
0.05
0.03
0.06
764
34
Boston
Mountains
Total Recoverable
Iron
mg/l
0.021
0.06
0.11
0.16
0.24
2.84
0.23
0.17
0.30
293
28
Boston
Mountains
Ammonia-
nitrogen
mg/l
0.002
0.01
0.02
0.02
0.03
0.27
0.02
0.02
0.02
1312
36
Boston
Mountains
Nitrite + Nitrate
as Nitrogen
mg/l
0.005
0.01
0.02
0.06
0.15
2.08
0.13
0.06
0.19
1311
36
Boston
Mountains
Total Kjeldahl
nitrogen
mg/l
0.025
0.04
0.07
0.14
0.24
1.42
0.18
0.13
0.16
911
34
Boston
Mountains
Total Phosphorus
mg/l
0.005
0.01
0.01
0.02
0.04
4.81
0.04
0.02
0.17
1119
35
Boston
Mountains
Orthophosphate
mg/l
0.003
0.01
0.01
0.01
0.02
1.51
0.02
0.01
0.08
1313
36
Boston
Mountains
Total suspended
solids
mg/l
0.500
0.50
0.50
1.50
3.85
462.00
5.62
1.82
22.35
1097
35
Boston
Mountains
PH
NA
4.780
6.39
6.77
7.16
7.56
10.50
7.16
7.13
0.62
1240
36
Boston
Mountains
Temperature
°C
0.800
7.00
10.18
16.05
23.40
33.80
16.74
14.73
7.57
1279
36
Mississippi
Alluvial Plain
Alkalinity, total
mg/l
CaCC>3
2.500
36.15
55.90
99.00
153.00
263.00
105.10
87.78
55.12
680
7
42
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Mississippi
Alluvial Plain
Bicarbonate
mg/l
3.384
45.03
69.71
122.70
190.70
329.95
130.36
108.88
68.53
674
7
Mississippi
Alluvial Plain
Chloride
mg/l
1.010
2.44
3.07
6.02
11.46
124.00
11.58
6.76
16.06
1186
7
Mississippi
Alluvial Plain
Hardness, Ca, Mg
mg/l
4.000
40.12
62.88
120.00
163.00
772.00
119.27
100.50
66.67
548
7
Mississippi
Alluvial Plain
Sulfate
mg/l
0.020
3.00
4.04
6.12
11.23
113.00
8.68
6.64
7.51
1193
7
Mississippi
Alluvial Plain
Calcium
mg/l
1.000
10.38
16.80
28.70
36.40
284.00
28.75
24.48
17.83
549
7
Mississippi
Alluvial Plain
Total Recoverable
Calcium
mg/l
2.310
11.12
17.00
26.30
36.80
258.00
28.43
23.85
21.55
193
5
Mississippi
Alluvial Plain
Sodium
mg/l
0.020
1.62
2.26
4.50
8.95
297.00
8.41
4.53
20.70
549
7
Mississippi
Alluvial Plain
Potassium
mg/l
0.230
0.90
1.50
3.10
4.65
13.10
3.43
2.65
2.31
549
7
Mississippi
Alluvial Plain
Magnesium
mg/l
0.434
3.37
5.38
10.80
17.20
31.60
11.58
9.38
6.74
549
7
Mississippi
Alluvial Plain
Specific
Conductivity
US/cm
57.400
110.40
152.00
239.00
329.00
891.00
259.08
224.37
146.05
253
5
Mississippi
Alluvial Plain
Total dissolved
solids
mg/l
48.000
97.00
129.00
159.25
188.38
717.00
167.94
156.55
67.66
1194
7
Mississippi
Alluvial Plain
Aluminum
mg/l
0.008
0.01
0.01
0.05
0.15
2.14
0.11
0.04
0.20
549
7
Mississippi
Alluvial Plain
Total Recoverable
Aluminum
mg/l
0.005
0.05
0.10
0.27
0.68
5.38
0.54
0.25
0.73
193
5
Mississippi
Alluvial Plain
Iron
mg/l
0.001
0.01
0.02
0.05
0.14
2.37
0.13
0.06
0.23
548
7
Mississippi
Alluvial Plain
Total Recoverable
Iron
mg/l
0.039
0.15
0.24
0.57
1.01
6.76
0.81
0.52
0.86
193
5
Mississippi
Alluvial Plain
Ammonia-
nitrogen
mg/l
0.003
0.01
0.02
0.02
0.04
0.76
0.04
0.02
0.05
1183
7
Mississippi
Alluvial Plain
Nitrite + Nitrate
as Nitrogen
mg/l
0.005
0.01
0.03
0.09
0.22
1.57
0.15
0.07
0.17
1188
7
43
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Mississippi
Alluvial Plain
Total Kjeldahl
nitrogen
mg/l
0.025
0.14
0.40
0.63
0.83
2.51
0.63
0.50
0.35
857
7
Mississippi
Alluvial Plain
Total Phosphorus
mg/l
0.006
0.03
0.07
0.14
0.21
1.05
0.15
0.11
0.11
1167
7
Mississippi
Alluvial Plain
Orthophosphate
mg/l
0.003
0.01
0.02
0.07
0.10
0.60
0.08
0.05
0.06
1190
7
Mississippi
Alluvial Plain
Total suspended
solids
mg/l
0.500
2.00
4.50
10.00
24.50
492.00
22.69
10.46
37.43
1185
7
Mississippi
Alluvial Plain
PH
NA
3.420
6.73
7.09
7.49
7.91
10.40
7.46
7.43
0.62
1169
7
Mississippi
Alluvial Plain
Temperature
°C
0.400
7.30
11.00
18.00
24.30
34.00
17.69
15.52
7.65
1170
7
Ouachita
Mountains
Alkalinity, total
mg/l
CaC03
0.050
5.59
10.30
20.40
42.28
115.00
27.78
19.07
22.04
2178
39
Ouachita
Mountains
Bicarbonate
mg/l
0.026
6.94
12.81
24.87
52.25
144.29
34.36
23.55
27.45
2104
39
Ouachita
Mountains
Chloride
mg/l
0.250
1.37
1.58
1.94
2.43
37.40
2.20
2.01
1.60
3048
41
Ouachita
Mountains
Hardness, Ca, Mg
mg/l
0.500
8.81
13.00
23.00
47.08
406.00
31.75
23.78
25.08
1686
40
Ouachita
Mountains
Sulfate
mg/l
0.350
2.74
3.41
4.52
6.11
458.00
5.61
4.68
9.10
3041
41
Ouachita
Mountains
Calcium
mg/l
0.010
1.83
3.09
6.50
14.30
115.00
9.34
6.32
8.04
1693
40
Ouachita
Mountains
Total Recoverable
Calcium
mg/l
0.279
1.83
2.86
6.44
14.50
40.90
9.13
6.13
7.48
1036
38
Ouachita
Mountains
Sodium
mg/l
0.010
1.10
1.48
1.87
2.40
160.00
2.26
1.79
4.34
1686
40
Ouachita
Mountains
Potassium
mg/l
0.010
0.45
0.57
0.75
1.07
5.76
0.87
0.77
0.49
1693
40
Ouachita
Mountains
Magnesium
mg/l
0.010
0.89
1.15
1.60
2.50
28.90
2.03
1.72
1.38
1693
40
Ouachita
Mountains
Specific
Conductivity
US/cm
2.960
29.00
38.00
58.00
105.00
472.00
73.43
60.92
46.95
1203
32
44
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Ouachita
Mountains
Total dissolved
solids
mg/l
12.000
28.00
34.50
45.00
63.00
712.00
51.21
46.68
25.89
3037
41
Ouachita
Mountains
Aluminum
mg/l
0.008
0.01
0.01
0.02
0.06
0.59
0.04
0.03
0.05
1693
40
Ouachita
Mountains
Total Recoverable
Aluminum
mg/l
0.005
0.01
0.02
0.04
0.11
2.11
0.09
0.05
0.16
1036
38
Ouachita
Mountains
Iron
mg/l
0.007
0.01
0.03
0.04
0.08
0.59
0.06
0.04
0.06
1693
40
Ouachita
Mountains
Total Recoverable
Iron
mg/l
0.010
0.03
0.03
0.08
0.16
2.90
0.13
0.07
0.20
1036
38
Ouachita
Mountains
Ammonia-
nitrogen
mg/l
0.003
0.02
0.02
0.02
0.03
0.64
0.02
0.02
0.03
3031
41
Ouachita
Mountains
Nitrite + Nitrate
as Nitrogen
mg/l
0.003
0.01
0.03
0.05
0.14
13.30
0.12
0.05
0.36
3031
41
Ouachita
Mountains
Total Kjeldahl
nitrogen
mg/l
0.015
0.05
0.09
0.16
0.26
4.68
0.20
0.15
0.20
2468
40
Ouachita
Mountains
Total Phosphorus
mg/l
0.005
0.01
0.01
0.02
0.04
2.03
0.04
0.02
0.08
2986
41
Ouachita
Mountains
Orthophosphate
mg/l
0.003
0.01
0.01
0.01
0.02
1.50
0.02
0.01
0.06
3041
41
Ouachita
Mountains
Total suspended
solids
mg/l
0.500
0.50
0.50
1.50
3.25
868.00
4.36
1.69
21.01
3036
41
Ouachita
Mountains
PH
NA
4.010
6.41
6.79
7.11
7.37
13.20
7.05
7.03
0.51
2957
42
Ouachita
Mountains
Temperature
°C
1.000
7.80
11.70
17.90
24.00
34.00
17.69
15.86
7.23
2997
42
Ozark
Highlands
Alkalinity, total
mg/l
CaC03
1.800
69.48
100.51
133.00
199.00
428.00
144.39
130.47
58.98
3725
65
Ozark
Highlands
Bicarbonate
mg/l
2.245
86.82
125.20
166.19
247.79
527.42
179.70
162.48
73.02
3630
64
Ozark
Highlands
Chloride
mg/l
0.250
2.06
2.65
3.39
5.30
62.60
5.65
4.10
6.15
6446
71
Ozark
Highlands
Hardness, Ca, Mg
mg/l
1.000
79.00
113.00
145.00
211.00
2160.00
162.43
144.38
100.75
2403
70
45
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Ozark
Highlands
Sulfate
mg/l
0.020
3.21
3.98
5.35
7.84
75.00
7.08
5.87
5.46
6454
71
Ozark
Highlands
Calcium
mg/l
0.025
25.90
35.00
42.30
47.90
452.00
43.42
39.60
29.94
2430
70
Ozark
Highlands
Total Recoverable
Calcium
mg/l
5.120
25.68
34.80
42.75
48.80
468.00
48.10
41.00
48.74
860
45
Ozark
Highlands
Sodium
mg/l
0.020
1.18
1.60
2.10
5.26
210.00
5.28
2.68
9.06
2425
70
Ozark
Highlands
Potassium
mg/l
0.045
0.80
1.19
1.60
2.50
25.10
2.11
1.70
1.72
2430
70
Ozark
Highlands
Magnesium
mg/l
0.065
1.80
2.19
7.38
24.90
253.00
13.16
7.14
13.25
2430
70
Ozark
Highlands
Specific
Conductivity
US/cm
4.570
148.00
200.75
279.00
381.00
568.00
285.37
262.58
105.22
1847
55
Ozark
Highlands
Total dissolved
solids
mg/l
6.000
108.00
145.00
191.00
224.00
477.00
182.87
174.34
51.02
5027
68
Ozark
Highlands
Aluminum
mg/l
0.008
0.01
0.01
0.02
0.13
1.03
0.08
0.03
0.10
2430
70
Ozark
Highlands
Total Recoverable
Aluminum
mg/l
0.005
0.02
0.05
0.07
0.13
8.40
0.18
0.08
0.53
860
45
Ozark
Highlands
Iron
mg/l
0.001
0.01
0.01
0.02
0.04
0.88
0.03
0.02
0.04
2430
70
Ozark
Highlands
Total Recoverable
Iron
mg/l
0.010
0.03
0.05
0.08
0.15
10.20
0.18
0.09
0.56
860
45
Ozark
Highlands
Ammonia-
nitrogen
mg/l
0.000
0.01
0.02
0.02
0.03
1.50
0.02
0.02
0.03
6449
71
Ozark
Highlands
Nitrite + Nitrate
as Nitrogen
mg/l
0.001
0.03
0.12
0.41
1.04
24.80
0.78
0.32
1.02
6401
71
Ozark
Highlands
Total Kjeldahl
nitrogen
mg/l
0.005
0.05
0.12
0.19
0.30
44.00
0.26
0.18
0.76
3757
57
Ozark
Highlands
Total Phosphorus
mg/l
0.005
0.01
0.02
0.04
0.08
4.86
0.09
0.05
0.20
5148
68
Ozark
Highlands
Orthophosphate
mg/l
0.002
0.01
0.01
0.02
0.04
4.39
0.06
0.02
0.18
6444
71
46
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Ozark
Highlands
Total suspended
solids
mg/l
0.500
1.00
2.00
4.00
8.00
1130.00
11.79
3.97
46.14
5024
68
Ozark
Highlands
PH
NA
4.580
7.26
7.62
7.91
8.15
10.23
7.85
7.84
0.45
6188
78
Ozark
Highlands
Temperature
°C
0.460
7.50
11.00
15.70
22.00
34.00
16.41
14.76
6.88
6365
78
South Central
Plains
Alkalinity, total
mg/l
CaC03
0.050
3.00
8.29
20.40
40.95
342.00
28.40
16.65
30.82
1802
28
South Central
Plains
Bicarbonate
mg/l
0.063
3.77
10.43
25.50
51.76
428.52
35.69
20.90
38.93
1753
28
South Central
Plains
Chloride
mg/l
0.190
2.29
3.07
4.13
5.60
165.00
5.35
4.33
6.29
2914
28
South Central
Plains
Hardness, Ca, Mg
mg/l
0.025
10.00
16.00
29.00
44.00
1710.00
36.40
26.20
65.93
1468
28
South Central
Plains
Sulfate
mg/l
0.250
3.43
5.27
10.50
20.60
255.00
15.68
10.44
15.83
2921
28
South Central
Plains
Calcium
mg/l
0.010
2.48
4.17
7.50
11.95
680.00
10.38
6.90
25.40
1483
28
South Central
Plains
Total Recoverable
Calcium
mg/l
0.909
2.03
3.69
6.89
11.10
627.00
12.28
6.78
35.26
746
28
South Central
Plains
Sodium
mg/l
0.010
1.98
3.30
6.05
10.20
187.00
8.36
5.55
9.84
1482
28
South Central
Plains
Potassium
mg/l
0.010
0.80
1.15
1.56
2.20
16.50
1.82
1.52
1.24
1483
28
South Central
Plains
Magnesium
mg/l
0.010
0.80
1.30
2.31
3.40
33.00
2.52
2.01
1.94
1483
28
South Central
Plains
Specific
Conductivity
US/cm
5.940
37.90
56.80
92.00
134.00
1150.00
118.97
90.62
114.07
841
22
South Central
Plains
Total dissolved
solids
mg/l
12.000
56.00
69.00
83.00
99.75
642.00
90.25
83.90
42.94
2923
28
South Central
Plains
Aluminum
mg/l
0.008
0.02
0.04
0.06
0.14
1.09
0.11
0.07
0.12
1483
28
South Central
Plains
Total Recoverable
Aluminum
mg/l
0.020
0.11
0.19
0.37
0.66
4.39
0.50
0.35
0.48
746
28
47
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
South Central
Plains
Iron
mg/l
0.005
0.07
0.14
0.30
0.53
2.79
0.38
0.26
0.34
1483
28
South Central
Plains
Total Recoverable
Iron
mg/l
0.050
0.29
0.59
0.94
1.47
6.25
1.16
0.88
0.90
746
28
South Central
Plains
Ammonia-
nitrogen
mg/l
0.003
0.02
0.02
0.03
0.04
10.20
0.04
0.02
0.20
2911
28
South Central
Plains
Nitrite + Nitrate
as Nitrogen
mg/l
0.001
0.01
0.03
0.10
0.18
7.75
0.15
0.08
0.31
2916
28
South Central
Plains
Total Kjeldahl
nitrogen
mg/l
0.025
0.22
0.30
0.43
0.63
11.10
0.49
0.42
0.36
2307
28
South Central
Plains
Total Phosphorus
mg/l
0.005
0.02
0.04
0.06
0.09
3.87
0.08
0.06
0.16
2871
28
South Central
Plains
Orthophosphate
mg/l
0.003
0.01
0.01
0.02
0.03
2.94
0.04
0.02
0.13
2909
28
South Central
Plains
Total suspended
solids
mg/l
0.500
2.50
4.50
7.50
12.50
414.00
11.63
7.42
17.77
2922
28
South Central
Plains
PH
NA
4.300
6.00
6.41
6.83
7.17
9.39
6.77
6.74
0.62
2875
28
South Central
Plains
Temperature
°C
1.000
8.00
12.00
18.00
25.00
36.60
18.23
16.36
7.56
2879
28
48
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Table A.2.2. Summary station median statistics for all stations.
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Arkansas Valley
Alkalinity, total
mg/l
CaCC>3
0.050
3.00
10.60
24.50
75.90
418.00
43.09
24.27
41.48
4324
72
Arkansas Valley
Bicarbonate
mg/l
0.063
3.77
13.07
30.61
94.74
321.33
53.69
30.17
51.38
4177
72
Arkansas Valley
Chloride
mg/l
0.035
1.65
2.84
5.70
45.90
1890.00
31.01
9.83
56.24
6795
86
Arkansas Valley
Hardness, Ca, Mg
mg/l
0.500
10.00
13.00
24.00
92.35
1050.00
54.43
32.00
62.65
3307
85
Arkansas Valley
Sulfate
mg/l
0.020
2.51
3.42
10.30
38.11
338.00
23.91
11.47
29.30
6812
86
Arkansas Valley
Calcium
mg/l
0.066
2.00
2.70
5.10
23.05
406.00
14.10
7.15
19.98
3319
85
Arkansas Valley
Total Recoverable
Calcium
mg/l
1.040
2.04
2.72
4.94
21.60
413.00
14.55
7.08
26.18
1387
48
Arkansas Valley
Sodium
mg/l
0.020
1.41
2.43
6.29
36.40
679.30
24.80
8.44
39.77
3319
85
Arkansas Valley
Potassium
mg/l
0.010
0.70
1.20
2.28
3.80
46.00
3.28
2.15
4.53
3319
85
Arkansas Valley
Magnesium
mg/l
0.010
1.02
1.48
2.90
7.00
58.90
4.66
3.12
4.47
3319
85
Arkansas Valley
Specific Conductivity
US/cm
2.380
29.00
38.43
61.15
262.75
1300.00
184.10
95.02
228.93
2098
43
Arkansas Valley
Total dissolved solids
mg/l
7.000
30.50
41.00
77.00
248.00
5020.00
154.48
96.82
166.56
6836
86
Arkansas Valley
Aluminum
mg/l
0.008
0.01
0.02
0.05
0.07
1.21
0.08
0.04
0.10
3319
85
Arkansas Valley
Total Recoverable
Aluminum
mg/l
0.010
0.07
0.12
0.25
0.48
4.47
0.37
0.24
0.39
1387
48
Arkansas Valley
Iron
mg/l
0.001
0.02
0.05
0.11
0.21
7.65
0.17
0.09
0.25
3319
85
Arkansas Valley
Total Recoverable
Iron
mg/l
0.025
0.17
0.26
0.44
0.78
6.88
0.59
0.44
0.50
1387
48
Arkansas Valley
Ammonia-nitrogen
mg/l
0.001
0.02
0.02
0.03
0.06
19.00
0.21
0.03
1.15
6760
86
Arkansas Valley
Nitrite + Nitrate as
Nitrogen
mg/l
0.005
0.02
0.06
0.21
0.45
51.50
0.67
0.17
2.13
6786
86
Arkansas Valley
Total Kjeldahl
nitrogen
mg/l
0.025
0.11
0.24
0.46
0.65
25.80
0.72
0.39
1.55
5645
86
Arkansas Valley
Total Phosphorus
mg/l
0.001
0.02
0.03
0.07
0.12
25.76
0.26
0.07
1.07
6701
86
Arkansas Valley
Orthophosphate
mg/l
0.003
0.01
0.01
0.02
0.06
17.80
0.20
0.03
0.95
6778
86
Arkansas Valley
Total suspended
solids
mg/l
0.250
1.00
2.50
6.30
13.00
960.00
11.95
5.58
24.47
6845
86
49
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Arkansas Valley
PH
NA
4.000
6.35
6.68
7.05
7.50
9.91
7.08
7.05
0.63
6570
86
Arkansas Valley
Temperature
°C
0.200
7.00
10.90
17.80
25.00
36.30
17.87
15.60
8.16
6794
86
Boston Mountains
Alkalinity, total
mg/l
CaCC>3
0.299
10.10
19.80
35.00
56.00
178.00
43.64
31.86
33.36
1649
63
Boston Mountains
Bicarbonate
mg/l
0.376
12.63
24.99
43.80
70.21
219.82
54.69
39.90
41.77
1593
63
Boston Mountains
Chloride
mg/l
0.250
1.27
1.75
2.45
3.52
239.00
3.85
2.66
8.33
2723
90
Boston Mountains
Hardness, Ca, Mg
mg/l
1.000
10.30
17.00
35.80
57.00
622.00
45.05
32.65
39.87
1413
68
Boston Mountains
Sulfate
mg/l
0.020
2.62
3.66
5.87
10.90
110.48
10.22
6.75
12.08
2724
89
Boston Mountains
Calcium
mg/l
0.100
2.52
4.72
11.60
19.00
243.00
14.67
9.87
14.15
1419
68
Boston Mountains
Total Recoverable
Calcium
mg/l
1.110
2.39
4.08
11.00
17.30
234.00
13.06
8.88
14.27
477
37
Boston Mountains
Sodium
mg/l
0.020
1.03
1.42
2.04
3.56
161.00
4.00
2.36
8.28
1421
68
Boston Mountains
Potassium
mg/l
0.100
0.60
0.80
1.12
1.60
21.50
1.40
1.13
1.25
1421
68
Boston Mountains
Magnesium
mg/l
0.200
0.90
1.20
1.67
2.40
24.70
2.07
1.76
1.44
1421
68
Boston Mountains
Specific Conductivity
US/cm
18.000
34.20
52.80
91.00
144.00
437.00
112.52
89.06
79.47
723
39
Boston Mountains
Total dissolved solids
mg/l
13.000
32.00
43.38
62.00
92.00
564.00
77.92
65.47
52.79
2312
80
Boston Mountains
Aluminum
mg/l
0.008
0.01
0.01
0.02
0.06
1.28
0.05
0.03
0.08
1421
68
Boston Mountains
Total Recoverable
Aluminum
mg/l
0.010
0.04
0.06
0.13
0.27
14.70
0.29
0.13
0.79
477
37
Boston Mountains
Iron
mg/l
0.001
0.01
0.01
0.03
0.05
0.94
0.04
0.03
0.05
1421
68
Boston Mountains
Total Recoverable
Iron
mg/l
0.021
0.06
0.10
0.18
0.29
20.80
0.31
0.18
1.02
477
37
Boston Mountains
Ammonia-nitrogen
mg/l
0.002
0.01
0.02
0.02
0.03
0.39
0.02
0.02
0.03
2739
90
Boston Mountains
Nitrite + Nitrate as
Nitrogen
mg/l
0.005
0.02
0.04
0.12
0.39
4.86
0.29
0.11
0.45
2723
90
Boston Mountains
Total Kjeldahl
nitrogen
mg/l
0.025
0.05
0.09
0.18
0.32
2.94
0.25
0.17
0.24
1893
67
Boston Mountains
Total Phosphorus
mg/l
0.004
0.01
0.02
0.03
0.05
4.81
0.05
0.03
0.13
2331
78
Boston Mountains
Orthophosphate
mg/l
0.001
0.01
0.01
0.01
0.02
1.51
0.02
0.01
0.07
2741
90
Boston Mountains
Total suspended
solids
mg/l
0.500
0.50
1.00
2.50
7.50
744.00
9.68
2.78
34.46
2305
80
50
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Boston Mountains
PH
NA
4.780
6.57
6.98
7.37
7.69
11.30
7.32
7.30
0.59
2584
89
Boston Mountains
Temperature
°C
0.400
7.00
10.00
16.00
23.00
34.00
16.50
14.47
7.54
2647
89
Mississippi Alluvial
Plain
Alkalinity, total
mg/l
CaCC>3
0.203
16.20
36.10
79.30
121.00
456.00
82.92
60.75
54.22
4703
95
Mississippi Alluvial
Plain
Bicarbonate
mg/l
0.255
20.49
45.50
98.95
150.84
570.10
103.36
75.90
67.45
4572
95
Mississippi Alluvial
Plain
Chloride
mg/l
0.100
2.67
3.79
6.75
21.60
683.00
20.21
9.40
30.76
8077
102
Mississippi Alluvial
Plain
Hardness, Ca, Mg
mg/l
0.500
21.00
42.00
94.00
139.00
970.00
97.26
74.01
65.97
4270
102
Mississippi Alluvial
Plain
Sulfate
mg/l
0.020
3.71
5.23
7.85
15.40
221.00
13.89
9.22
15.47
8101
102
Mississippi Alluvial
Plain
Calcium
mg/l
0.010
5.40
11.00
23.60
34.60
374.00
24.46
18.60
17.70
4287
102
Mississippi Alluvial
Plain
Total Recoverable
Calcium
mg/l
1.490
5.49
11.40
22.75
34.30
505.00
25.51
18.84
28.05
1550
68
Mississippi Alluvial
Plain
Sodium
mg/l
0.010
1.93
3.08
6.29
16.88
481.00
15.23
7.11
24.26
4278
102
Mississippi Alluvial
Plain
Potassium
mg/l
0.010
1.16
1.80
3.00
4.38
61.50
3.38
2.74
2.47
4287
102
Mississippi Alluvial
Plain
Magnesium
mg/l
0.010
1.80
3.50
7.80
12.60
46.40
8.83
6.52
6.19
4287
102
Mississippi Alluvial
Plain
Specific Conductivity
US/cm
8.440
63.36
116.00
215.00
306.13
957.00
237.13
188.66
158.28
1740
36
Mississippi Alluvial
Plain
Total dissolved solids
mg/l
8.000
77.00
117.00
154.50
200.00
1287.50
171.24
149.45
91.52
8119
102
Mississippi Alluvial
Plain
Aluminum
mg/l
0.008
0.01
0.01
0.06
0.16
2.66
0.12
0.05
0.21
4286
102
Mississippi Alluvial
Plain
Total Recoverable
Aluminum
mg/l
0.005
0.09
0.20
0.43
0.92
18.60
0.85
0.42
1.46
1550
68
Mississippi Alluvial
Plain
Iron
mg/l
0.001
0.01
0.02
0.06
0.22
5.35
0.17
0.07
0.28
4285
102
Mississippi Alluvial
Plain
Total Recoverable
Iron
mg/l
0.025
0.19
0.36
0.69
1.30
21.10
1.12
0.68
1.66
1550
68
51
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Mississippi Alluvial
Plain
Ammonia-nitrogen
mg/l
0.001
0.02
0.02
0.03
0.07
8.67
0.06
0.03
0.15
8056
102
Mississippi Alluvial
Plain
Nitrite + Nitrate as
Nitrogen
mg/l
0.005
0.02
0.07
0.18
0.31
12.10
0.27
0.14
0.54
8075
102
Mississippi Alluvial
Plain
Total Kjeldahl
nitrogen
mg/l
0.022
0.24
0.41
0.63
0.89
9.76
0.70
0.58
0.48
6360
101
Mississippi Alluvial
Plain
Total Phosphorus
mg/l
0.006
0.04
0.07
0.13
0.21
5.38
0.18
0.12
0.26
7962
102
Mississippi Alluvial
Plain
Orthophosphate
mg/l
0.003
0.01
0.02
0.06
0.10
5.39
0.10
0.05
0.21
8077
102
Mississippi Alluvial
Plain
Total suspended
solids
mg/l
0.500
4.00
7.50
15.50
30.00
1170.00
26.75
14.87
47.04
8073
102
Mississippi Alluvial
Plain
PH
NA
2.620
6.55
6.93
7.40
7.80
10.40
7.35
7.33
0.63
7885
102
Mississippi Alluvial
Plain
Temperature
°C
0.400
7.50
11.10
18.00
25.00
35.00
18.02
15.89
7.83
7945
102
Mississippi Valley Loess
Plains
Alkalinity, total
mg/l
CaCC>3
3.000
19.57
30.03
64.70
102.00
275.00
70.03
54.78
44.51
168
1
Mississippi Valley Loess
Plains
Bicarbonate
mg/l
3.770
24.56
37.61
80.95
126.81
341.35
87.24
68.39
55.17
168
1
Mississippi Valley Loess
Plains
Chloride
mg/l
1.080
2.70
4.58
11.40
21.83
186.00
15.31
10.03
17.77
168
1
Mississippi Valley Loess
Plains
Hardness, Ca, Mg
mg/l
2.000
26.85
38.78
69.50
96.50
173.00
71.69
60.31
37.04
80
1
Mississippi Valley Loess
Plains
Sulfate
mg/l
2.600
4.90
7.08
12.40
17.83
66.70
13.47
11.23
8.52
168
1
Mississippi Valley Loess
Plains
Calcium
mg/l
0.352
6.14
9.37
17.60
24.30
46.80
17.50
14.60
9.18
81
1
Mississippi Valley Loess
Plains
Total Recoverable
Calcium
mg/l
4.140
6.99
9.30
17.20
22.20
53.50
17.24
14.87
9.51
51
1
Mississippi Valley Loess
Plains
Sodium
mg/l
1.050
3.42
4.72
14.40
22.80
77.70
16.18
11.07
13.83
81
1
Mississippi Valley Loess
Plains
Potassium
mg/l
0.694
2.24
2.71
3.56
4.22
38.00
4.92
3.77
6.01
81
1
52
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Mississippi Valley Loess
Plains
Magnesium
mg/l
0.182
2.89
3.53
6.69
8.72
20.70
6.77
5.73
3.54
81
1
Mississippi Valley Loess
Plains
Specific Conductivity
US/cm
45.400
70.88
89.00
164.00
318.00
558.00
203.87
165.58
125.91
73
1
Mississippi Valley Loess
Plains
Total dissolved solids
mg/l
65.500
91.70
109.00
151.00
210.50
515.00
162.51
150.64
66.66
168
1
Mississippi Valley Loess
Plains
Aluminum
mg/l
0.010
0.01
0.01
0.05
0.20
2.47
0.15
0.06
0.30
81
1
Mississippi Valley Loess
Plains
Total Recoverable
Aluminum
mg/l
0.045
0.11
0.28
0.71
1.27
6.91
1.01
0.55
1.31
51
1
Mississippi Valley Loess
Plains
Iron
mg/l
0.010
0.05
0.06
0.13
0.23
2.57
0.18
0.12
0.29
81
1
Mississippi Valley Loess
Plains
Total Recoverable
Iron
mg/l
0.093
0.36
0.52
0.79
1.30
10.50
1.20
0.82
1.61
51
1
Mississippi Valley Loess
Plains
Ammonia-nitrogen
mg/l
0.015
0.02
0.02
0.04
0.11
8.56
0.17
0.05
0.72
167
1
Mississippi Valley Loess
Plains
Nitrite + Nitrate as
Nitrogen
mg/l
0.005
0.02
0.03
0.14
0.31
5.06
0.26
0.11
0.45
168
1
Mississippi Valley Loess
Plains
Total Kjeldahl
nitrogen
mg/l
0.224
0.48
0.62
0.73
0.97
21.00
1.05
0.82
1.68
165
1
Mississippi Valley Loess
Plains
Total Phosphorus
mg/l
0.060
0.12
0.14
0.27
0.50
5.35
0.53
0.31
0.76
168
1
Mississippi Valley Loess
Plains
Orthophosphate
mg/l
0.010
0.03
0.06
0.13
0.29
5.35
0.38
0.15
0.70
168
1
Mississippi Valley Loess
Plains
Total suspended
solids
mg/l
1.000
3.50
7.38
14.75
29.85
903.00
48.06
16.21
120.06
168
1
Mississippi Valley Loess
Plains
PH
NA
6.400
6.91
7.21
7.49
7.81
9.33
7.50
7.48
0.48
168
1
Mississippi Valley Loess
Plains
Temperature
°C
0.200
4.81
9.65
16.15
24.70
33.40
16.94
13.57
8.74
168
1
Ouachita Mountains
Alkalinity, total
mg/l
CaC03
0.050
3.00
8.36
14.80
29.35
190.00
21.86
14.48
19.98
4925
102
Ouachita Mountains
Bicarbonate
mg/l
0.026
3.77
10.43
18.45
36.25
238.43
27.09
17.97
24.78
4791
102
Ouachita Mountains
Chloride
mg/l
0.100
1.47
1.77
2.38
3.39
129.00
3.89
2.72
6.23
6913
129
53
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Ouachita Mountains
Hardness, Ca, Mg
mg/l
0.234
8.00
11.10
19.00
39.00
1400.00
39.47
21.94
88.04
4147
113
Ouachita Mountains
Sulfate
mg/l
0.020
2.43
3.29
4.71
7.60
1380.00
17.51
5.72
80.43
6910
129
Ouachita Mountains
Calcium
mg/l
0.010
1.30
2.18
4.80
10.90
372.00
10.49
5.01
24.40
4156
113
Ouachita Mountains
Total Recoverable
Calcium
mg/l
0.020
1.18
2.22
4.97
12.40
331.00
10.76
5.17
23.72
2508
94
Ouachita Mountains
Sodium
mg/l
0.010
1.26
1.69
2.31
3.47
196.00
3.92
2.51
7.99
4148
113
Ouachita Mountains
Potassium
mg/l
0.010
0.50
0.67
0.96
1.47
71.30
1.49
1.03
2.89
4156
113
Ouachita Mountains
Magnesium
mg/l
0.010
0.94
1.27
1.74
2.68
90.00
3.18
1.99
6.80
4158
113
Ouachita Mountains
Specific Conductivity
US/cm
2.960
27.72
38.00
56.00
105.00
6370.00
99.73
64.80
195.66
2853
85
Ouachita Mountains
Total dissolved solids
mg/l
6.000
30.00
36.00
46.00
65.00
2100.00
69.61
51.71
119.32
6891
129
Ouachita Mountains
Aluminum
mg/l
0.008
0.01
0.02
0.02
0.06
29.90
0.23
0.03
1.66
4158
113
Ouachita Mountains
Total Recoverable
Aluminum
mg/l
0.005
0.02
0.02
0.08
0.22
30.70
0.35
0.09
1.54
2508
94
Ouachita Mountains
Iron
mg/l
0.007
0.02
0.03
0.06
0.14
6.19
0.12
0.06
0.22
4158
113
Ouachita Mountains
Total Recoverable
Iron
mg/l
0.010
0.03
0.06
0.15
0.32
7.57
0.27
0.14
0.42
2508
94
Ouachita Mountains
Ammonia-nitrogen
mg/l
0.001
0.02
0.02
0.02
0.03
3.52
0.04
0.02
0.11
6889
129
Ouachita Mountains
Nitrite + Nitrate as
Nitrogen
mg/l
0.003
0.01
0.03
0.07
0.19
43.40
0.31
0.08
1.48
6891
129
Ouachita Mountains
Total Kjeldahl
nitrogen
mg/l
0.015
0.05
0.11
0.20
0.35
9.05
0.29
0.19
0.35
5756
112
Ouachita Mountains
Total Phosphorus
mg/l
0.004
0.01
0.02
0.03
0.05
23.00
0.15
0.03
1.03
6786
129
Ouachita Mountains
Orthophosphate
mg/l
0.003
0.01
0.01
0.01
0.02
27.52
0.13
0.02
1.05
6907
129
Ouachita Mountains
Total suspended
solids
mg/l
0.500
0.50
1.00
2.00
4.00
868.00
4.76
2.00
17.47
6916
129
Ouachita Mountains
PH
NA
3.270
6.12
6.52
6.93
7.24
14.00
6.85
6.82
0.68
6750
132
Ouachita Mountains
Temperature
°C
1.000
7.70
11.40
17.70
24.00
35.00
17.58
15.71
7.34
6814
132
Ozark Highlands
Alkalinity, total
mg/l
CaC03
1.800
83.10
112.00
135.38
177.00
428.00
143.11
133.46
48.90
8236
172
Ozark Highlands
Bicarbonate
mg/l
1.614
103.40
139.03
169.36
220.00
527.42
178.12
166.15
60.49
7997
171
Ozark Highlands
Chloride
mg/l
0.015
2.34
3.10
4.74
9.07
900.00
9.87
5.79
19.39
13583
215
54
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Ozark Highlands
Hardness, Ca, Mg
mg/l
0.500
94.00
122.00
145.00
186.00
2160.00
158.48
144.36
96.66
5493
189
Ozark Highlands
Sulfate
mg/l
0.020
3.69
4.78
6.66
10.80
109.00
9.65
7.51
8.70
13573
215
Ozark Highlands
Calcium
mg/l
0.025
29.00
37.00
45.10
52.30
663.00
47.47
43.19
34.93
5570
190
Ozark Highlands
Total Recoverable
Calcium
mg/l
1.690
29.72
37.30
46.18
54.70
707.00
54.76
45.69
61.84
2082
104
Ozark Highlands
Sodium
mg/l
0.020
1.39
1.93
3.32
8.08
2515.00
9.21
4.06
37.15
5562
190
Ozark Highlands
Potassium
mg/l
0.045
0.90
1.29
1.82
3.08
47.40
2.69
1.99
2.81
5571
190
Ozark Highlands
Magnesium
mg/l
0.065
1.79
2.17
3.60
17.38
253.00
9.83
5.59
10.86
5571
190
Ozark Highlands
Specific Conductivity
US/cm
2.000
181.00
248.75
318.00
383.00
1540.00
315.29
294.38
107.75
4124
128
Ozark Highlands
Total dissolved solids
mg/l
6.000
126.00
156.00
190.50
224.00
7701.50
195.35
184.94
97.11
11165
200
Ozark Highlands
Aluminum
mg/l
0.008
0.01
0.01
0.02
0.13
1.03
0.08
0.03
0.10
5571
190
Ozark Highlands
Total Recoverable
Aluminum
mg/l
0.005
0.02
0.02
0.05
0.10
8.40
0.16
0.06
0.52
2084
104
Ozark Highlands
Iron
mg/l
0.001
0.01
0.01
0.02
0.04
0.88
0.03
0.02
0.04
5571
190
Ozark Highlands
Total Recoverable
Iron
mg/l
0.010
0.03
0.03
0.05
0.11
10.20
0.15
0.06
0.55
2084
104
Ozark Highlands
Ammonia-nitrogen
mg/l
0.000
0.01
0.02
0.02
0.03
5.55
0.03
0.02
0.10
13584
214
Ozark Highlands
Nitrite + Nitrate as
Nitrogen
mg/l
0.001
0.06
0.22
0.61
1.74
25.30
1.24
0.52
1.66
13468
214
Ozark Highlands
Total Kjeldahl
nitrogen
mg/l
0.005
0.06
0.14
0.23
0.37
60.00
0.32
0.21
0.87
8896
161
Ozark Highlands
Total Phosphorus
mg/l
0.003
0.01
0.03
0.05
0.11
13.83
0.22
0.06
0.68
11310
200
Ozark Highlands
Orthophosphate
mg/l
0.002
0.01
0.01
0.02
0.06
12.50
0.17
0.03
0.61
13572
214
Ozark Highlands
Total suspended
solids
mg/l
0.500
0.50
1.00
2.50
6.00
1130.00
9.09
2.81
39.89
11166
200
Ozark Highlands
PH
NA
4.580
7.28
7.60
7.90
8.15
11.93
7.85
7.84
0.45
13128
221
Ozark Highlands
Temperature
°C
0.460
8.00
11.00
15.90
21.86
37.20
16.36
14.84
6.63
13448
221
South Central Plains
Alkalinity, total
mg/l
CaC03
0.050
3.00
10.40
20.40
42.25
1040.00
34.65
19.49
43.62
8935
180
South Central Plains
Bicarbonate
mg/l
0.063
3.77
12.82
25.10
53.29
1306.38
43.59
24.27
55.32
8597
180
South Central Plains
Chloride
mg/l
0.035
2.39
3.37
5.83
17.90
2970.00
24.61
8.53
68.50
13969
195
55
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
South Central Plains
Hardness, Ca, Mg
mg/l
0.025
11.00
17.99
27.30
48.00
1710.00
44.78
29.77
62.23
7296
193
South Central Plains
Sulfate
mg/l
0.020
3.19
4.92
9.30
21.20
817.00
26.11
11.29
51.67
13983
195
South Central Plains
Calcium
mg/l
0.010
2.68
4.59
7.50
13.50
680.00
12.70
7.99
19.68
7457
193
South Central Plains
Total Recoverable
Calcium
mg/l
0.250
2.50
4.30
7.00
13.30
696.00
13.48
7.81
28.60
3608
165
South Central Plains
Sodium
mg/l
0.010
2.35
3.50
6.36
15.00
566.20
19.31
7.95
39.80
7449
193
South Central Plains
Potassium
mg/l
0.010
0.95
1.30
2.00
3.30
48.70
2.85
2.09
3.02
7457
193
South Central Plains
Magnesium
mg/l
0.010
0.98
1.37
1.98
3.12
64.10
3.10
2.11
4.37
7457
193
South Central Plains
Specific Conductivity
US/cm
5.350
41.42
60.00
93.70
160.00
1760.00
161.36
105.01
210.06
4313
129
South Central Plains
Total dissolved solids
mg/l
7.000
45.50
62.00
86.50
147.00
5231.00
149.96
103.57
187.59
14022
195
South Central Plains
Aluminum
mg/l
0.008
0.01
0.03
0.06
0.12
1.51
0.10
0.06
0.11
7457
193
South Central Plains
Total Recoverable
Aluminum
mg/l
0.005
0.07
0.16
0.32
0.62
14.10
0.50
0.31
0.65
3608
165
South Central Plains
Iron
mg/l
0.001
0.05
0.13
0.28
0.52
7.46
0.40
0.24
0.43
7457
193
South Central Plains
Total Recoverable
Iron
mg/l
0.025
0.30
0.57
0.94
1.45
12.50
1.16
0.86
0.95
3608
165
South Central Plains
Ammonia-nitrogen
mg/l
0.001
0.02
0.02
0.03
0.07
151.50
0.18
0.04
1.94
13954
195
South Central Plains
Nitrite + Nitrate as
Nitrogen
mg/l
0.001
0.02
0.06
0.15
0.28
211.00
0.71
0.14
3.69
13969
195
South Central Plains
Total Kjeldahl
nitrogen
mg/l
0.002
0.24
0.34
0.53
0.79
64.64
0.71
0.53
1.17
11401
191
South Central Plains
Total Phosphorus
mg/l
0.005
0.03
0.04
0.07
0.13
16.50
0.17
0.08
0.49
13729
195
South Central Plains
Orthophosphate
mg/l
0.002
0.01
0.01
0.02
0.05
9.35
0.10
0.03
0.42
13967
195
South Central Plains
Total suspended
solids
mg/l
0.500
1.80
3.50
7.00
14.85
3232.00
17.92
7.52
49.95
14016
195
South Central Plains
PH
NA
0.650
6.05
6.46
6.85
7.20
10.05
6.82
6.79
0.63
13668
195
South Central Plains
Temperature
°C
1.000
8.50
12.00
18.10
25.00
39.00
18.37
16.66
7.31
13691
195
56
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Table A.2.3. Summary statistics using multiple samples for least disturbed stations.
Ecoregion
Parameter
Unit
Fraction
Min
PerlO
Per25
Per50
Per75
Max
Mean
GeoMean
SD
N
Sites
Arkansas Valley
Alkalinity, total
mg/l CaCC>3
Dissolved
0.050
3.00
7.20
11.80
18.30
117.18
14.72
10.68
12.41
1022
18
Arkansas Valley
Bicarbonate
mg/l
Dissolved
0.063
3.77
8.92
14.82
22.90
147.17
18.43
13.33
15.70
987
18
Arkansas Valley
Chloride
mg/l
Dissolved
0.190
1.50
2.20
3.33
4.54
112.00
3.76
3.13
4.16
1497
18
Arkansas Valley
Hardness, Ca, Mg
mg/l
Dissolved
4.000
9.00
11.00
13.30
19.00
95.00
16.52
14.73
9.44
753
18
Arkansas Valley
Sulfate
mg/l
Dissolved
0.020
2.14
2.73
3.66
5.87
71.60
4.91
3.95
4.75
1495
18
Arkansas Valley
Calcium
mg/l
Dissolved
1.200
1.90
2.27
2.81
3.71
26.80
3.33
2.99
2.08
757
18
Arkansas Valley
Total Recoverable
Calcium
mg/l
Total
Recoverable
1.330
2.00
2.33
2.93
3.88
28.30
3.41
3.10
2.01
348
12
Arkansas Valley
Sodium
mg/l
Dissolved
0.020
1.30
1.80
2.70
4.30
167.00
3.63
2.71
6.41
757
18
Arkansas Valley
Potassium
mg/l
Dissolved
0.010
0.63
0.95
1.45
2.20
11.00
1.77
1.43
1.20
757
18
Arkansas Valley
Magnesium
mg/l
Dissolved
0.212
0.90
1.12
1.59
2.42
7.30
1.99
1.70
1.24
757
18
Arkansas Valley
Specific Conductivity
US/cm
Total
18.500
30.00
35.00
45.80
63.30
423.00
54.29
48.82
31.34
604
13
Arkansas Valley
Total dissolved solids
mg/l
Total
8.000
27.50
33.00
44.00
58.50
387.00
48.09
44.41
22.64
1497
18
Arkansas Valley
Aluminum
mg/l
Dissolved
0.008
0.01
0.02
0.05
0.06
0.93
0.07
0.04
0.09
757
18
Arkansas Valley
Total Recoverable
Aluminum
mg/l
Total
Recoverable
0.020
0.05
0.08
0.19
0.43
2.14
0.30
0.19
0.31
348
12
Arkansas Valley
Iron
mg/l
Dissolved
0.008
0.05
0.09
0.18
0.31
2.16
0.25
0.17
0.27
757
18
Arkansas Valley
fotal Recoverable Iron
mg/l
Total
Recoverable
0.103
0.22
0.31
0.52
0.86
3.27
0.67
0.52
0.51
348
12
Arkansas Valley
Ammonia-nitrogen
mg/l
Dissolved
0.002
0.02
0.02
0.02
0.04
3.44
0.04
0.02
0.11
1490
18
Arkansas Valley
Nitrite + Nitrate as
Nitrogen
mg/l
Dissolved
0.005
0.03
0.06
0.15
0.28
3.01
0.21
0.12
0.23
1496
18
Arkansas Valley
Fotal Kjeldahl nitrogen
mg/l
Total
0.025
0.10
0.18
0.31
0.53
3.88
0.39
0.29
0.32
1249
18
Arkansas Valley
Total Phosphorus
mg/l
Total
0.002
0.01
0.02
0.04
0.07
1.17
0.06
0.04
0.07
1472
18
Arkansas Valley
Orthophosphate
mg/l
Dissolved
0.003
0.01
0.01
0.01
0.02
0.45
0.02
0.01
0.03
1493
18
Arkansas Valley
fotal suspended solids
mg/l
Total
0.500
0.50
1.50
4.00
9.80
392.50
10.20
4.10
23.25
1495
18
Arkansas Valley
PH
NA
Total
4.000
6.23
6.50
6.75
7.06
8.82
6.78
6.76
0.51
1480
18
57
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Fraction
Min
PerlO
Per25
Per50
Per75
Max
Mean
GeoMean
SD
N
Sites
Arkansas Valley
Temperature
°C
Total
1.000
6.90
10.70
17.90
25.00
36.30
17.83
15.60
8.08
1525
18
Boston Mountains
Alkalinity, total
mg/l CaCC>3
Dissolved
0.299
8.30
14.00
24.80
35.63
164.00
28.18
21.51
20.89
888
35
Boston Mountains
Bicarbonate
mg/l
Dissolved
0.376
10.43
17.74
31.19
44.87
204.47
35.54
27.12
26.28
863
35
Boston Mountains
Chloride
mg/l
Dissolved
0.250
1.19
1.51
1.97
2.46
239.00
2.56
1.98
7.74
1359
36
Boston Mountains
Hardness, Ca, Mg
mg/l
Dissolved
0.500
9.04
13.00
23.00
35.80
231.00
27.33
21.84
21.20
787
34
Boston Mountains
Sulfate
mg/l
Dissolved
0.500
2.38
3.04
4.05
5.63
82.60
4.81
4.18
3.63
1362
36
Boston Mountains
Calcium
mg/l
Dissolved
0.039
2.29
3.40
6.95
11.60
83.20
8.56
6.32
7.39
788
34
Boston Mountains
Total Recoverable
Calcium
mg/l
Total
Recoverable
1.110
2.19
3.14
6.48
11.50
42.00
7.93
6.00
6.11
294
28
Boston Mountains
Sodium
mg/l
Dissolved
0.010
0.97
1.26
1.63
2.21
161.00
2.51
1.68
6.98
789
34
Boston Mountains
Potassium
mg/l
Dissolved
0.010
0.50
0.70
0.92
1.24
6.12
1.05
0.91
0.60
789
34
Boston Mountains
Magnesium
mg/l
Dissolved
0.010
0.83
1.04
1.30
1.69
24.70
1.46
1.32
1.03
789
34
Boston Mountains
Specific Conductivity
US/cm
Total
18.000
31.00
45.00
70.10
92.45
259.00
75.30
65.04
43.18
418
26
Boston Mountains
Total dissolved solids
mg/l
Total
14.500
28.00
35.00
47.00
58.00
519.00
49.89
45.96
26.48
1144
35
Boston Mountains
Aluminum
mg/l
Dissolved
0.008
0.01
0.01
0.02
0.06
1.28
0.05
0.03
0.08
789
34
Boston Mountains
Total Recoverable
Aluminum
mg/l
Total
Recoverable
0.010
0.04
0.06
0.11
0.26
3.57
0.25
0.13
0.41
294
28
Boston Mountains
Iron
mg/l
Dissolved
0.006
0.01
0.02
0.04
0.06
0.94
0.05
0.03
0.06
789
34
Boston Mountains
fotal Recoverable Iron
mg/l
Total
Recoverable
0.021
0.06
0.11
0.16
0.24
2.84
0.23
0.17
0.30
294
28
Boston Mountains
Ammonia-nitrogen
mg/l
Dissolved
0.002
0.01
0.02
0.02
0.03
0.27
0.02
0.02
0.02
1362
36
Boston Mountains
Nitrite + Nitrate as
Nitrogen
mg/l
Dissolved
0.005
0.01
0.02
0.06
0.15
2.08
0.13
0.05
0.19
1361
36
Boston Mountains
Fotal Kjeldahl nitrogen
mg/l
Total
0.025
0.03
0.07
0.14
0.24
1.84
0.18
0.13
0.18
949
34
Boston Mountains
Total Phosphorus
mg/l
Total
0.005
0.01
0.01
0.02
0.04
4.81
0.04
0.02
0.17
1167
35
Boston Mountains
Orthophosphate
mg/l
Dissolved
0.003
0.01
0.01
0.01
0.02
1.51
0.02
0.01
0.08
1361
36
Boston Mountains
fotal suspended solids
mg/l
Total
0.500
0.50
0.50
1.50
3.90
598.00
6.34
1.82
30.36
1143
35
Boston Mountains
PH
NA
Total
4.780
6.39
6.76
7.16
7.56
10.50
7.16
7.13
0.62
1304
36
Boston Mountains
Temperature
°C
Total
0.800
6.93
10.20
16.00
23.50
33.80
16.70
14.66
7.62
1344
36
58
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Fraction
Min
PerlO
Per25
Per50
Per75
Max
Mean
GeoMean
SD
N
Sites
Mississippi Alluvial
Plain
Alkalinity, total
mg/l CaCC>3
Dissolved
2.500
36.34
55.90
99.70
153.00
263.00
105.22
87.87
55.26
698
7
Mississippi Alluvial
Plain
Bicarbonate
mg/l
Dissolved
3.384
45.52
69.69
122.94
191.00
329.95
130.52
109.02
68.69
692
7
Mississippi Alluvial
Plain
Chloride
mg/l
Dissolved
1.010
2.44
3.06
5.99
11.20
124.00
11.51
6.73
15.96
1225
7
Mississippi Alluvial
Plain
Hardness, Ca, Mg
mg/l
Dissolved
4.000
39.00
63.00
120.00
165.00
772.00
119.46
100.52
66.83
556
7
Mississippi Alluvial
Plain
Sulfate
mg/l
Dissolved
0.020
3.00
4.03
6.07
11.27
113.00
8.67
6.63
7.48
1231
7
Mississippi Alluvial
Plain
Calcium
mg/l
Dissolved
1.000
10.22
16.80
28.70
36.50
284.00
28.76
24.47
17.83
557
7
Mississippi Alluvial
Plain
Total Recoverable
Calcium
mg/l
Total
Recoverable
2.310
11.14
17.00
26.30
36.65
258.00
28.49
23.93
21.45
195
5
Mississippi Alluvial
Plain
Sodium
mg/l
Dissolved
0.020
1.61
2.26
4.50
8.90
297.00
8.34
4.50
20.56
557
7
Mississippi Alluvial
Plain
Potassium
mg/l
Dissolved
0.230
0.89
1.40
3.10
4.65
13.10
3.41
2.63
2.30
557
7
Mississippi Alluvial
Plain
Magnesium
mg/l
Dissolved
0.434
3.32
5.40
10.80
17.30
31.60
11.61
9.39
6.78
557
7
Mississippi Alluvial
Plain
Specific Conductivity
US/cm
Total
57.400
110.40
152.00
239.00
329.00
891.00
259.08
224.37
146.05
253
5
Mississippi Alluvial
Plain
Total dissolved solids
mg/l
Total
48.000
96.65
128.00
159.25
188.00
717.00
167.53
156.14
67.45
1234
7
Mississippi Alluvial
Plain
Aluminum
mg/l
Dissolved
0.008
0.01
0.01
0.05
0.15
2.14
0.11
0.04
0.21
557
7
Mississippi Alluvial
Plain
Total Recoverable
Aluminum
mg/l
Total
Recoverable
0.005
0.05
0.10
0.26
0.67
5.38
0.53
0.25
0.73
195
5
Mississippi Alluvial
Plain
Iron
mg/l
Dissolved
0.001
0.01
0.02
0.05
0.14
2.37
0.13
0.06
0.23
556
7
Mississippi Alluvial
Plain
fotal Recoverable Iron
mg/l
Total
Recoverable
0.039
0.15
0.23
0.55
1.00
6.76
0.80
0.51
0.86
195
5
Mississippi Alluvial
Plain
Ammonia-nitrogen
mg/l
Dissolved
0.003
0.01
0.02
0.02
0.04
0.76
0.04
0.02
0.05
1221
7
59
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Fraction
Min
PerlO
Per25
Per50
Per75
Max
Mean
GeoMean
SD
N
Sites
Mississippi Alluvial
Plain
Nitrite + Nitrate as
Nitrogen
mg/l
Dissolved
0.005
0.01
0.03
0.09
0.22
1.57
0.15
0.07
0.17
1229
7
Mississippi Alluvial
Plain
fotal Kjeldahl nitrogen
mg/l
Total
0.025
0.14
0.40
0.63
0.83
2.51
0.63
0.50
0.35
872
7
Mississippi Alluvial
Plain
Total Phosphorus
mg/l
Total
0.006
0.03
0.07
0.14
0.21
1.05
0.15
0.11
0.11
1202
7
Mississippi Alluvial
Plain
Orthophosphate
mg/l
Dissolved
0.003
0.01
0.02
0.07
0.10
0.60
0.08
0.05
0.06
1227
7
Mississippi Alluvial
Plain
fotal suspended solids
mg/l
Total
0.500
2.00
4.50
10.00
24.50
598.70
23.07
10.50
40.40
1225
7
Mississippi Alluvial
Plain
PH
NA
Total
3.420
6.73
7.09
7.49
7.90
10.40
7.46
7.43
0.61
1208
7
Mississippi Alluvial
Plain
Temperature
°C
Total
0.400
7.50
11.00
18.00
24.30
34.00
17.74
15.59
7.62
1207
7
Ouachita Mountains
Alkalinity, total
mg/l CaC03
Dissolved
0.050
4.90
10.30
20.40
42.60
115.00
27.97
19.07
22.36
2267
39
Ouachita Mountains
Bicarbonate
mg/l
Dissolved
0.026
3.79
12.81
24.85
52.44
144.29
34.57
23.51
27.86
2190
39
Ouachita Mountains
Chloride
mg/l
Dissolved
0.250
1.36
1.58
1.94
2.44
37.40
2.20
2.01
1.58
3187
41
Ouachita Mountains
Hardness, Ca, Mg
mg/l
Dissolved
0.500
8.50
12.70
23.00
47.05
406.00
31.83
23.68
25.51
1763
40
Ouachita Mountains
Sulfate
mg/l
Dissolved
0.350
2.71
3.40
4.50
6.11
458.00
5.62
4.67
9.18
3180
41
Ouachita Mountains
Calcium
mg/l
Dissolved
0.010
1.79
3.03
6.48
14.40
115.00
9.37
6.27
8.18
1770
40
Ouachita Mountains
Total Recoverable
Calcium
mg/l
Total
Recoverable
0.279
1.68
2.74
6.39
14.50
40.90
9.08
6.01
7.54
1088
38
Ouachita Mountains
Sodium
mg/l
Dissolved
0.010
1.08
1.47
1.86
2.39
160.00
2.26
1.79
4.26
1763
40
Ouachita Mountains
Potassium
mg/l
Dissolved
0.010
0.45
0.56
0.74
1.06
5.76
0.87
0.77
0.50
1770
40
Ouachita Mountains
Magnesium
mg/l
Dissolved
0.010
0.88
1.14
1.59
2.50
28.90
2.03
1.71
1.40
1770
40
Ouachita Mountains
Specific Conductivity
US/cm
Total
2.960
28.00
38.00
58.00
104.00
840.00
73.22
60.20
50.77
1325
32
Ouachita Mountains
Total dissolved solids
mg/l
Total
12.000
28.00
34.50
45.00
63.88
712.00
51.45
46.78
26.39
3178
41
Ouachita Mountains
Aluminum
mg/l
Dissolved
0.008
0.01
0.01
0.02
0.05
0.59
0.04
0.03
0.05
1770
40
Ouachita Mountains
Total Recoverable
Aluminum
mg/l
Total
Recoverable
0.005
0.01
0.02
0.04
0.11
2.11
0.10
0.05
0.17
1088
38
Ouachita Mountains
Iron
mg/l
Dissolved
0.003
0.01
0.03
0.04
0.08
0.59
0.06
0.04
0.06
1770
40
60
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Fraction
Min
PerlO
Per25
Per50
Per75
Max
Mean
GeoMean
SD
N
Sites
Ouachita Mountains
Total Recoverable
Iron
mg/l
Total
Recoverable
0.010
0.03
0.03
0.08
0.16
2.90
0.13
0.07
0.20
1088
38
Ouachita Mountains
Ammonia-nitrogen
mg/l
Dissolved
0.001
0.02
0.02
0.02
0.03
0.64
0.02
0.02
0.03
3170
41
Ouachita Mountains
Nitrite + Nitrate as
Nitrogen
mg/l
Dissolved
0.003
0.01
0.03
0.05
0.14
13.30
0.12
0.05
0.37
3166
41
Ouachita Mountains
fotal Kjeldahl nitrogen
mg/l
Total
0.015
0.05
0.09
0.16
0.26
4.68
0.20
0.15
0.20
2565
40
Ouachita Mountains
Total Phosphorus
mg/l
Total
0.005
0.01
0.01
0.03
0.04
2.03
0.04
0.02
0.09
3120
41
Ouachita Mountains
Orthophosphate
mg/l
Dissolved
0.003
0.01
0.01
0.01
0.02
1.50
0.02
0.01
0.06
3180
41
Ouachita Mountains
fotal suspended solids
mg/l
Total
0.500
0.50
0.50
1.50
3.25
868.00
4.37
1.68
20.76
3178
41
Ouachita Mountains
PH
NA
Total
4.010
6.39
6.77
7.10
7.36
13.20
7.04
7.02
0.53
3183
42
Ouachita Mountains
Temperature
°C
Total
1.000
8.00
12.00
18.00
24.00
37.80
17.95
16.13
7.22
3209
42
Ozark Highlands
Alkalinity, total
mg/l CaC03
Dissolved
1.800
68.90
100.00
133.00
199.00
428.00
144.12
129.92
59.34
3817
65
Ozark Highlands
Bicarbonate
mg/l
Dissolved
2.245
85.71
124.95
166.16
247.80
527.42
179.39
161.79
73.47
3720
64
Ozark Highlands
Chloride
mg/l
Dissolved
0.250
2.06
2.65
3.40
5.34
62.60
5.68
4.11
6.19
6629
71
Ozark Highlands
Hardness, Ca, Mg
mg/l
Dissolved
1.000
78.00
113.00
145.00
211.00
2160.00
162.20
143.78
100.8C
2459
70
Ozark Highlands
Sulfate
mg/l
Dissolved
0.020
3.21
3.98
5.34
7.83
75.00
7.08
5.87
5.48
6630
71
Ozark Highlands
Calcium
mg/l
Dissolved
0.025
25.80
35.00
42.30
47.88
452.00
43.32
39.37
30.06
2486
70
Ozark Highlands
Total Recoverable
Calcium
mg/l
Total
Recoverable
5.120
25.48
34.80
42.70
48.80
468.00
48.31
41.04
49.39
869
45
Ozark Highlands
Sodium
mg/l
Dissolved
0.020
1.18
1.59
2.10
5.27
210.00
5.26
2.66
9.02
2476
70
Ozark Highlands
Potassium
mg/l
Dissolved
0.045
0.80
1.17
1.60
2.50
25.10
2.10
1.68
1.71
2486
70
Ozark Highlands
Magnesium
mg/l
Dissolved
0.065
1.80
2.19
7.24
24.90
253.00
13.16
7.15
13.21
2486
70
Ozark Highlands
Specific Conductivity
US/cm
Total
4.570
147.00
200.50
277.00
380.50
568.00
284.36
261.51
105.3C
1899
55
Ozark Highlands
Total dissolved solids
mg/l
Total
6.000
108.00
145.00
190.50
224.00
477.00
182.70
174.14
51.11
5185
68
Ozark Highlands
Aluminum
mg/l
Dissolved
0.008
0.01
0.01
0.02
0.13
1.03
0.08
0.03
0.10
2486
70
Ozark Highlands
Total Recoverable
Aluminum
mg/l
Total
Recoverable
0.005
0.02
0.05
0.07
0.13
8.40
0.18
0.08
0.58
869
45
Ozark Highlands
Iron
mg/l
Dissolved
0.001
0.01
0.01
0.02
0.04
0.88
0.03
0.02
0.05
2486
70
Ozark Highlands
fotal Recoverable Iron
mg/l
Total
Recoverable
0.010
0.03
0.05
0.08
0.15
10.20
0.19
0.09
0.63
869
45
61
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Fraction
Min
PerlO
Per25
Per50
Per75
Max
Mean
GeoMean
SD
N
Sites
Ozark Highlands
Ammonia-nitrogen
mg/l
Dissolved
0.000
0.01
0.02
0.02
0.03
1.50
0.02
0.02
0.04
6633
71
Ozark Highlands
Nitrite + Nitrate as
Nitrogen
mg/l
Dissolved
0.001
0.03
0.12
0.41
1.04
24.80
0.78
0.32
1.02
6578
71
Ozark Highlands
fotal Kjeldahl nitrogen
mg/l
Total
0.005
0.05
0.12
0.19
0.30
44.00
0.27
0.18
0.76
3870
57
Ozark Highlands
Total Phosphorus
mg/l
Total
0.005
0.01
0.02
0.04
0.08
4.86
0.09
0.05
0.20
5306
68
Ozark Highlands
Orthophosphate
mg/l
Dissolved
0.002
0.01
0.01
0.02
0.04
4.39
0.06
0.02
0.18
6622
71
Ozark Highlands
fotal suspended solids
mg/l
Total
0.500
1.00
2.00
4.00
8.00
1130.00
12.53
3.97
49.97
5179
68
Ozark Highlands
PH
NA
Total
4.580
7.26
7.62
7.91
8.15
10.23
7.85
7.84
0.45
6397
78
Ozark Highlands
Temperature
°C
Total
0.460
7.50
11.00
15.90
22.10
34.00
16.49
14.83
6.92
6570
78
South Central Plains
Alkalinity, total
mg/l CaC03
Dissolved
0.050
3.00
8.27
20.45
40.85
342.00
28.40
16.67
30.75
1820
28
South Central Plains
Bicarbonate
mg/l
Dissolved
0.063
3.77
10.42
25.57
51.76
428.52
35.69
20.93
38.84
1771
28
South Central Plains
Chloride
mg/l
Dissolved
0.190
2.29
3.07
4.13
5.61
165.00
5.35
4.33
6.29
2975
28
South Central Plains
Hardness, Ca, Mg
mg/l
Dissolved
0.025
10.00
16.03
29.00
44.00
1710.00
36.44
26.25
65.66
1482
28
South Central Plains
Sulfate
mg/l
Dissolved
0.250
3.46
5.29
10.50
20.74
255.00
15.75
10.47
16.08
2981
28
South Central Plains
Calcium
mg/l
Dissolved
0.010
2.49
4.19
7.50
12.00
680.00
10.39
6.92
25.29
1497
28
South Central Plains
Total Recoverable
Calcium
mg/l
Total
Recoverable
0.909
2.06
3.70
6.83
11.15
627.00
12.26
6.78
35.15
751
28
South Central Plains
Sodium
mg/l
Dissolved
0.010
1.98
3.30
6.03
10.20
187.00
8.37
5.55
9.90
1496
28
South Central Plains
Potassium
mg/l
Dissolved
0.010
0.80
1.15
1.56
2.20
16.50
1.82
1.52
1.24
1497
28
South Central Plains
Magnesium
mg/l
Dissolved
0.010
0.80
1.30
2.32
3.40
33.00
2.52
2.02
1.95
1497
28
South Central Plains
Specific Conductivity
US/cm
Total
5.940
37.36
57.50
91.50
134.00
1150.00
118.90
90.49
114.19
845
22
South Central Plains
Total dissolved solids
mg/l
Total
12.000
56.00
69.00
83.50
100.00
642.00
90.48
84.09
43.11
2989
28
South Central Plains
Aluminum
mg/l
Dissolved
0.008
0.02
0.04
0.06
0.14
1.09
0.11
0.07
0.12
1497
28
South Central Plains
Total Recoverable
Aluminum
mg/l
Total
Recoverable
0.020
0.11
0.19
0.37
0.66
4.39
0.50
0.35
0.48
751
28
South Central Plains
Iron
mg/l
Dissolved
0.005
0.07
0.14
0.30
0.53
2.79
0.38
0.26
0.34
1497
28
South Central Plains
fotal Recoverable Iron
mg/l
Total
Recoverable
0.050
0.29
0.59
0.94
1.48
6.25
1.17
0.89
0.90
751
28
South Central Plains
Ammonia-nitrogen
mg/l
Dissolved
0.003
0.02
0.02
0.03
0.04
10.20
0.04
0.02
0.20
2972
28
62
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Fraction
Min
PerlO
Per25
Per50
Per75
Max
Mean
GeoMean
SD
N
Sites
South Central Plains
Nitrite + Nitrate as
Nitrogen
mg/l
Dissolved
0.001
0.01
0.03
0.10
0.18
7.75
0.15
0.08
0.31
2979
28
South Central Plains
Fotal Kjeldahl nitrogen
mg/l
Total
0.025
0.22
0.30
0.43
0.63
11.10
0.50
0.42
0.36
2343
28
South Central Plains
Total Phosphorus
mg/l
Total
0.005
0.02
0.04
0.06
0.09
3.87
0.08
0.06
0.16
2927
28
South Central Plains
Orthophosphate
mg/l
Dissolved
0.003
0.01
0.01
0.02
0.03
2.94
0.04
0.02
0.13
2971
28
South Central Plains
fotal suspended solids
mg/l
Total
0.500
2.50
4.30
7.50
12.50
414.00
11.65
7.40
17.76
2990
28
South Central Plains
PH
NA
Total
4.300
6.00
6.41
6.84
7.17
9.39
6.77
6.74
0.62
2943
28
South Central Plains
Temperature
°C
Total
1.000
8.00
12.00
18.00
25.00
36.60
18.27
16.40
7.56
2944
28
63
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Table A.2.4. Summary statistics using multiple samples for all stations.
Ecoregion
Parameter
Unit
Fraction
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Arkansas Valley
Alkalinity, total
mg/l CaCC>3
Dissolved
0.050
3.00
10.63
24.65
76.20
418.00
43.20
24.35
41.49
4354
72
Arkansas Valley
Bicarbonate
mg/l
Dissolved
0.063
3.77
13.18
30.65
95.37
321.33
53.82
30.26
51.40
4207
72
Arkansas Valley
Chloride
mg/l
Dissolved
0.035
1.65
2.85
5.70
46.50
1890.00
31.10
9.86
56.14
6976
86
Arkansas Valley
Hardness, Ca, Mg
mg/l
Dissolved
0.500
10.00
13.00
24.00
92.00
1050.00
54.37
31.95
62.56
3346
85
Arkansas Valley
Sulfate
mg/l
Dissolved
0.020
2.51
3.43
10.40
38.20
338.00
23.99
11.51
29.38
6993
86
Arkansas Valley
Calcium
mg/l
Dissolved
0.066
2.00
2.70
5.09
23.00
406.00
14.08
7.13
19.94
3358
85
Arkansas Valley
Total Recoverable
Calcium
mg/l
Total
Recoverable
1.040
2.04
2.72
4.94
21.45
413.00
14.53
7.07
26.15
1391
48
Arkansas Valley
Sodium
mg/l
Dissolved
0.020
1.41
2.46
6.23
36.30
679.30
24.85
8.46
39.79
3358
85
Arkansas Valley
Potassium
mg/l
Dissolved
0.010
0.70
1.20
2.28
3.80
46.00
3.27
2.15
4.53
3358
85
Arkansas Valley
Magnesium
mg/l
Dissolved
0.010
1.01
1.48
2.90
6.97
58.90
4.65
3.12
4.47
3358
85
Arkansas Valley
Specific Conductivity
US/cm
Total
2.380
29.00
38.50
61.00
260.00
1300.00
183.47
94.78
>28.4!
2113
43
Arkansas Valley
Total dissolved solids
mg/l
Total
6.000
30.50
41.00
77.00
249.00
5020.00
154.82
97.05
L66.1"
7027
86
Arkansas Valley
Aluminum
mg/l
Dissolved
0.008
0.01
0.02
0.05
0.07
1.21
0.08
0.04
0.10
3358
85
Arkansas Valley
Total Recoverable
Aluminum
mg/l
Total
Recoverable
0.010
0.07
0.12
0.25
0.48
4.47
0.37
0.24
0.39
1391
48
Arkansas Valley
Iron
mg/l
Dissolved
0.001
0.02
0.05
0.11
0.21
7.65
0.17
0.09
0.25
3358
85
Arkansas Valley
Total Recoverable
Iron
mg/l
Total
Recoverable
0.025
0.17
0.26
0.44
0.78
6.88
0.59
0.44
0.50
1391
48
Arkansas Valley
Ammonia-nitrogen
mg/l
Dissolved
0.001
0.02
0.02
0.03
0.06
19.00
0.21
0.03
1.14
6943
86
Arkansas Valley
Nitrite + Nitrate as
Nitrogen
mg/l
Dissolved
0.005
0.02
0.06
0.21
0.45
51.50
0.67
0.17
2.12
6961
86
Arkansas Valley
Total Kjeldahl
nitrogen
mg/l
Total
0.025
0.11
0.24
0.46
0.65
25.80
0.73
0.39
1.56
5742
86
Arkansas Valley
Total Phosphorus
mg/l
Total
0.001
0.02
0.03
0.07
0.12
25.76
0.27
0.07
1.07
6865
86
Arkansas Valley
Orthophosphate
mg/l
Dissolved
0.003
0.01
0.01
0.02
0.06
17.80
0.20
0.03
0.96
6960
86
Arkansas Valley
Total suspended
solids
mg/l
Total
0.250
1.00
2.50
6.50
13.00
960.00
12.05
5.60
24.68
7029
86
64
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Fraction
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Arkansas Valley
PH
NA
Total
4.000
6.35
6.68
7.05
7.50
9.91
7.08
7.05
0.63
6735
86
Arkansas Valley
Temperature
°C
Total
0.200
7.00
11.00
17.90
25.00
36.30
17.91
15.63
8.16
6988
86
Boston Mountains
Alkalinity, total
mg/l CaCC>3
Dissolved
0.299
10.30
20.22
35.20
55.68
178.00
43.74
32.03
33.44
1706
63
Boston Mountains
Bicarbonate
mg/l
Dissolved
0.376
12.78
25.61
44.00
69.83
219.82
54.82
40.11
41.86
1650
63
Boston Mountains
Chloride
mg/l
Dissolved
0.250
1.28
1.76
2.45
3.52
239.00
3.84
2.66
8.31
2831
90
Boston Mountains
Hardness, Ca, Mg
mg/l
Dissolved
0.500
10.00
17.00
35.80
56.70
622.00
44.89
32.43
39.90
1461
68
Boston Mountains
Sulfate
mg/l
Dissolved
0.020
2.62
3.68
5.91
10.90
110.48
10.18
6.74
12.02
2832
89
Boston Mountains
Calcium
mg/l
Dissolved
0.039
2.50
4.71
11.60
18.90
243.00
14.63
9.80
14.18
1467
68
Boston Mountains
Total Recoverable
Calcium
mg/l
Total
Recoverable
1.110
2.40
4.08
11.00
17.28
234.00
13.06
8.88
14.26
478
37
Boston Mountains
Sodium
mg/l
Dissolved
0.010
1.03
1.42
2.05
3.54
161.00
3.97
2.35
8.17
1469
68
Boston Mountains
Potassium
mg/l
Dissolved
0.010
0.60
0.80
1.11
1.60
21.50
1.40
1.12
1.25
1469
68
Boston Mountains
Magnesium
mg/l
Dissolved
0.010
0.90
1.20
1.65
2.39
24.70
2.05
1.73
1.43
1469
68
Boston Mountains
Specific Conductivity
US/cm
Total
L8.00(
35.00
54.00
90.30
142.75
437.00
112.28
89.17
79.06
742
39
Boston Mountains
Total dissolved solids
mg/l
Total
L3.00(
32.00
43.50
62.00
92.00
564.00
77.85
65.49
52.66
2415
80
Boston Mountains
Aluminum
mg/l
Dissolved
0.008
0.01
0.01
0.02
0.06
1.28
0.05
0.03
0.08
1469
68
Boston Mountains
Total Recoverable
Aluminum
mg/l
Total
Recoverable
0.010
0.04
0.06
0.13
0.27
14.70
0.29
0.13
0.79
478
37
Boston Mountains
Iron
mg/l
Dissolved
0.001
0.01
0.01
0.03
0.05
0.94
0.04
0.03
0.06
1469
68
Boston Mountains
Total Recoverable
Iron
mg/l
Total
Recoverable
0.021
0.06
0.10
0.18
0.29
20.80
0.31
0.18
1.02
478
37
Boston Mountains
Ammonia-nitrogen
mg/l
Dissolved
0.002
0.01
0.02
0.02
0.03
0.39
0.02
0.02
0.03
2848
90
Boston Mountains
Nitrite + Nitrate as
Nitrogen
mg/l
Dissolved
0.005
0.01
0.04
0.12
0.39
4.94
0.29
0.11
0.45
2831
90
Boston Mountains
Total Kjeldahl
nitrogen
mg/l
Total
0.025
0.05
0.09
0.18
0.32
2.94
0.25
0.16
0.25
1980
67
Boston Mountains
Total Phosphorus
mg/l
Total
0.004
0.01
0.02
0.03
0.05
4.81
0.05
0.03
0.13
2437
78
Boston Mountains
Orthophosphate
mg/l
Dissolved
0.001
0.01
0.01
0.01
0.02
1.51
0.02
0.01
0.07
2846
90
Boston Mountains
Total suspended
solids
mg/l
Total
0.500
0.50
1.00
2.50
7.50
744.00
10.39
2.76
40.14
2409
80
65
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Fraction
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Boston Mountains
PH
NA
Total
4.780
6.56
6.97
7.37
7.69
11.30
7.32
7.29
0.59
2698
89
Boston Mountains
Temperature
°C
Total
0.400
6.60
10.00
16.00
23.00
34.00
16.47
14.40
7.60
2764
89
Mississippi Alluvial Plair
Alkalinity, total
mg/l CaCC>3
Dissolved
0.203
16.20
36.00
79.37
122.00
456.00
83.06
60.81
54.31
4817
95
Mississippi Alluvial Plair
Bicarbonate
mg/l
Dissolved
0.255
20.48
45.44
99.07
151.47
570.10
103.53
75.97
67.58
4684
95
Mississippi Alluvial Plair
Chloride
mg/l
Dissolved
0.100
2.67
3.78
6.74
21.80
683.00
20.25
9.41
30.71
8326
102
Mississippi Alluvial Plair
Hardness, Ca, Mg
mg/l
Dissolved
0.500
21.00
42.00
94.00
140.00
970.00
97.37
74.08
65.95
4337
102
Mississippi Alluvial Plair
Sulfate
mg/l
Dissolved
0.020
3.71
5.22
7.85
15.39
221.00
13.89
9.22
15.44
8351
102
Mississippi Alluvial Plair
Calcium
mg/l
Dissolved
0.010
5.40
10.93
23.65
34.60
374.00
24.48
18.61
17.68
4354
102
Mississippi Alluvial Plair
Total Recoverable
Calcium
mg/l
Total
Recoverable
1.490
5.34
11.40
22.80
34.30
505.00
25.48
18.81
27.93
1567
68
Mississippi Alluvial Plair
Sodium
mg/l
Dissolved
0.010
1.93
3.07
6.27
16.90
481.00
15.24
7.12
24.20
4345
102
Mississippi Alluvial Plair
Potassium
mg/l
Dissolved
0.010
1.16
1.78
3.00
4.39
61.50
3.38
2.74
2.46
4354
102
Mississippi Alluvial Plair
Magnesium
mg/l
Dissolved
0.010
1.80
3.50
7.80
12.70
46.40
8.84
6.53
6.20
4354
102
Mississippi Alluvial Plair
Specific Conductivity
US/cm
Total
8.440
61.68
114.00
214.00
305.25
957.00
235.62
186.57
L58.2^
1780
36
Mississippi Alluvial Plair
Total dissolved solids
mg/l
Total
8.000
77.00
116.00
154.75
200.00
1287.50
171.22
149.48
91.29
8376
102
Mississippi Alluvial Plair
Aluminum
mg/l
Dissolved
0.008
0.01
0.01
0.06
0.16
2.66
0.12
0.05
0.21
4353
102
Mississippi Alluvial Plair
Total Recoverable
Aluminum
mg/l
Total
Recoverable
0.005
0.09
0.20
0.42
0.92
18.60
0.85
0.42
1.47
1567
68
Mississippi Alluvial Plair
Iron
mg/l
Dissolved
0.001
0.01
0.02
0.06
0.22
5.35
0.17
0.07
0.28
4352
102
Mississippi Alluvial Plair
Total Recoverable
Iron
mg/l
Total
Recoverable
0.025
0.19
0.36
0.69
1.30
21.10
1.12
0.68
1.66
1567
68
Mississippi Alluvial Plair
Ammonia-nitrogen
mg/l
Dissolved
0.001
0.02
0.02
0.03
0.07
8.67
0.06
0.03
0.16
8305
102
Mississippi Alluvial Plair
Nitrite + Nitrate as
Nitrogen
mg/l
Dissolved
0.005
0.02
0.07
0.18
0.31
12.10
0.27
0.13
0.55
8337
102
Mississippi Alluvial Plair
Total Kjeldahl
nitrogen
mg/l
Total
0.022
0.24
0.41
0.63
0.88
9.76
0.71
0.58
0.49
6517
101
Mississippi Alluvial Plair
Total Phosphorus
mg/l
Total
0.006
0.04
0.07
0.13
0.21
7.06
0.18
0.12
0.28
8198
102
Mississippi Alluvial Plair
Orthophosphate
mg/l
Dissolved
0.003
0.01
0.02
0.06
0.10
7.33
0.10
0.05
0.23
8331
102
Mississippi Alluvial Plair
Total suspended
solids
mg/l
Total
0.500
4.00
7.50
15.30
30.00
1170.00
26.67
14.81
46.95
8334
102
66
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Fraction
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Mississippi Alluvial Plair
PH
NA
Total
2.620
6.54
6.92
7.39
7.80
10.40
7.35
7.32
0.64
8151
102
Mississippi Alluvial Plair
Temperature
°C
Total
0.400
7.50
11.20
18.10
25.00
35.00
18.07
15.95
7.82
8210
102
Mississippi Valley
Loess Plains
Alkalinity, total
mg/l CaCC>3
Dissolved
3.000
19.59
30.40
65.35
101.75
275.00
70.27
55.10
44.33
170
1
Mississippi Valley
Loess Plains
Bicarbonate
mg/l
Dissolved
3.770
24.61
38.18
81.84
126.79
341.35
87.56
68.79
54.95
170
1
Mississippi Valley
Loess Plains
Chloride
mg/l
Dissolved
1.080
2.70
4.60
11.45
21.50
186.00
15.29
10.07
17.67
170
1
Mississippi Valley
Loess Plains
Hardness, Ca, Mg
mg/l
Dissolved
2.000
26.85
38.78
69.50
96.50
173.00
71.69
60.31
37.04
80
1
Mississippi Valley
Loess Plains
Sulfate
mg/l
Dissolved
2.600
4.97
7.12
12.40
17.75
66.70
13.48
11.26
8.48
170
1
Mississippi Valley
Loess Plains
Calcium
mg/l
Dissolved
0.352
6.14
9.37
17.60
24.30
46.80
17.50
14.60
9.18
81
1
Mississippi Valley
Loess Plains
Total Recoverable
Calcium
mg/l
Total
Recoverable
4.140
6.99
9.30
17.20
22.20
53.50
17.24
14.87
9.51
51
1
Mississippi Valley
Loess Plains
Sodium
mg/l
Dissolved
1.050
3.42
4.72
14.40
22.80
77.70
16.18
11.07
13.83
81
1
Mississippi Valley
Loess Plains
Potassium
mg/l
Dissolved
0.694
2.24
2.71
3.56
4.22
38.00
4.92
3.77
6.01
81
1
Mississippi Valley
Loess Plains
Magnesium
mg/l
Dissolved
0.182
2.89
3.53
6.69
8.72
20.70
6.77
5.73
3.54
81
1
Mississippi Valley
Loess Plains
Specific Conductivity
US/cm
Total
15.40(
70.88
89.00
164.00
318.00
558.00
203.87
165.58
L25.9:
73
1
Mississippi Valley
Loess Plains
Total dissolved solids
mg/l
Total
55.50(
91.90
109.75
151.00
210.00
515.00
162.59
150.84
66.28
170
1
Mississippi Valley
Loess Plains
Aluminum
mg/l
Dissolved
0.010
0.01
0.01
0.05
0.20
2.47
0.15
0.06
0.30
81
1
Mississippi Valley
Loess Plains
Total Recoverable
Aluminum
mg/l
Total
Recoverable
0.045
0.11
0.28
0.71
1.27
6.91
1.01
0.55
1.31
51
1
Mississippi Valley
Loess Plains
Iron
mg/l
Dissolved
0.010
0.05
0.06
0.13
0.23
2.57
0.18
0.12
0.29
81
1
Mississippi Valley
Loess Plains
Total Recoverable
Iron
mg/l
Total
Recoverable
0.093
0.36
0.52
0.79
1.30
10.50
1.20
0.82
1.61
51
1
67
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Fraction
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Mississippi Valley
Loess Plains
Ammonia-nitrogen
mg/l
Dissolved
0.015
0.02
0.02
0.04
0.11
8.56
0.17
0.05
0.71
169
1
Mississippi Valley
Loess Plains
Nitrite + Nitrate as
Nitrogen
mg/l
Dissolved
0.005
0.02
0.03
0.14
0.31
5.06
0.26
0.11
0.45
170
1
Mississippi Valley
Loess Plains
Total Kjeldahl
nitrogen
mg/l
Total
0.224
0.48
0.62
0.73
0.97
21.00
1.05
0.82
1.67
167
1
Mississippi Valley
Loess Plains
Total Phosphorus
mg/l
Total
0.060
0.12
0.14
0.27
0.53
5.35
0.53
0.31
0.75
170
1
Mississippi Valley
Loess Plains
Orthophosphate
mg/l
Dissolved
0.010
0.03
0.06
0.13
0.29
5.35
0.38
0.15
0.70
170
1
Mississippi Valley
Loess Plains
Total suspended
solids
mg/l
Total
1.000
3.50
7.13
14.40
29.95
903.00
47.66
16.11
L19.4;
170
1
Mississippi Valley
Loess Plains
PH
NA
Total
6.400
6.92
7.21
7.48
7.81
9.33
7.49
7.48
0.48
170
1
Mississippi Valley
Loess Plains
Temperature
°C
Total
0.200
4.87
9.55
16.15
24.70
33.40
16.94
13.58
8.72
170
1
Ouachita
Mountains
Alkalinity, total
mg/l CaC03
Dissolved
0.050
3.00
8.40
14.80
29.65
190.00
22.02
14.53
20.19
5231
102
Ouachita
Mountains
Bicarbonate
mg/l
Dissolved
0.026
3.77
10.43
18.47
36.36
238.43
27.25
17.99
25.04
5084
102
Ouachita
Mountains
Chloride
mg/l
Dissolved
0.100
1.47
1.78
2.40
3.45
129.00
3.93
2.76
6.16
7396
129
Ouachita
Mountains
Hardness, Ca, Mg
mg/l
Dissolved
0.234
8.00
11.40
20.00
39.08
1400.00
39.20
22.17
85.44
4486
113
Ouachita
Mountains
Sulfate
mg/l
Dissolved
0.020
2.44
3.30
4.77
7.75
1380.00
17.66
5.79
79.79
7384
129
Ouachita
Mountains
Calcium
mg/l
Dissolved
0.010
1.30
2.20
4.97
10.90
372.00
10.41
5.06
23.72
4496
113
Ouachita
Mountains
Total Recoverable
Calcium
mg/l
Total
Recoverable
0.020
1.17
2.22
5.02
12.30
331.00
10.60
5.17
22.94
2731
94
Ouachita
Mountains
Sodium
mg/l
Dissolved
0.010
1.26
1.70
2.34
3.60
196.00
3.91
2.53
7.78
4488
113
Ouachita
Mountains
Potassium
mg/l
Dissolved
0.010
0.50
0.67
0.97
1.49
71.30
1.50
1.04
2.83
4496
113
68
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Fraction
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Ouachita
Mountains
Magnesium
mg/l
Dissolved
0.010
0.95
1.28
1.77
2.79
90.00
3.17
2.01
6.60
4498
113
Ouachita
Mountains
Specific Conductivity
US/cm
Total
2.960
27.03
38.00
57.00
105.00
6370.00
97.59
64.78
L84.92
3254
85
Ouachita
Mountains
Total dissolved solids
mg/l
Total
6.000
29.50
36.00
47.00
66.00
2100.00
69.99
52.15
L17.5"
7370
129
Ouachita
Mountains
Aluminum
mg/l
Dissolved
0.008
0.01
0.02
0.02
0.06
29.90
0.22
0.03
1.62
4498
113
Ouachita
Mountains
Total Recoverable
Aluminum
mg/l
Total
Recoverable
0.005
0.02
0.02
0.08
0.22
30.70
0.35
0.09
1.49
2731
94
Ouachita
Mountains
Iron
mg/l
Dissolved
0.003
0.02
0.03
0.06
0.15
6.19
0.12
0.07
0.22
4497
113
Ouachita
Mountains
Total Recoverable
Iron
mg/l
Total
Recoverable
0.010
0.03
0.06
0.15
0.33
7.57
0.28
0.14
0.43
2731
94
Ouachita
Mountains
Ammonia-nitrogen
mg/l
Dissolved
0.001
0.02
0.02
0.02
0.03
3.52
0.04
0.02
0.11
7372
129
Ouachita
Mountains
Nitrite + Nitrate as
Nitrogen
mg/l
Dissolved
0.003
0.01
0.03
0.07
0.19
43.70
0.31
0.07
1.53
7354
129
Ouachita
Mountains
Total Kjeldahl
nitrogen
mg/l
Total
0.015
0.05
0.11
0.20
0.34
9.05
0.29
0.19
0.35
6123
112
Ouachita
Mountains
Total Phosphorus
mg/l
Total
0.004
0.01
0.02
0.03
0.05
23.00
0.15
0.03
1.02
7223
129
Ouachita
Mountains
Orthophosphate
mg/l
Dissolved
0.003
0.01
0.01
0.01
0.02
27.52
0.13
0.02
1.04
7393
129
Ouachita
Mountains
Total suspended
solids
mg/l
Total
0.500
0.50
1.00
2.00
4.00
868.00
4.82
1.99
17.47
7406
129
Ouachita
Mountains
PH
NA
Total
3.160
6.09
6.51
6.92
7.24
14.00
6.84
6.80
0.69
7418
132
Ouachita
Mountains
Temperature
°C
Total
0.100
8.00
11.80
18.00
24.00
37.80
17.82
15.94
7.31
7480
132
Ozark Highlands
Alkalinity, total
mg/l CaC03
Dissolved
1.800
82.40
112.00
135.44
177.00
428.00
142.93
132.99
49.30
8425
172
Ozark Highlands
Bicarbonate
mg/l
Dissolved
1.614
102.49
138.91
169.43
220.06
527.42
177.90
165.55
60.99
8183
171
Ozark Highlands
Chloride
mg/l
Dissolved
0.015
2.34
3.10
4.75
9.12
900.00
9.93
5.81
19.42
13959
215
69
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Fraction
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
Ozark Highlands
Hardness, Ca, Mg
mg/l
Dissolved
0.500
93.00
122.00
145.00
186.00
2160.00
158.75
144.05
99.25
5604
189
Ozark Highlands
Sulfate
mg/l
Dissolved
0.020
3.69
4.78
6.65
10.71
109.00
9.66
7.51
8.72
13929
215
Ozark Highlands
Calcium
mg/l
Dissolved
0.010
28.90
36.90
45.10
52.30
663.00
47.53
42.95
36.11
5684
190
Ozark Highlands
Total Recoverable
Calcium
mg/l
Total
Recoverable
1.690
29.62
37.05
46.10
54.75
707.00
54.72
45.60
61.85
2103
104
Ozark Highlands
Sodium
mg/l
Dissolved
0.020
1.37
1.90
3.30
8.00
2515.00
9.17
4.01
36.88
5668
190
Ozark Highlands
Potassium
mg/l
Dissolved
0.010
0.90
1.28
1.82
3.04
47.40
2.68
1.97
2.81
5685
190
Ozark Highlands
Magnesium
mg/l
Dissolved
0.030
1.78
2.17
3.65
17.60
253.00
9.85
5.60
10.85
5685
190
Ozark Highlands
Specific Conductivity
US/cm
Total
2.000
180.00
249.00
319.00
384.00
1540.00
317.00
295.58
L09.5:
4319
128
Ozark Highlands
Total dissolved solids
mg/l
Total
6.000
125.00
156.00
190.00
224.00
7701.50
195.33
184.84
96.68
11491
200
Ozark Highlands
Aluminum
mg/l
Dissolved
0.008
0.01
0.01
0.02
0.13
1.03
0.08
0.03
0.10
5685
190
Ozark Highlands
Total Recoverable
Aluminum
mg/l
Total
Recoverable
0.005
0.02
0.02
0.05
0.10
8.40
0.16
0.06
0.55
2105
104
Ozark Highlands
Iron
mg/l
Dissolved
0.001
0.01
0.01
0.02
0.04
0.88
0.03
0.02
0.05
5685
190
Ozark Highlands
Total Recoverable
Iron
mg/l
Total
Recoverable
0.010
0.03
0.03
0.05
0.11
10.20
0.16
0.06
0.59
2105
104
Ozark Highlands
Ammonia-nitrogen
mg/l
Dissolved
0.000
0.01
0.02
0.02
0.03
5.55
0.03
0.02
0.10
13986
214
Ozark Highlands
Nitrite + Nitrate as
Nitrogen
mg/l
Dissolved
0.001
0.06
0.22
0.61
1.74
28.40
1.24
0.52
1.69
13857
214
Ozark Highlands
Total Kjeldahl
nitrogen
mg/l
Total
0.005
0.06
0.14
0.23
0.38
60.00
0.33
0.21
0.86
9149
161
Ozark Highlands
Total Phosphorus
mg/l
Total
0.003
0.01
0.03
0.05
0.11
24.62
0.23
0.06
0.71
11633
200
Ozark Highlands
Orthophosphate
mg/l
Dissolved
0.002
0.01
0.01
0.02
0.06
16.00
0.17
0.03
0.63
13962
214
Ozark Highlands
Total suspended
solids
mg/l
Total
0.500
0.50
1.00
2.50
6.00
1130.00
9.68
2.82
43.34
11489
200
Ozark Highlands
PH
NA
Total
4.580
7.28
7.60
7.90
8.14
11.93
7.85
7.83
0.45
13644
221
Ozark Highlands
Temperature
°C
Total
0.460
8.00
11.00
16.00
22.00
37.20
16.43
14.90
6.65
13956
221
South Central Plains
Alkalinity, total
mg/l CaC03
Dissolved
0.050
3.00
10.40
20.40
42.20
1040.00
34.68
19.50
43.61
9081
180
South Central Plains
Bicarbonate
mg/l
Dissolved
0.063
3.77
12.81
25.13
53.28
1306.38
43.64
24.29
55.30
8739
180
South Central Plains
Chloride
mg/l
Dissolved
0.035
2.39
3.38
5.83
18.00
2970.00
24.63
8.53
68.13
14298
195
70
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Ecoregion
Parameter
Unit
Fraction
Min
10th
25th
50th
75th
Max
Mean
GeoMean
SD
Samples
Stations
South Central Plains
Hardness, Ca, Mg
mg/l
Dissolved
0.025
11.00
18.00
27.60
48.00
1710.00
44.76
29.80
62.10
7404
193
South Central Plains
Sulfate
mg/l
Dissolved
0.020
3.20
4.94
9.33
21.30
817.00
26.41
11.35
52.55
14307
195
South Central Plains
Calcium
mg/l
Dissolved
0.010
2.69
4.60
7.55
13.50
680.00
12.70
8.00
19.63
7566
193
South Central Plains
Total Recoverable
Calcium
mg/l
Total
Recoverable
0.250
2.50
4.29
7.00
13.30
696.00
13.44
7.80
28.51
3633
165
South Central Plains
Sodium
mg/l
Dissolved
0.010
2.34
3.50
6.34
15.00
566.20
19.37
7.94
39.93
7558
193
South Central Plains
Potassium
mg/l
Dissolved
0.010
0.95
1.30
2.00
3.30
48.70
2.85
2.09
3.05
7566
193
South Central Plains
Magnesium
mg/l
Dissolved
0.010
0.99
1.37
1.99
3.12
64.10
3.11
2.11
4.37
7566
193
South Central Plains
Specific Conductivity
US/cm
Total
5.350
41.49
60.00
93.65
159.00
1760.00
160.73
104.85
!08.9(
4390
129
South Central Plains
Total dissolved solids
mg/l
Total
7.000
45.50
62.00
86.50
148.00
5231.00
150.69
103.85
L88.5;
14362
195
South Central Plains
Aluminum
mg/l
Dissolved
0.008
0.01
0.03
0.06
0.12
1.51
0.10
0.06
0.11
7566
193
South Central Plains
Total Recoverable
Aluminum
mg/l
Total
Recoverable
0.005
0.07
0.16
0.32
0.63
14.10
0.50
0.31
0.65
3633
165
South Central Plains
Iron
mg/l
Dissolved
0.001
0.05
0.13
0.28
0.52
7.46
0.40
0.24
0.43
7566
193
South Central Plains
Total Recoverable
Iron
mg/l
Total
Recoverable
0.025
0.30
0.57
0.95
1.45
12.50
1.16
0.86
0.95
3633
165
South Central Plains
Ammonia-nitrogen
mg/l
Dissolved
0.001
0.02
0.02
0.03
0.07
151.50
0.18
0.04
1.92
14270
195
South Central Plains
Nitrite + Nitrate as
Nitrogen
mg/l
Dissolved
0.001
0.02
0.06
0.15
0.28
211.00
0.71
0.13
3.66
14299
195
South Central Plains
Total Kjeldahl
nitrogen
mg/l
Total
0.002
0.24
0.34
0.53
0.80
64.64
0.71
0.53
1.17
11628
191
South Central Plains
Total Phosphorus
mg/l
Total
0.005
0.03
0.04
0.07
0.13
16.50
0.17
0.08
0.51
14036
195
South Central Plains
Orthophosphate
mg/l
Dissolved
0.002
0.01
0.01
0.02
0.05
9.97
0.10
0.03
0.44
14291
195
South Central Plains
Total suspended
solids
mg/l
Total
0.500
1.80
3.50
7.00
14.80
3232.00
17.88
7.49
49.66
14377
195
South Central Plains
PH
NA
Total
0.650
6.05
6.45
6.85
7.20
10.05
6.82
6.79
0.63
14035
195
South Central Plains
Temperature
°C
Total
1.000
8.50
12.10
18.30
25.00
39.00
18.43
16.71
7.32
14048
195
71
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Table A.2.5. Summary station median statistics for relative cation dominance based on mg/1 least disturbed stations ([Ca2+]
+ [Mg2+])/([Na+] + [K+]). Molar ratio or microequivalent ratios would differ.
Centile
Ecoregion
Percentage"
Min
10
25
50
75
90
Max
Stations
Arkansas Valley
67
0.94
0.95
0.97
1.05
1.58
1.69
2.16
18
Boston Mountains
100
1.02
1.37
1.72
2.05
3.90
5.35
9.69
34
Mississippi Alluvial
Plain
100
1.64
1.94
2.27
2.59
11.54
19.70
20.43
7
Ouachita Mountains
95
0.86
1.43
2.14
4.56
6.96
9.57
15.03
40
Ozark Highlands
100
1.86
3.99
8.33
15.17
23.40
38.74
156.79
70
South Central Plains
43
0.39
0.73
0.82
0.97
1.13
2.95
6.15
28
([Ca2+] + [Mg2+])/([Na+]
+ [K+])>1
Table A.2.6. Summary station median statistics for relative anion dominance based on mg/1 least disturbed stations
([HCOs] + [SO42 ])/[Cl ]).
Molar ratio or microequivalent ratios would differ.
Centile
Ecoregion
Percentage"
Min
10
25
50
75
90
Max
Stations
Arkansas Valley
100
2.72
3.47
4.06
4.55
7.42
8.70
10.88
18
Boston Mountains
100
4.46
6.28
8.69
11.75
19.61
24.81
44.13
35
Mississippi Alluvial
Plain
100
4.87
5.21
7.68
13.01
42.44
70.35
72.18
7
Ouachita Mountains
100
4.79
6.50
10.10
18.20
27.52
35.55
40.13
39
Ozark Highlands
100
7.72
16.71
34.81
50.82
77.96
103.40
203.09
59
South Central Plains
96.43
0.70
1.91
2.45
3.92
9.95
19.98
28.13
28
a([HC03"] + [S042-])/[Cl-]) > 1
72
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
A-3. Temporal comparisons of predicted and observed SC
The predictive performance of the empirical model (Olson and Cormier, 2019) was assessed by
comparing long-term predicted and observed SC at USGS gaging stations that had multiple SC
measurements. The following factors were considered:
(1) the coincidence of seasonal variation,
(2) the range of seasonal variation (low variation is associated with little or no anthropogenic
loadings or continuous anthropogenic inputs),
(3) magnitude of difference between median and minimum observed and predicted values,
and
(4) statistics on the general performance of the model in the ecoregion reported in Olson and
Cormier (2019).
We identified two long-term records in the Arkansas Valley, five in the Boston Mountains and
four in the Mississippi Alluvial Plain.
The two records for the Arkansas Valley coincide with the seasonal variation, but the empirical
model over-predicts background SC (USGS-07257500, USGS-07261090).
In the Boston Mountains, three records did not appear to represent background, owing to either
high SC or a large SC range—indicative of anthropogenic inputs. One had a median background
of twice the predicted background and was located in an area of mixed land use (USGS-
07048495). Downstream from that location at another gaging station (USGS-07048550), the
observed median was within 12 |iS/cm of the predicted value, but the annual monthly range was
>100 to 200 |iS/cm, indicative of source inputs or influence from the upstream location. A third
station occasionally exhibited large annual monthly observed SC (USGS- 07048600), indicative
of inputs. Two gage stations appeared to exhibit minimally affected conditions in the Boston
Mountains. They were located in forested uplands and had low SC and modest monthly changes
in SC (USGS-07075250; 07075270), indicative of background SC regimes; but predicted SC
was at least 10X greater than observed SC, indicative that the model overestimates background
SC.
Inspection of these example records comparing predicted SC and observed SC indicate that the
Arkansas Valley (Figure A.3.1) and Boston Mountains (Figure A.3.2) predicted background
estimates are consistently greater than observed values. Therefore, where background is less than
the ecoregional background, observations at the station itself best represent minimally affected
or least disturbed background. For stations with no observed values, background may be best
represented by the observed ecoregional least disturbed background estimate.
For the Mississippi Alluvial Plain (Figure A.3.3), the predicted least disturbed background
appears to be reliable without calibration. Using a different least disturbed background estimate
for the Mississippi Alluvial Plain is not justified by these data.
73
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
07261090
2014
2015
2011
2012
2013
2014
2015
2016
Date
250-
200-
Type
Observed
Predicted
Figure A-3.1. Arkansas Valley, USGS gaging stations and predicted SC.
The monthly averaged predicted SC coincides with the seasonal variation, but over-predicts SC.
74
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
07048495
rt li
¦
350-
* "i i i,«
400-
I ill" I ! *; 1
¦
300-
350-
y !
¦
250-
300-
250-
« % i
It f \ f * #
¦
4 4
200-
150-
2018 2017 2013 2019 2020 2021
07048600
o
o
o
c
ra
o
D
¦o
c
o
o
o
o
CD
Q.
C/3
250-
200-
150-
100-
50-
2016 2017 2018 2019 2020 2021
07075270
400-
350-
300-
250-
200-
150-
100-
50-
0-
it |if f#
%|§ %|
Ig! l|it
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
FRA0012
FRA0013
3
"D
C
o
350
o
o
(D
(f) 300
200
100
WHI0004
2015 2016 2017 2018 2019 2020 2021
700
650
600
550
500
450
400
350
300
250
200
150
100
50
2015
Date
WHI0074
2016 2017 2018 2019 2020 2021
900-
850-
800-
750-
700-
650-
600-
550-
500-
450-
400-
350-
___ 300-
£ 250-
200-
3 150"
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
A.4. Histograms of number of stations and SC
The distribution and range of SC can influence the estimation of XC95 values and XCD05
values. For the entire Arkansas data set (Figure A.4.1) and the Group 1 (Figure A.4.2) paired
data set, the range is broad, with some stations between 500 and 1,000 jiS/cm. The number of
samples and the range is more restricted in Group 2 (Figure A.4.3).
35 H
30 -
25 -
20 -
15
10
5 -
0 -
d.
n 1 r~
50 100 500
Specific Conductivity, pS/cm
1000
Figure A.4.1. Distribution and range of SC Arkansas data set.
77
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
25 -
20 -
15 -
10 -
5 -
0 -
Q.
n.
50
100
500
1000
Specific Conductivity, MS/cm
Figure A.4.2. Distribution and range of SC Group 1 data set.
78
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
25 -
20 -
c 15 -
3
O
O
10 -
5 -
—! ! 1 1 1
100 200 300 400 500
Specific Conductivity, pS/cm
Figure A.4.3. Distribution and range of SC Group 2 data set. The SC range is marginal for
estimating XC95 values. Most of the XC95 values were characterized as greater than the
estimated value. The permutation test for the XCD05 was also weak and the critical value
overlapped the upper 95% Confidence limit (Figure A.5.3).
79
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Extirpation Levels Based on Field-Based Method
This section describes the estimation of 5% extirpation using the field-based extirpation
distribution (XCD05) approach for two combined groups of ecoregions. Ecoregions were
grouped together to increase sample size and range of exposures (USEPA, 2011, 2016; Cormier
et al., 2020). These results are not recommended except as a screening tool because the scale
was coarse and not representative of local background SC conditions. However, the XC95
values for individual genera may be useful for causal assessments.
Data sets for field-based method
To increase the number of stations in the data set, both TDS and SC water chemistry
measurements were accepted. At stations where TDS data were available, but SC data were not,
SC was estimated using regression models (Table 4 and Figure 4). Then, biological records were
paired with SC at stations with the nearest spatial and temporal proximity and within two
kilometers and sampled within 30 days to maximize the data set. Ecoregions were grouped into
two larger regional data sets for subsequent analysis. Ecoregion Group 1, a lower background
SC area, consists of the Arkansas Valley, Boston Mountains, Ouachita Mountains, and South
Central Plains (Table A.5.1). Ecoregion Group 2, a higher expected background area, consisted
of the Ozark Highlands, the Mississippi Alluvial Plain, and the Mississippi Valley Loess Plains
(Table A.5.1). Merging ecoregions also helped to increase the sample size. XC95 values were
calculated for genera with a minimum of 25 occurrences.
Field-based method application
Analyses were performed using paired data for the entire state of Arkansas, and Ecoregion
Groups 1 and 2. The 5th centile extirpation values (XCD05) were calculated in a two-step
process following a field-based method using an extirpation concentration distribution (XCD)
(USEPA, 2011b; Cormier and Suter, 2013; Cormier et al., 2013). First, the extirpation
concentration (XC95) values for each genus with >25 occurrences were calculated as the 95th
centile of a weighted cumulative frequency distribution (CFD) of SC levels at sites where a
genus had been collected. Then, a frequency distribution of XC95 values was constructed, and
the 5th centile was identified from the CFD by a log linear 2-point interpolation.
Confidence of individual XC95 values was assessed using generalized additive models (GAM)
with 3 degrees of freedom that estimate the probability of a genus being observed with
increasing SC. The probability of observing a genus is the percentage of sampled stations in a
given SC bin of stations with the genus present. The uncertainty bounds of a GAM indicate the
confidence in the calculated XC95 and whether the value is greater than the observed SC range
(USEPA, 2011). Confidence in the XC95 was qualitatively scored as follows. If the GAM mean-
fitted curve at maximum SC was approximately equal to zero (defined as less than 1% of the
maximum modeled probability), then the XC95 was listed without qualification; otherwise, the
XC95 was designated as approximate or greater than the calculated value. If the generalized
80
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
additive model mean curve at maximum SC was > 0 but the lower confidence limit
approximated 0 (<1% of the maximum mean modeled probability), then the value was listed as
approximate. If the generalized additive model lower confidence limit was > 0, then the XC95
was listed as greater than the 95th centile.
Confidence in the XCD05 was evaluated using two simulation methods (Cormier et al., 2020).
(1) Bootstrapping with replacement was used to estimate the 95% confidence interval for XC95
and XCD05 values. (2) A permutation test was used to estimate the probability that the XCD05
values could have arisen by chance. The permutation test evaluated whether the data set was
large enough and the sample distribution covered a wide enough range of SC levels to provide a
defensible 5% extirpation value. The permutation test simulates conditions where the SC has no
influence on occurrence of a genus. Confidence in the observed XCD05 value depends upon the
lack or degree of overlap with the permutated XCD05 values. Details for both methods can be
found in Cormier et al. (2020).
For bootstrapping, in the paired biological and SC data sets for Arkansas and Ecoregions Groups
1 and 2, samples were randomly selected with replacement from the original set of samples.
Next, the XC95 values were calculated for each genus in the bootstrapped data set by the same
methods applied to the original data, and the XCD05 was calculated. The process was repeated
to generate 1,000 bootstrapped data sets. Two-tailed 95% confidence bounds were generated for
these bootstrap-derived XC95 values and the distribution of the 1,000 XCD05 values.
The permutation test evaluated whether the XCD05 values could have occurred by chance.
XC95 values were recalculated using the observed sample sizes for the 86, 72 and 24 genera in
Arkansas, and Ecoregion Groups 1 and 2, respectively—as if they occurred randomly, with
respect to SC, across the sites in the original three data sets. Genera met the inclusion criteria of
> 25 occurrences in a data set. This randomization process was repeated 1,000 times, generating
XC95 values for each genus and 1,000 sensitivity distributions. Thus, each of the new 1,000
permutated data sets maintained the number of occurrences of a genus but randomized their SC
exposure. The results are shown in Figures A.5.1, A.5.12, and A.5.3. Also, to estimate the
probability that the XCD05 values could have arisen by chance, we fitted the 1,000 permutation
XCD05 values to a normal distribution in each run. Then, the probability of an observed XCD05
value occurring by chance was calculated based on centiles of the fitted normal distribution. The
5th centile of the 1,000 permutated XCD05 values corresponds to the 1-tailed critical value that
defines the XCD05 value that may have occurred by chance, with an alpha of 0.05.
Field-based XC95 and XCD05 values results
XC95 values and CFD and GAM plots are available from the authors at cormier.susan@epa.gov.
The Arkansas and Ecoregion Group 1 and 2 XCD05 values pass the permutation test, but due to
the low number of genera and modest number of stations, the estimate for Ecoregion Group 2 is
much less confident (Figures A.5.1, A.5.2, A.5.3, and Table A.5.1). The XCD05 for Group 2 is
only 10 |iS/cm less than the critical value for accepting the hypothesis that the XCD05 could
81
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
have arisen by chance. Also, the two genera with the lowest XC95 values in Ecoregion Group 2
met the minimum sample size for inclusion, i.e., N > 25, but their XC95 values are ambiguous
based on uncertainty bounds of their GAM plots as indicated as triangles in the XCD plot
(Figure A.5.3).
For the Arkansas data set, the observed XCD05 is 156.2 [j,S/cm with a two-tailed 95% CI of
97.5-195.94 [j,S/cm (Figure A.5.1). The 5th centile of the permutation values is 428.24 [j,S/cm.
This corresponds to the 1-tailed critical value that defines the upper limit of the rejection region
(alpha = 0.05). Therefore, for an observed XCD05 value of <428.4 [j,S/cm, the hypothesis that
the XCD05 is not associated with SC is rejected.
For the Ecoregion Group 1 data set, the observed XCD05 is 125.5 [j,S/cm with a two-tailed 95%
CI of 72.7-170.54 [j,S/cm (Figure A.5.2). The 5th centile of the permutation values is 382 [j,S/cm.
This corresponds to the 1-tailed critical value that defines the upper limit of the rejection region
(alpha=0.05). Therefore, for an observed XCD05 value of < 380 [j,S/cm, the hypothesis that the
XCD05 is not associated with SC is rejected.
For the Ecoregion Group 2 data set, the observed XCD05 is 403.5 [j,S/cm with a two-tailed 95%
CI of 390.7-455.8 [j,S/cm (Figure A.5.3). The 5th centile of the permutation values is 413.2
[j,S/cm. This corresponds to the 1-tailed critical value that defines the upper limit of the rejection
region (alpha=0.05). Therefore, for an observed XCD05 value of < 413.2 [j,S/cm, the hypothesis
that the XCD05 is not associated with SC is rejected. However, confidence is not strong because
the 95% CI of the observed XCD05 from the bootstrapping procedure (upper 95% CL=457.8
(j,S/cm) overlaps with the 5th centile permutation value (413.2 (j,S/cm). Furthermore, there were
only 24 genera and most of their XC95 values were undefined, that is greater than the estimated
XC95 (USEPA 2011, 2016).
Summary: Field-based Extirpation Estimates
The derivation of a field-based SC benchmark using the currently available paired biological and
chemical data are a reasonable XCD05 estimate for Arkansas as a whole and for Ecoregion
Group 1. The XCD05 value for Ecoregion Group 2 is best considered as a screening benchmark
because the statistical test for confidence was weaker, probably because the SC exposure range
was too narrow and the number of genera were so few. Group 1 and Group 2 data sets had only
72 and 24 genera, respectively. For greater confidence, a data set that yields around 90 genera or
species gives a more consistent XCDC05 (Cormier et al., 2020). The number of genera may be
increased either with identification of all individuals in fewer sites or with fewer individuals
identified from more sites (-500 samples). The data set could be increased by using data from
the entire ecoregion outside of Arkansas or collected by other entities.
For all three XCD05 estimates, there is potential unmeasured variation due to the necessity to
match water chemistry samples with macroinvertebrate data that were not collected at the same
river mile or date. As a result, the water chemistry at a station may not be optimally matched
with the biological sample in space or time. For example, a low SC measurement may be
82
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
upstream of a point source and the biological data downstream of a source where the SC could
be high. Also, biology may have been obtained from one tributary and chemistry from a nearby
tributary or main stem.
The XCD05 values are not recommended because the B-C method provided comparable results
and had the benefit of being at the stream segment scale for most of Arkansas, a more reliable
scale than the grouped ecoregions.
—i i | i i i i |
100 500 1000
Specific Conductivity, (jS/cm
Figure A.5.1. Benthic invertebrate genus extirpation concentration distributions (XCD) in
Arkansas. XCD05 = 156.2 (95% CI 97.5-195.9) (iS/cm. Each open circle or triangle is an
observed extirpation concentration (XC95) value for one genus forming an XCD. Approximate
or greater-than XC95 values are designated by triangles. The observed 5% extirpation level
(XCD05) is the SC at the intersection of the solid horizontal line at the 5th centile. The XCDs
simulate when there is no influence of SC by randomly shuffling the occurrence of taxa among
stations independent of actual SC. A thousand permutated simulated XCDs are the solid gray
lines on the right. The critical value is shown at the intersection of the dashed horizontal line at
the 5th centile. There is a clear delineation between the observed XCD05 (156.2 (iS/cm) and the
permutation test critical-XCD05 value (428.24 |j,S/cm, alpha = 0.05), indicating that the
observed XCD05 did not arise by chance.
83
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
100 500 1000
Specific Conductivity, (jS/cm
Figure A.5.2. Bent hie invertebrate genus extirpation concentration distributions (XCD) for
Region 1. XCD05 = 125 (95% CI 72.7-1670.5) jiS/cm. Each open circle or triangle is an
observed extirpation concentration (XC95) value for one genus forming an XCD. Approximate
or greater-than XC95 values are designated by triangles. The observed 5% extirpation level
(XCD05) is the SC at the intersection of the solid horizontal line at the 5th centile. The XCDs
simulate when there is no influence of SC by randomly shuffling the occurrence of taxa among
stations independent of actual SC. A thousand permutated simulated XCDs are the solid gray
lines on the right. The critical value is shown at the intersection of the dashed horizontal line at
the 5th centile. There is a clear delineation between the observed XCD05 (125 (iS/cm) and the
permutation test critical-XCD05 value (382 }j,S/cm, alpha=0.05), indicating that the observed
XCD05 did not arise by chance.
84
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
«.•
c
0)
<3>
T3
d)
+-»
ro
Q.
O
C
o
+-»
o
Q.
O
1.0 -
0.8 -
0.6 -
0.4 -
0.2 -
0.0 -
i i i i i i i i i i i i—i i i i i i—i—i i , i . i i i i i i i i i i i i i i i i i i
400
500
Specific Conductivity, |jS/cm
Figure A.5.3. Bent hie invertebrate genus extirpation concentration distributions (XCD) for
Region 2. XCD05 = 403.5(95% CI 390.7-455.8). Each open triangle is an observed extirpation
concentration (XC95) value for one genus; most are greater-than XC95 values and therefore not
confidently assigned. The observed 5% extirpation level (XCD05) is the SC at the intersection of
the solid horizontal line at the 5th centile. The XCDs generated by permutation simulating no
influence of SC by randomly shuffling of taxa occurrences are solid gray lines on the right. The
critical value is shown at the intersection of the dashed horizontal line at the 5th centile.
Observed XCD05 (403.5 jiS/cm) and the permutation test critical-XCD05 value (413.2 }j,S/cm,
alpha=0.05) are statistically different, but discrimination is not as strong as in Region 2. Number
of genera is small; thus, the uncertainty in the XCD05 value is greater, as indicated by the
overlap of the observed and simulated XCDs and the small difference between the observed and
simulated XCD05s.
85
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Table A.5.1. XCDos values in different regions, sample sizes, number of genera, XCD 95%
confidence intervals, and XCD05 critical values.
SC = specific conductivity
USEPA Technical Document
2016a
Present study
N
Samples
N
Genera
XCD05o5
SC
(|iS/cm)
N
Samples
N
Stations
N
Genera
XCD05os SC
(|iS/cm)
(95% CI).
[Critical valuel
Entire
ADEQ
paired
dataset
380
64
328.5
465
198
86
156.2
(97.5-195.9)
[428.2]
ADEQ
Region 1
(35, 36, 37,
38)
187
31
204
319
121
72
125.0
(72.7-170.5)
[382.0]
ADEQ
Region 2
(39, 73, 74)
193
27
358
146
77
24
403.5
(390.7-455.8)
T413.2]
aUSEPA, 2016. Final Report: EPA Technical Support: Evaluation of Several Approaches to
Develop Mineral Criteria in Arkansas. February 2016. Pp. 85.
86
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Map A.6. Maps of Arkansas
Map A.6.1. Oil and Gas wells 2017. Active (green circles), Inactive (light green), Permitted
(orange), Spud (read). Blue lines are stream network. South Central Plains (35), Ouachita
Mountains (36), Arkansas Valley (37), Boston Mountains (38), Ozark Highlands (39),
Mississippi Alluvial Plain (73), Mississippi Valley Loess Plains (74). Source:
https://services.arcgis.com/iDGu08tYggdCCnUJ/arcgis/rest/services/Arkansas Oil Gas Map/F
eature Server
87
-------
Background Specific Conductivity and
Associated 5% Extirpation Estimates in Arkansas
Map A.6.2. Geologic Map of Arkansas
Haley BR. and Arkansas Geological Commission staff. 1993. Geologic Map of Arkansas.
Download map at: https://www.geologv.arkansas.gov/maps-and-data/geologic maps/geologic-
map-of-arkansas-1993-revised-from-1976-edition.html
88
------- |