HRL-SAC-Chla-Criteria-Rec

A Chlorophyll a Criterion for High Rock Lake

by the

North Carolina Nutrient Criteria Science Advisory Council

(Marcelo Ardon, Clifton Bell, James Bowen, Linda Ehrlich, Nathan Hall, William Hall,
Martin Lebo, Michael O’Driscoll, Deanna Osmond, Hans Paerl, Lauren Petter, Astrid Schnetzer)

May 26, 2020

Table of Contents

Executive Summary ....................................................................................................................... 1

1. Introduction ............................................................................................................................. 4

2. Literature Review of Chl a and Use Attainment ................................................................ 11

3. Current Conditions in High Rock Lake .............................................................................. 24

4. A Proposed Site-Specific Chlorophyll a Criterion for High Rock Lake .......................... 58

5. Potential Elements of a Framework for Deriving Site-Specific Criteria .......................... 76
1

Executive Summary

This document provides a description and technical background for a site-specific chlorophyll a
(chl a) criterion for High Rock Lake, North Carolina, a freshwater reservoir in the Yadkin-Pee
Dee river basin of North and South Carolina. The work by the North Carolina Science Advisory
Council (SAC) to establish this criterion is part of larger effort in North Carolina to develop
nutrient criteria throughout North Carolina on a site-specific basis for three separate water body
types: 1) reservoirs/lakes, 2) rivers/streams and 3) estuaries. The existing numeric chl a criterion
of 40 µg/L is assessed on a “not-to-exceed” basis as part of a narrative standard for lakes,
sounds, estuaries, reservoirs, and other slow-moving waters not designated as trout waters. The
criterion is exceeded when there is greater than a 90% statistical confidence that more than 10%
of samples will exceed a 40 µg/L photic zone average concentration. The efficacy of applying a
single chl a criterion to protect the wide variety of surface water habitats in North Carolina has
been debated, and development of site-specific Chl a criteria have been promoted by the US EPA
and authorized for North Carolina by the Environmental Management Commission.

This newly-developed, site-specific Chl a criterion has been developed according to a process
that considered the designated uses (aesthetics, water supply, aquatic habitat, and recreation) of
High Rock Lake. The criterion was developed to protect these designated uses. Multiple lines of
evidence (e.g. literature review, water quality monitoring results, assessments of designated use
attainments) were used to determine the appropriate chl a concentration, its averaging period,
and the frequency of criterion exceedance that would be protective of the designated uses.

The literature review found that increases in chl a concentration decrease water clarity and
correlate strongly with increasing primary production in phytoplankton dominated
systems. Freshwater fisheries production generally responds positively to increases in chl a. An
upper threshold exists, however, to the positive relationship between chl a and overall fisheries
production. At chl a levels beyond the threshold, negative impacts of excessive algal production
on water quality (e.g. dissolved oxygen concentrations, water clarity) may reduce fish
production, or cause substantial shifts toward less desirable fish species. Higher chl a values
may also increase risks from phytotoxins. Several genera of bloom-forming cyanobacteria can
produce a potent suite of secondary metabolites that are hepatotoxic and neurotoxic and can
harm aquatic life. There is not a simple relationship between chl a and toxin concentration.
Despite a considerable literature on phytotoxins in lakes, given current information available, the
SAC does not advise establishing chl a standards based solely on cyanotoxin risk to aquatic life.

The SAC reviewed water quality monitoring studies conducted by a number of research groups
in High Rock Lake from 1973 – 2016. Designated use assessments were also reviewed. Based
on nutrient and chl a concentrations, previous studies have consistently characterized High Rock
Lake as a eutrophic reservoir. The lake has been considered to be like many “run-of-the-river”
reservoirs that have distinct riverine, transitional, and lacustrine zones. Chl a concentrations
were generally highest in the transitional zone of the lake and have frequently exceeded the
existing 40 µg/L chl a standard. There are no clear long-term trends in chl a concentration.
2

Data on other indicators of water quality such as dissolved oxygen (DO), pH, water clarity, algal
abundance, and phytotoxin concentration were also reviewed. Chl a concentrations in High
Rock Lake are correlated with relatively high DO surface concentrations, mixed effects on
bottom DO concentrations, and relatively high DO percent saturation values. The reservoir
attains water quality criteria for DO under existing chl a conditions. The pH of surface waters in
High Rock Lake (<0.2 m) was found to be highest during the months of June through
September. Measured pH exceeded 9.0 in 24-38% of the measurements. Exceedances of a pH of
9.0 occurred over the entire range of chl a values, but were more common when chl a exceeded
30 μg/L. The water clarity in High Rock Lake, based upon the most recent assessment using
turbidity measurements, is considered impaired in the upper riverine portion of the lake. Algal
abundances and taxonomy were found to vary seasonally in a fashion typical for temperate
eutrophic reservoirs with summer maxima and winter minima. In-situ phytotoxin tracking
devices deployed as part of a special sampling study in 2016 indicated that microcystin,
anatoxin, and cylindrospermopsin were present throughout much of the summer in High Rock
Lake and were often detected simultaneously. Bulk water analysis indicated that toxin
concentrations were below action limits or health advisory concentrations.

Based on assessments made by the NC Wildlife Resources Commission, current water quality
conditions appear to be supportive of a sport fishery focused on largemouth bass, striped bass,
and crappie, sunfish, and catfish. The largemouth fishery has been consistently evaluated as a
“quality fishery” sustained by adequate recruitment and non-excessive mortality. Fish kills are
uncommon in HRL, and large fish kills have only been noted during the major drought of 2002
when low flows, low water levels, high summer temperatures, and low dissolved oxygen caused
major fish kills. The SAC is not aware of any aesthetic or swimming use impairment of the lake,
even though chl a concentrations routinely exceed 50 µg/L.

The SAC used a literature review and the reservoir-specific water quality and use assessment
observations to develop the recommended site-specific chl a criterion. The proposed chl a
criterion for High Rock Lake is a seasonal geomean of 35 µg/L, not to be exceeded more than
once in three years, for growing season months of April-October based on protection of all uses
while maintaining the productivity of the sport fishery. In terms of spatial considerations, all
monitoring data from open waters within assessment units collected during the months of April
through October would be used to compute a geomean to compare with the proposed
criterion. The criterion would apply to all months of the year, with attainment of the standard
assessed with data from the growing season months. The SAC recommended maximum
exceedance frequency is not to exceed more than one in three calculated seasonal geomean
values.

The SAC also considered how lessons learned from the reservoir pilot might inform a statewide
framework for deriving lake-specific chl a criteria. Such a framework should produce criteria
that minimize both type I (false finding of use impairment) and type II (false finding of use
attainment) errors. Several framework elements could streamline the criteria development
process while still making use of both the scientific literature and lake-specific information.
3

These elements include: (1) using similar duration and frequency components as recommended
for the lake pilot; (2) a chlorophyll a screening range to inform lake use attainment status; (3) a
predetermined list of numeric and narrative indicators of use attainment; and (4) decision
guidelines for translating lake evaluation results into site-specific criteria. The SAC and DEQ
could revisit these concepts during the statewide criteria development phase of the NCDP.

4

1. Introduction

As described in the North Carolina Nutrient Criteria Development Plan (NCDP), (NCDWR
2014) and its revised version (NCDWR 2019), North Carolina is working towards developing
scientifically defensible numeric nutrient criteria throughout the state on a site-specific basis.
According to the plan, numeric nutrient criteria will be developed initially for one example each
of three distinct water body types. The water bodies and the water body types are as follows:

1.0 High Rock Lake (reservoirs/lakes)
2.0 Central portion of the Cape Fear River (rivers/streams)
3.0 Chowan River/Albemarle Sound (estuaries)

An important component of the NCDP has been the creation of a twelve-member scientific
advisory council (SAC) to advise and assist the North Carolina Division of Water Resources
(NCDWR) in the development of numeric nutrient criteria. This document represents the work
of the SAC done with the cooperation and assistance of the NCDWR. In this document the SAC
provides a description, a rationale, and technical background for site-specific chlorophyll a (chl
a) criterion for High Rock Lake, North Carolina, a freshwater reservoir in the Yadkin-Pee Dee
river basin of North and South Carolina.

High Rock Lake is a freshwater reservoir in the piedmont region of North Carolina. It is a
15,180-acre reservoir with a 3,974 mi2 drainage area located within the upper portion of the
Yadkin River basin (Figure 1.1). It is the first of a chain of four lakes (High Rock, Tuckertown
Badin, and Falls) that were created between 1917 and 1962 by Alcoa to provide hydroelectric
power for aluminum production (Cube Hydro Carolinas 2019).

According to a 2004 review of water quality data (Tetra Tech 2004), High Rock Lake has been
characterized as eutrophic since the 1970’s. EPA assessed water quality conditions in 1973 in
sixteen North Carolina lakes as part of a national eutrophication survey (USEPA 1975), finding
High Rock Lake to be the most eutrophic of the North Carolina lakes studied. At the time, EPA
noted that High Rock Lake’s variable but relatively short residence time (estimated at 27 days
for mean flow) produced a lake that operates more like a slow-moving river than a typical lake.

Tetra Tech summarized several additional water quality assessments in ensuing years that have
each shown High Rock Lake to have relatively high levels of turbidity, nutrients, and
phytoplankton abundance (i.e. high chl a concentration) (Tetra Tech 2004). High Rock Lake is
currently on North Carolina’s list of impaired or threatened waters as required under Section
303(d) of the Clean Water Act. Based upon the current numeric chl a criterion, the entire lake is
impaired for chl a and parts of the lake are impaired for pH and turbidity. Additional
information on the current numeric North Carolina chl a criterion, and policies for listing and
delisting waterbodies as impaired is provided in the following section of this chapter.

5

The work of the SAC on a new numeric nutrient criterion for High Rock Lake has had multiple
objectives. While the immediate, primary objective of the work has been to develop a site-
specific criterion for High Rock Lake, a secondary objective has been to develop a methodology
for criteria development that can be applied to other North Carolina lakes and reservoirs, and
perhaps to other water body types within the state. The final section of this chapter describes the
general approach that the SAC has used to develop a site-specific chl a criterion for High Rock
Lake. One aspect of the approach is to utilize the scientific literature as a basis for the site-
specific criterion. A review of the relevant literature relating important eutrophication response
variables such as water clarity and chl a concentrations to relevant designated uses such as
aesthetics, water supply, aquatic habitat, and recreation is provided in chapter 2. Chapter 3 then

Figure 1.1. High Rock Lake and watershed. (figure taken from the North Carolina Nutrient Criteria
Development Plan (NCDWR 2019))
6

looks specifically at the extent to which designated uses in High Rock Lake are supported given
the current water quality conditions. Chapter 4 then describes the proposed site-specific numeric
chl a criterion for High Rock Lake. The concluding chapter of this document (Chapter 5) then
returns to the larger task of developing numeric nutrient criteria for all the lakes and reservoirs in
North Carolina. The chapter proposes elements of a framework that the SAC believes could be
the basis for a general approach for developing site specific nutrient criteria across the range of
water body types in North Carolina.

1.1 Description of the Current North Carolina Chl a Criterion

As described in Division of Water Resources’ (DWR) May 2017 chl a description document (NC
Division of Water Resources 2017), the existing chl a criterion “arose through an advisory group
process and was informed by lake and reservoir research including the 1976 report by Charles
Weiss and Edward J. Kuenzler ‘The Trophic States of North Carolina Lakes (Weiss and
Kuenzler 1976) .’” The current approved regulatory text for the State’s chl a criterion, located at
15A NCAC 02B .0211(4), states:

Chlorophyll-a (corrected): not greater than 40 µg/l for lakes, reservoirs, and other
waters subject to growths of macroscopic or microscopic vegetation not designated as
trout waters, and not greater than 15 µg/l for lakes, reservoirs, and other waters subject
to growths of macroscopic or microscopic vegetation designated as trout waters (not
applicable to lakes or reservoirs less than 10 acres in surface area). The Commission or
its designee may prohibit or limit any discharge of waste into surface waters if the
surface waters experience or the discharge would result in growths of microscopic or
macroscopic vegetation such that the standards established pursuant to this Rule would
be violated or the intended best usage of the waters would be impaired; (Emphasis
added)

The 2017 Summary and 1976 report characterize 40 µg/L as the “upper range for alpha-
eutrophic waters (15µg/l to 40 µg/L).” Weiss and Kuenzler indicated that the scale of quality is
an interpretation that is not about whether the water should or should not be used; but rather the
interpretation that some attributes of more eutrophic waters are more acceptable – plenty of fish
– while others are less acceptable – swimming in algal blooms. The 2017 Summary also
references excerpts from another historical document, the records provided by Mike McGhee,
former EPA and NC DEM employee (McGhee 1983). The comments included in those notes
point to the importance of a chl a criterion to limit point and nonpoint discharges of nutrients,
including nitrogen and phosphorus.

For additional information, excerpts (shown in italics) from the NCDWR 2018 303(d)
listing/delisting procedures document are included. (NCDWR 2018). The excerpts summarize
how the state completes assessments of the existing chl a criterion based on collected ambient
data for determining whether a waterbody should be listed on the North Carolina Section 303(d)
7

list. The flowcharts (Fig. 1.2) for listing and delisting waters when assessing numeric criteria are
also provided for reference.

ASSESSING CHLOROPHYLL-A NUMERIC CRITERIA
The following sets of evaluations will be used for the 2018 assessment for these parameters:
chlorophyll-a, dissolved oxygen, MBAS, mercury, nitrate/nitrite, pH, temperature, toxic
substances, and turbidity. For each parameter there is a brief discussion of the standard used for
assessment of the parameter including any parameter-specific good causes for not assessing in
Category 5. Note Category 5 is the 303(d) list.

The true frequency of criteria exceedances cannot be measured. It must be estimated from a set
of samples, which introduces statistical uncertainty. The degree of uncertainty depends on the
sample size. NC will use a nonparametric hypothesis testing approach based on the binomial
distribution. The binomial method allows a quantifiable level of statistical confidence (90%) for
listing decisions, which provides a 10% probability of listing an assessment unit when it should
not be listed. The null hypothesis is that the overall exceedance probability is less than or equal
to the 10% exceedance allowance. NC will also consider the number of excursions of criterion
for newer data that have not been assessed before. For 2018 assessment, newer data are defined
as data collected during calendar years 2015 and 2016.

Exceeding Criteria-Category 5
● Sample size is greater than nine.
● Greater than 10% exceedance with greater than or equal to 90% confidence,
or
● Greater than 10% exceedance, but less than 90% statistical confidence, and
at least 4 excursions in newer data that have previously not been assessed.

DELISTING WATERS

NC will review the final 2016 303(d) list as the starting point for the development of the 2018
303(d) list. All waters on the 2016 303(d) listing will be evaluated for appropriate inclusion on
the 2018 303(d) list as defined in 40 CFR 130.2(j). NC will apply a combination of
nonparametric hypotheses testing based on the binomial distribution as well as an analysis of the
dates of excursions to determine if there is good cause to delist a water. An analysis of newer
data that have not been previously assessed is included in the delisting procedure to allow the
state to determine if criterion excursions are more recent.

For delisting waters, if the 2018 assessment results in greater than 10% exceedance rate with
less than 90% statistical confidence and the water was on the 2016 303(d) list, the water will be
delisted if there are less than 2 excursions of the criterion in newer data that have not been
previously assessed. If the 2018 assessment results in less than 10% exceedance rate and the
water was on the 2016 303(d) list, the water will be delisted if there is greater than 40%
8

statistical confidence that there is less than a 10% exceedance of the criterion or if there are less
than 3 excursions of the criterion in newer data that have not been previously assessed.

Flow chart for listing a waterbody

Flow chart for delisting a waterbody

Figure 1.2. Numeric Criteria Assessment Flowcharts

9

1.2 Overview of Science Advisory Council Approach

The SAC was charged with recommending new numeric nutrient criteria so that High Rock Lake
meets its designated uses, which include public water supply, recreation, and aquatic life. An
important designated use in High Rock Lake is fisheries due to the quality of the bass fishing.
The focus of the SAC discussions were around designated uses, but also included lake use
protection into the future as the climate changes or if information on the lake’s health changes.

The SAC has proceeded in a step-wise fashion to recommend a new chl a criterion for High
Rock Lake. The first phase of the SAC’s work was information gathering so that the diverse
group could understand water quality standards, numeric nutrient criteria development, and learn
about High Rock Lake. Information was diverse and included uses and attainments, historical
water quality data, modeling, and other pertinent material collected from and about High Rock
Lake.

Additional information on multiple topics, such as the relationship between lake pH or chl a
values and fisheries, was developed by various SAC members from literature reviews. These
data were often tabulated in a “database” that was contextualized geographically for better
comparison to High Rock Lake conditions. Literature and data were shared among and discussed
between SAC members.

Numerous proposals for pH and chl a criteria were then developed by various SAC members
using multiple lines of evidence. Proposal discussions focused on averaging period and the
frequency of criterion exceedance that would be protective of the designated uses. Once fully
discussed, votes were taken on the different proposals until consensus was reached for a new
recommended chl a criterion. At a two-day SAC meeting in December 2018, the group’s
conclusions were substantively captured and that content was used to produce the current
document, including additional refinement on certain components as the document was finalized.
The newly developed proposed criterion has used the best science available and multiple lines of
evidence to determine the appropriate chl a criterion that is protective of the water quality
standards. As with most water quality decisions, numeric outcome-points consist of both data
and best scientific judgement.

Since the role of the SAC is to recommend standards protective of the water resource, the
committee tried not to discuss criteria relative to their feasibility and/or attainability. A
companion committee, the Criteria Implementation Committee or CIC, was formed to focus on
implementation of the recommended nutrient criteria as determined by the SAC. The CIC group
meets after the SAC committee proposes criteria. Their job is to refer clarifying questions back
to the SAC and also make determinations relative to the feasibility that the water resource can
meet these criteria. The process between the SAC and the CIC is iterative. Many of the CIC
members have attended the SAC meetings to better understand the deliberations that occur
around the new nutrient criteria recommendations for the High Rock Lake.

10

1.2 References

Cube Hydro Carolinas. (2019). "The Yadkin Project." 2019, from http://cubecarolinas.com/the-
yadkin-project/. (retrieved May 25, 2020).
McGhee, R. (1983). "Experiences in developing a chlorophyll a. standard in the Southeast to
protect lakes, reservoirs, and estuaries." Lake Restoration Protection and Management: 163-
165.
NC Division of Water Resources (2017). Surface Water Quality Standards History Document:
Chlorophyll a. Raleigh, NC, NC Division of Water Resources: 2.
NCDWR, N. C. D. o. W. R. (2014). North Carolina Nutrient Criteria Development Plan, June
20, 2014. Raleigh, NC, North Carolina Department of Environment and Natural Resources.
NCDWR, N. C. D. o. W. R. (2018). 2018 303(d) LISTING AND DELISTING
METHODOLOGY, Approved by the North Carolina Environmental Management
Commission on March 8, 2018. N. C. D. o. E. Quality. Raleigh, NC.
NCDWR, N. C. D. o. W. R. (2019). North Carolina Nutrient Criteria Development Plan, v.2,
May 16, 2019. Raleigh, NC.
Tetra Tech (2004). Water Quality Data Review for High Rock Lake, North Carolina. Research
Triangle Park, NC 27709.
USEPA (1975). Report on High Rock Lake, Davidson and Rowan Counties, North Carolina,
EPA Region IV. National Eutrophication Survey Working Paper No. 381. Pacific Northwest
Environmental Research Laboratory, Corvallis, OR.
Weiss, C. M. and E. J. Kuenzler (1976). The trophic state of North Carolina lakes, Water
Resources Research Institute of the University of North Carolina.

11

2. Literature Review of Chl a and Use Attainment

This chapter reviews the existing literature relating chlorophyll a (chl a) concentrations and the
attainment of designated uses in surface water bodies. Surface water chl a concentration is a
proxy for phytoplankton biomass and correlates strongly with primary production in systems
where phytoplankton are the dominant primary producers (Cloern et al. 1995). Thus, chl a is a
strong indicator of trophic status. The principal function of chl a is to absorb visible sunlight
within the photosynthetically active radiation band (PAR, 400 - 700 nm), and convert PAR into
chemical energy needed to fuel carbon fixation. Through PAR absorption, chl a can directly
impact light levels necessary for other plants (e.g., submerged aquatic vegetation) to grow and
for animals including humans to see. Indirect impacts of elevated chl a on aquatic life include
excessive organic matter production and subsequent water quality degradation (e.g. high/low pH,
high/low dissolved oxygen), and toxicity from secondary metabolites that co-occur with chl a in
phytoplankton cells (e.g. cyanotoxins).

2.1. Chl a and Water Clarity

On average, chl a in phytoplankton cause about a 0.02/m attenuation of photosynthetically active
radiation (PAR) for every μg/L chl a (Koseff et al. 1993). A phytoplankton bloom of about 40
μg/L would result in a light attenuation value that approximately halves the light availability with
every meter depth. The relative importance of phytoplankton chl a in attenuating light depends
on the concentrations of other light attenuating substances including suspended mineral
sediment, suspended organic detritus (terrestrial or aquatic), and colored dissolved organic
matter (CDOM) (Biber et al. 2008).

For water bodies with submerged aquatic vegetation (SAV), the amount of light that penetrates
to the bottom often determines the maximum depth that SAV can grow, and can limit SAV areal
coverage of shallow, nearshore areas in waters with elevated chl a concentration. Because of the
importance of SAVs for stabilizing sediments, trapping nutrients, and serving as a structured
habitat for fish and invertebrate communities, maintaining SAV coverage is often an important
component protecting aquatic life uses. Determining a chl a criterion that is protective of SAV
coverage requires knowledge of the light requirement for SAV growth, the maximum depth of
SAV beds to be protected, and the amount of background light attenuation from substances other
than chl a. Light requirements for SAV growth vary modestly among species and sediment
characteristics, but usually range from 10-20% of incident sunlight. Concentrations and relative
importance of light-attenuating substances vary greatly across aquatic systems and result in chl a
targets for SAV protection being highly site specific. For example, within different regions of
Chesapeake Bay, chl a targets to maintain SAV growth at depths from 0.5 to 2 m ranged by more
than an order of magnitude from 2.7 to 43 μg/L (EPA 2007). For some waters, concentrations of
sediments or CDOM are so high that SAV cannot grow, even though chl a in these waters is
often negligible, and otherwise suitable substrates exist (Bachmann et al. 2002). A standard of 20
μg/L was approved for Lake Winona, Minnesota to protect SAV coverage (MN PCA 2014).
Although protection of SAV is often a consideration in developing chl a criteria, SAV have
12

apparently never been established in HRL, and with high suspended sediment concentrations and
widely fluctuating water levels, it is unclear whether even drastic chl a reductions would allow
for SAV establishment (see Chapter 3.4.1). Decreases in water clarity associated with high chl a
can also affect aquatic life uses by impacting predator/prey (Manning et al. 2013) and
competitive (Stasko et al. 2015) interactions, and altering heat budgets with resultant changes in
temperature and oxygen solubility (Rose et al. 2016; Heiskanen et al. 2015). These indirect
effects of water clarity are becoming better understood, but at present have not been used to
establish chl a thresholds for protecting aquatic life.

2.2. Fisheries Effects

In general, freshwater fisheries production responds positively to increases in chl a due to higher
rates of phytoplankton based primary production (Deines et al. 2015) ) that fuels production at
higher trophic levels. Bachmann et al. (1996) found a clear positive relation between standing
stock fish biomass and annual average chl a across 60 Florida Lakes with chl a levels ranging
from 1 to about 100 μg chl a (Fig. 2.1 A). For crappie, an optimal range 20-60 μg/L has been
reported (Schupp and Wilson 1993), which is slightly lower than optimal for bass and sunfish
production (40-60 μg/L, Maceina and Bayne 2001). Similar results were found in a comparison
of fish and chl a in Iowa reservoirs (Egerston and Downing 2004). In a meta analysis of over 700
freshwater systems worldwide, Deines et al. (2015) also found consistent positive relationships
between chl a and several metrics of fish production (production, yield, catch per unit effort, and
density), with coefficients of variation averaging 0.71 (95% confidence interval = 0.59-0.80).
Their study also included examination of climate impacts on fisheries but measures of
autotrophic production were consistently more important predictors of fish production metrics.

Across four Alabama and Georgia reservoirs, biomass and growth rates of black bass, the apex
predator, were positively related to average growing season chl a across the range 2 – 27 μg/L
(Bayne et al. 1994). Higher production of top predators in the eutrophic reservoirs was partly
related to increased efficiency of trophic transfer that was driven by a shortened food chain. In
the more eutrophic lakes, large phytoplankton were consumed directly by herbivorous shad
while in the mesotrophic reservoir, crustacean zooplankton served as a more important trophic
link between phytoplankton and planktivorous fish. Lower relative abundance of crustacean
zooplankton in the eutrophic reservoirs was linked to lower relative abundance of Lepomis
sunfish that prey largely on crustacean zooplankton in their early developmental stages. Thus,
higher productivity may favor planktivorous fish and their predators over other guilds of fish
(Bayne et al. 1994; Allen et al. 1999). There is also indication that very high productivity may
increase the predominance of benthic species such as catfish and roughfish that may or may not
be desirable (Egertson and Downing 2004; Michaletz et al. 2012. The types of fish communities
that are desired should be considered when designing nutrient management strategies to support
both fishery and water quality related uses.

In addition to causing shifts in composition of fish communities, an upper threshold to the
positive relationship between chl a and overall fisheries production is expected. At chl a levels
13

beyond the threshold, negative impacts of excessive algal production on water quality (e.g.
dissolved oxygen concentrations, water clarity) may reduce fish production, or cause substantial
shifts toward less desirable fish species. Yurk and Ney (1989) found that across 22 southeastern
US reservoirs, chl a correlated positively with total fish abundance, but suitable habitats for
desirable walleye and striped bass occurred where reduced algal production allowed
hypolimnetic waters to remain oxygenated. In Westpoint Reservoir, Georgia, a 50% reduction in
chl a from approximately 40 to 20 led to shifts in the dominant species of black bass (Maceina
and Bayne 2001). The smaller spotted bass replaced largemouth bass with an overall increase in
number of fish, but a decrease in total black bass biomass.

Boucek et al. (2017) found some evidence for an upper threshold in the relationship between
largemouth bass condition (mass divided by length) and chl a such that condition improved up to
a chl a level of about 80-100, but subsequently decreased at higher chl a (Fig. 2.1 B). It is worth
noting that the decrease in body condition at higher chl a levels was driven only by two data
points with the highest chl a. In general, evidence for declines in fisheries production at the
highest chl a levels is weaker than evidence for a monotonic, positive relationship (Deines et al.
2015), and if a threshold exists it is most likely at a chl a level greater than 80 μg/L.

2.3. Chl a Relationships to Toxins

Several genera of bloom-forming cyanobacteria can produce a potent suite of secondary
metabolites that are hepatoxic and neurotoxic and can harm aquatic life (Chorus and Bartram
1999). Some freshwater eukaryotes (e.g. Prymnesium parvum, Roelke 2016) also produce toxins
and have caused massive fish kills in reservoirs of the southeast U.S. but these occurrences are
much less common than incidences of toxic cyanobacterial blooms. Microcystins (MCYs) are

Figure 2.1. Cross lake comparisons of chl a concentration versus total fish standing crop (A) redrawn
from Bachmann et al. (1996), and largemouth bass condition factor (B) redrawn from Boucek et al.
(2018).

14

the most common cyanobacterial toxins measured in freshwaters, and far more is known about
MCYs than the other cyanotoxins. There are many congeners of MCYs that vary greatly in their
toxicity, but all primarily affect the liver and digestive function. Direct consumption of MCY
containing algal cells by feeding on toxic cyanobacteria cells, or by drinking bloom-
contaminated waters are the primary exposure pathways for animals (Ibelings and Havens 2008).
Acute microcystin exposure causes necrosis of the liver and death (Tencalla et al. 1994).
However, the sensitivity of aquatic organisms varies significantly, and organisms from eutrophic
freshwater systems where elevated microcystins are more common tend to be less affected by
microcystins than those from oligotrophic systems (Malbrouck and Kestemont 2006). Toxins
accumulated by zooplankton and bivalve filter feeders can be passed up the foodweb, but MCYs
are not known to biomagnify at higher trophic levels (Kozlowsky-Suzuki et al. 2012; Ibelings et
al. 2005). Rather, biodilution occurs, and animals at the top of freshwater aquatic food chains
(e.g. predatory fish) are least likely to accumulate MCYs to levels that cause liver damage to the
fish (Ibelings et al. 2005) or to humans that may eat their flesh (Hardy et al. 2015; Wilson et al.
2008). Emerging evidence indicates that MCYs may also have neurotoxic activity at
concentrations lower than those known to cause liver damage. Dissolved MCY concentrations of
0.5 μg/L or prepared in food at 10 ppb have been shown to alter behaviors of fish diurnal
swimming activity (Baganz et al. 1998; 2006) and refuge seeking and escape behaviors of
crayfish (Clearwater et al. 2014).

Two pieces of information are needed to determine a chl a level that is protective of aquatic life
from the threat imposed by MCYs. First, a toxin threshold below which negative impacts are
unlikely to occur must be established. Second, a sufficiently strong linkage between chl a and
MCY must be established to estimate the chl a level below which MCY concentrations remain
below harmful levels. The wide range of susceptibility of aquatic organisms to impacts from
MCYs, as well as uncertainties associated with impacts of low level, chronic exposures to MCYs
makes establishing a safe MCY level very difficult (Bukaveckas et al. 2017). In water bodies
where the dominant bloom forming phytoplankton are MCY producing cyanobacteria, strong
temporal and spatial relationships between chl a and MCYs have been documented (Otten et al.
2012; Gagala et al. 2014). For these water bodies, chl a may serve as a useful indicator for toxin
related risks to aquatic life (e.g. Otten et al. 2014). However, correlations of chl a with MCYs are
usually weak both for studies of individual water bodies (Vaitomaa et al. 2003; Ha et al. 2009)
and for intersystem comparisons (Yuan et al. 2014). The general lack of correlation between
cyanotoxins and chl a is primarily due to variability in chl a driven by eukaryotes and non MCY
producing cyanobacteria (Ha et al. 2009) but additional variation in MCYs relative to chl a is
produced by changes in environmental growth conditions (Orr and Jones 1998), and selection of
cyanobacterial strains genetically equipped for greater/lesser MCY production (Orr et al. 2004;
Otten et al. 2012). Given the difficulties in establishing the necessary threshold MCY
concentration for protecting aquatic life or a corresponding chl a value associated with any
particular MCY level, designing a chl a criterion to be protective of cyanotoxin exposure for
aquatic life would contain a very large amount of uncertainty. Therefore, given current
information available, establishing a chl a criterion based on cyanotoxin risk to aquatic life is not
advised.
15

2.4. Chl a and Potable Water Supply Use

High Rock Lake is designated as Class WS-IV (waters protected as water supplies). (See, 15A
NCAC 02B .0301). In determining whether a water is suitable as a potable water supply, the
physical, chemical, and bacteriological maximum contaminant levels specified by Environmental
Protection Agency regulations are used as a guide. In other words, the requirements of EPA’s
Safe Drinking Water Act are used as a guide to determine the water quality necessary to ensure
this use is protected. The North Carolina Administrative Code also provides that the suitability of
water supplies are evaluated after treatment. In practice, potable water supplies are evaluated at
the point of a potable water intake and take into account the treatment provided in evaluating
whether uses are attained in the finished water. At a minimum, these treatment requirements
include filtration and disinfection for surface water supplies. (See, 40 CFR 141.70).

The Safe Drinking Water Act establishes primary and secondary standards for contaminants in
drinking water. (See, https://www.epa.gov/ground-water-and-drinking-water/national-primary-
drinking-water-regulations) The Primary Drinking Water Regulations (Primary Standards)
establish legally enforceable contaminant level concentrations and treatment techniques that
apply to public water systems to protect public health. Primary Standards include disease-
causing organisms, turbidity (an indicator of whether disease-causing organisms may be
present), and various chemical substances. The Secondary Drinking Water Regulations
(Secondary Standards) are non-enforceable guidelines for regulating contaminants that may
cause cosmetic effects or aesthetic effects (taste, odor, and color) in drinking water, but does not
prevent its use. The Secondary Standards include chemical contaminant concentrations, color,
odor, and other standards.

Under the SDWA, EPA may also publish health advisories for contaminants that are not subject
to any Primary Standards. In 2015, EPA developed such health advisories for two cyanotoxins,
microcystins and cylindrospermopsin. (See, https://www.epa.gov/cyanohabs/epa-drinking-water-
health-advisories-cyanotoxins) EPA also published guidance on managing cyanotoxins in public
drinking water systems. (See, https://www.epa.gov/ground-water-and-drinking-water/managing-
cyanotoxins-public-drinking-water-systems) This guidance generally discusses cyanobacteria,
hazardous algal blooms (HABs) of cyanobacteria, and the potential for cyanotoxins to be present
when HABs occur. The guidance notes that HABs can create taste and odor problems in drinking
water. Conventional water treatment (coagulation, sedimentation, filtration, and chlorination) can
generally remove cyanobacterial cells and low levels of cyanotoxins. Risks associated with
HABs can also be reduced through active management of public water systems.

The chl a concentration of water does not directly affect its use as a potable water supply. Rather,
chl a or the presence of algal cells would be considered in a similar fashion to secondary
drinking water standards. Secondary drinking water standards apply to contaminants that are not
health threatening but may affect color, taste and odor, or have other undesirable effects.
Conventional potable water treatment facilities include processes to remove algal cells and their
16

associated chl a prior to use. Consequently, even if chl a levels are elevated, adjustments in
treatment can generally be made without the need for additional facilities. However, operations
and maintenance (O&M) costs may be affected but this is not an impairment of the use.
Source water chl a concentration, at the point of intake to a potable water treatment system,
influences the potential cost of treatment to prepare the water for potable use, but does not affect
its use as a potable water supply. Treatment requirements for potable water supplies that
originate from surface waters, such as lakes and rivers, are highly regulated by USEPA. Under
the Safe Drinking Water Act (SDWA), the EPA Office of Water (EPA-OW) is charged with
setting water quality standards and regulations to protect the public drinking water supply. These
requirements impose treatment strategies at all potable water treatment facilities that are readily
able to control particulates (including algal cells). The regulatory basis for these treatment
strategies is presented in Attachment B of the pH criteria proposed by the SAC for HRL.
As discussed in Attachment B to the proposed pH criteria, potable water supplies, which use
surface water as a source, must provide treatment to settle and filter waterborne disease-causing
contaminants, and provide disinfection. The chemicals used in treatment to enhance particulate
removal will remove algal cells/chlorophyll before the treated water is provided for use.
Additional treatment, such as that required to minimize the formation of disinfection byproducts
under the Disinfection Byproducts Rule, would typically require the use of activated carbon to
reduce the amount of naturally occurring dissolved organic material. Activated carbon is also
very effective in removing taste and odor-causing compounds (2-methylisoborneol (MID) and
geosmin) and cyanotoxins. (EPA, 2015)

A review of the literature on chl a concentration necessary to protect drinking water uses yields a
mixture of reports that confound chl a with the actual cause of concern. Several of these studies
were identified during the meeting of the SAC in April, 2016. The meeting minutes and
presentation slides for this meeting identified several literature references related to development
of a chl a criterion to protect drinking water uses. These include the following specific references
(Table 2.1).

Table 2.1. A Review of Chl a Concentrations Necessary to Protect Drinking Water Uses
Chl a Target
(µg/L) Source/Notes
30 Values above 30 µg/L increase the risk of algal-related health problems. (Heath et
al., 1998)
9 – 10
15 – 20
20 – 80

Taste and Odor problems become noticeable
Water supply uses impaired
Consumptive uses severely impaired
(Carney, 1998)
10
50

Relatively low probability of adverse health effects
Moderate probability of adverse health effects (assumes cyanobacteria dominance)
(Chorus and Bartram, 1999)
15 To keep geosmin < 5 ng/L. (Smith et al, 2002)

17

A review of these citations shows that the parameter associated with the impairment of the
drinking water use is not chl a but some other parameter. Heath et al (1998) and Chorus and
Bartram (1999) were primarily concerned with cyanobacteria and cyanotoxins. Carney (1998)
and Smith et al (2002) focused on taste and odor issues. These are separate issues that would
require a two-step process to generate a chl a criterion for the protection of drinking water (EPA,
2010). The first step involves identification of an impairment threshold for the agent causing the
impairment (e.g., cyanotoxin, geosmin). Then the causative agent must be related to chl a
concentration. This relationship typically results in low predictive capability.

For example, the State of Illinois prepared a literature review on taste and odor issues in potable
water supplies (Lin, 1977; https://www.isws.illinois.edu/pubdoc/C/ISWSC-127.pdf). Taste and
odor issues are attributed to chemical substances released by algae during the growth phase of
algal cell development, with about 60 species identified as producers of substances leading to
taste and odors in water. One such substance, geosmin, is produced by certain algae, including
cyanobacteria. In addition, taste and odor problems may also be caused by actinomycetes. This
literature review identifies other sources of taste and odor issues, various characteristics of taste
and odors, as well as methods for controlling taste and odor issues.

The Kansas Department of Health and Environment (KDHE) prepared a white paper on
Chlorophyll-a Criteria for Public Water Supply Lakes or Reservoirs (2011)
(http://www.kdheks.gov/water/download/tech/Chlorophylla_final_Jan27.pdf). They note that
excessive algal growth can have undesirable effects on drinking water supplies including taste
and odor problems, increased levels of cyanotoxins, higher levels of trihalomethane precursors,
and increased turbidity levels in source water. Treatment costs for dealing with issues caused by
excessive algal growth can be very high. KDHE noted, for example, that the City of Wichita
spent $8.5 million on an ozone facility to control taste and odor problems in the Cheney
Reservoir, and massive algal blooms have triggered the shutdown of drinking water intakes at
several other reservoirs. They conclude, prevention is one of the most cost-effective ways for
dealing with nutrient related problems for lakes and reservoirs. Problems associated with
excessive algal growth are specific to the types of algae present, but direct counting of algal
communities is time-consuming and labor-intensive, while chl a measurement is a good practical
alternative for assessing algal biomass. For Kansas reservoirs, taste and odor problems begin
occurring once chl a values reach 10 µg/L. KDHE subsequently adopted a chl a criterion of 10
µg/L to protect domestic water supply uses (See,
http://www.kdheks.gov/tmdl/download/Unofficial_Copy_SURFACE_WATER_QUALITY_ST
ANDARDS_04.11.18.pdf).

As discussed above, Kansas adopted chl a criterion of 10 µg/L to protect drinking water supplies
from taste and odor problems. Taste and odor problems are secondary drinking water standards
that do not preclude the use as a potable water supply under the SDWA. This is readily apparent
given that the use of High Rock Lake water as a potable water supply for a downstream
municipality has not been impaired by chl a concentrations that are significantly higher.
Moreover, based on modeling of High Rock Lake, it would be impossible to consistently achieve
18

10 µg/L as a seasonal mean concentration. Consequently, the application of this criterion to High
Rock Lake is not recommended. As described by KDHE, dealing with taste and odor issues is a
cost-effectiveness problem. In this case, the cost to lower chl a concentrations in the lake should
be weighed against the cost of treatment to provide drinking water from this source.

KDHE also noted the relationship between chl a and the likelihood of cyanobacteria dominance,
the occurrence of cyanotoxins, precursors to disinfection byproducts, and turbidity. For these
parameters to serve as a basis for setting a chl a criterion, an impairment threshold for the
specific condition must be identified and then related back to chl a concentration, with
consideration for the removal that occurs during treatment at the water treatment plant. Since
these parameters are all subject to removal at the treatment works by the currently mandated
treatment processes, the analysis will become a cost-effectiveness evaluation to set an
appropriate criterion.

2.5. Chlorophyll a and Recreation Use

Clearer water is valued more highly for recreation than turbid waters (Andradi et al. 2018;
Smeltzer and Heiskary 1990), and therefore chl a-rich, turbid waters are generally perceived as
having poorer recreational value compared to waters with less chl a (Andradi et al. 2018; Smith
et al. 2015; Smeltzer and Heiskary 1990). It is important to recognize, however, that water clarity
is also controlled by suspended sediment and CDOM, and it is mainly water clarity rather than
chl a that relates to recreational value (Andradi et al. 2018). Waders and swimmers value water
clarity because the ability to see the bottom provides increased perception of safety pertaining to
physical hazards, a greater perception that the water is “clean”, and an increased aesthetic appeal
(Angradi et al. 2018). The aesthetic value of low chl a waters also extends to non-contact
recreational activities such as boating, fishing, or just lake viewing (Andradi et al. 2018).
However, other factors including surround land use (e.g. forested, cleared/ developed shorelines)
and abundance of litter play equal roles in a water body’s aesthetic appeal (House 1996; Andradi
et al. 2018). Aesthetic values are not explicitly protected as a designated use for NC waters but
implicitly are protected due to this strong relationship with recreational value. High algal
biomass can also generate unsightly scums that may also produce odors, and or toxins.
Increasing public recognition of toxin production by some bloom-forming phytoplankton may
further strengthen the perception of the safety of recreating in clearer waters. However, as
discussed in Section 2.3, the relationships between chl a and toxin production is too uncertain at
this time to derive a meaningful, quantitative chl a criterion for High Rock Lake.

Although water clarity is a strong determinant of perceived recreational value, the quantitative
water clarity judged by water users to be acceptable for recreation displays strong regional
variation that depends on the water clarity to which recreators are accustomed (Andradi et al.
2018; Smeltzer and Heiskary 1990). In regions that generally have high water clarity with Secchi
depths extending down 5-10 meters, a lake with a 2 m deep Secchi depth might be judged to
have impaired recreational value. At the same time, a lake with a 2 m deep Secchi depth might
be judged as having outstanding recreational value in the piedmont of NC where water clarity is
19

generally poor due to a combination of high phytoplankton and suspended sediment. In regions
with very poor water clarity, water clarity also becomes a less useful predictor of recreational
value (Smeltzer and Heiskary 1990). These regional variations in user perceptions of acceptable
water clarity lessen the usefulness of recreational chl a criteria outside of the region where they
were developed. When translating survey results across regions, it is important that the average
water clarity in the survey region matches the average clarity of the region where the criterion is
being developed. Surveys of recreators on eight Texas reservoirs with water clarity similar to
North Carolina reservoirs indicated that lakes with annual average chl a values between 35-40
mg/L, about 30% of respondents judged the water quality to be impaired to some degree for
recreation (Glass 2006).

20

2.6. References

Andradi, T.R., Ringold, P.L., Hall, K. 2018. Water clarity measures as indicators of recreational
benefits provided by U.S. Lakes: Swimming and aesthetics. Ecological Indicators 93: 1005-
1019.
Allen, M.S. Greene, J.C. Snow, F.J. Maceina, M.J. DeVries, D.R. 1999. Recruitment of
Largemouth Bass in Alabama Reservoirs: Relations to Trophic State and Larval Shad
Occurrence. North American Journal of Fisheries Management 19(1):67-77.
Bachmann, R.W., Horsburgh, C.A., Hoyer, M.V., Mataraza, L.K., Canfield, D.E. Jr. 2002.
Relations between trophic state indicators and plant biomass in Florida lakes. Hydrobiologia
470: 219-234.
Bachmann, R.W., Jones, B.L., Fox, D.D., Hoyer, M., Bull, L.A., Canfield, D.E. Jr. 1996.
Relations between trophic state indicators and fish in Florida (U.S.A.) lakes. Canadian
Journal of Fisheries and Aquatic Science 53: 842-855.
Baganz, D., Siegmund, R., Staaks, G., Pflugmacher, S., Steinberg, C.E.W. 2005. Temporal
pattern in swimming activity of two fish species ( Danio rerio and Leucaspius delineatus )
under chemical stress conditions, Biological Rhythm Research 36: 263-276, DOI:
10.1080/09291010500103112
Bayne, D.R., Maceina, M.J., Reeves, W.C. 1994. Zooplankton, fish, and sport fishing quality
among four Alabama and Georgia reservoirs of varying trophic status. Lake and Reservoir
Management 8: 153-163.
Beganz, D., Staaks, G., Steinberg, C. 1998. Impact of the cyanobacteria toxin, microcystin-LR
on behavior of zebrafish, Danio rerio. Water Research 32: 948-192.
Biber, P.D., Gallegost, C.L.., Kenworth, W.J. 2008. Calibration of a bio-optical model in the
North River, North Carolina (Albemarle-Pamlico Sound): A tool to evaluate water quality
impacts on seagrasses. Estuaries and Coasts 31: 177-191.
Boucek, R., Barrientos, C., Bush, M.R, Gandy, D.A., Wilson, K.L., Young, J.M. 2017. Trophic
state indicators are a better predictor of Florida bass condition compared to temperature in
Florida’s freshwater bodies. Environmental Biology of Fishes 100: 1181-1192.
Bukaveckas, P.A., Lesutiene, J., Gasiunaite, Z.R., Lozys, L., Olenina, I., Pilkaityte, R., Putys, Z.,
Tassone, S. Wood, J. 2017. Microcystin in aquatic food webs of the Baltic and Chesapeake
Bay regions. Estuarine Coastal and Shelf Science 191: 50-59.
Carney, C.E. 1998. A primer on lake eutrophication and related pollution problems: Kansas
Department of Health and Envirnoment. Bureau of Environmental Field Services.
Chorus, I., Bartram, J. 1999. Toxic Cyanobacteria in Water: A Guide to Their Public Health
Consequences, Monitoring and Management. London, United Kingdom. E. and F.N.
Spon/Chapman and Hall.
Clearwater, S.J., Wood, S.A., Phillips, N.R., Parkyn, S.M., Van Ginkel, R., Thompson, K.J.
2014. Toxicity thresholds for juvenile freshwater mussels Echyridella menziesii and crayfish
Paranephrops planifrons, after acute or chronic exposure to Microcystis sp. Environmental
Toxicology 29: 487-502.
21

Cloern, J.E., Grenz, C., Videgar-Lucas, L. 1995. An empirical model of the phytoplankton
chlorophyll: carbon ratio-the conversion factor between productivity and growth rate.
Limnology and Oceanography 40: 1313-1321.
Deines, A.M., Bunnell, D.B., Rogers, M.W., Beard, T.D. Jr., Taylor, W.W. 2015. A review of
the global relationship among freshwater fish, autotrophic activity, and regional climate.
Review in Fish Biology and Fisheries 25: 323-336.
Egerston, C.J. Downing, J.A. 2004. Relationship of fish catch and composition to water quality
in a suite of agriculturally eutrophic lakes. Canadian Journal of Fisheries and Aquatic
Science 61: 1784-1796.
EPA. 2007. Ambient water quality criteria for dissolved oxygen, water clarity, and chlorophyll a
for the Chesapeake Bay and its tidal tributaries: 2007 chlorophyll criteria addendum. U.S.
Environmental Protection Agency Region III. Chesapeake Bay Program Office. Annapolis,
Maryland. November 2007.
EPA. 2010. Using Stressor-Response Relationships to Derive Numeric Nutrient Criteria.
https://www.epa.gov/sites/production/files/2018-10/documents/using-stressor-response-
relationships-nnc.pdf
EPA. 2015. Recommendations for Public Water Systems to Manage Cyanotoxins in Drinking
Water. https://www.epa.gov/sites/production/files/2018-11/documents/cyanotoxin-
management-drinking-water.pdf
Gagala, I., K. Izydorczyk, T. Jurczak, J. Pawełczyk, J. Dziadek, A. Wojtal-Frankiewicz, A.
Jóźwik & A. Jaskulska,and J.Mankiewicz-Boczek. 2014. Role of environmental factors and
toxic genotypes in the regulation of microcystins-producing cyanobacterial blooms.
Microbial Ecology 67:465–479.
Glass, P. W. 2006. Development of use-based chlorophyll criteria for recreational uses of
reservoirs. Proceedings of the Water Environment Federation 2006 (8): 4038-4050.
Ha, J.H., Hidaka, T., Tsuno, H. 2009. Analysis of factors affecting the ratio of microcystin to
chlorophyll-a in cyanobacterial blooms using real-time polymerase chain reaction.
Environmental Toxicology 26: 21-228.
Hardy, F.J., Johnson, A., Hamel, K., Preece, E. 2015. Cyanotoxin bioaccumulation in freshwater
fish, Washington State, USA. Environmental Monitoring and Assessment 187: 667. DOI
10.1007/s10661-015-4875-x
Heath, R.G., Steynberg, M.C., Guglielmi, R., Maritz, A.L. 1998. The implications of point
source phosphorus management to potable water treatment. Water Science and Technology
37(2): 343–350.
Heiskanen, J.J., Mammarella, I., Ojala, A., Stepanenko, V., Erkkila, K.M., Miettinen, H.,
Sandstrom, H., Eugster, W., Lepparanta, M., Jarvinen, J., Vesala, T., Nordbo, A. 2015.
Effects of water clarity on lake stratification and lake-atmosphere heat exchange. Journal of
Geophysical Research-Atmospheres 120: 7412-7428.
Hollister, J.W., Kreakie, B.J. 2016. Associations between chlorophyll a and various microcystin-
LR health advisory concentrations. F1000 Research 5:151. DOI:
10.12688/f1000research.7955.1
House, M.A. 1996. Public perception and water quality management. Water Science and
Technology 34: 25-32.
22

Ibelings, B.W., Bruning, K., de Jonge, J., Wolfstein, K., Pires, L.M.D., Postma, J., Burger, T.,
2005. Distribution of microcystins in a lake foodweb: No evidence for biomagnification.
Microbial Ecology 49, 487-500.
Ibelings, B. W., Havens, K. E. 2008. Cyanobacterial toxins: a qualitative meta-analysis of
concentrations, dosage and effects in freshwater estuarine and marine biota. In Hudnell H. K.
(ed.), Cyanobacterial Harmful Algal Blooms: State of the Science and Research Needs, Vol.
619. Advances in Experimental Medicine and Biology. Springer, New York: 675–732.
Koseff, J.R., Holen, J.K., Monismith, S.G., Cloern, J.E. 1993. Couple effects of vertical mixing
and benthic grazing on phytoplankton populations in shallow, turbid estuaries. Journal of
Marine Research 51: 843-868.
Kozlowsky-Suzuki, B., Wilson, A.E., Ferrao-Filho A.S. 2012. Biomagnification or biodilution of
microcystins in aquatic foodwebs? Meta-analyses of laboratory and field studies. Harmful
Algae 18: 47-55.
Maceina, M.J., Bayne, D.R. 2001. Changes in the black bass community and fishery with
oligotrophication in West Point Reservoir, Georgia. North American Journal of Fisheries
Management 21: 745-755.
Malbrouck, C., Kestemont, P. 2006. Effects of microcystins on fish. Environmental Toxicology
and Chemistry 25: 72-86.
Manning, N.F., Mayer, C.M., Bossenbroek, J.M., Tyson, J.T. 2013. Effects of water clarity on
the length and abundance of age-0 yellow perch in the Western Basin of Lake Erie. Journal
of Great Lake Research 39: 295-302.
MN PCA. 2014. Minnesota Pollution Control Agency. Site Specific Water Quality Standards.
Winona Lake. https://www.pca.state.mn.us/water/site-specific-water-quality-standards
Orr, P.T. and G.J. Jones. 1998. Relationship between microcystin production and cell division
rates in nitrogen-limited Microcystis aeruginosa cultures. Limnology and Oceanography
43:1604−1614.
Orr, P.T., G.J. Jones, and G.B. Douglas. 2004. Response of cultured Microcystis aeruginosa
from the Swan River, Australia, to elevated salt concentration and consequences for bloom
and toxin management in estuaries. Marine and Freshwater Research 55: 277–283.
Otten, T.G., H. Xu, B. Qin., G. Zhu, and H.W. Paerl. 2012. Spatiotemporal patterns and
ecophysiology of toxigenic Microcystis blooms in Lake Taihu, China: Implications for water
quality management. Environmental Science & Technology 46: 3480−3488.
Roelke, D.L, Barkoh, A., Brooks, B.W., Grover, J.P, Hambright, K.D., LaClaire, J.W., Moeller,
P.D.R., Patino, R. 2016. A chronicle of a killer alga in the west: ecology, assessment, and
management of Prymnesium parvum blooms. Hydrobiologia 764: 29-50.
Rose, K.C., Winslow, L.A., Read, J.S., Hansen, G.J.A. 2016. Climate-induced warming of lakes
can be either amplified or suppressed by trends in water clarity. Limnology and
Oceanography Letters 1: 44-53.
Schupp, D., Wilson, D. 1993. Developing lake goals for water quality and fisheries. LakeLine
13(4): 18-21.
Smeltzer, E., Heiskary, S.A. 1990. Analysis and applications of lake user survey data. Lake and
Reservoir Management 6: 109-118.
23

Smith, A. J., Duffy, B.T., Novak, M.A. 2015. Observer rating of recreational use in wadeable
streams of New York State, USA: implications for nutrient criteria development. Water
Research 69: 195-209.
Smith, V. H., Sieber-Denlinger, J., deNoyelles, Jr., F., Campbell, S., Pan, S., Randtke, S.J.,
Blain, G.T., Strasser, A.A. 2002. Managing taste and odor problems in a eutrophic drinking
water reservoir. Lake & Reservoir Management 18(4): 319-323.
Stasko, A.D., Johnston, T.A., Gunn, J.M., 2015. Effects of water clarity and other environmental
factors on trophic niches of two sympatric piscivores. Freshwater Biology 60: 1459-1472.
Tencalla, F., Dietrich, D., Schlatter, C. 1994. Toxicity of Microcystis aeruginosa peptide toxin to
yearling rainbow trout (Oncorhynchus mykiss). Aquatic Toxicology 30:215–224.
Vaitomaa J, Rantala A, Halinen K, Rouhiainen L, Tallberg P, Mokelke L, Sivonen K. 2003.
Quantitative real-time PCR for determination of microcystin synthetase E copy numbers for
Microcystis and Anabaena in Lakes. Appl Environ Microbiol 69:7289–7297.
Wilson, A.E., Gossiaux, D.C., Hook, T.O., Berry, J.P., Landrum, P.F., Dyble, J., Guildford, S.J.
2008. Evaluation of the human health threat associated with the hepatotoxin microcystin in
the muscle and liver tissues of yellow perch (Perca flavescens). Canadian Journal of
Fisheries and Aquatic Sciences 65: 1487-1497.
Yuan, L.L., Pollard, A.I., Pather, S., Oliver, J.L., D’Anglada, L. 2014. Managing microcystin:
identifying national-scale thresholds for total nitrogen and chlorophyll a. Freshwater Biology
59: 1970-1981.
Yurk, J.J., Ney, J.J. 1989. Phosphorus-fish community biomass relationships in Southern
Appalachian reservoirs: Can lakes be too clean for fish? Lake and Reservoir Management 5:
83-90.

24

3. Current Conditions in High Rock Lake

The first two sections of this chapter present chlorophyll a (chl a) conditions in High Rock Lake
and relationships between chl a and other parameters of interest such as dissolved oxygen, pH,
water clarity. Later sections describe High Rock Lake conditions with respect to algal
abundance and species composition, and algal toxins. The final section of this chapter reviews
the current state of designated use attainment in High Rock. Separate evaluations of use
attainment are provided for fisheries and aquatic life, potable water supply, and
aesthetics/swimming. Included with each evaluation is a discussion of how the findings were
considered to indicate support or nonsupport of designated uses under High Rock Lake’s
prevailing chl a conditions.

High Rock Lake is one of the most studied reservoirs in North Carolina. Tetra Tech (2004)
summarized the results of six separate water quality monitoring programs conducted by the EPA,
NC DEM, a UNC research team, and contractors to Alcoa Power Generating Inc. that took place
between 1973 and 2001. These studies had various project objectives and sampling designs, and
consistently characterized High Rock Lake as a eutrophic reservoir based on nutrient and chl a
concentrations. More recently, the NC Division of Water Resources (DWR) conducted two
rounds (2005-2006 and 2008-2010) of intensive water quality investigations that collected
“photic-zone” composites (defined as twice the Secchi depth) that were analyzed for chl a and
other water quality constituents. Twelve stations (Figure 3.1) across the lake and its tributaries
Figure 3.1. DWR Monitoring Stations in High Rock Lake during the 2005-2006, 2008-2010, and 2016
monitoring programs. Not all stations were sampled in all of the monitoring programs.
2

Figure 1. Monitoring Stations in High Rock Lake

Crane
Creek
Abbotts
Creek
Swearing
Creek
Yadkin River
Grants
Creek
South
Yadkin
River
Lake Characteristics:
Mean Depth: 17 feet
Max Depth: 62 feet
Surface Area: 15,180 acres
Drainage Area: 3,974 square miles
Volume: 239,672 acre-feet
Retention Time: 4 to 50 days
25

were sampled as part of the High Rock Lake Scoping Study of 2005-2006 and the 2008-2010
intensive monitoring study. As part of the nutrient criteria development process, an
additional round of water quality sampling and analysis was performed the NC Department of
Environmental Quality (DEQ) in 2016. Results of these sampling efforts are described in detail
in the following sections.

3.1 Spatial and Temporal Patterns of Chl a Concentrations in High Rock Lake

An examination of the spatial and temporal patterns of chl a in High Rock Lake provides a
foundation for understanding the algal dynamics within the reservoir. Spatially, the reservoir
exhibits a consistent upstream-to-downstream pattern in relative chl a concentrations. A useful
conceptual model of the lake is that it operates like many “run-of-the-river” reservoirs that have
distinct riverine, transitional, and lacustrine zones. (Figure 3.2). The boundaries separating
these zones can shift upstream or downstream with river discharge, and the extent of the zone
can expand or contract in response to watershed runoff events, operation of the dam, and other
changes within the reservoir that influence the flow and water residence time (Cooke et al.
2005).

The riverine zone is located furthest upstream from the dam where the major river flows into the
lake. The riverine zone is characterized by the highest velocity and shortest hydraulic residence
time. This region tends to receive relatively high levels of nutrients and particulate matter. The
turbidity within this zone limits light penetration so primary production can be influenced by
light limitation. The transitional zone is marked by an increase in lake width, which can cause
decreased velocity and an increase in residence time. As the water slows, the suspended
sediment tends to settle out of the water and deposit on the lakebed. As turbidity decreases, light
penetration increases, and irradiance levels in the epilimnion increase. The transitional zone can
be a more productive region of the reservoir because light limitation plays less of a role there.
Bio-available nutrient concentrations decrease through the transitional zone while turbidity
decreases and irradiance levels increase. Controls on phytoplankton production transition from

Figure 3.2. Common lake zones (riverine, transitional, and lacustrine) observed for run-of-the
river reservoirs, such as High Rock Lake (modified from Cooke et al 2005).

Figure 3.2. Common lake zones (riverine, transitional, and lacustrine) observed for run-of-the river
reservoirs, such as High Rock Lake (modified from Cooke et al 2005).
26

light-limited production in the riverine zone to nutrient limited production in the downstream
lacustrine zone (Rudd 2018). In addition, internal nutrient recycling can play a larger role in the
transition and lacustrine zones (Cooke et al. 2005).

DWR monitoring stations are located in each of three zones within High Rock Lake (Figure 3.3).
Consistent spatial differences have been seen between chl a concentrations located in different
zones, for samples collected between 2008 and 2012. Two stations in the transitional zone of the
lake frequently exceeded the existing 40 µg/L chl a criterion (Figure 3.4). YAD152C and
YAD152 are the sites that have most frequently exceeded the 40 µg/L chl a criterion (Figure
3.5).

Figure 3.3. High Rock Lake monitoring station locations and lake zones.
Seasonal patterns in chl a are difficult to determine, because the majority of the samples
collected over the long-term have been collected during the growing season only. Monthly
sampling during 2008-2011 at station YAD152C showed that chl a concentrations were highest
during July and August, but could remain relatively high even in December (Figure 3.6). The
samples with the highest chl a tended to be dominated by cyanobacteria (Figure 3.6) in terms of
number of cells, although other taxa still comprised significant proportions of the algal biomass
or biovolume. Samples from a site located in one of the arms of the reservoir (Abbotts Creek)
also tended to show higher chl a values during the summer, but in this location high values
during September and October were also observed (Figure 3.7).

27

Figure 3.4. Map of percentage of water samples with chl a concentrations greater than 40 µg/L in the
time period 2008-2012 in High Rock Lake (from Behm presentation to SAC May 6, 2015).

Figure 3.5. Distribution (% of samples) of chl a concentrations across different stations in High Rock
Lake sampled between 2008-2011 (from Behm presentation to SAC May 6, 2015).

The variation in sampling frequency over the various High Rock Lake monitoring programs (e.g.
monthly, yearly, every five years) makes it challenging to draw conclusions on temporal trends
in the monitoring data. There are no clear long-term trends in chl a concentrations (Figures 3.8
and 3.9). Plots are shown for two of the sites with most data over the long-term sampling period
(1980-2011). There is no statistically significant trend, examined using linear regression.
28

Figure 3.6. Seasonal patterns of chl a in station YAD152C High Rock Lake 2008-2010. (from Behm
presentation to SAC May 6, 2015).

Figure 3.7. Seasonal pattern of chl a in station HRL052 in High Rock Lake 2008-2010. (from Behm
presentation to SAC May 6, 2015).
29

Figure 3.8. Long-term (1980-2011) chl a concentrations in YAD152C station in High Rock Lake. Years
in which more than 1 sample were collected were averaged and error bars represent standard error. There
has not been a clear increase or decrease in chl a concentration.

Figure 3.9. Long-term (1980-2011) chl a concentrations in YAD156A station in High Rock Lake. Years
in which more than 1 sample were collected were averaged and error bars represent standard error. There
has not been a clear increase or decrease in chl a concentration.
30

3.2 Chl a Relationships with Other Indicators

The subsections below present evaluations of the relationships of chl a with other key parameters
such as dissolved oxygen, pH, and water clarity. These parameters are useful indicators of
attainment of aquatic life and recreational uses, so their relationship with chl a has direct bearing
on the selection of a chl a criterion for High Rock Lake. Specifically, if chl a has a strong
relationship with a key parameter, it would be desired to set chl a criteria at levels at which that
parameter is within use-supporting ranges, considering both magnitude and temporal aspects of
the parameter goals. If a parameter lacks strong relationships with chl a, or the parameter lacks
clear thresholds of attainment/non-attainment, it would have less bearing on the chl a criteria
selected.

3.2.1 Dissolved Oxygen

High Rock Lake generally experiences favorable dissolved oxygen (DO) concentrations in the
epilimnion and is not 303(d)-listed for this parameter. However, DO concentration is one of the
most direct indicators of aquatic life support, and so the relationship between chl a and DO
should be considered when setting site-specific criteria. The North Carolina Administrative Code
(15A NCAC 02B) identifies the DO water quality criteria applicable to High Rock Lake based
on the designated uses of the lake. The DO criteria for Class C waters (15A NCAC 02B.0211(6))
provides: for non-trout waters, not less than a daily average of 5.0 mg/l with a minimum
instantaneous value of not less than 4.0 mg/l; swamp waters, lake coves, or backwaters, and lake
bottom waters may have lower values if caused by natural conditions.

Lin (2015) previously evaluated the relations between chl a and DO in High Rock Lake based on
the historical fixed station monitoring record. This evaluation determined that surface DO
concentration and DO percent saturation was positively correlated with chl a in spring and
summer, but negatively correlated with chlorophyll in the winter (Figure 3. 10). Bottom DO
was negatively correlated with chl a in the winter and spring.

The positive correlation between chl a and surface DO in growing season months is expected due
to algal photosynthesis, especially considering that most fixed station data were collected during
daytime hours. Weaker correlations were detected between chl a and deeper DO. While some of
the DO from surface algal photosynthesis can reach hypolimnetic waters by diffusion or
advective mixing, increases in organic matter may also increase the decay of algal biomass, thus
depleting DO in bottom waters. Some hypolimnetic oxygen depletion is considered a natural
process in lakes and reservoirs (such as High Rock Lake) especially during temperature-driven
stratification in warm months. For this reason, compliance with DO standards is normally
assessed using surface measurements, and the present evaluation did not consider hypolimnetic
DO depletion as an impairment of designated uses.

In 2016, the North Carolina Division of Water Quality (DWQ) also deployed monitoring sondes
in High Rock Lake to measure short-term variations in chl a, DO, and DO percent saturation,
31

among other variables. The sondes were deployed from July 13 to October 5, 2016. Surface and
bottom sondes were deployed at station YAD152C for the entire period, whereas the other sonde

Figure 3.10. Relation between fractional DO saturation and chl a in High Rock Lake fixed station data.
Source: Lin (2015).
pair was moved between three stations (YAD169A, YAD169B, and HRL051). The chl a
concentrations from the sondes were not similar in magnitude to chl a concentrations measured
in grab samples (extraction method), and so are of questionable reliability. However, the sonde
data are still considered useful for exploring the DO conditions that High Rock Lake experiences
under the prevailing chl a conditions. For reference, chl a concentrations measured in grab
samples in July-October 2016 ranged from 11 to 47 µg/L station HRL051, 58 to 75 µg/L at
YAD152C, and 31 to 56 µg/L at YAD169B.

The sonde data reveal generally favorable DO concentrations at the surface, with >99 percent of
individual measurements above North Carolina’s minimum DO criterion of 4 mg/L for Class B
waters, and almost 100% percent of daily average DO measurements exceeding the daily average
criterion of 5 mg/L (Table 3.1).

Table 3.1. Proportion of High Rock Lake Surface Sonde DO Measurements at or Above DO Criteria
Station
Proportion of
Observations
≥ 4 mg/L
Proportion of Daily
Averages
≥ 5 mg/L
YAD152C 100% 100%
YAD169A ~98% 100%
YAD169B 100% ~100%
HRL051 ~100% 100%
All >99% ~100%

32

The sonde data also revealed relatively high diel variability in surface DO concentration (Figure
3.11) and surface DO saturation (Figure 3.12) associated with diurnal cycles in algal
photosynthesis and respiration. Table 3.2 presents a statistical summary of the sonde chl a and

Figure 3.11. Surface and bottom DO concentrations at YAD152C during a portion of the 2016 sonde
data collection period.

Figure 3.12. Surface and bottom DO percent saturation during a portion of the 2016 sonde data collection
period.
DO data by station. The surface DO percent saturation averaged 122% for all the sonde data
combined, but exceeded 175% about 10 percent of the time and was less than 71% about 10
percent of the time. The surface DO percent saturation occasionally exceeded 225%, although
33

this occurred in only about one percent of the individual measurements. The single highest DO
percent saturation measurement (265%) was observed at station YAD152C.
Chl a was positively correlated with DO concentration and DO percent saturation in both surface
and bottom sonde measurements (Table 3.2). The positive correlation with bottom DO
demonstrates the possibility of downward diffusion/mixing of high DO at the surface, at
Table 3.2. Spearman Rank Correlation Coefficients of Daily Average Surface Chl a vs DO Metrics [Data
source: DWQ 2016 sonde data from High Rock Lake]
DO Metric Depth
Zone Statistic n
Spearman
Rank
Correlation
Coefficient
p-value
DO Concentration Surface Daily Minimum 164 +0.303 <0.001
Daily Average 164 +0.371 <0.001
Daily Maximum 164 +0.404 <0.001
Bottom Daily Minimum 145 +0.313 <0.001
Daily Average 145 +0.472 <0.001
Daily Maximum 145 +0.512 <0.001
DO percent
saturation
Surface Daily Minimum 164 +0.279 <0.001
Daily Average 164 +0.345 <0.001
Daily Maximum 164 +0.375 <0.001
Bottom Daily Minimum 145 +0.322 <0.001
Daily Average 145 +0.472 <0.001
Daily Maximum 145 +0.504 <0.001

least under certain conditions. North Carolina does not have a water quality criterion for DO
percent saturation and utilizes DO concentration criteria to protect against low DO conditions.
This approach for protection against low DO conditions is consistent with federal guidance
(USEPA, 1986) which states that concentration-based DO criteria are more direct and easier to
administer than percent saturation-based criteria and that percent saturation-based criteria could
be either over or under protective based on temperature and elevation. North Carolina does have
a criterion of not more than 110 percent saturation of total dissolved gas saturation, intended to
prevent over-aeration of water and subsequent gas bubble disease in aquatic life, as can occur in
hydroelectric dam tailwaters. However, percent saturation of total gases cannot be directly
translated to a goal for DO percent saturation, and gas bubble disease is usually caused by excess
nitrogen rather than excess oxygen (Weitkamp and Katz, 1980).

The effects of oxygen supersaturation on aquatic life is not as well understood as that of total
dissolved gases or nitrogen. Under most circumstances, fish can tolerate short periods of oxygen
supersaturation relatively well, partly because (unlike nitrogen) oxygen can be removed from
tissue via metabolic activity (Weitkamp and Katz, 1980). However, some studies have attributed
gas bubble disease to oxygen supersaturation (Renfro, 1963; McKee and Wolf, 1963; Woodbury,
1942; Lassleben, 1951; Faruqui, 1975), albeit at higher percent saturation values than would
apply to nitrogen or total dissolved gases. Mortality has been attributed with DO percent
34

saturation values of 200 – 410%, depending on study. However, other authors point out that
despite the frequency occurrence of oxygen supersaturation in eutrophic lakes and aquaculture
facilities, fish mortality from oxygen supersaturation is very rare (Boyd and Tucker, 1998).

Chronic effects have been noted at lower DO percent saturations under laboratory conditions
when the supersaturated condition was maintained for extended periods. For example, Doulos
and Kindschi (1990) found signs of gas bubble disease in cutthroat trout when percent DO
saturation was maintained at levels as high as 172%. Espmark and others (2010) found signs of
gas bubble disease in Atlantic salmon with continuous, multi-day exposures to DO percent
saturation levels of 160 – 220%, and McKee and Wolf (1963) cite a greater incidence of disease
in carp exposed to 150% DO saturation, compared with carp exposed to 100-125% DO
saturation. Based on these studies, a DO percent saturation of 150% is sometimes cited in the
aquaculture literature as the maximum safe level for continuous, long-term exposures. It is
unclear if similar chronic effects occur in the field, where conditions of >150% DO saturation
tend to be more variable in space and time, and fish can migrate vertically within the epilimnion.
High Rock Lake has not been observed to experience fish kills associated with gas bubble
disease, and the North Carolina Wildlife Resources Commission reports no signs of gas bubble
disease in fish from the reservoir (L. Dorsey, pers. comm., 18 Nov 2015).

In conclusion, the current chl a concentrations in High Rock Lake are correlated with favorable
surface DO concentrations, mixed effects on bottom DO concentrations, and relatively high DO
percent saturation values under some conditions. The reservoir attains water quality criteria for
DO under existing chl a conditions. However, based on the limited scientific literature available,
exceedances of 150% DO saturation for extended periods—or 200-250% for shorter periods—
might be cited as a reason for concern. Because this parameter correlates with chl a, chl a
reduction would probably also reduce the DO percent saturation values and daily variability in
this parameter.

3.2.2 pH

The acidity or alkalinity of water as measured by pH is considered a eutrophication-related
parameter because algal photosynthesis can elevate pH, especially during the day. North
Carolina’s existing pH criteria are expressed as range of 6.0 to 9.0 and lack an explicit averaging
period or return frequency. North Carolina DEQ’s current practice is to only use surface pH
measurements to assess reservoirs for pH impairment.

For the present evaluation, variation in the measured pH of surface waters in High Rock Lake
was assessed using data collected by NCDWR staff from 1981 to 2016. Monitoring typically
includes multiple measurements at different depths at established ambient monitoring stations.
On a seasonal basis, the pH of surface waters (<0.2 m) was highest during the months of June
through September (days 150-270), and pH exceeded 9.0 in 24-38% of the measurements,
depending on month (Figure 3.13, top panel). Exceedances of a pH of 9.0 occurred over the
entire range of chl a values, but were more common when chl a exceeded 30 µg/L. The line in
35

the bottom panel of Figure 3.13 connects the median pH value for each interval of 10 µg/L chl a
(0-10, >10-20, etc). The median pH value was 8.6-8.9 for chl a concentration intervals greater
than 30 µg/L. However, the frequency of pH values greater than 9.0 increased from 21.6% for
the >30-40 µg/L chl a interval to 37.5% for the >50-60 µg/L chl a interval. The frequency of pH
value greater than 9.0 for chl a intervals below 30 µg/L ranged from 4.8% to 15.2%.

The pH of waters in High Rock Lake varied with depth, consistent with the expectation that
maximum rates of photosynthesis occur near the surface of the reservoir. Figure 3.14 displays
depth versus pH based on 2011-2016 monitoring, with the dataset filter to only include pH
observations at stations and dates on which the chl a concentration exceeded 40 µg/L. For the
profiles shown, the maximum pH value occurred near the surface of the reservoir to a depth of
about 3 m for some dates and locations. The majority of the water column at the open water
stations had a pH below the existing criterion of 9.0 for all profiles of pH reported from the
ambient monitoring. Thus, there is available habitat in the mid-depth portion of the reservoir
even when the surface reading is >9.0. As part of the evaluation of the pH criterion, the Science

Figure 3.13. Measured pH in the surface layer for 1981-2016 by day of year and by chl a. The line in the
lower panel connects the median pH by chl a intervals of 10 µg/L.

36

Advisory Council evaluated the availability of habitat for aquatic life where pH was below the
existing criterion and DO was sufficient (>4 mg/L). Habitat meeting both the pH and DO
criteria was available for all dates and locations on which NCDWR conducted ambient
monitoring (SAC, 2019). This is relevant to the selection of a chl a criterion, because the oxic
zone-average pH could be maintained below 9.0 at moderate to high chl a concentrations,
whereas Figure 3.13 would indicate that maintaining the surface pH below 9.0 might not be
practicable even with very large chl a reductions.

Figure 3.14. Measured pH by depth in 2011 and 2016 for stations with reported chl a > 40 µg/L in High
Rock Lake.
3.2.3 Water Clarity
Water clarity is a measure of how deep into the water column light can penetrate. Suspended
mineral and organic particles and dissolved organic matter can affect light attenuation in surface
waters. Reduced water clarity associated with suspended sediments and algal blooms can affect
lake ecosystems by reducing the visual range in water and the light available for photosynthesis.
Impacts associated with poor water clarity include reduced visual range (fish feeding), reduced
light availability for increased water treatment costs, diminished aesthetics and recreation value,
and reduced property values (Dodds et al. 2009 and Borok, 2014). Indicators of water clarity
such as turbidity or Secchi depth can be early response variables that can indicate nutrient-related
changes to the system, particularly when algal growth affects light penetration. However,
because these indicators are also sensitive to suspended mineral sediment, increased turbidity
37

and decreased Secchi depth can also indicate sediment transport from the watershed upstream,
particularly during wet weather conditions.

Turbidity is a metric of light scattering by suspended particles that can be used as a proxy for
suspended sediment and water clarity. Secchi depth is a direct metric of visual clarity attained by
quantifying the depth of transparency in the water column. A Secchi disk is lowered into the
water column, and the Secchi depth is recorded as the depth at which the disk is no longer
visible. Thus, Secchi depth provides an indication of the transparency of the water column.
Secchi depth can be directly relevant to aesthetics, recreational uses, and fish habitat (Davies-
Colley and Smith 2001). Turbidity and Secchi depth are typically inversely related, as shown for
High Rock Lake (Figure 3.15). Currently, there are no Secchi depth criteria for NC lakes but
there is a turbidity criterion (25 nephelometric turbidity units or NTU:
https://deq.nc.gov/documents/nc-stdstable-06102019). Based on the relationship between Secchi
depth (m) and turbidity in High Rock Lake (2.12 (Turbidity)-0.47), a Secchi depth value of
approximately 0.47 m or 1.54 ft. would be similar to a turbidity value of 25 NTU, the NC lake
turbidity standard (Figure 3.15).

Figure 3.15. Secchi depth (m) vs. turbidity (NTU) in High Rock Lake based on the 2008-2009 and 2016
water quality sampling campaign.

Based on the most recent 2018 NC Category 5 Assessments "303(d) List" (approved by EPA
May 22, 2019), the water clarity in High Rock Lake is considered impaired based on turbidity
measurements in portions of the lake and its tributaries (Figure 3.16). The Yadkin River and
secchi depth (m) = 2.12(turbidity (ntu))-0.47
R² = 0.65
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
0 20 40 60 80 100 120 140
Se
c
c
h
i

d
e
p
t
h

(
m
)
Turbidity (NTU)
38

upper portion of the lake, the lower portion of the lake to Second Creek Arm, the Abbotts Creek
Arm, Second Creek, and the Yadkin River are listed as impaired for turbidity. The turbidity
impairment in High Rock Lake has been partially attributed to sediment loads, although algal
growth also contributes to the increased turbidity (Tetra Tech, 2012), particularly in the
transitional and lacustrine (downstream) segments of the lake (Rudd 2018).

The most recent assessment of High Rock Lake was based on 2016 data and included Secchi
depth and turbidity data for eight stations (HRL051, YAD152A, YAD152C, YAD156A,
YAD169A, YAD169B, YAD169E, and YAD169F) with monitoring data collected on 10 dates
from May 11, 2016- October 5, 2016 (NC DEQ, 2018) (Figure 3.15). The Secchi depth data for

Figure 3.16. Segments of High Rock Lake that are currently listed as impaired due to elevated (> 25
NTU) turbidity based on the 2018 NC 303(d) list:
https://files.nc.gov/ncdeq/Water%20Quality/Planning/TMDL/303d/2018/2018-NC-303-d--List-Final.pdf

High Rock Lake for this period ranged from 0.2-1.3 m, indicating that the clarity of the water
ranged from good to poor. The lowest Secchi depths (0.2-0.6 m) were observed at the most
upstream sampling site (HRL051), in the riverine segment of the lake. At this site, turbidity
averaged 44 NTU and was above the 25 NTU lake criterion for most (8 out of 9) of the sampling
dates, except for May 11, 2016, when turbidity levels were 23 NTU. The report stated that the
soils in the watershed are highly erodible and high sediment inputs to the lake have resulted in
deposition of sediments in the upper section of the lake that have reduced lake depth and affected
boat navigation (NC DEQ 2018).

In addition to the lake assessment, the Yadkin-Pee Dee River Basin Ambient Monitoring System
Report (NC DEQ 2012) provided a synthesis of turbidity data collected in rivers in the
watershed. The NC turbidity criterion for rivers is 50 NTU. This study found that the turbidity
39

criterion was exceeded more than ten percent of the time at 32 of the 103 monitoring stations in
the study area. Of the 103 stations monitored, only six stations had no samples that exceeded the
50 NTU threshold. They noted that episodic high turbidity values can often be associated with
rainfall events (NC DEQ, 2012). The monitoring data for stations on streams draining to High
Rock Lake showed that turbidity in the streams draining to the upper segments of the lake were
commonly elevated above the state standard. These data and the recent synthesis by Rudd (2018)
suggest that riverine sediment inputs have a large influence on lake water clarity, particularly
during storm events and in the upstream segments of the lake near the HRL 051 monitoring site.
The literature on run-of-the river reservoirs suggests that reservoirs often exhibit a longitudinal
gradient of water clarity from the riverine inflow to the outflow at the dam, as the system
transitions from riverine to lacustrine conditions. As discussed earlier, based on this gradient,
reservoirs can be divided into three zones: riverine, transitional, and lacustrine (Cooke et al.
2005) (Figure 3.2). This lake zone framework could be useful to categorize High Rock Lake
sampling stations (Figure 3.3) and assist with data interpretation of water clarity measurements
(see section 4.4.2 for additional discussion on spatial considerations regarding chl a
measurements). Longitudinal patterns in water clarity become evident when the turbidity data are
plotted versus the distance upstream from the dam (Figure 3.17). The turbidity and data suggest
that the uppermost stations: HRL051 and YAD1391A, are in the riverine zone. During high
flows YAD152A may also be in the riverine zone. The transition zone generally occurs from
YAD 152C until the YAD169A station, where the lacustrine zone begins. However, during
extreme streamflow events the riverine and transition zones may extend closer to the dam.

Figure 3.17. Lake turbidity vs. distance to the dam (2016 lake survey data).

60411653447521562707682875912669
70
60
50
40
30
20
10
0
Turbidity (NTU)
Distance to the dam (m)
40

In High Rock Lake, the relationship between chl a and water clarity is complex due to variations
in nutrient inputs, residence time and the influence of riverine sediment inputs on clarity and
light limitation in the riverine and transitional zones (Rudd 2018). The relationship between chl
a and water clarity can be more direct in the transitional and lacustrine sections of the reservoir,
during time periods when riverine inputs are low and the residence time is longer. For instance,
a comparison between chl a concentrations and Secchi depth in the transition and lacustrine zone
revealed a decline in Secchi depth with increased chl a concentrations in this zone (Figure 3.18).
However, in the riverine zone an inverse relationship between Secchi depth and chl a was present
presumably due to the influence of riverine sediment inputs on Secchi depth in that zone. In
general, in the transition and lacustrine zones, the Secchi depth was lowest during periods when
chl a was elevated, but data from some years show the opposite pattern, presumably due to
higher sediment concentration reaching these zones. These data suggest that decreased nutrient
concentrations and reduced chl a concentrations can result in an increased water clarity in the
lake, but that the improvement potential varies based on year and hydrologic conditions. For
example, reducing the chl a from the high of 73 μg/l to the current criterion of 40 μg/l would
increase the Secchi depth by approximately 0.3 m, based on the chl a-Secchi depth relation
observed in the 2016 303(d) assessment dataset (primarily 2011 data).

Figure 3.18. A comparison of the relationship between chl a and secchi depth for the upstream riverine
zone vs. the downstream transition and lacustrine zones. The data are from the 2016 HRL assessment,
which included growing season data from 2011.
Overall, these data suggest that streamflow variations have a strong influence on chl a and water
clarity in the reservoir. Riverine discharge and residence time are important variables to consider
when developing nutrient criteria for this and other NC reservoirs. In the future, modeling efforts
may help to elucidate more of the complex inter-relationships associated with discharge, nutrient
concentrations, chl a, and water clarity variability. Because of the influence of low flows on
secchi depth =
-0.0094(chlorophyll-a) + 1.24
R² = 0.46
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0 20 40 60 80
Se
c
c
h
i

d
e
p
t
h

(
m
)
Chlorophyll-a (µg/l)
Transition and
Lacustrine Zone
Riverine
41

increased residences time and elevated chl a levels, it will be important to understand the role of
dam operations and climate change on streamflow to the lake, residence time, and potential
influences on chl a exceedances.

3.3 Algal Taxonomy

3.3.1 Background and Rationale

Algal species composition is a potential indirect indicator of use attainment in High Rock Lake.
North Carolina defines biological integrity as “the ability of an ecosystem to support and
maintain a balanced and indigenous community of organisms having species composition,
diversity, population densities and functional organization similar to that of reference conditions”
(15A NCAC 02B.02020). This definition lacks a specific meaning for an artificial reservoir for
which no reference conditions are available, and North Carolina has not adopted an index of
biotic integrity (IBI) for algal assemblages. However, the SAC identified biovolume and algal
assemblage as one of the intermediate components of the conceptual model relating nutrients to
use impairment, adopted at its February 17, 2016 meeting (Fig. 3.19) (Hall, 2018). Algal data are
integral in trend analysis and in the development of NC DWR nutrient response models.

Figure 3.19. Conceptual model relating nutrients to use impairment (NC DEQ, Feb. 17, 2016)
42

Moreover, an understanding of the qualitative nature of algal blooms is essential for assessment
of their potential toxicity (Touchette et al, 2007; Vanderborgh, 2015; Hall, 2018). This section
summarizes available information on algal assemblages in High Rock Lake, and how they vary
with chl a concentration.

3.3.2 Methods and Sampling Sites

Unless otherwise noted, phytoplankton analyses were performed on whole water samples
collected from NC DEQ Ambient Lakes Monitoring Program designated sites on High Rock
Lake (Lin, 2015b) (Fig. 3.20). Algal studies were conducted by NC DEQ in the following years:
2004, 2004-2006, 2008-2010, and 2011, encompassing a total of 181 assessments. Additionally,
NC DWR staff requested a supplemental analysis of High Rock Lake samples by SAC member,
Dr. Linda C. Ehrlich, of Spirogyra Diversified Environmental Services. This analysis was
conducted by Dr. Linda C. Ehrlich on samples collected by NC DEQ on August 30, 2017. NC
DEQ staff collected whole water phytoplankton samples (fresh and Lugol’s iodine-preserved) at
the following lake sites: HRL151, YAD152C, and YAD169F. Additionally, a fresh sample was
collected in an un-named arm of the reservoir (N35.64.430 W80.28816). Phytoplankton samples
are collected according to the standard procedure for Lake Water Sample Collection described in
the NC DEQ Intensive Survey Branch SOP document (NC DEQ, 2013).

Figure 3.20. Designated NC DEQ algal sampling sites on High Rock Lake (Lin, 2015b).
43

3.3.3 Results, NC DEQ

Over the totality of its studies, NC DEQ taxonomists documented 140 unique taxa, identified to
genus or to species when possible. Although all of the major algal phyla were represented at
various levels, the three most commonly observed phyla were the Bacillariophyta, the
Cryptophyta, and the overwhelmingly predominant Cyanobacteria (Fig. 3.21) (Vanderborgh,
2015; Lin, 2015b). There was seasonal variance, though summers were consistently dominated
by high densities of cyanobacteria (up to 177,000 units/mL), comprising 69% - 96% of the total
unit density in July-September. Through the other months, January - March, unit densities were
consistently much lower (as low as 100 units/mL), and were dominated by the Cryptophyta, the
Bacillariophyta, and Ochrophyta1 (Chrysophyta), comprising 40% - 50% of total unit density.
Figure 3.23 clearly reveals the positive relationship between chl a and cyanobacterial unit density
versus the negative relationships for diatoms and green algae (Lin, 2015b).
The most common genera within the Cryptophyta were Komma and Cryptomonas, whereas,
within the Bacillariophyta, the most common genera were centric diatoms and Synedra.
However, the distinctly most influential genus was the cyanobacterium, Pseudanabaena, found
in 83% of the assessments, often comprising > 60% of total unit density (Fig. 3.24).
Possible toxigenic cyanobacteria that were observed included Pseudanabaena (83% of samples),
Microcystis (7% of samples), Aphanizomenon (17% of samples), Anabaena (Dolichospermum)
(22% of samples), and Cylindrospermopsis (44% of samples).

1 See www.algaebase.org for current taxonomic hierarchies.
44

Figure 3.21. Algal unit density of the major algal phyla in High Rock Lake, 2008-2010 (Lin, 2015b).

Figure 3.22. Correlation between algal unit density and chl a in High Rock Lake, 2005-2010 (Lin,
2015b).

Figure 3.23. Chl a and percent algal unit density in High Rock Lake, 2008-2010 (Lin, 2015b).
45

Figure 3.24. The filamentous cyanobacterium, Pseudanabaena, 1000X (Spirogyra Diversified
Environmental Services, JP Optical).

Because cyanobacteria cells are smaller than those of most other algae taxa, cyanobacteria by
density will generally be lower than by biovolume. Analysis of the 2004-2011 NC DEQ algal
data by Rudd (2018) revealed that although cyanobacteria were often dominant by density in
High Rock Lake, the sum of non-cyanobacteria algal taxa usually comprised the majority of the
algal biovolume. Cyanobacteria were a relatively minor component of the biovolume in the
samples from the riverine stations (see Fig. 3.3), but on average were over a third of the
biovolume in the samples from transitional and lacustrine stations.

3.3.4 Results, Spirogyra Diversified Environmental Services

3.3.4.1 Qualitative Observations

The phytoplankton assemblages at all four sites were mixed, though there was an immediately
observable dominance of the filamentous cyanobacterium, Pseudanabaena limnetica
(Lemmerman) Komarek C, corroborating NC DEQ results (Fig. 3.25). Algal taxa representing all
of the major algal groups (phyla), except for the Haptophyta (haptophyte flagellates) were
observed at all four sites, including the Cyanobacteria (blue-green algae), Chlorophyta (green
algae), Bacillariophyta (diatoms), Ochrophyta (Chrysophyta) (golden algae), Cryptophyta
(cryptomonad flagellates), Euglenophyta (euglenoids), and Dinophyta/Pyyrhophyta
(dinoflagellates). There were only minor spatial differences in the total numbers of taxa
observed, even though there was considerable spatial difference in sediment content, with
notably high levels of sediment in the HRL051 sample. There also appeared to be some spatial
differences in physiological health of the phytoplankton. Many of the cells in the highly turbid
HRL051 sample appeared small, deformed, and chlorotic (reduced green coloration); whereas,
cells in the un-named arm sample appeared more robust. Other visibly important, though
considerably less abundant, cyanobacterial taxa included Komvophoron sp. K. Anagnostidis & J.
Komarek, 1988 and Cylindrospermopsis phillippinensis (W.R. Taylor).

46

3.3.4.2 Quantitative Results

There were clear spatial differences in abundance of the three taxa (Fig. 3.25). Abundance was
clearly highest at YAD152C and lowest at the highly turbid HRL051. However, abundance was
only slightly higher at the low turbidity dam site than at HRL051. At the dam, nutrient limitation
may have become influential.

Figure 3.25. Phytoplankton community structure at three sites in High Rock Lake, August 30, 2017
(Spirogyra Diversified Env. Svcs.).
In conclusion, cyanobacteria dominated High Rock Lake’s algal assemblage by cell density
during the summer, but non-cyanobacteria usually comprised a majority of the biovolume.
Cyanobacteria densities and dominance were positively correlated with chl a. The algal
assemblage contained several potential toxin formers, and several of these were frequently
detected. High cyanobacteria counts are not a direct indication of impairment. Potentially
toxigenic cyanobacteria do not always produce high concentrations of toxins, and algal toxin
concentrations (addressed in the following section) are a more direct measure of potential toxic
effects. Similarly, there is no evidence that the prevailing algal assemblage is incapable of
supporting higher tropic levels, and measures of fishery health (addressed in a following section)
would be a more direct measure in that regard. Based on relationships such as that shown in
Figure 3.21), it can be stated that cyanobacteria are likely to remain a significant component of
High Rock Lake’s algal assemblage over a wide range of chl a concentrations (30 or 40 µg/L and
higher).

0
10
20
30
40
50
60
HR
Dam
HR
152C
HR
151
Un
i
t
s
/
m
L
Th
o
u
s
a
n
d
s
Spatial Distribution of Three
Cyanobacterial Taxa in High Rock Lake,
August 30, 2017
C. phiillipinensis
Komvophoron
Pseudanabaena
47

3.4 Algal Toxins
At least a dozen cyanobacterial genera have been implicated with toxin production, and at least
eight toxin groups have been characterized, of which microcystin (MCY) has been studied most
extensively (Cheung et al. 2013). However, cyanobacterial abundances (or chl a
concentrations) are not reliable indicators for the presence of cyanotoxins since not all species
within a genus produce these substances and those that can, do not do so continuously (e.g.,
Kaebernick and Neilan 2001; Loftin et al. 2016). Toxin production can be associated with
specific environmental conditions but these conditions are likely species-dependent. For
instance, MCY concentrations may be linked to increased dissolved inorganic nutrients (mainly
N and P), or more strongly associated with temperature and light levels (Codd et al. 2005; Davis
et al. 2009). A recent US-wide survey of over 1,100 lakes showed that at least one of four
common cyanotoxins could be detected in 92% of the States; all of which can harm fish,
livestock, pets and humans in varying ways (Loftin et al. 2016). Understanding the conditions
that favor cyanobacterial growth and/or toxin production is of key importance to guarantee the
safe use of freshwater systems and lakes.

For High Rock Lake, the presence and distribution of cyanotoxins was examined in a subset of
the water quality sampling stations (Fig. 3.26) during the most recent water quality assessment in
summer of 2016 (NC DEQ, 2018). Here, the common toxins that were investigated included
MCY, cylindrospermopsin (CYL), anatoxin (ANA), N-methylamino-L-alanine (BMAA) and
Saxitoxin (STX). Exposure to MCY and
CYL can impair liver function and at high
doses be lethal (Carmichael and Boyer
2016; Chorus 2000; Råbergh et al. 1991).
ANA and STX are both neurotoxins
(Cheung et al. 2013; Falconer and Humpage
2006). ANA causes an overstimulation in
neuromuscular junctions, leading to
respiratory failure (Falconer 2008). STX is
responsible for paralytic shellfish poisoning
(PSP), a condition that can cause paralysis
and death in humans (Acres and Gray 1978;
Kaas and Henriksen 2000). More recently,
BMAA has been investigated for its
connection to neurological diseases,
including amyotrophic lateral sclerosis
(ALS), Alzheimer’s disease and Parkinson’s
disease (Banack et al. 2010; Murch et al.
2004).
For the assessment, a combination of in-situ
toxin tracking devices (Solid Phase

Figure 3.26: Toxins Assessment, Sampling
locations.

48

Adsorption Toxin Tracking or SPATT; (Kudela 2011) and the collection of surface water grab
samples was used. In contrast to “grabbing” a sample and analyzing for toxins in a finite volume
of water at one specific time, the advantages of employing SPATTs comes from their higher
sensitivity in detecting low toxin levels via a time-integrative signal. Moreover, SPATTs can be
used in freshwater to marine environments, they facilitate testing for multiple toxins (depending
on the resin used), and they are easily deployed and recovered (Howard et al. 2018; Kudela
2011; Wiltsie et al., 2018). The disadvantage of using SPATTS is that the method is semi-
quantitative and average accumulation values cannot yet be linked to absolute concentrations and
therefore health risk guidelines. All cyanotoxin analyses for SPATT extracts and dissolved
samples were conducted using Enzyme-Linked ImmunoSorbent Assays or ELISAs (Abraxis
Inc.,Warminster, PA, USA). Each toxin kit allows for the detection of a specific suite of
congeners and has its specific lower detection limit (LDL): 1) MCY-ADDA (#520011) sensitive
to MCY-LR, -YR, -LF, -RR, LW, and nodularin; LDL = 0.10 µg L-1, 2) CYL (#522011)
sensitive to CYL and deoxy-CYL; LDL = 0.04 µg L-1, ) ANA (#520060); sensitive to anatoxin-a
and homoanatoxin-a; LDL = 0.1 µg L-1, 4) STX (#52255B; sensitive to STX and other paralytic
shellfish poison [PSP] toxins; LDL = 0.015 µg L-1, and (5) BMAA (#520040) sensitive to
BMAA and other amino acids; limit of quantitation = 4 µg L-1.
SPATTs were deployed at stations 051 (n =1), 152C (n = 5), 169A (n = 8), 169B (n = 2), 169E
(n = 6), and at Q6120 (n = 7) and typically replaced on a biweekly to weekly schedule (Fig.
3.26). Q6120 was located close to the intake for the Denton Water Treatment Plant south of the
dam. SPATT sampling revealed that MCY, ANA and CYL were present throughout much of the
summer and often detected simultaneously (Fig. 3.27). MCY was found across the lake while
CYL and ANA were observed at 4 and 3 out of 6 SPATT locations, respectively (Fig. 3.27).

In addition to SPATT sampling, grab samples were analyzed for absolute dissolved and
intracellular toxin concentrations at each of the stations (shown for dissolved fraction in Fig.
3.28). Running several intracellular extracts for all five toxins did not result in detectable levels
for any of the substances (n = 10), despite SPATT data indicating at least the presence of MCY,
ANA and CYL for several of the dates and locations. For the dissolved fraction, MCY and ANA
could be confirmed at a subset of stations and sampling events (Fig 3.28) but considerable
discrepancies between toxin dynamics based on SPATT versus grab samples were indicated due
Figure 3.27. SPATT toxin values for
MCY, CYL and ANA in ng toxin (g
resin) −1 d−1. Averages are shown for
multiple deployments throughout the
summer. Standard deviations (SD) absent
if less than 2 observations were made.
Note: y-axes is log-transformed due to
differing concentration ranges.

0.0
0.1
1.0
10.0
100.0
051 152C 169A 169B 169E Q6120
ng

(
g

r
e
s
i
n
)
-1
d-1
MCY
CYL
ANA
49

to detection limits. Dissolved BMAA and STX were not present during our study period based
on a subset of grab samples (n = 30 across varying sites).

The discrepancies between SPATT and grab sampling are partially explained by continued flow
that transports algae and by the “boom and bust nature” of algal blooms since both make grab
sampling a “hit or miss affair” compared to in-situ tracking. While the dissolved MCY and CYL
concentrations (Fig. 3.28) never reached EPA recreational guidelines
(https://www.epa.gov/cyanohabs), an increasing number of studies do raise questions about the
risks that might be associated with recreational exposure to chronic low-level toxins (e.g.,
swimming, boating and wading) (Backer et al. 2010; Stewart et al. 2006). This issue together
with the potential poisoning of wildlife and humans that consume toxified fish and shellfish
(Ibelings and Chorus 2007; Lehman et al. 2010) has yet to be addressed in High Rock Lake.

3.5 Other Indicators of Use Attainment in High Rock Lake

Whereas section 3.2 explored relationships between chl a and specific quantitative indicators,
this section examines other useful information on use support in High Rock, including available
knowledge on fisheries and aquatic life, potable water supply, and recreation/aesthetics. The
types of information presented in this section do not necessarily lend themselves to direct
graphical or statistical comparison with chl a concentrations. However, the associated
conclusions regarding use support (or lack thereof) can be considered in light of the reservoir’s
existing trophic status and chl a concentrations, along with other lines of evidence presented in
this document. If a use currently appears to be met, it would support the conclusion that the
reservoir’s existing chl a concentrations are supportive of that use. Conversely, information that
a use is not supported could lead to the conclusion that lower chl a concentrations would be
beneficial, if a cause-effect linkage between algal biomass and the use can be reasonably
assumed. While not a part of the sampling and analysis of pelagic algae presented here, benthic
algae are also present in High Rock Lake. At the time of writing this document a bloom of
benthic cyanobacteria, Lyngbya wollei, has been reported in HRL, which may warrant further
assessment in the upcoming years.

0.00
0.04
0.08
0.12
0.16
0.20
0 5 10
ng

t
o
x
i
n

L
-1
MCY
ANA
Figure 3.28. SPATT toxin values for MCY,
CYL and ANA in ng toxin (g resin) −1 d−1.
Averages are shown for multiple deployments
throughout the summer. Standard deviations
(SD) absent if less than 2 observations were
made. Note: y-axes is log-transformed due to
differences among concentrations for each of
the toxin types.

50

3.5.1 Fisheries and Aquatic Life

In HRL, aquatic life is managed primarily to support a sport fishery focused on largemouth bass,
striped bass, and crappie, though fishing for sunfish and catfish also occur. Support for the
fishery includes ensuring healthy populations of fish that are also safe for human consumption.
Based on assessments made by the NC Wildlife Resources Commission (NC WRC), current
water quality conditions appear to be supportive of the sport fishery. Table 3.3 summarizes the
findings of sportfish population assessments in HRL over the last decade. The largemouth
fishery has been consistently evaluated as a “quality fishery” sustained by adequate recruitment
and non-excessive mortality. Body condition of young fish has been observed to be lower than
ideal but within the normal range for other Piedmont reservoirs. Crappie also showed high
abundances with slightly lower than average body condition. Lower average body condition of
both crappie and largemouth bass is believed due to intraspecific competition that results from
high fish densities (Table 3.3), and therefore, is likely more related to fisheries management than

Table 3.3. Summary of conclusions from fisheries assessments conducted by the North
Carolina Wildlife Resources Commission for High Rock Lake over the past decade.
Species
(reference)
Survey
Year
Fishery status Growth/ Condition Recruitment/Mortality
Largemouth
bass
(NC WRC
2007)
2006 Quality fishery Relative weight of some year classes
not ideal but within normal range for
piedmont reservoirs
As expected, and no
apparent negative
impacts on population
Crappie
(NC WRC
2008)
2006 High densities of
black and white
crappie
Good body condition but somewhat
slow growth for black crappie,
potentially due to high density and
intraspecific competition
Weak recruitment
during 2002 during
drought
Striped bass
(NC WRC
2009)
2006 Fast growth with excellent body
condition
Recruitment due to
stocking. Few large (>
year 3) fish caught,
believed due to small
gill net size used
Largemouth
bass
(NC WRC
2011)
2009 Quality fishery Average growth for piedmont
reservoirs. Relative weight of younger
fish not ideal, but at or above levels in
other Piedmont reservoirs. Suspected
cause intraspecific competition from
higher than average density
As expected, and no
apparent negative
impacts on population
Crappie
(NC WRC
2012)
2009 Survey catch below
normal, suspected
cause was high
turbidity from high
river inputs
Slower than average growth,
suspected due to high density and
intraspecific competition

Largemouth
bass
(NC WRC
2013)
2012 Quality fishery Relative weights of younger fish
slightly less than expected. Suspected
due to high density and intraspecific
competition
Well balanced age
structure. Adequate
reproduction and
mortality is not
excessive
51

to water quality conditions. As in most NC piedmont reservoirs, striped bass do not reproduce in
HRL due to high temperature and low hypolimnetic dissolved oxygen conditions (L. Dorsey, NC
WRC personal communication). Annual stocking of 89,000 fingerlings maintain the HRL
population of striped bass. The 2006 striped bass survey indicated that striped bass grow fast in
HRL and maintain a high body condition for longer than average as they age compared to other
piedmont reservoirs. Estimation of the number of older (> 3 year) striped bass abundance has
been hampered by sampling biases. Fish kills are uncommon in HRL, and large fish kills have
only been noted during the major drought of 2002 when low flows, low water levels, high
summer temperatures, and low dissolved oxygen caused major fish kills (L. Dorsey, NC WRC
personal communication).

As noted in chapter 2, the relationship between fishery production and chl a is generally positive
between 0 and about 100 μg/L (Bachmann et al. 1996; Deines et al. 2015). Currently, chl a
averages about 50 μg/L in the most production region of HRL near station YAD152C. Reducing
chl a to meet a new criterion may cause some decrease in fisheries production. However, there is
a huge degree of variation in the relationship between lake productivity and fisheries, and there
are many examples of lakes with highly productive fisheries with chl a concentrations much
lower than 40 μg/L. Studies of fisheries in Alabama and Georgia reservoirs have found that chl a
concentrations of 10-15 μg/L supported fisheries that were as productive as more eutrophic lakes
and also maintained high water clarity desirable for recreation (Maceina et al. 1996; Bayne et al.
1994). The SAC views the risk of a potential modest reduction in fisheries production an
acceptable tradeoff for the reduction in risks associated with the current high level of
phytoplankton biomass (e.g. potential for cyanobacterial blooms and toxin production).

Harmful effects on fish by cyanotoxins with subsequent consumption by fishermen is also a
potential concern, particularly due to the high levels of cyanobacteria biomass. In HRL, this risk
has not been fully assessed. Low resolution sampling (monthly) for total MCY (intracellular and
dissolved) in summer of 2002 (Touchette et al. 2007) and for accumulated dissolved toxins using
a field tracking approach in 2016 (see 3.2.5.) indicated concentrations < 1 μg MCY /L. Limited
data on toxin ranges, maxima and temporal dynamics are the presumed reason for a virtual
absence of a relationship between chl a and any of the cyanotoxins observed in the southeast US
(Chapter 2). Any refinement of chl a criterion, established to minimize the risks posed by
cyanotoxins including fish intoxication, will depend on more comprehensive measurements of
toxins in lake water as well as animals.

Submerged aquatic vegetation (SAV) is an aquatic life that is commonly protected by chl a
criteria. SAV, however, are not present in HRL probably due to a combination of poor water
clarity and highly variable water level. High phytoplankton biomass contributes significantly to
poor water clarity in HRL with Secchi disk depths rarely more than 1 m (see section 3.3.3).
However, high concentrations of suspended sediment also contribute significantly to low water
clarity and large fluctuations in water level would likely inhibit SAV colonization in the absence
of high phytoplankton biomass due to periodic desiccation of suitable benthic habitats. Lack of
52

existing SAV and a hydrologic regime unfavorable for their development renders a chl a
criterion to protect SAV irrelevant for HRL.

3.5.2 Potable Water Supply

High Rock Lake is designated as Class WS-IV (waters protected as water supplies). (See, 15A
NCAC 02B .0301). In determining the suitability of waters for use as a source of water supply
for drinking, culinary or food processing purposes after approved treatment, the Commission will
be guided by the physical, chemical, and bacteriological maximum contaminant levels specified
by Environmental Protection Agency regulations. As noted, the suitability of water supplies are
evaluated after treatment. In practice, potable water supplies are evaluated at the point of a
potable water intake and take into account the treatment provided in evaluating whether uses are
attained.

There are no potable water intakes in HRL. Consequently, HRL is not being used as a potable
water supply. Consequently, a direct assessment of use attainment is not possible. However,
there is a potable water intake located downstream. The Town of Denton Water Treatment Plant
(WTP) is located downstream of the dam on HRL and takes its water supply from the
Tuckertown Reservoir, the next downstream lake on the Yadkin River. The intake is located only
about 0.5 mile downstream of the High Rock Lake dam, and much of the water at that location
was recently released from High Rock Lake. The WTP employs conventional water treatment
processes including coagulation, flocculation, settling, activated carbon filtration, and
disinfection. Although chl a levels in HRL are routinely elevated during the growing season,
staff at the Denton WTP do not report that the reservoir has been unavailable as a source for
potable water due to chl a level. Rather, the conventional treatment processes have been capable
producing a high quality potable water. The Town does report the need to carefully monitor the
quality of the raw water supply—especially with regard to turbidity from high flow and seasonal
turnover—and adjust treatment processes accordingly.

The chl a concentration of water does not directly affect its use as a potable water supply. Rather,
chl a or the presence of algal cells would be considered in a similar fashion to secondary
drinking water standards. Secondary drinking water standards apply to contaminants that are not
health threatening but may affect color, taste and odor, or have other undesirable effects.
Conventional potable water treatment facilities include processes to remove algal cells and their
associated chl a prior to use. Consequently, even if chl a levels are elevated, adjustments can be
made without the need for additional facilities. Operations and maintenance (O&M) costs may
be affected.

Source water chl a concentration, at the point of intake to a potable water treatment system,
influences the potential cost of treatment to prepare the water for potable use, but normally does
not prevent its use as a potable water supply. Treatment requirements for potable water supplies
that originate from surface waters, such as lakes and rivers, are highly regulated by USEPA.
Under the Safe Drinking Water Act (SDWA) the EPA Office of Water (EPA-OW) is charged
53

with setting water quality standards and regulations to protect the public drinking water supply.
These requirements impose treatment strategies at all potable water treatment facilities that are
readily able to control particulates. The regulatory basis for these treatment strategies is
presented in Attachment B of the pH criteria document proposed by the SAC for HRL (NC SAC,
2018).

As discussed in Attachment B to the proposed pH criteria, potable water supplies, which use
surface water as a source, must provide treatment to settle and filter waterborne disease-causing
contaminants, and provide disinfection. The chemicals used in treatment to enhance particulate
removal will remove chl a before the treated water is provided for use.

3.5.3 Aesthetics, Swimming

Aesthetic and swimming uses may be adversely affected by chl a concentrations due to the
recreating public’s perception of color, turbidity, and/or water clarity (Secchi depth) associated
with specific concentrations of chl a. Information provided to the SAC suggests that public
perception is highly dependent upon the experience of the population using the lake. More
generally, the literature shows that public expectations of lake clarity and color have very large
regional variations, based on the conditions to which users are accustomed (e.g., Burden and
Malone, 1987; Smeltzer and Heiskary, 1988). It can also be reasonably expected that user
perceptions would be influenced by the form of algal growth in a reservoir; i.e., highly visible
scums or mats could elicit more user complaints than dispersed growths of the same biomass.

In the case of HRL, the SAC is not aware of any aesthetic or swimming use impairment of the
lake, even though chl a concentrations routinely exceed 50 µg/L. Most phytoplankton growth in
the reservoir is relatively dispersed rather than occurring as highly visible scums or mats, and
SAC was not provided with any information to indicate that user complaints are common. In
September 2019, the Davidson County Health Department investigated a complaint and
confirmed the presence of a benthic cyanobacteria (Lyngbya wollei) in the reservoir. Information
on the location and extent of the taxa was not available to the SAC, so it could not be determined
whether it was restricted to a single cove area versus more widely-occurring. Regardless,
because Lyngbya is a benthic alga, it would not be directly measured by water column chl a.

The contribution of chl a to water clarity was discussed in section 3.2.3. This section concluded
that although water clarity was dominated by suspended sediment in much of the reservoir, chl a
reduction from ~70 to ~40 ug/L could cause modest increases (0.1 – 0.3 m) in Secchi depth in
parts of the reservoirs in some years or hydrologic conditions.

54

3.6 References

Acres, J., and J. Gray. 1978. Paralytic shellfish poisoning. Canadian Medical Association journal
119: 1195-1197.
AlgaeBase.org. 2019. Global algal web database.
https://www.algaebase.org/browse/taxonomy/detail/?taxonid=97243&-
session=abv4:AC1F036516a2f00ABCLR8F6B32A0 accessed 9/26/19.
American Public Health Association (APHA), American Water Works Association (AWWA),
and Water Environment Federation (WEF). 2017. Standard Methods for the Examination of
Water and Wastewater. E.W. Rice, R.B. Baird, and A.D. Eaton (eds.). 23rd Edition,
Washington, D.C.: APHA.
Bachmann, R.W., Jones, B.L., Fox, D.D., Hoyer, M.V., Bull, L.A., Canfield, D.E. Jr. 1996.
Relations between trophic state indicators and fish in Florida (U.S.A.) lakes. Canadian
Journal of Fisheries and Aquatic Science 53: 842-855
Backer, L. C. and others 2010. Recreational exposure to microcystins during algal blooms in two
California lakes. Toxicon 55: 909-921.
Banack, S. A., T. A. Caller, and E. W. Stommel. 2010. The cyanobacteria derived toxin Beta-N-
methylamino-L-alanine and amyotrophic lateral sclerosis. Toxins (Basel) 2: 2837-2850.
Bayne, D.R., Maceina, M.J., Reeves, W.C. 1994. Zooplankton, fish, and sport fishing quality
among four Alabama and Georgia reservoirs of varying trophic status. Lake and Reservoir
Management 8: 153-163.
Borok, A. 2014. Turbidity Technical Review: Summary of Sources, Effects, and Issues Related
to Revising the Statewide Water Quality Standard for Turbidity. Oregon Department of
Environmental Quality. Portland, OR.
Boyd, C.E., and Tucker, C.S. 1998. Pond Aquaculture Water Quality Management. Kluwer
Academic Publishers, Boston MA. 711 p.
Burden, D.G., and Malone, R.F. 1987. A classification of freshwater Louisiana lakes based on
water quality and user perception data. Environ Monit Assess. 1987 Sep;9(2):179-93.
Carmichael, W. W., and G. L. Boyer. 2016. Health impacts from cyanobacteria harmful algae
blooms: Implications for the North American Great Lakes. Harm. Algae 54: 194-212.
Cheung, M. Y., S. Liang, and J. Lee. 2013. Toxin-producing cyanobacteria in freshwater: A
review of the problems, impact on drinking water safety, and efforts for protecting public
health. Journal of Microbiology 51: 1-10.
Chorus, I. 2000. Health risks caused by freshwater cyanobacteria in recreational waters. Journal
of Toxicology and Environmental Health, Part B 3: 323-347.
Codd, G. A., L. F. Morrison, and J. S. Metcalf. 2005. Cyanobacterial toxins: risk management
for health protection. Toxicol. Appl. Pharmacol. 203: 264-272.
Cooke, D., E.B. Welch, S.A. Peterson, & Nichols, S.A. (2005). Restoration and Management of
Lakes and Reservoirs. Taylor and Francis Group. Boca Raton, FL.
Davis, T. W., D. L. Berry, G. L. Boyer, and C. J. Gobler. 2009. The effects of temperature and
nutrients on the growth and dynamics of toxic and non-toxic strains of Microcystis during
cyanobacteria blooms. 8: 715-725.
55

Davies-Colley, R. and Smith, D. 2001. Turbidity, suspended sediment, and water clarity: A
review. Journal of the American Water Resources Association 37(5):1085-1101.
Deines, A.M., Bunnell, D.B., Rogers, M.W., Beard, T.D. Jr., Taylor, W.H. 2015. A review of the
global relationship among freshwater fish, autotrophic activity, and regional climate. Rev
Fish Biol Fisheries 25:323-336.
Dodds, W., Bouska, W., Eitzmann, J., Pilger, T., Pitts, K., Riley, A., Schloesser, J., and
Thornbrugh, A. 2009.Eutrophication of U.S.freshwaters: Analysis of potential economic
damages. Environ. Sci. Technol. 43 (1): 12-19.
Doulos, S.K., and Kindschi, G.A. 1990. Effects of oxygen supersaturation on the culture of
cutthroat trout, Oncorhynchus clarki Richardson, and rainbow trout, Oncorhynchus mykiss
Richardson. Aquaculture Research 21 (1), p. 39-46.
Espmark, A.M., Hjelde, K., and Baeverfjord, G. 2010. Development of gas bubble disease in
juvenile Atlantic salmon exposed to water supersaturated with oxygen. Aquaculture 306(1):
198-204.
Falconer, I. R. 2008. Health effects associated with controlled exposures to cyanobacterial
toxins, p. 607-612. In H. K. Hudnell [ed.], Cyanobacterial Harmful Algal Blooms: State of
the Science and Research Needs. Springer New York.
Falconer, I. R., and A. R. Humpage. 2006. Cyanobacterial (blue-green algal) toxins in water
supplies: Cylindrospermopsins. Environmental Toxicology 21: 299-304.
Faruqui, A. M. 1975. Fluctuations in oxygen concentration and occurrence of mortality of carp
hatchlings in a hatchery pond at Parta Fish Farm, Bhopal. Broteria Series Trimester Ciencias
Naturale 44:67-79.
Hall, W. 2018. Evaluation of Recommendations for High Rock Lake Criteria – Chl a by Clifton
Bell (March 13, 2018). Technical document. Hall & Assoc., Washington, DC 9 pp.
Howard, D. A., K. Hayashi, J. Smith, R. M. Kudela, and D. Caron. 2018. Standard operating
procedure for Solid Phase Adsorption Toxin Testing (SPATT) assemblage and extraction of
HAB toxins.
Ibelings, B. W., and I. Chorus. 2007. Accumulation of cyanobacterial toxins in freshwater
“seafood” and its consequences for public health: A review. Environ. Pollut. 150: 177-192.
Kaas, H., and P. Henriksen. 2000. Saxitoxins (PSP toxins) in Danish lakes. Water Research 34:
2089-2097.
Kaebernick, M., and B. A. Neilan. 2001. Ecological and molecular investigations of cyanotoxin
production. FEMS Microbiol. Ecol. 35: 1-9.
Komarek, J. and K. Anagnostidis. 2005. ‘Cyanoprokaryota II. Oscillatoriales’, in Süβwasserflora
von Mitteleuropa, Band 19/2, Koeltz, Scientific Books, Koenigstein.
Kudela, R. M. 2011. Characterization and deployment of Solid Phase Adsorption Toxin Tracking
(SPATT) resin for monitoring of microcystins in fresh and saltwater. Harm. Algae 11: 117-
125.
Lassleben, P. 1951. Is supersaturation with oxygen dangerous? Fischbauer 2, 105; Water
Pollution Abstracts, 25:6.
Lehman, P. W., S. J. Teh, G. L. Boyer, M. L. Nobriga, E. Bass, and C. Hogle. 2010. Initial
impacts of Microcystisaeruginosa blooms on the aquatic food web in the San Francisco
Estuary. Hydrobiol. 637: 229-248.
56

Lin, J. 2015. High Rock Lake Data Review. Presentation to North Carolina Nutrient Science
Advisory Council, August 18, 2015. 95 p.
Lin, J. 2015a. High Rock Lake Nutrient Response Model. PowerPoint presentation given at
NCDP, August 18, 2015. Division of Water Resources – Water Planning, NC Department of
Environment and Natural Resources, Raleigh, 6 pp.
Lin, J. 2015b. High Rock Lake Data Review. PowerPoint presentation given at NCDP, August
18, 2015. Division of Water Resources – Water Planning, NC Department of Environment
and Natural Resources, Raleigh, 31 pp.
Loftin, K. A. and others 2016. Cyanotoxins in inland lakes of the United States: Occurrence and
potential recreational health risks in the EPA National Lakes Assessment 2007. Harm. Algae
56: 77-90.
Maceina, M.J., Bayne, D.R., Hendricks, A.S., Reeves, W.C., Black, W.P., DiCenzo, V.J. 1996.
Compatibility between water clarity and quality black bass and crappie fisheries in Alabama.
American Fisheries Society Symposium 16: 296-305.
McKee, J. E., and H. W. Wolf. 1963. Water quality criteria, second edition. Publication No. 3-A,
State Water Quality Control Board; Sacramento, CA.
Murch, S. J., P. A. Cox, S. A. Banack, J. C. Steele, and O. W. Sacks. 2004. Occurrence of β-
methylamino-l-alanine (BMAA) in ALS/PDC patients from Guam. Acta Neurologica
Scandinavica 110: 267-269.
NC DEQ, Division of Water Quality. 2012. Yadkin-Pee Dee River Basin Ambient Monitoring
System Report. Raleigh, North Carolina: North Carolina Department of Environment and
Natural Resources.
NC DEQ. 2018. Lake and Reservoir Assessments: Yadkin-Pee Dee River Basin. North Carolina
Department of Environmental Quality. Raleigh, North Carolina.
NC DEQ. 2013. Intensive Survey Branch Standard Operating Procedures Manual: Physical and
Chemical Monitoring. Technical Document. NC Division of Environmental Quality, Intensive
Survey Branch, Raleigh, NC 132 pp.
https://files.nc.gov/ncdeq/Water%20Quality/Environmental%20Sciences/ISU/ISB%20SOP%
20Version2.1%20%20FINAL.pdf accessed 9/26/19.
NC DEQ 2019. Algal and Aquatic Plant Program. Webpage description of the Ambient Lakes
Monitoring Program, Water Sciences Section, NC DEQ, Raleigh.
https://deq.nc.gov/about/divisions/water-resources/water-resources-data/water-sciences-
home-page/intensive-survey-branch/ambient-lakes-monitoring accessed 9/26/19.
NC SAC. 2018. Recommendations for pH Criteria in High Rock Lake. 44 p.
NC WRC. 2009. High Rock Lake Striped Bass Survey, 2006. Thompson, T. North Carolina
Wildlife Resources Commission Division of Inland Fisheries. Raleigh, NC. 2009.
NC WRC. 2009. High Rock Lake Crappie Survey, 2006. Thompson, T. North Carolina Wildlife
Resources Commission Division of Inland Fisheries. Raleigh, NC. 2009.
NC WRC. 2007. High Rock Lake Largemouth Bass Survey, 2006. Dorsey, L. North Carolina
Wildlife Resources Commission Division of Inland Fisheries. Raleigh, NC. 2007.
NC WRC. 2011. High Rock Lake Largemouth Bass Survey, 2009. Dorsey, L. North Carolina
Wildlife Resources Commission Division of Inland Fisheries. Raleigh, NC. 2011.
57

NC WRC. 2012. High Rock Lake Crappie Survey, 2009. Thompson, T. North Carolina Wildlife
Resources Commission Division of Inland Fisheries. Raleigh, NC. 2012.
NC WRC. 2013. High Rock Lake Largemouth Bass Survey, 2012. Dorsey, L. North Carolina
Wildlife Resources Commission Division of Inland Fisheries. Raleigh, NC. 2013.
Råbergh, C. M. I., G. Bylund, and J. E. Eriksson. 1991. Histopathological effects of microcystin-
LR, a cyclic peptide toxin from the cyanobacterium (blue-green alga) Microcystis aeruginosa
on common carp (Cyprinus carpio L.). Aquatic Toxicology 20: 131-145.
Renfro, W. C. 1963. Gas-bubble mortality of fishes in Galveston Bay, Texas. Trans. Am. Fish.
Soc. 92:320–322.
Rudd, M. 2018. An Evaluation of Water Quality Parameters and Flow Dynamics in High Rock
Lake, North Carolina to Assist in the Development of Nutrient Criteria for Lakes and
Reservoirs in the State. Masters in Environmental Management Report. Duke University,
Durham, NC.
Smeltzer, E. and Heiskary, S. A. 1990. Analysis and applications of lake user survey data. Lake
and Reservoir Management 6(1), 109-118.
Stewart, I. and others 2006. Epidemiology of recreational exposure to freshwater cyanobacteria –
an international prospective cohort study. BMC Public Health 6: 93.
SWCS Environmental Consultants. 2010. East Canyon Reservoir and East Canyon Creek
TMDL. Report prepared for the Utah Division of Water Quality. 304 p
Tetra Tech 2012. High Rock Lake Watershed Model. Tetra Tech, Research Triangle Park, NC.
Tetra Tech 2015. Analysis Report For Classification and Exploratory Analysis of North Carolina
Lakes Data for the Nutrient Scientific Technical Exchange Partnership and Support (N-
STEPS). Tetra Tech, Research Triangle Park, NC.
Touchette, B.W., Burkholder, J.M., Allen, E.H., Alexander, J.L., Kinder, C.A., Brownie, C.,
James, J., Britton, C.H. 2007. Eutrophication and cyanobacteria blooms in run-of-river
impoundments in North Carolina, U.S.A. Lake and Reservoir Management, 23: 179-192.
USEPA. 1986. Ambient Water Quality Criteria for Dissolved Oxygen. EPA 440/5-86-003. 46 p.
Weitkamp, D.E. and Katz, M. 1980. A Review of Dissolved Gas Supersaturation Literature,
Transactions of the American Fisheries Society, 109:6, 659-702.
Wiltsie, D., A. Schnetzer, J. Green, M. Vander Borgh, and E. Fensin. 2018. Algal Blooms and
Cyanotoxins in Jordan Lake, North Carolina. Toxins (Basel) 10: 92.
Woodbury. L. A. 1942. A sudden mortality of fishes accompanying a supersaturation of oxygen
in Lake Waubesa, Wisconsin. Transactions of the American Fisheries Society 71:112–117.

58

4. A Proposed Site-Specific Chlorophyll a Criterion for High Rock Lake
This section presents the SAC’s recommendation for a site-specific chlorophyll a (chl a) criterion
to protect the designated uses of High Rock Lake from excessive nutrient-driven enhanced
primary productivity. The proposed criterion would minimize potential nutrient-driven adverse
effects over short- and long-time scales, equating to impacts that are acute and chronic in this
man-made reservoir (see Section 4.2.1.). Literature presented in chapter 2 and the reservoir-
specific observations in chapter 3 were used to develop the recommended site-specific chl a
criterion.

Water quality standards consist of designated uses, parameter-specific criteria to protect those
uses, and antidegradation policies. The SAC is not recommending changes to the designated
uses or antidegradation policies that currently apply to the waters of High Rock Lake; rather, the
focus of this proposal is on a site-specific chl a criterion. Subsections below describe the
designated uses of waters of High Rock Lake and recommendations on how the chl a criterion is
expressed in terms of the temporal (e.g. duration, frequency), spatial, and magnitude components
of the criterion.

4.1 Designated Uses for High Rock Lake

The waters of High Rock Lake are classified in the water quality standards regulations of North
Carolina as WS-V Class B waters in upstream reaches or WS-IV Class B waters in downstream
reaches (15A NCAC 02B .0309). The Class B designation requires protection of primary and
secondary recreation, fishing, aquatic life including propagation and survival, and wildlife (15A
NCAC 02B .0219). The water supply designations (WS-IV and WS-V) protect waters as water
supplies in moderately to highly developed watersheds. The water supply designations require
local programs to control nonpoint sources and stormwater discharges for WS-IV waters and
may apply appropriate management requirements in WS-V waters, as deemed necessary, for the
protection of downstream receiving waters per 15A NCAC 2B .0203.

The key components of the designated uses for classifications applied to High Rock Lake that
may be impacted by nutrient-driven enhanced primary production are primary recreation,
fishing/aquatic life, and water supply. For recreational activities, protection of primary
recreation activities, which includes swimming on a frequent or organized basis, also would be
protective of secondary recreation and fishing activities. Further, protection of primary
recreation and aquatic life would be protective of wildlife uses around the margins of High Rock
Lake. For the aquatic life use, propagation of species naturally occurring in the man-made
system and the overall productivity and diversity of the sport fishery, as an indication of healthy
transfer of primary production to apex predators, are the primary considerations. The use of
apex predator species as an indicator of overall aquatic life protection was used in the rationale
developed for protection of aquatic life in Missouri reservoirs (MDNR, 2017). For this
application for High Rock Lake, the productivity and diversity of multiple trophic levels were
considered in combination with available information on the site-specific fisheries described in
section 4.4.2.
59

4.2 Temporal Components

The temporal components of how a water quality criterion is expressed include both duration
(averaging period, which for chl a focused on seasonal considerations) and allowable frequency
of exceedance. These components are discussed in subsections below.

4.2.1 Duration Components

Water quality studies to assess nutrient-driven enhanced productivity in natural and man-made
systems have shown that both short-term acute impacts (fish kills, algal toxins, etc.) and long-
term enhanced productivity, with potential shifts in the species assemblage present, can occur in
different systems (e.g. USEPA, 2000). The development of the temporal component for a site-
specific chl a criterion should consider how the key designated uses described above in Section
4.1 may be impacted on an acute and chronic basis. In general, acute effects can be associated
with algal toxins or with depletion of dissolved oxygen due to the decay of large algal blooms.
For High Rock Lake, the algal assemblage during the growing season often has a high proportion
of cells contributed by species of cyanobacteria (see Section 3.2.4). A limited number of
measurements to date indicate that algal toxins are present but at a relatively low concentration
(see Section 3.2.5). As discussed in Chapter 3, it is important to note that these observations are
mainly limited to biweekly measurements of dissolved toxins during the summer of 2016. The
sample resolution may not be representative of peak bloom conditions when toxin concentrations
(both dissolved and intracellular) tend to reach their maxima nor can any conclusions be drawn
in regard to year to year variability. The abundance of algae during the growing season is
typically high, and periods of depleted dissolved oxygen in deeper waters of the reservoir have
been reported when bottom waters become isolated from surface waters due to thermal
stratification (see Section 3.2.1). It is unclear, however, as to the extent that elevated levels of
nutrient-driven productivity contribute to dissolved oxygen depletion compared to the thermal
isolation of bottom waters during warm season conditions. Due to a lack of clear nutrient-driven
acute effects in High Rock Lake, the SAC chose to focus criterion development efforts on
longer-term measures of the reservoir’s trophic state.

The potential long-term or chronic effects of nutrient-driven enhancement of primary production
would be evaluated with a seasonal geometric mean (geomean). The objective of the criterion
would be to assess the central tendency of chl a concentrations over time for stations included in
each assessment unit. The use of a geomean for the proposed criterion is due to the geomean
being the best measure of central tendency for log-normally distributed parameters such as chl a
(USEPA, 2010). It is proposed that the geomean be calculated with data collected during the
growing season (April-October), as an indication of overall algal production and representative
of the time of maximum productivity in High Rock Lake, since chl a concentrations in High
Rock Lake are typically higher during the growing season than in other months of the year (see
Section 3.2.4). Utilizing data from the growing season is appropriate to assess reservoir trophic
60

status and the general potential for algal-related effects. Overall, the reduction of the long-term
central tendency for chl a would also reduce the frequency of elevated chl a values over time.

Use of a geomean statistic to express the proposed chl a criterion is also consistent with approved
water quality criteria for chl a in other states. Examples of states that have adopted chl a criteria
expressed as a geomean include Arkansas, Florida, Texas, and Virginia. While the current
SAC’s analysis focused on the current science supporting the development of a geomean
criterion, the expression of the criterion as a geomean is also consistent with the historical
discussions related to the development of the existing instantaneous chl a criteria for North
Carolina.2

The proposed chl a criterion is intended to serve as an indicator of average algal growth during
the growing season. Therefore, the SAC recommends sufficient data be collected to provide a
representative average for the growing season, including samples collected in at least five
different growing season months for each year of data included in the analysis. Additional
discussion and SAC recommendations on the use of data from more than one year is included in
the following section.

4.2.2 Frequency of Exceedance

Water quality criteria have allowable frequencies of exceedance to acknowledge natural
variability and the fact that aquatic life can recover from periodic exceedances. Some states have
adopted specific allowable frequencies of exceedance for chl a criteria expressed as a geometric
mean (geomean). For example, Florida’s criteria for lakes, reservoirs, and estuaries may not be
exceeded more than once in three years. Florida adopted chl a criteria with an 20 percent
probability of exceedance in any given year, and used binomial statistics to demonstrate that a 1-
in-3 exceedance frequency would limit the probability of a Type I error (false finding of
impairment) to 10 percent (FDEP, 2012).

Similarly, Virginia and Missouri use a version of a once in three-year exceedance frequency
approach for chl a criteria in lakes and reservoirs (e.g. 9VAC25-260-187), which is based on a
magnitude tied to a single year’s computed mean. Minnesota has adopted multi-year average
criteria for total phosphorus, chl a, and Secchi depth in lakes and reservoirs (MAR 7050.0222).
Water bodies are considered impaired for phosphorus if the phosphorus criterion is exceeded and
either the chl a criterion or Secchi depth criterion (or both) are exceeded. Because the criteria are
expressed as long-term summer averages, values are computed by aggregating summer data
collected over multiple years. Minnesota uses a period as long as ten years for assessments
because it provides reasonable assurance that data will have been collected over a range of
weather and flow conditions and that all seasons will be adequately represented (MPCA, 2018).
All of the criteria components of these approaches have been approved by USEPA.

2 The chair of the advisory group that recommended North Carolina’s existing chl a criterion confirmed the intent of
the 40/15 standards were based on “growing season” averages and not any time / any place standards (Mike
McGhee, elec. comm., May 10, 2009).
61

The SAC considered the existing data collection efforts by NCDWR in considering potential
frequency approaches for the proposed chl a criterion. For many lakes and reservoirs in North
Carolina, monitoring data are collected approximately monthly during the growing season as part
of the ongoing ambient monitoring program in a single year during each five-year assessment
period. Limited available data with which to assess compliance with a seasonal geomean
criterion for chl a presents an obvious challenge to considering a frequency component to the
criterion. The most common frequencies used by states are instantaneous or a frequency based
on some limited number of exceedances, which as described above, is typical for chl a criteria.

The SAC recommends data incorporated into the assessment be collected in two or more years to
incorporate year-to-year variability in chl a concentrations (see Table 4.1). The SAC considered
two options to evaluate compliance with the seasonal geomean criterion: (1) computing the
geometric mean for each year of individual data and applying a frequency component of not
more than one exceedance out of three years of data; or (2) computing a multi-year geometric
mean by aggregating data from at least two years within the assessment period. The multi-year
geometric mean would be considered a not-to-exceed value. The SAC’s criterion discussions did
not include an explicit maximum number of years to be included in a calculated multi-year
geometric mean. The SAC’s agreement from December 2018 cited the use of data from “the
assessment period,” which corresponds to an implicit maximum of five years. Some SAC
members expressed concerns that if multi-year averaging periods were too long, the assessment
would have a more difficult time detecting eutrophication-related problems in the reservoir.
Some SAC members also discussed the fact that a three-year averaging period would have the
closest statistical correspondence to a single-season, 1-in-3 year allowable exceedance approach.
The recommendation from the SAC is to utilize the exceedance frequency approach, and
recommended a maximum exceedance frequency of no more than one-in-three.

In cases when data are only available for a single year within an assessment period, data from
previous assessment periods could be used in order to complete the assessment. This is
consistent with North Carolina’s existing practice for some other parameters, and the SAC would
support this practice up to a total assessment period of 10 years. The SAC also recommends
additional sampling be undertaken to add a third year of sampling when the data are needed to
assess the maximum one-in-three exceedance frequency. The additional year of sampling would
provide nearer term information regarding the current health of the lake to help conclude whether
the criterion is met (i.e. only one of the three geometric mean year values exceed 35 µg/L) or not
(i.e. two of the three geometric mean year values are greater than 35 µg/L). No additional
sampling would be added if both existing seasonal geomean chl a values are below 35 µg/L or
both existing seasonal geomean values are above 35 µg/L. This approach is recommended by
the SAC in that it adds additional sampling only in instances when the data are needed to assess
the one-in-three maximum exceedance frequency.

62

4.3 Criterion Magnitude

The magnitude component of the chl a criterion is more challenging to derive than for
constituents that display a simple dose-response relationship with designated uses. In some
settings, development of precise, quantitative relationships between chl a and indicators of
designated use impairment may be possible, and a magnitude could be selected to limit identified
response indicators from exceeding specific thresholds. However, the SAC’s comprehensive
examination of relationships between chl a and potential indicators in High Rock Lake (see
Section 3) did not identify dose-response relationships upon which a chl a criterion could be
based. In fact, High Rock Lake exhibits a combination of favorable indicators and indicators of
potential concern. With the understanding that scientific judgment would be required, the SAC
adopted the following general approach for deriving a site-specific chl a criterion magnitude for
High Rock Lake:

1. An extensive review of literature was conducted to define the ranges of chl a
concentration in natural and man-made systems that have been interpreted to be protective
of designated uses potentially impacted by a high abundance of algae (see Section 2). This
review culminated in the decisions made by the SAC at its December 2018 meeting. 3

2. The current conditions of High Rock Lake were evaluated, with an emphasis on current
chl a levels, on relationships between chl a and indicator parameters, and on evidence for
algal-related impacts to designated uses (see Section 3).

3. The results of steps 1 and 2 were synthesized to develop chl a concentration range that
was deemed to support designated uses in water bodies similar to High Rock Lake. At the
December 2018 SAC meeting, a chl a criterion magnitude was selected from this range. 4

4. A Monte Carlo analysis was performed to confirm that attainment of the recommended
criterion would protect the reservoir’s fishery and result in a low rate of exceedance of the
upper end of the acceptable chl a range.

The results of steps 3 (range derivation) and 4 (Monte Carlo analysis) are provided in the
following subsections along with the specification of the SAC recommended criterion
magnitude.

3 The summary of the group’s basis states that the “literature supports recreation, aquatic life and drinking waters
uses are achieved when chla is 20-40 µg/L.”
4 The magnitude summary states “35 µg/L to support average chl a levels throughout High Rock Lake of 20-25
µg/L, derived from 25-40 µg/L range for warmwater reservoirs.” The 35 µg/L was “near the upper end of the range
selected due to mostly favorable use indicators.”
63

4.3.1 Derivation of a Chl a Protective Range

The literature review identified relatively wide ranges for chl a that have been supportive of
designated uses in different aquatic systems (Figure 4.1). The target range highlighted represents
a range for High Rock Lake that protects the water supply use, primary recreation (if algal toxins
can be presumed low; see Chapter 3), and apex predator productivity as an indication of aquatic
life protection (see below). In terms of site-specific observations, the existing condition of High
Rock Lake supports a thriving sport fishery for apex predators and with no surface, scum-
forming algal species (see Chapter 3 for details). These observations, in combination with the
literature review, were used to derive a chl a protective range of 25-40 µg/L for warmwater
reservoirs similar to High Rock Lake.

Figure 4.1. Proposed chl a concentration (µg/L) ranges by designated use. The green arrow for Water
Supply acknowledges treatment can remove chl a at higher concentrations.

An important indicator for protection of the aquatic life designated use in High Rock Lake is the
productivity of apex predators, including the forage trophic levels. Studies on the productivity of
apex predators in reservoirs have shown increased abundance of apex predators, prey species,
and zooplankton for chl a concentrations of 35-40 µg/L (Allen et al., 1998; Bayne et al, 1994)
typically reported as growing season mean values. In terms of changes with lower nutrient
levels, Maceina and Bayne (2006) showed a decrease in largemouth bass recruitment and growth
rate when chl a concentration was reduced from greater than 40 µg/L to 9-17 µg/L. The lower
end of the proposed range of chl a concentrations is set at 25 µg/L to provide sufficient algal
64

production to support abundant apex predators in High Rock Lake and avoid the potential impact
to the fishery noted by Maceina and Bayne (2006). Based on the literature review (see Section 2;
Figure 4.1), a chl a value of 25 µg/L would be protective of the water supply and primary
recreation uses, assuming the associated presence of cyanobacteria is not linked to algal toxin
levels that can pose a risk to animal and human health (Chapter 3), and since observation of
surface algal scums have generally been absent at High Rock Lake (see Chapter 3 for details).

The upper end of the chl a range to support aquatic life is based on reservoir research
documenting abundant apex predators, prey species, and zooplankton at average chl a
concentrations of 35-40 µg/L (Allen et al., 1998; Bayne et al., 1994). Overall fish production has
been shown to increase even with chl a concentrations greater than 100 µg/L, although there is
indication of more benthic species (e.g. carp and flathead catfish) at very high chl a levels (e.g.
Egertson and Downing, 2004; Michaletz et al., 2012). The selection of 40 µg/L as the upper end
of the range is to maintain a balanced overall aquatic community considering the apex predators,
prey species, and zooplankton. The literature indicates overall apex predator abundance would
be higher at chl a concentrations >40 µg/L, but there likely would be a shift in species toward
bottom-dwelling species and the diversity of prey species and zooplankton may be affected.
Further, frequent high chl a concentrations in High Rock Lake could be associated with a higher
risk of toxin exposure potentially above proposed thresholds protective of human health, which
was also a factor in setting the upper end of the chl a range at 40 µg/L. Literature and
observations from High Rock Lake indicate primary recreation and public water supply would be
supported at a chl a concentration of 40 µg/L.

4.3.2 The SAC Recommended Criterion Magnitude

The SAC recommends a criterion magnitude of 35 ug/L, from the derived range of 25-40 µg/L,
expressed as a seasonal geomean. In developing the recommendation, the SAC considered
proposals as low as 25 µg/L and as high as 40 µg/L. Ultimately, the criterion magnitude was set
in the upper half of the potential range in acknowledgement of the favorable indicators of use
attainment in High Rock Lake, such as a thriving fishery and low algal toxin levels observed in
summer of 2016. The maximum value was not selected based on site-specific fisheries
information presented to the SAC indicating abundant benthic species, possible overall decreased
fish species diversity, and decreased catch rate of striped bass compared with other North
Carolina Piedmont reservoirs. Implementation of the proposed criterion of 35 µg/L in High
Rock Lake would require a reduction in the level of chl a in the reservoir from the existing
condition (see Figure 4.1). Total productivity of the fishery would be expected to decrease,
which may increase diversity and shift species abundance toward pelagic species.

4.4 Spatial Components

The spatial variation in biological and physical properties in man-made reservoirs follows a
regular spatial pattern (see Figure 3.2). The most upstream reach reflects primarily river
conditions as the river flows into the impoundment in which water level is controlled by the
65

downstream dam structure. In the case of High Rock Lake, waters at HRL051 reflect turbid river
conditions, and the average chl a is lower than in downstream waters. In the middle reach or
transitional zone, water velocity slows down, mineral turbidity settles to the bottom, and a peak
in algal abundance typically occurs. In the case of High Rock Lake, waters at YAD152A and
YAD152C would be in the transitional zone. Waters downstream of the transitional zone in the
lacustrine zone above the dam (YAD169B and YAD169F) would typically have decreased algal
abundance compared with the transitional zone.

Available chl a data for monitoring stations listed above for the three reservoir zones are
summarized in Table 4.1. Monitoring during the growing season was conducted approximately
monthly in five individual years during the period of 2006 through 2016. Substantial variation,
expressed as the coefficient of variation (COV), is evident in the river reach (HRL051), but
variability is lower in the transitional and lacustrine zones. In terms of protection of uses, the chl
a criterion’s geometric mean calculated as the geomean of samples collected during the growing
season (April-October) will normally be protective of all designated uses even though winter
months are not part of the calculation, since chl a is typically lower in winter months (see Sec.
3.2.4).
Table 4.1. Growing Season (April-Oct) chl a geomean (µg/L) by sampling
location (COV = coefficient of variation)
Year HRL051 YAD152A YAD152C YAD169B YAD169F
2006 27.3 51.2 59.6 38.3 34.6
2008 34.1 49.2 53.4 40.3 32.5
2009 16.9 42.1 53.0 43.4 36.0
2011 30.7 50.1 55.6 42.5 36.5
2016 20.8 52.3 58.7 44.3 36.1
Overall 24.1 47.9 55.2 42.0 34.8
COV 29.2% 8.3% 5.5% 5.8% 4.7%

It is recommended that the spatial assessment scale for the site-specific chl a criterion be
consistent with the derivation of the criterion magnitude (see Section 4.4.2.) and expressed as a
seasonal geomean. It is recommended that all observations for the assessment period from open
waters within an assessment unit would be incorporated into the computation of the geomean of
available data from the growing season months (April-October). Monitoring locations in
backwaters, isolated coves, or where water depth is typically shallow (e.g. <10 feet) would be
evaluated based on narrative criteria but excluded from the calculation of the chl a geomean for
open waters based on the expectation that such data are not representative of the data used to
develop the criterion itself. The SAC also recommends that compliance with the chl a criterion
be evaluated with samples collected as photic zone composite samples (e.g. from the water
surface down to twice the Secchi depth).

66

4.4.1 Monte Carlo Spatial Analysis

Evaluation of chl a data for High Rock Lake has shown a consistent spatial pattern with
maximum values in the transition zone for the reservoir and lower values in the lacustrine zone
and downstream tributaries (see Table 4.1). A Monte Carlo analysis was performed to evaluate
how spatial grouping of sampling locations could affect three specific implementation scenarios
relative to a seasonal geomean criterion of 35 µg/L. The Monte Carlo approach was used for the
analysis to extend conditions simulated to include the five primary years in which regular
monitoring was done and to include conditions that could have occurred in other years. Data
from monitoring efforts during 2006-2016 were used in the analysis. The objective of the
analysis was to evaluate how the seasonal geomean for chl a varies at target stations in the
transitional and in the lacustrine reservoir zones relative to the selected range for protection of
nutrient-sensitive uses (25-40 µg/L; see section 4.4.1) based on whether the seasonal geomean of
35 µg/L is achieved at all locations individually or for multiple locations aggregated together.

For the evaluation, the Monte Carlo approach was used to create 100 potential datasets for each
of four monitoring locations evaluated based on reported chl a concentrations for the growing
season for the five primary years in which regular monitoring was done (see Table 4.1).
Monitoring stations simulated were HRL051 (riverine), YAD152C (transitional), YAD169B
(lacustrine), and YAD169A (tributary embayment) (see Figure 3.2). Figure 4.2 plots the

Figure 4.2. Cumulative distributions of measured data by stations utilized in Monte Carlo analysis
67

cumulative distribution of reported individual sampling chl a concentration values for the four
simulated stations for growing season samples from the five monitoring years listed in Table 4.1.

The datasets derived through a Monte Carlo analysis for the four locations simulated were
developed with a sampling design comparable to the current NCDWR ambient monitoring effort
of monthly sampling during the growing season. Five monthly samples were derived from the
cumulative distribution for a given location for two separate years, yielding a total of 10 data
points from which to calculate the seasonal geomean. Each point was derived by selecting
randomly a probability between 0 and 100%, and then converting the probability to a chl a value
by linear interpolation from the respective distribution in Figure 4.2 for the location. This
process of creating a dataset of 10 chl a values was performed 100 times for each location.
Figure 4.3 provides the distribution of geomean values for each location based on the 100 Monte
Carlo simulations for existing conditions.

Figure 4.3. Distributions of growing season geomean chl a concentration (µg/L) by location derived from Monte
Carlo analysis

The Monte Carlo simulation results were used in conjunction with a target seasonal geomean for
chl a of 35 µg/L selected to be above the midpoint of the range highlighted in Figure 4.1 but
below the maximum value of 40 µg/L (see Section 4.4.1). Three potential approaches to
applying the target chl a geomean criterion were simulated: (1) each individual station meets the
criterion as a long-term geomean; (2) each reservoir zone meets the criterion as a long-term
68

geomean; and (3) the transitional and lacustrine zones collectively meet the criterion as a long-
term geomean. A long-term geomean was used in the analysis for reduction scenarios to reduce
the influence of year-to-year variation in the seasonal geomean on predicted results. The
analysis to support the evaluation is summarized in Table 4.2. The reduction percentage in long-
term geomean for chl a to achieve the criterion of 35 µg/L varied from 36.6% for Approach 1
based on YAD152C to 18.7% for the combined transitional and lacustrine zones approach.
Approach 2 is based on reducing chl a in the transitional zone to 35 µg/L. The potential impact
of each approach on the chl a levels to support the currently healthy fishery was evaluated by
reducing the chl a distributions derived for YAD152C, YAD169B, and YAD169A (Tributary) by
the required reduction for each approach to achieve the criterion. The analysis assumed
reductions in chl a would be the same percentage throughout the reservoir stations.

Table 4.2. Influence of assessment unit approach on results
Unit Existing Long-Term
chl a Geomean (µg/L)
Range for Individual
Years (see Table 4.1)
Reduction to
35 µg/L (%)
YAD152A 47.9 42.1 - 52.3 27.0%
YAD152C 55.2 53.0 - 59.6 36.6%
YAD169B 42.0 38.3 - 44.3 16.6%
YAD169F 34.8 32.5 - 36.5 N/A
Transitional 48.8 44.0 - 55.5 28.3%
Lacustrine 37.8 35.8 - 40.0 7.5%
Reservoir 43.0 41.1 - 47.3 18.7%
Notes: (1) Reduction percent is to reduce long-term geomean to 35.0 µg/L; (2) Transitional zone
assessed as YAD152A and YAD152C; (3) Lacustrine zone assessed as YAD169B and
YAD169F; (4) Reservoir assessed as YAD152A, YAD152C, YAD169B, and YAD169F.

Cumulative distributions for chl a at YAD152C, YAD169B, and YAD169A for the three
approaches evaluated are provided in separate panels of Figure 4.4. Evaluation of the three
approaches, in terms of protection of aquatic life, was determined by the frequency of overall
data points for each approach that were between 25 and 40 µg/L. The analysis also considered
whether data points outside the target range were below 25 µg/L or greater than 40 µg/L. In
terms of a frequency comparison with the target range, Approaches 2 and 3 were comparable at
72.7% and 73.7%, respectively, while Approach 1 had only 60.3% of data points in the target
range (see Table 4.3). Data points outside the target range were primarily <25 µg/L for
Approach 1 and primarily >40 µg/L for Approach 3, with data points <25 and >40 µg/L for
Approach 2. Approach 1 would likely cause the seasonal geomean of portions of the reservoir to
frequently fall below 25 µg/L, which could impact the valued fishery. Approach 2 provides a
balance between limiting chl a values <25 µg/L, which may impact the fishery, and limiting chl a
values >40 µg/L that could contribute to acute nutrient-dependent impacts in the future.
69

Approach 3 would likely continue seasonal geomean chl a in the transitional zone >40 µg/L on a
frequent basis.

Figure 4.4. Distributions of growing season geomeans (µg/L) by location and assessment approach (top
panel: each individual station meets the criterion as a long-term geomean; middle panel: each reservoir
70

zone meets the criterion as a long-term geomean; and bottom panel: the transitional and lacustrine zones
collectively meet the criterion as a long-term geomean). The box indicates target chl a concentration
range of 25-40 µg/L.

Table 4.3. Distribution of chl a geomean by spatial assessment
approach.
Assessment <25 µg/L (%) 25 – 40 µg/L (%) >40 µg/L (%)
Approach 1 39.0 60.3 0.7
Approach 2 15.7 72.7 11.7
Approach 3 0.3 73.7 26.0

Of note is the difference between the temporal averaging used in the Monte Carlo analysis (ten
randomly selected chl a values used to compute a geometric mean) and the temporal averaging in
the proposed chl a criterion (all growing season chl a values from a single year used to calculate
a seasonal geometric mean). The normal lake sampling plan of the NC DWR is to collect five
such chl a samples each growing season. Also not included in the Monte Carlo analysis is the
consideration of a maximum allowable exceedance frequency (the proposed criterion is that one-
in-three seasonal geomeans may exceed the chl a criterion). It is believed that these two
differences between the Monte Carlo analysis and the proposed chl a criterion offset one another,
so that the analysis presented is usable as-is for comparing the implications of the three
assessment unit approaches analyzed with the Monte Carlo analysis. Repeating the Monte Carlo
analysis with a different set of assumptions would likely not have significantly changed the
analysis outcome and would have led to an additional delay in completing the proposed High
Rock Lake chl a criterion development, and was therefore not pursued.

4.4.2 Considerations for Delineating Assessment Units

In the Clean Water Act framework, an assessment unit (AU) is the basic spatial component that
states use for evaluating attainment status of water bodies. States use various bases to delineate
AU boundaries, including hydrography datasets, hydrologic unit codes, maps of water body
names, major junctions, morphology, or limnological zones. Although assessment units can be
delineated in different manners, USEPA (2005) offers the following guidance on segmentation:

Segmentation may reflect an a priori knowledge of factors such as flow, channel
morphology, substrate, riparian condition, adjoining land uses, confluence with
other waterbodies, and potential sources of pollutant loadings…Segments should…
represent a relatively homogenous parcel of water (with regard to hydrology, land
use influences, point and nonpoint source loadings, etc.)

71

States also vary widely with regard to how chl a is assessed spatially within reservoirs, and the
procedures often differ from those used for toxics. For example, Alabama uses an assessment
methodology that varied based upon the size of the waterbody. Some relatively small lakes that
are most easily monitored near the forebay use only this location for assessment. When a lake is
considered large enough to have more than one station, separate criteria are generally applied to
these separate stations (A.A.C. 335-6-10-.11). Georgia assigns specific chl a criteria to
individual stations within large reservoirs. Criteria can vary between stations to recognize
different expectations for different parts of the reservoir. Florida applies chl a criteria for most
lakes as a lake-wide or lake segment-wide average (F.A.C. 62-302). Virginia recognizes three
limnologically-defined zones within reservoirs (riverine, transitional, and lacustrine), but only
applies numeric nutrient criteria to the lacustrine zone (Virginia DEQ, 2009).

Despite considerable discussion, the SAC did not come to a consensus regarding how spatial
assessment units should be defined for High Rock Lake or other water bodies. However, the
manner in which assessment units are spatially defined for chl a has implications for the
stringency/conservativeness of the criterion, and also for how different uses or risks are balanced
within a reservoir. For that reason, this section provides a general discussion of the two basic
approaches discussed and considered by the SAC: (1) delineating AUs based on individual
monitoring stations, similar to NC’s existing or default approach; and (2) delineating AUs by
three major limnological zones.

4.4.2.1 Defining Chl a Assessment Units by Individual Stations

In large reservoirs, many DWR monitoring stations are more than 1 mile apart. For example, in
High Rock Lake, the distance between neighboring monitoring stations varies between 0.3 and
3.6 miles. Hence, most of the AUs delineated around individual stations in High Rock Lake are
still relatively large. Compared with other approaches, the use of individual stations increases the
homogeneity of water within an AU, which is an important characteristic of AUs as
recommended by USEPA (2005). The single station approach also avoids averaging that can
mask temporal and/or spatial changes in chl a concentration. Accordingly, an individual-station
approach will generally be more sensitive to detecting chl a related changes that occur at specific
locations within the reservoir. The individual station approach will also be better able to detect
chl a related problems that result from changes in the spatial distribution of nutrient loading to
the lake from loading hot spots or changing development patterns in the watershed.

Because the highest-chlorophyll station would tend to control a reservoir TMDL, an individual-
station approach for delineating AUs will generally require higher levels of nutrient reduction
than approaches that would average the chl a goal over larger segments. To this extent, the
individual station approach is more environmentally conservative with respect to potential
harmful effects of excess algae (e.g, toxins, bloom events, etc.). An estimate from the Monte
Carlo analysis is that applying the criterion using individual stations for AU specification rather
than the limnological AU specification will decrease the prevalence of chl a values above 40
µg/L from 11.7% to 0.7% (Table 4.3). Another practical advantage of the individual station
72

approach is consistency with North Carolina’s existing approach and assessment data processing
procedures.

4.4.2.2 Defining Chl a Assessment Units by Limnological Zones

In contrast to delineating AUs around individual stations, this approach would define AUs using
a priori knowledge of major reservoir zones that are functionally different and represent logical
units for water quality management. The concept that reservoirs exhibit three major spatial zones
(riverine, transitional, lacustrine) is well established in the scientific literature and consistent with
observed water quality in High Rock lake (see section 3.1). In practice, the three-zone approach
would only involve aggregating data from DWR monitoring stations that are relatively close to
each other (e.g., YAD152A and YAD152C) and would not involve a dramatic change in overall
segmentation, but would avoid the delineation of small segments around individual stations such
as the AU currently associated with YAD152C.

A potential advantage of the limnological zone approach to AUs is the protection of current
levels of fish production in High Rock Lake, as demonstrated by the Monte Carlo analysis
(section 4.4.1). The limnological zone approach for AU specification raises the percentage of
chl a values within the fully protective range from 60.3% to 70.2%, when compared to the
individual station approach. The percentage of chl a values below the protective range also
decreases from 39.0% to 15.7% (Table 4.3). Attainment of the recommended criterion will
require significant chl a reductions in High Rock Lake, regardless of whether AUs are individual
station or three limnological zones. The three-zone approach reduces the risk of harmful effects
associated with high chl a, relative to existing levels, but provides a higher level of protection of
the fishery use compared to the individual station approach.

4.5 Consideration of Statistical Confidence

The SAC discussed the concept of incorporating a statistical test of confidence that the chl a
criterion had been exceeded in a given assessment period, as a potential means to reduce false
findings of non-attainment (for 303d listing of water bodies) or false findings of attainment (for
delisting water bodies). North Carolina currently uses a non-parametric statistical test (the
binomial method) for not-to-exceed criteria. Although the binomial method is not appropriate
for a seasonal geometric mean, other methods could be developed, such as the calculation of
confidence limits on the geometric mean. An argument against the use of a statistical test is the
primary purpose of these test is to prevent a very small number of data from controlling the
listing/delisting decision, but seasonal geometric mean chl a values (calculated for at least two
years) would be based on at least 10 data points. Also, if only 10 data points were available for a
given assessment period, confidence limits could be relatively wide, which could make it very
difficult to either list or delist water bodies. Although the SAC is not recommending a specific
statistic test at this time, this topic could be re-examined at the time of statewide criteria
development.

73

4.6 Summary of Proposed Criterion

The proposed chl a criterion for High Rock Lake is a seasonal geomean of 35 µg/L, not to be
exceeded more than once in three years, for growing season months of April-October based on
protection of all uses while maintaining the productivity of the sport fishery (Table 4.4). In
terms of spatial considerations, all monitoring data from open waters within assessment units
collected during the months of April through October would be used to compute a geomean to
compare with the proposed criterion. The criterion would apply to all months of the year, with
attainment of the criterion assessed with data from the growing season months. The SAC
recommends the exceedance frequency assessment approach. The SAC recommended frequency
is not to exceed more than one in three calculated seasonal geomean values.

The SAC recognizes that several considerations remain in establishing the site specific chl a
criterion for High Rock Lake. These considerations include how much data to include and what
data might be excluded during assessment, spatial aggregation of data, and whether the criterion
should include a statistical confidence test (Table 4.5). Furthermore, the SAC encourages
continued monitoring of cyanobacterial toxin levels paired with chl a assessments to better
evaluate potential exposure risks and toxin dynamics in High Rock Lake. The SAC refers these
implementation questions to the CIC for further consideration.

Table 4.4. Proposed Chl a Criterion for High Rock Lake.
Component Selection Notes on Selection
Magnitude 35 µg/L None
Period/Duration Seasonal
Geomean Calculated Geomean based on all data from growing season
Season/Duration April-October Include samples collected in at least five different growing season
months for each year of data included in the analysis
Frequency
Maximum
Exceedance
Frequency of
One-in-three
Compute the geometric mean for each year of individual data and apply
a frequency component of not more than one exceedance out of three
years of data

Spatial
Considerations Open Waters
Photic zone composite based on twice the Secchi depth; shallow waters
and isolated coves to be addressed through narrative criteria; all data
within each assessment unit would be incorporated into the calculated
geomean

74

Table 4.5 SAC’s Additional Topics for Specific Consideration by CIC
Component Alternatives or Additional Information included
in this document
Sample Size/Filtering of
Monitoring Data
SAC encourages CIC to offer implementation thoughts on whether data should
be collected from at least five different months within the growing season or if
there are other bounds or minimums on data density that may be acceptable.
Spatial Assessment Whether or not to include multiple stations in an assessment unit
Statistical Test of
Confidence
Whether or not to consider a statistical test of confidence that the chl a criterion
was exceeded in a given assessment period

75

4.7 References

Allen, M.S. Greene, J.C. Snow, F.J. Maceina, M.J. DeVries, D.R. 1999. Recruitment of
Largemouth Bass in Alabama Reservoirs: Relations to Trophic State and Larval Shad
Occurrence. North American Journal of Fisheries Management 19(1):67-77.
Bayne, D.R. Maceina, M.J. Reeves, W.C. 1994. Zooplankton, fish and sport fishing quality
among four Alabama and Georgia reservoirs of varying trophic status. Lake and Reservoir
Management 8(2):153-163.
Egertson, C.J. Downing, J.A. 2004. Relationship of fish catch and composition to water quality
in a suite of agriculturally eutrophic lakes. Canadian Journal of Fisheries and Aquatic
Science 61: 1784-1796.
Florida Department of Environmental Protection. 2012. Overview of Approaches for Numeric
Nutrient Criteria Development in Marine Waters. 110 p.
Maceina, M.J. Bayne, D.R. 2001. Changes in the black bass community and fishery with
oligotrophication in West Point Reservoir, Georgia. North American Journal of Fisheries
Management 21(4):745-755.
Missouri Department of Natural Resources (MDNR). 2017. Rationale for Missouri Lake
Numeric Nutrient Criteria. December 2017.
Michaletz, P.H. Obrecht, D.V. Jones J.R. 2012. Influence of Environmental Variables and
Species Interactions on Sport Fish Communities in Small Missouri Impoundments. North
American Journal of Fisheries Management, 32:6, 1146-1159. First Published November 1,
2012.
North Carolina Division of Water Resources. 2019. 2020 303(d) Listing and Delisting
Methodology.
https://files.nc.gov/ncdeq/Water%20Quality/Planning/TMDL/303d/2020/2020-Listing-
Methodology-approved.pdf. 14 p.
U.S. Environmental Protection Agency (USEPA). 2000. Nutrient Criteria Technical Guidance
Manual – Lakes and Reservoirs. EPA 822-B00-001. Washington, DC: USEPA, Office of
Water.
USEPA. 2010. Ambient Water Quality Criteria for Dissolved Oxygen, Water Clarity and
Chlorophyll a for the Chesapeake Bay and Its Tidal Tributaries: 2010 Technical Support for
Criteria Assessment Protocols Addendum. EPA 903-R-10-002. 63 p.
U.S. Environmental Protection Agency. 2005. Guidance for 2006 Assessment, Listing and
Reporting Requirements Pursuant to Sections 303(d), 305(b) and 314 of the Clean Water
Act. Memorandum from Diane Regas to Water Division Directors. 89 p.
Virginia Department of Environmental Quality. 2009. Monitoring and Assessment of Lakes and
Reservoirs. Water Guidance memo No. 09-2005. 31 p.

76

5. Potential Elements of a Framework for Deriving Site-Specific Criteria

North Carolina’s Nutrient Criteria Development Plan (NCDP) states a commitment to develop
nutrient-related criteria (causal and/or response variables) throughout the state on a site-specific
basis. High Rock Lake has served as the pilot water body for reservoirs and lakes, and the
chlorophyll a (chl a) criterion recommendation of this technical support document apply to that
specific water body. However, the NCDP schedule calls for the adoption of nutrient-related
criteria on a statewide basis during the 2023-2028 timeframe. Part of this process will be to
“confirm the approach proposed during the adoption of the nutrient criteria in [High Rock Lake]
with SAC involvement”. The purpose of this section is to discuss how lessons learned during the
reservoir pilot might apply to the future effort to derive chl a criteria for other reservoirs and
lakes.

The SAC has not yet developed a detailed framework for deriving reservoir-specific chl a
criteria. However, many elements of the SAC’s approach for High Rock Lake would be
transferable to other water bodies. At various times, the SAC also discussed potential elements of
a more formal framework for site-specific criteria derivation. This section attempts to document
some of those concepts in case they are useful during the future, statewide effort. Any of the
framework elements discussed herein are subject to additional discussion by the SAC and DWR.

5.1. Desired Characteristics of a Framework

North Carolina’s intent to develop nutrient criteria throughout the state on a site-specific basis is
challenging from a scientific and regulatory perspective. Site-specific chl a criteria have the
advantage of reflecting water body-specific responses to nutrient inputs, and to avoid the
misallocation of resources that can result from one-size-fits-all criteria. However, the derivation
of site-specific criteria can be resource-intensive because it requires evaluation of water body-
specific conditions and nutrient-response relations. It is not practical for North Carolina DEQ to
develop complex nutrient-response models for every water body in the state, nor to devote the
level of time and resources that were devoted to the High Rock Lake pilot. Ideally, a framework
for deriving site-specific criteria would be streamlined enough for practical application with
datasets of moderate size, while also including enough water body-specific information to make
the correct criteria decisions.

With this background, the SAC cites the following characteristics as desirable for a framework
for developing site-specific chl a criteria:

1. The framework produces site-specific chl a criteria that are protective of
designated uses. This is a minimum requirement of any criteria derivation process. All
uses of the reservoir should be considered, including public water supply, recreation, and
aquatic life. The site-specific nature of the desired framework is explicit in the NCDP,
and is based in the understanding that different water bodies can respond to nutrient
inputs in different manners.
77

2. The framework should minimize assessment and management errors. Both type I
(false findings of impairment) and type II (false finding of attainment) errors are of
concern and should be minimized to the extent possible. Overprotective criteria would
lead to type I errors, whereas underprotective criteria would lead to type II errors.
Although some degree of conservativeness is appropriate for water quality criteria, highly
overprotective criteria would misdirect TMDL and implementation resources.

3. The framework should consider both literature and reservoir-specific
information. The SAC’s chl a recommendations for High Rock Lake were derived using
both literature-based and reservoir-specific information. Research for the reservoir pilot
revealed that targets based on the literature and reservoir-specific data can be very
different. The scientific and lake management literature includes a wide range of
potential chl a targets associated with different regions, reservoir/lake types, and uses.
Many of the studies from the literature focus on water bodies that have experienced algal-
related problems that might or might not occur in other reservoirs being considered for
site-specific criteria. The literature also includes many chl a targets from higher latitudes
or altitudes, many of which could be unrealistically low for southeastern lakes and
reservoirs. Some literature-based chl a targets are based in concepts such as user
perception, which are difficult to transfer from one region to the next.

Reservoir-specific data or models can help determine whether uses are currently being met, and
also provide insights into the empirical relations between chl a and other use indicators. But like
the scientific literature, reservoir-specific information also has limitations for deriving site-
specific criteria. Some water bodies may have relatively few water quality data and little
narrative information on use attainment (e.g., fishery status, water treatability issues, algal toxins,
etc.). Even for a relatively data-rich water body such as High Rock Lake, the SAC did not find it
simple to identify chl a thresholds above which specific uses were met or not met. Rather, much
of the information pointed to a continuum of risk, where the concern over potential impacts
increased with chl a.

Ultimately, the SAC recommended a chl a criterion from within a range of candidate values (25
– 40 µg/L), as described in section 4. That range was determined from both literature and High
Rock Lake-specific information. The lower end of the range was more strongly influenced by the
literature and the desire to limit potential impacts to the fishery, whereas the upper end of the
range was from multiple lines of evidence that include the literature and High Rock Lake’s
existing chlorophyll-indicator relations. Similar consideration of both literature and reservoir-
specific information is likely to be useful for a statewide framework. The framework could
emphasize reservoir-specific information for water bodies with more definitive chlorophyll-use
indicator relations. The literature will remain informative of the chl a concentration at which
some lakes/reservoirs experience algal-related problems.

78

5.2. Potential Common Elements

Some elements of the proposed chl a criterion for High Rock might be directly transferred to
other lakes and reservoirs without site-specific deliberations. This could be the case for criteria
elements whose technical justification for High Rock Lake would apply equally to other
reservoirs, or criteria elements for which it would be unnecessarily problematic to use different
approaches for different water bodies during the assessment process. The basis for the following
criteria elements is provided in section 4, and much of the reasoning for High Rock Lake would
also apply to other reservoirs:

● Geometric mean
● April – October growing season
● A 1-in-3 year allowable exceedance frequency
● Photic zone grab sample at 2X Secchi depth

The magnitude of the chl a criterion is the element most likely to change between
lakes/reservoirs. Factors to consider in the adjusting the magnitude of chl a criterion between
water bodies include warmwater vs. coldwater classification, historical and recent chl a
concentrations, designated uses, and various narrative and numeric indicators of use support.
Following are major steps of a potential framework to derive site-specific criteria:

1. Application of a chl a screening range as the initial evaluation of impairment status.
2. Consideration of other numeric and narrative indicators.
3. Application of decision rules on impairment status.
4. Application of decision rules on site-specific criteria.

These factors are discussed in subsections below.

5.3. Chl a Screening Range Concept

With any framework for deriving site-specific criteria, one of the first steps would be to
determine whether the reservoir is effectively meeting designated uses vs. experiencing tangible
nutrient-related impairments. Results of this determination would be a major factor in deciding
if the site-specific criteria should be lower than existing conditions. The use of readily-available
water quality data such as chl a concentrations could streamline this determination. As discussed
in previous sections, the SAC did not identify a one-size-fits-all chl a criteria that could be used
in a pass-fail manner to answer this question. However, the SAC did consider it more practical to
identify a range of chl a concentrations that was associated with increasing risk of impairment.

With this background, a potential first step of a framework could be to compare a reservoir’s
existing chl a concentration (seasonal geometric mean) to a screening range, with the goal of
determining whether the reservoir can be categorized as likely attaining vs. likely impaired based
on chl a alone. The upper end of the range would represent a value above which nutrient
79

impairment is likely, and the lower end of the range would represent a value below which
nutrient impairment is unlikely. Reservoirs in the “gray area” (i.e, within the range) would
require additional narrative assessment (step 2) to determine if they experience nutrient-related
impairments. Figure 5.1 illustrates the chlorophyll-based screening range with a range developed
by the SAC (25-40 μg/L) during the High Rock Lake pilot.

Figure 5.1 – Illustration of the chl a screening range concept.

The use of a chl a screening range is conceptually similar to an approach published by Arizona
(Arizona DEQ, 2008), and is also similar to criteria recently adopted by Missouri [10 CSR 20-
7.031(5)(N)1.C.(I)] and approved by USEPA. However, the screening range concept described is
specifically discussed herein as a step to streamline the derivation of site-specific chl a criteria
rather than a long-term assessment method.

5.4. Consideration of Narrative and Numeric Indicators

In the second step of a potential framework for deriving site-specific criteria, various other types
of reservoir-specific information would be considered to support impairment categorization.
Although many types of information might be considered during this step, the framework could
be applied more consistently if it included a pre-defined list of useful indicators with associated
thresholds. Table 5.1 provides an example of such a checklist. The list includes both narrative
indicators (e.g., presence/absence of fish kills, nuisance conditions, fishery status) and numeric
indicators (pH, DO, cyanotoxin concentrations). It would not necessarily be required to have
information for every indicator to perform the categorization.

An important aspect of the indicator list is that indicators are categorized as either primary or
secondary. Primary indicators are those that are more direct indicators of nutrient impairments,
80

whereas secondary indicators may indicate concerns but are not direct indicators of impairments.
For example, a high cyanobacteria density would be a secondary indicator, whereas persistent
exceedance of cyanotoxin thresholds would be a primary indicator. This distinction is important
because decision guidelines for impairment determinations would weight primary indicators
more than secondary indicators.

Table 5.1: Examples of Potential Indicators for Narrative Evaluation
Use
Category Indicator Primary or Secondary
Indicator
Narrative or
Numeric
Indicator
Aquatic
Life
DO concentration Primary Numeric
DO saturation Secondary Numeric
Ph Primary Numeric
Algal toxins Primary Numeric
%Cyanobacteria Secondary Numeric
Fishery status Primary Narrative
Fish kills Primary Narrative
Fish abnormalities Secondary Narrative
Public
water
supply
Algal toxins Primary Numeric
T&O-causing compounds Secondary Numeric
Treatability challenges Primary Narrative
Recreation Algal toxins Primary Numeric
Secchi depth Secondary Numeric
Nuisance blooms; mats or
extensive scums
Primary Narrative

Under a potential framework, each indicator could be categorized as green (full use support
indicated), yellow (potential concerns), or red (strong evidence of use impairment). Associated
guidance would provide numeric ranges or other guidelines for these determinations. The
guidance could also include decision rules for how multiple or mixed-result indicators would be
used to interpret existing use support.

If sufficient data were available, this step 2 could also involve direct examination of the relations
between chl a and other indicators such as water clarity, pH, cyanotoxins, etc. Such empirical
relations could lead to the selection of chlorophyll targets to achieve specific responses.
Examples of chlorophyll-indicator relations for High Rock Lake are provided in Section 3 of this
document.

81

5.5. Decision Guidelines for Site-Specific Criteria
After application of the chlorophyll-based screening range and narrative numeric evaluation, the
final steps would be to make the appropriate site-specific chl a criterion. Although professional
judgment would be required, DWR’s decisions would be more transparent and defensible if clear
decision guidelines were developed. The associated decision guidelines could be organized as a
matrix based on the existing chl a concentration (below, within, or above screening range) and
outcome of the narrative evaluation (narrative evidence of use attainment, non-attainment, or
inconclusive). For example, if a reservoir’s chl a concentration was above the screening range
but the reservoir did not show clear signs of impairment from the narrative/numeric evaluation,
the criterion could likely be set at or near the top of the screening range. But a reservoir within
the screening range that failed the narrative/numeric evaluation might receive criteria in the
lower half of the screening range. The formulation of specific decision guidelines would require
additional discussion by the SAC and DEQ.

In some cases, criteria could be set to protect a reservoir’s existing condition. For example, a
lake with chl a in the 15-20 ug/L range (below the screening range) but with some exceedances
of secondary indicators might receive a criterion of 20 ug/L to prevent impairments. If robust
chlorophyll-response linkages were available, they could also be applied to set a specific chl a
target during this step, and these linkages might support criteria outside of the screening range.

5.6. References

Arizona Department of Environmental Quality. 2008. Narrative Nutrient Standard
Implementation Procedures for Lakes and Reservoirs. Available at
https://legacy.azdeq.gov/environ/water/standards/download/draft_nutrient.pdf. 21 p.