Next Steps to Develop Nutrient Criteria
Lakes and Reservoirs in West Virginia
Evan Hansen and Martin Christ
West Virginia Rivers Coalition
June 20, 2003
Table of Contents
In late 2002, the Nutrient Criteria Committee (NCC) submitted a plan to the U.S. Environmental Protection Agency (USEPA) that outlined steps to be taken toward the development of nutrient criteria in West Virginia. This plan sets a first priority on developing criteria for lakes and reservoirs (NCC, 2002). The 108 publicly owned lakes in West Virginia cover a total of 22,373 surface acres (DEP, 2000).
The plan includes preferences regarding how to develop nutrient criteria:
“Depending on the availability of data of sufficient quantity and quality, and funds for research and model development, the state will consider the following methods, in the following order of preference:
· Empirical and/or cause and effect analyses based on West Virginia data.
· Empirical and/or cause and effect analyses based on other data.
· Alternatives to the first two approaches are to define when and under what circumstances reference-based or other methods might be appropriate.” (NCC, 2002, p. 1)
In April 2003, Mike Arcuri of the West Virginia Department of Environmental Protection (DEP) gave a presentation to the NCC regarding the approach that the agency has used to classify lakes as impaired on the state’s 303(d) list, and to set targets for nutrient-related total maximum daily loads (TMDLs).
This paper reviews the status of the NCC’s discussions of lake criteria, summarizes some of the literature related to lake impairment, and suggests next steps as the NCC begins developing nutrient criteria for lakes.
USEPA has recommended nutrient criteria for lakes and reservoirs in Nutrient Ecoregion XI, based on the reference approach (USEPA, 2000). If adopted, these criteria would apply to the state of West Virginia, which sits entirely within this ecoregion. These criteria, together with ranges for sub-ecoregions, are shown in Table 1.
Table 1: Nutrient criteria for West Virginia lakes and reservoirs recommended by USEPA
Notes: Data from USEPA, 2000, p. vi. Suggested criteria are the aggregate Nutrient Ecoregion XI reference conditions based on 25th percentile only. Chlorophyll a is based on spectrophotometric method.
NCC members have expressed skepticism about adopting these criteria for several reasons, including the following:
Because of these and other concerns, the NCC’s plan includes this reference-based approach only as a last resort, and instead prefers the use of empirical and/or cause and effect analyses.
Although West Virginia’s water quality standards do not include nutrient criteria, DEP has needed de facto nutrient criteria for several purposes, including:
· making impairment decisions for the state’s 303(d) list,
· making decisions about use support for the state’s 305(b) report, and
· setting numeric targets for TMDLs for waters impaired by nutrients.
Mike Arcuri’s presentation to the NCC in April 2003 summarized the agency’s sampling protocol and method for evaluating these data to make determinations of impairment and to set TMDL targets.
DEP bases these decisions on the trophic state index (TSI), a standard method for evaluating the condition of lakes (Carlson, 1977). TSI has been used widely since it was first proposed in the 1970s (see literature review below).
TSI uses three parameters to estimate algal production and to determine the trophic status of lakes: secchi depth, chlorophyll a, and total phosphorus. Measurements of these three parameters are plugged into regression equations to calculate three separate TSI values, as shown in Table 2. TSIs range from 0 to 100. Based on the TSI, a trophic state is assigned, as shown in Table 3.
Table 2: Regression equations used by DEP to calculate trophic state indices
Note: Equations from USEPA, 1999, p.4-2. TP and CHL are expressed in ug/L. SD is expressed in meters. TSI is unitless.
Table 3: Trophic states used by DEP
Notes: Table from USEPA, 1999, p.4-2.
DEP uses a value of 65 to define lake impairment. To derive this threshold, DEP first calculated TSIs for all lakes, and then used best professional judgment to determine which lakes seemed to be impaired by nutrients. Lakes with TSI greater than 65 corresponded to those lakes judged to be impaired by DEP; therefore, a threshold of 65 was chosen. Impairment decisions for the 303(d) list were based on TSIs exceeding 65.
A TSI of 65 is in the eutrophic range shown in Table 3, closer to the hypereutrophic than the mesotrophic boundary. Using this value as a threshold implies that DEP judges that some degree of eutrophy is acceptable and not necessarily indicative of impairment.
Numeric targets for nutrient TMDLs are based on the same TSI threshold of 65 for chlorophyll a and total phosphorus. Secchi depth is ignored. In other words, the regression equations in Table 2 are solved for CHL and TP by setting TSI(CHL) and TSI(TP) to 65. These result in numeric targets for chlorophyll a of 33.3 ug/L and for total phosphorus of 68.0 ug/L. These targets, plus TSI-based targets based on a threshold of 50—the cutoff between mesotrophic and eutrophic—are summarized in Table 4.
Table 4: Chlorophyll a and total phosphorus targets based on TSI (ug/L)
Notes: Calculated from equations in Table 2. DEP uses TMDL targets based on TSI = 65, but the threshold between mesotrophic and eutrophic in Table 3 is 50.
Figures 1 and 2 plot the relationships between chlorophyll a, total phosphorus, and Secchi depth and TSI. These curves illustrate that, if TSIs are to be used as the basis for setting nutrient criteria, the criteria will vary widely depending on the chosen TSI threshold. The circles show the values from Table 4, based on TSI values of 65 and 50. The triangles illustrate USEPA’s reference conditions in Table 1. The reference conditions correspond to TSIs of 41 for chlorophyll a, 34 for total phosphorus, and 45 for Secchi depth, all significantly lower than the TSI threshold of 65 used by DEP, and all lower than the TSI that defines the mesotrophic/ eutrophic boundary.
In West Virginia’s 305(b) report, the three TSIs shown in Table 2 are calculated separately, and then averaged for each lake. A trophic state is then assigned by comparing the average TSI to the thresholds in Table 3 (DEP, 2000). The report also classifies these lakes according to their degree of use support: fully supporting, supporting but threatened, partially supporting, not supporting, or not attainable. But the report does not explain whether the TSIs are used directly to place lakes in these categories, and if so, by what method.
Figure 2: Relationships between Secchi depth and TSI
Many peer-reviewed articles have been published regarding the use of TSIs. This report reviews just a few. Carlson (1977) first proposed calculating TSIs for lakes, and his equations are still in use today.
Osgood (1982) considers the variability often found among the three TSI values, and suggests methods for transforming TSI values to be consistent with regional relationships between chlorophyll a and Secchi depth, and between chlorophyll a and total phosphorus. Also, when TSI differences are large, Osgood proposes diagnosing the processes that cause these variations based on a regional investigation, and choosing the most appropriate TSI value or values based on that investigation.
In response to Osgood’s suggestions, Carlson (1983) maintains that the three TSI values are not to be averaged. Chlorophyll a should be used to provide lake classifications, and the other two indices provide ancillary information. He presents several interpretations of situations where TSIs for the three parameters vary.
Since the 1960s, researchers have created plots that classify lakes into trophic states based on lake depth and annual phosphorus loads. Figure 3 illustrates one of these original loading plots. The two lines show the thresholds between oligotrophic and mesotrophic, and between mesotrophic and eutrophic. Similar to the thresholds for the TSI, these thresholds are based on professional judgment and observation. According to Chapra (1997), Vollenweider and others suggest that they correspond to phosphorus concentrations of 10 ug/L and 20 ug/L, respectively.
As an extension of these plots, researchers developed mass balance equations that relate phosphorus concentrations to loads, lake volume, outflow rates, and settling (Chapra, 1997; Vollenweider, 1975 and 1976; Chapra, 1975; Dillon and Rigler, 1975; Thomann and Mueller, 1987). A sample plot, shown in Figure 4, differs from the previous plot because hydraulic overflow rate rather than mean depth is plotted on the x-axis.
This loading plot, by incorporating these extra factors, helps explain the mechanisms that shape the curve. In particular, it shows how slow flushers (lakes with small hydraulic overflow rates) are expected to be eutrophic even with relatively small phosphorus loads, while fast flushers can assimilate larger loads while remaining mesotrophic.
Figure 3: A phosphorus loading plot
In his discussion of phosphorus loading plots, Chapra (1997) suggests thresholds for several nutrient related parameters regarding trophic state; these thresholds are reproduced in Table 5. The threshold for total phosphorus, 20 ug/L, is lower than either of the TSI thresholds for total phosphorus in Table 4. The threshold for chlorophyll a, 10 ug/L, falls between the thresholds based on TSI values of 50 and 65, as shown in Table 4. The threshold for Secchi depth, 2 m, is exactly equal to the threshold based on a TSI of 50.
Chapra also proposes thresholds based on the percent oxygen saturation of the hypolimnion, the deep portion of stratified lakes. There are some biogeochemical consequences of low oxygen in the hypolimnion; for example, low oxygen can cause the release of heavy metals from sediments (Brick and Moore, 1996).
Figure 4: A phosphorus loading plot based on mass balance equations
Table 5: Trophic state thresholds suggested by Chapra for several parameters
Notes: Copied from Chapra, 1997, Table 29.1. Chapra does not cite specific references for these classifications.
Since at least the 1960s, researchers have proposed thresholds for classifying lakes as eutrophic. These thresholds typically involve total phosphorus, chlorophyll a, and Secchi depth. Further research is needed for the authors and the NCC to determine if anything other than professional judgment was used to set these thresholds.
USEPA’s recent suggested reference conditions are based on an entirely different method, and are generally more stringent than the TSI thresholds that define eutrophy.
Unless further research demonstrates otherwise, none of these methods seem to conform to the NCC’s preferred method for setting criteria by using empirical and/or cause and effect analyses. Still, these thresholds are commonly used even now, about 30 years after their original publication, suggesting a degree of confidence among scientists.
When setting lake criteria, the challenge for the NCC will be to decide which, if any, of these thresholds are sufficiently justified by empirical and/or cause and effect analyses to allow their use. If none are found to be acceptable, the NCC may then have to commission new research to provide a solid scientific foundation for West Virginia’s nutrient criteria for lakes and reservoirs. Alternatively, the NCC could simply choose TSI thresholds based on best professional judgment.
Additional questions to be considered by the NCC include:
Brick, C. M. and J. N. Moore. 1996. Diel variation of trace metals in the upper Clark Fork River, Montana. Environ. Sci. Technol. 30:1953-60.
Carlson, Robert E. 1977. A trophic state index for lakes. Limnology and Oceanogaphy. 22:361-9.
________________. 1983. Discussion: Using differences among Carlson’s trophic state index values in regional water quality assessment. Water Resources Bulletin. 19(2):307-8.
Chapra, Steven C. 1997. Surface Water-Quality Modeling. Boston: WCB McGraw-Hill.
________________. 1975. Comment on: An empirical method of estimating the retention of phosphorus in lakes, by W.B. Kirchner and P.J Dillon. Water Resources Res. 1033-4.
Dillon, P.J. and F.H. Rigler. 1975. A simple method for predicting the capacity of a lake for development based on lake trophic status. J. Fish. Res. Bd. Can. 3(19):1519-31.
Osgood, Richard A. 1982. Using differences among Carlson’s trophic state index values in regional water quality assessment. Water Resources Bulletin. 18(1):67-74.
Thomann, R.V. and J.A. Mueller. 1987. Principles of Surface Water Quality Modeling and Control. New York: Harper & Row.
U.S. Environmental Protection Agency (USEPA). 2000. Ambient Water Quality Criteria Recommendations, Information Supporting the Development of State and Tribal Nutrient Criteria, Lakes and Reservoirs in Nutrient Ecoregion XI. Office of Water. EPA 822-B-00-012. December.
U.S. Environmental Protection Agency (USEPA). 1999. Total Maximum Daily Loads for Bear Lake, West Virginia. Region 3. September.
Vollenweider, R.A. 1968. The Scientific Basis of Lake and Stream Eutrophication with Particular Reference to Phosphorus and Nitrogen as Eutrophication Factors. Technical Report DAS/DS1/68.27. Paris: Organization of Economic Cooperation and Development.
________________. 1975. Input-Output Models with Special Reference to the Phosphorus Loading Concept in Limnology. Schweiz. Z. Hydrol. 37:53-84.
________________. 1976. Advances in defining critical loading levels for phosphorus in lake eutrophication. Mem. Ist. Ital. Idrobiol. 33:53-83.
West Virginia Division of Environmental Protection. 2000. West Virginia’s Water Quality Assessment, 305(b) Report, 2000. Office of Water Resources.
West Virginia Nutrient Criteria Committee (NCC). 2002. Nutrient Criteria Development Plan for West Virginia. October 24.
 Several nutrient-related criteria are actually included in West Virginia’s water quality standards (e.g., un-ionized ammonia, nitrate, and turbidity), but these criteria are not sufficient to satisfy USEPA’s current requirements.
 There is some debate about whether or not it is acceptable to average TSIs into a single index, or if the TSIs are meant to be used independently. DEP averages the values into a single index. This issue is discussed below in the literature review.
 The Bear Lake TMDL sets the target TSI for total phosphorus at 63.2 and the target TSI for chlorophyll a at 64.1, which are slightly more conservative than 65. The TMDL provides no explanation for these differences.