Natural and Anthropogenic Influences on the Mount Hope Bay Ecosystem 2006 Northeastern Naturalist 13(Special Issue 4):71–94 Effects of Brayton Point Station’s Thermal Discharge on Mount Hope Bay Winter Flounder Robert J. O’Neill1,*, Thomas L. Englert1, and Jee K. Ko1 Abstract - On behalf of Brayton Point Station, an electrical generating plant located on the shore of Mount Hope Bay, the authors performed an innovative biothermal modeling assessment to evaluate effects of heat load from the Station on 10 bay-resident fish and shellfish species. The assessment linked several biological functions (growth, reproduction, avoidance, migratory blockage, and thermal mortality) to hydrothermal simulations of the Station’s thermal plume under two plantoperating scenarios and the no-plant scenario. The assessment methodology is described, and results are presented for Pseudopleuronectes americanus (winter flounder), the species with the lowest thermal tolerance temperatures of those studied. Based on the modeling approach and input assumptions, the effects of the Station’s thermal discharge on the winter flounder life stages and functions studied were found to be negligible, especially when compared to other effects such as fishing pressure. The largest plant effect observed was only 3.9 percentage points more than for the no-plant scenario (compared to a fishing effect of approximately 40–50%). Limitations of the model and potential future refinements to address additional biological effects are discussed and evaluated. The modeling methodology used to complete this study represents a novel and scientifically grounded approach to quantifying the Station’s thermal impacts on the biota of Mount Hope Bay. Introduction A number of independent trawl and beach-seine surveys of finfish populations have been conducted in the Narragansett Bay complex over the past several decades (Collie and Delong 2002, Lynch 2000, MRI 2000). They indicate that catches of several commercially and recreationally important species, including Pseudopleuronectes americanus Waldbaum (winter flounder), have declined region-wide since the late 1970s. Potential causes suggested for the decline include overfishing, increased abundance of finfish predators, and habitat degradation due to natural and anthropogenic effects. In the Mount Hope Bay part of the Narragansett Bay complex, Brayton Point Station operations have been cited as a possible stressor on the local finfish populations. Located in Somerset, MA, Brayton Point Station is a 1600-megawatt electrical generating station that uses Mount Hope Bay for withdrawal and discharge of condenser cooling water and plant-service water. Cooling water is heated as it circulates through condensers that cool spent steam, and thus its discharge from the Station introduces heat into the 1HDR|LMS, 1 Blue Hill Plaza, PO Box 1509, Pearl River, NY 10965. Corresponding author - Robert.ONeill@hdrinc.com. 72 Northeastern Naturalist Vol. 13, Special Issue 4 Bay. The heated discharge mixes with ambient bay water, expands into a thermal plume, and eventually dissipates to nearly ambient temperature (USGenNE 2001). To determine the potential for any direct impacts of Brayton Point Station’s thermal plume on biological functions of finfish in Mount Hope Bay, a bay-wide biothermal assessment was performed by HDR|LMS. The assessment effort was funded by Brayton Point Station in preparation for completing its NPDES permit renewal application. The purpose of the study was to quantify the direct effects of Station heat load on the biological functions of nine vertebrate (winter flounder, Cynoscion regalis Block and Schneider [weakfish], Anchoa mitchilli Valenciennes [bay anchovy], Pomatomus saltatrix L. [bluefish], Brevoortia tyrannus Latrobe [Atlantic menhaden], Menidia menidia L. [Atlantic silverside], Alosa pseudoharengus Wilson [alewife], Morone saxatilis Waldbaum [striped bass], and Morone Americana Gmelin [white perch]) and one invertebrate (Mercenaria mercenaria L. [quahog]) species that reside for all or part of the year in Mount Hope Bay. This paper describes the assessment process and presents results for winter flounder. Of the 10 species evaluated, winter flounder is the most thermally sensitive (USGenNE 2001) and has been an important part of the commercial and recreational fisheries in the region. Results for the other nine species are provided in Appendix B of USGenNE (2001). The biothermal assessment model is an innovative tool designed to mathematically link biological functions of selected species to hydrothermal simulations of the Station’s thermal plume under various operating scenarios. Hydrothermal simulations of the Station’s thermal plume were provided by Applied Science Associates, Inc. (ASA). ASA divided Mount Hope Bay into 1049 three-dimensional grid cells and modeled temperatures within each cell. ASA’s thermal data are presented in USGenNE (2001). Essential to this biological/thermal (thus “biothermal”) synthesis was the ability to portray thermal effects on biological functions mathematically— for example, to quantify the level of mortality that would result from a prolonged exposure over a continuum of possible temperatures. There was no comprehensive “cookbook” approach to this process in the scientific literature. Therefore, HDR|LMS developed a series of mathematical algorithms, each designed to determine temperature effects on a specific biological function. While modeling components were developed from various sources in the scientific literature, their subsequent organization and application in the biothermal model are unique. Model Description Assumptions An overarching goal of the assessment was to take a conservative approach— that is, to model “worst-case” effects. With respect to selecting temperatures for modeling, this goal translated into identifying a year with unusually warm temperatures. The year selected for modeling was 1999, 2006 R.J. O’Neill, T.L. Englert, and J.K. Ko 73 identified as a “reasonable worst-case” warm year based on hydrothermal modeling of water temperatures and a review of local air-temperature records. The hydrothermal model and recorded weather information showed that: (1) 1999 ranked as the second warmest with respect to water temperatures among 40 consecutive years for which ASA simulated ambient (i.e., no-plant) water temperatures, and (2) during the same 40-year period, 1999 ranked as the fifth warmest year with respect to average air temperature reported at Rhode Island’s T.F. Green County Airport, located 12 miles southwest of Brayton Point Station. Bay-water temperature sets under three different heat loads were evaluated for biothermal effects. The first set was predicted temperatures under current Station operation (average heat load of 3.18 tBTU per month, which was the actual monthly average during the 1999 model year). The second set was predicted temperatures under a hypothetical Station operating scenario that would yield a 33% reduction in heat (average heat load of 2.25 tBTU per month) compared to the currently permitted monthly average of 3.50 tBTU. Finally, as a baseline condition, effects of the ambient temperature regime were also determined (heat load of 0.0 tBTU per month) for the third set. Ambient temperatures are those that would occur naturally in the Bay under the no-plant scenario. Each temperature set contained ASA-simulated water temperatures at every hour of the 1999 year in each of the 1040 Mount Hope Bay grid cells. An intensive search of the scientific literature was conducted to identify those winter flounder biological functions for which temperature tolerance data were available. Based on this search, the following functions were selected for assessment: growth, reproduction, avoidance, migratory blockage, and chronic (72-hour) thermal mortality (heat shock and cold shock were evaluated in earlier assessments and shown to have no potential for harm under either the current or reduced-heat-load Station operating scenario [USGenNE 2001]). The corresponding winter flounder habitat interactions were examined for all life stages except larvae (due to insufficient data). More information about these functions is provided in the “Methods and Results” section of this paper. Approach The biothermal assessment consisted of six steps: 1. Map the distribution and temperatures of the thermal plume over time. ASA made extensive temperature readings in Mount Hope Bay over several years. These field observations were used to calibrate WQMAP, a hydrothermal modeling system developed by ASA and the University of Rhode Island to model the circulation and water quality of estuarine and coastal waters (Spaulding et al. 1999). ASA then used WQMAP to simulate temperature data within each of the Bay’s 1049 grid cells. As can be seen in Figure 1, grid-cell sizes varied. Typical cell dimensions were 200 by 300 m (650 by 1000 ft) in the southern portion of Mount Hope Bay and 50 by 100 m (160 by 330 ft) in the northern portion. The grid was 74 Northeastern Naturalist Vol. 13, Special Issue 4 finer in the area near the Station’s discharge venturi to better resolve the complex hydrodynamics affecting the recently discharged thermal plume. Variations in cell size made it necessary to area-weight assessment findings. This task was accomplished using the following equation: 􀂵 􀂵 􀂵 􀂵 􀂘 􀂗 􀂳 􀂳 􀂳 􀂳 􀂖 􀂕 × = 􀀭 􀀭 = = n i i n i i i Area Bt Area 1 1 ( ) Area - Weighted Average (1) where Bti = predicted biothermal effect for habitat grid cell i (n = total number of habitat cells), and Areai =area of habitat grid cell i. The assessment time-step is the amount of time that temperature conditions within a grid cell were assumed to be constant. Although ASA simulated hourly baywater temperatures, a 24-hour time-step was used for the biothermal assessment, resulting in a single averaged temperature value for each day. Several shorter time-steps were considered—including a one-hour Figure 1. Mount Hope Bay hydrothermal grid cells. 2006 R.J. O’Neill, T.L. Englert, and J.K. Ko 75 interval—to explore whether plume location and dissipation (i.e., mixing, which is directly related to predicted temperature) depend on changing phases of the tide. However, examination of ASA’s hydrothermal model results revealed little intra-day variation (during the summer period, the hourly bottom water temperatures characteristically oscillated within just ± 0.5 °C of the daily mean). In addition, the 24-hour interval is shorter than the 72-hour duration used in examining chronic exposure. Finally, use of a 24-hour timestep instead of a 1-hour interval reduced the number of records requiring processing from 246 million to 10.25 million. For one of the biological functions measured—migratory blockage of upper-bay tributaries—a one-hour time-step was used. This time-step is appropriate because juveniles moving up the river may pass through the tested hypothetical “door” at each tributary mouth within a relatively short period of time. From a data-processing standpoint, the one-hour time-step for migratory blockage was facilitated by the fact that the mouths of the tributaries are represented by relatively small numbers of grid cells in the hydrothermal model. 2. Determine acclimation temperatures within all grid cells. Ambient temperatures in Mount Hope Bay vary substantially from one season to the next. In response to such temperature changes, fish undergo physiological changes that alter their thermal preferences and tolerances. This adjustment process is called acclimation.2 Acclimation temperature is the temperature to which a fish has been exposed for a period of time sufficient to allow adjustment of physiological processes, e.g., metabolic rates (Brett 1956, Coutant 1972). The time required varies from several days to more than a week (Fry 1971). The acclimation process was incorporated into the winter flounder biothermal assessment. For each grid cell for each day of the 1999 year under the two Station operating scenarios and the no-plant scenario, an acclimation temperature was developed. It was calculated by averaging the temperatures within a given grid cell over the seven days prior to the day evaluated. This methodology is consistent with two key facts: (1) the home range of winter flounder juveniles is quite small, with 98% residing within a 100-m (328-ft) radius over a 1-week time frame (Saucerman and Deegan 1991), and (2) typical ASA grid-cell dimensions (200 by 300 m) accommodate a 100-m home-range radius. 3. Determine winter flounder’s temperature tolerance for each biothermal function evaluated. Acclimation/exposure temperature relationships were defined for winter flounder growth, avoidance, reproduction inhibition, and chronic thermal mortality. They are plotted in Figure 2, a temperature-tolerance polygon that shows how different combinations of exposure and acclimation temperatures affect winter flounder biological functioning. Use of the polygon makes it possible to depict a continuum of acclimation/exposure temperature relationships for multiple biological functions in a single graphic display. 76 Northeastern Naturalist Vol. 13, Special Issue 4 For each function assessed, the scientific literature reports a range of temperature tolerances (Beitinger and Bennett 2000). Consideration of the full range of responses is essential in a biothermal assessment because reliance on a single thermal threshold could grossly oversimplify the findings. For example, it could lead to the conclusion that an exposure temperature 0.1 °C above a single threshold would cause the entire population to be adversely affected or, conversely, that an exposure temperature 0.1 °C below the threshold would cause no discernable effect. Throughout the assessment of biothermal effects on the Mount Hope Bay RIS, variable response was considered in evaluating the effect of acclimation/exposure temperatures at each time-step in each bay grid cell. The temperature tolerance polygon reflects a synthesis of the range of population responses found for each function examined. For example, Figure 2 shows mean chronic thermal mortality thresholds—i.e., acclimation/ exposure-temperature combinations at which 50% mortality would occur. (The full range of reported threshold mortality temperatures above and below the mean are depicted in Figure 3). An explanation of the data presented in Figure 2 follows: Chronic (72-hour) thermal mortality under a prolonged exposure (dark blue line)—biothermal assessment of winter flounder chronic mortality is relevant because not all winter flounder avoid elevated temperatures. The scientific literature suggests that some portion of the winter flounder Figure 2. Temperature tolerance polygon for juvenile and adult winter flounder: how winter flounder temperature tolerances change in response to changing combinations of acclimation and exposure temperatures. 2006 R.J. O’Neill, T.L. Englert, and J.K. Ko 77 population might burrow in the bottom substrate rather than leave or avoid an area of elevated temperature, and thus be exposed for a prolonged period (Klein-MacPhee 1978, Olla et al. 1969). The 72-hr duration was chosen because it is essentially the longest laboratory test period for chronic mortality found in the scientific literature. Typically, no additional mortality is expected, even if the testing period were to be extended. To test this supposition, we developed a multiple regression model using the equations from Kilgour et al. (1985) to evaluate winter flounder mortality beyond the range of exposure durations available in the scientific literature (i.e., 1, 3, 6, 24, 48, and 72 hours). It showed that the predicted adult winter flounder TL50 for an extrapolated 25-day exposure of juveniles fully acclimated to 28.0 °C was 29.0 °C. The value used in the biothermal model (for winter flounder fully acclimated to 28 °C) was 29.1 °C. Thus, the effect of increasing exposure duration beyond 72 hours was found to be negligible. It should be noted that the 1972 USEPA Water Quality Criteria (NAS/NAE, 1973) indicate that chronic assessments should be done using weekly average temperatures. Since that approach would only serve to dampen the effect of the peak values used in the biothermal assessment, and therefore result in less potential for thresholds to be exceeded, it was rejected in favor of the 72-hour exposure duration. Data for the development of the chronic mortality line came from Hoff and Westman (1966). As can be seen in Figure 2, in general, the higher the acclimation temperature, the higher the exposure temperature that can be tolerated—until a maximum limit is reached, which is the point at which no Figure 3. Variable response of winter flounder to exposure temperatures around the mean chronic threshold. 78 Northeastern Naturalist Vol. 13, Special Issue 4 further increase in thermal tolerance is possible via acclimation. This limit is the point where the chronic mortality line plateaus (acclimation temperature of approximately 28 °C). Based on Coutant (1972) and USEPA (1976), and as depicted in Figure 3, the temperature at which winter flounder chronic thermal mortality approaches zero was set at 2 °C lower than the mean tolerance line (TL5072 hr) shown in the polygon. By extension, assuming a normal distribution, winter flounder chronic thermal mortality would effectively be 100% at 2 °C higher than the TL50, and at 0.5 °C above the mean temperature, approximately 75% of the population would respond. Avoidance (red line)—a thermal avoidance response occurs when mobile species evade stressful high temperatures by moving to water with lower, more acceptable temperatures (Meldrim et al. 1974). The avoidance response can deter a species from occupying otherwise useful habitat in the vicinity of a thermal plume. Because the percentage of winter flounder that burrow rather than avoid elevated temperature is unclear, both avoidance and chronic mortality were assessed. It is important to not sum the predicted results for these two biological functions—an action that would represent double-counting. Clearly, a fish can either burrow or avoid in response to elevated temperature. As shown in Figure 2 for chronic mortality, in general, the higher the acclimation temperature, the higher the exposure temperature that can be tolerated before avoidance occurs. No further increase in mean thermal tolerance is possible after the acclimation temperature reaches approximately 28 °C. Lower and upper bounds around the mean were set at ± 5 °C based on the extensive laboratory avoidance test results reported in Mathur et al. (1983). Multiple sources were available to support development of the avoidance line in Figure 2, including Gift and Westman (1971), IA (1978), Meldrim et al. (1974), and Terpin et al. (1977). Growth zone (green-blue area)—this area delineates the acclimation/exposure temperature combinations for which growth is predicted (McCullough 1999). The upper limit of the growth zone was defined as roughly half-way between the optimal growth temperature and the temperature producing netzero growth. The 1972 EPA Recommended Water Quality Criteria (NAS/NAE 1973) specify that, in the absence of zero-growth data, upper growth-zone limits can also be approximated via the following equation: Critical growthlimit = OT + (UILT - OT) / 3 (2) where the optimal temperature (OT) is the “temperature preference” of fish in a thermal gradient—an adaptive mechanism that allows the organisms to be positioned in an environment where they can achieve optimal physiological performance (Coutant 1977, Hutchison and Maness 1979), and the upper incipient lethal temperature (UILT) is the temperature that is survivable by 50% of the population for an extended exposure. It is relevant to note that maximum growth temperatures are not consistently maintained in nature on a daily basis, and delineation of a growth “zone” makes clear the fact that suboptimal temperatures are not necessarily 2006 R.J. O’Neill, T.L. Englert, and J.K. Ko 79 adverse.Data in support of the development of the growth zone depicted in Figure 2 came from Radle (1971).3 Temperature tolerance zone (yellow area)—this area is outside the growth zone. It delineates the temperature regime over which winter flounder can survive, but in which they are stressed and experience near-zero or negative growth, i.e., weight loss (Beitinger and Bennett 2000). Reproduction inhibition zone (pink area)—Buckley et al. (1990) note that adult winter flounder acclimation temperature before spawning is an important factor in egg thermal mortality. Their research showed that eggs and subsequent larvae produced by adults acclimated to seasonally warm temperatures (7 °C) are better adapted for survival at warmer exposure temperatures (10 °C). Thus, since spawning females are typically spent after the first three weeks of the spawning period (Stoner et al 1999), the pre-spawning acclimation temperature for adult winter flounder was approximated as the average over the three weeks (rather than the seven days) prior to the day evaluated. The lower and upper limits of the reproduction inhibition zone shown in Figure 2 depict the various combinations of acclimation and exposure temperatures for which zero and 100% winter flounder egg mortality, respectively, is predicted. Within the reproduction inhibition zone, a normal probability distribution is applied, where the midway point is representative of 50% egg mortality. Because Buckley et al. (1990) did not test for adult acclimation temperatures above 7.0 °C, there was no basis for evaluating acclimation effects on egg mortality at temperatures higher than that value (Fig. 2). Thus, for adult acclimation temperatures greater than 7.0 °C, the lower and upper limits of the reproduction inhibition zone shown at 7.0 °C in Figure 2 were applied (i.e., the far right-edge of the reproduction inhibition zone). In addition to Buckley et al. (1990), two other sources were available to support development of the reproduction inhibition zone: Rogers (1976) and Williams (1975). 4. Determine winter flounder’s horizontal and vertical habitats in Mount Hope Bay. Figure 4 delineates the life-stage habitats used in assessing winter flounder biothermal effects. Winter flounder prefers sand or sand-silt bottoms, which are common throughout Mount Hope Bay. Based on habitat data from Stoner et al. (2001) and bathymetric data documenting Mount Hope Bay depths from Swanson et al. (1998), the winter flounder horizontal habitat was delineated as < 5 m for juveniles, which occupy shallow areas, and > 5 m for adults, which occupy the remaining deeper waters. 5. Determine when winter flounder life stages inhabit Mount Hope Bay. Figure 4 shows the periods of occurrence in Mount Hope Bay for the winter flounder life stages examined. This information reflects the findings of decades-long Mount Hope Bay ichthyoplankton and trawl sampling programs (MRI 2000) combined with period-of-occurrence data found in the scientific literature (McCracken 1963, Pearcy 1962). 80 Northeastern Naturalist Vol. 13, Special Issue 4 As noted above, the winter flounder larval life stage could not be examined due to the lack of specific thermal tolerance data (i.e., paired acclimation/exposure-temperature information) in the scientific literature. 6. Apply the preceding inputs to predict the plume’s effects on winter flounder biological functions. For each grid cell within winter flounder’s habitat in Mount Hope Bay, the acclimation temperature (the average of the cell’s temperatures during the prior seven days) was determined from ASA hydrothermal modeling output using the model's bottom layer temperatures (since winter flouder is a benthic species). This was done for every day that winter flounder reside in the habitat. Then, for each day for each biothermal function evaluated, the cell’s acclimation temperature and exposure temperature were evaluated using the thresholds in the temperature tolerance polygon in order to predict the biothermal effect.4 This process was carried out for each Station operating scenario using the ASA scenario-dependent baywater temperature values. Table 1 summarizes the life stages, habitats, and periods of occurrence examined for winter flounder. Collectively, these parameters comprised the inputs to the analytical process. Figure 4. Juvenile and adult winter flounder habitats in Mount Hope Bay. 2006 R.J. O’Neill, T.L. Englert, and J.K. Ko 81 Table 1. Winter flounder assessment—biothermal metrics evaluated and model inputs applied. Habitat Period of occurrence in Mount Hope Bay Biothermal function Life stage (see Fig. 4) Dates Rationale Reproduction Eggs Adult Jan 25 to May 11 The period of occurrence for winter flounder egg production was determined from the catch records of MRI’s 1973–1999 Mount Hope Bay ichthyoplankton sampling program (MRI 2000). Growth, avoidance, and chronic mortality Juveniles Juvenile Apr 21 to Oct 31 Both MRI data as well as life-history information from the scientific literature were used. To avoid excluding juveniles whose larval Stage IV antecedents occurred as early as April 21, the period of occurrence for juvenile winter flounder was assumed to be April 21 through October 31. Juvenile winter flounder remain in the nursery area throughout the summer and move offshore when water temperatures drop below 8 ºC (Casterlin and Reynolds 1982). Growth, avoidance, and chronic mortality Adults Adult Periods when bay- McCracken (1963) notes that adult winter flounder leave bays water temperatures and estuaries when bottom temperatures exceed 15 ºC. are below 15 ºC Blockage Juveniles Juvenile April, October Winter flounder juveniles move into Mount Hope Bay’s tributaries during April, remain there until the fall, and then in October return to the Bay’s more saline water to overwinter. 82 Northeastern Naturalist Vol. 13, Special Issue 4 Methods and Results Results of the Mount Hope Bay winter flounder biothermal assessment are summarized in Table 2 and discussed below. For each biothermal function evaluated, details about data processing techniques are included. Growth For this assessment, it was assumed that the rate of growth is uniform throughout the growth zone (Fig. 2) and that sufficient food is available for growth when exposure temperatures are within the growth-zone range. The following steps were performed for winter flounder adults and juveniles: 1. For each day within winter flounder’s period of occurrence, the acclimation and exposure temperatures for each habitat grid cell were examined relative to the growth zone shown in the temperature tolerance polygon. If the acclimation/exposure point of intersection fell within the growth-zone, that portion of the habitat was found suitable for growth. 2. The areas of the individual grid cells suitable for growth on a given day were summed to derive the total habitat area available for growth on that day. 3. For each day evaluated, the total area available for growth was divided by the total habitat area to derive the portion of habitat suitable for growth. The summation of these values over the applied period of occurrence yields the cumulative number of days for which growth is expected. No change in growth of adult winter flounder was predicted (Table 2). For juveniles, which are present in the Bay 194 days of the year, growth days lost are 8.6, 6.3, and 1.1 under current Station operation, operation with reduced heat load, and the no-plant scenario, respectively (Fig. 5).5 Compared to the no-plant scenario, the effect of current operation is a decrease of only 3.9%. Table 2. Winter flounder assessment—summary of biothermal effects found. Mount Hope Bay environment Current Station station operation with No-plant Biothermal metric Life stage operation reduced heat load effect Growth Juveniles 4.4% 3.2% 0.6% (% growth days lost) (8.6 days lost) (6.3 days lost) (1.1 days lost) Adults None predicted Reproduction Eggs 5.2% 4.8% 2.5% (% thermal egg mortality) Avoidance Juveniles 3.0% 2.8% 1.9% (% of habitat avoided) Adults None predicted Potential for blockage at the Juveniles None predicted entrances of the Mount Hope Bay tributaries Chronic mortality Juveniles 4.6% 3.6% 1.5% (% mortality from Adults None predicted a 72-hour exposure) 2006 R.J. O’Neill, T.L. Englert, and J.K. Ko 83 Temperature during critical reproductive seasons The length of incubation is of key importance in assessing biothermal effects on reproduction—the longer the incubation period, the greater the likelihood that potentially lethal temperatures could be experienced. Typically, winter flounder incubate for 18 days. Thus, if a single egg cohort (a given day’s egg production) is to avoid thermal mortality, non-lethal temperatures must occur for 18 consecutive days. Egg catch data from the MRI 1973–1999 ichthyoplankton sampling program indicate that winter flounder spawn in Mount Hope Bay between January 25 and May 11. The catch data were used to approximate the number of eggs in development on a given day and the percentage of annual Baywide winter flounder egg production represented by that number. The daily production value is critical input to the biothermal model because it provides the basis for defining predicted temperature effects relative to the proportion of the total egg production exposed. For example, when the predicted temperature effect for a given day is 100% mortality, all the eggs present on that day, regardless of their stages of development, are lost. Using the predicted adult-habitat temperatures, the reproduction inhibition temperature zone shown in Figure 2, and the methods discussed earlier, the biothermal effect on reproduction was determined as follows: 1. The percentage of thermal egg mortality on a given day was determined within each adult winter flounder habitat grid cell (Fig. 4). 2. This value was then multiplied by the percentage of the standing egg density present on that day (i.e., the portion of the total egg production that could potentially be affected) to yield a daily value. 3. Daily values were then summed to determine the percentage of annual egg production subject to thermal egg mortality. Figure 5. Total number of growth days for juvenile winter flounder. 84 Northeastern Naturalist Vol. 13, Special Issue 4 Figure 6 presents the plume’s predicted effect on winter flounder egg mortality—5.2%, 4.8%, and 2.5% for current Station operation, reducedheat- load-operation, and the no-plant scenario, respectively. As can be seen, compared to the no-plant scenario, current operation increases thermal egg mortality by only 2.7 percentage points (i.e., 5.2% - 2.5%). Thermal avoidance and habitat loss Using the mean avoidance line in Figure 2, with the lower and upper bounds around the mean set at ± 5 °C and a normal distribution assumed, the percent avoidance in each grid cell within the adult or juvenile winter flounder habitat was determined for each day. The area-weighted percentage of total habitat avoided was then determined, per Equation 1, by summing together the products of predicted avoidance and planar area for all grid cells in the habitat, and dividing the summed result by the total winter flounder habitat area and then multiplying by 100. Thermal avoidance caused by the Station’s plume was predicted to be zero for adult winter flounder, largely because adults are not present in Mount Hope Bay during the warmer months of the year, and when adults are present, temperatures even in the warmest portions of the near-field thermal plume are below winter flounder avoidance temperatures. It is well documented that adult winter flounder leave shoreline waters when the temperatures reach approximately 15 °C (McCracken 1963, USEPA 2002). This is not an avoidance response, since there is no laboratory correlation between 15 °C and a short-term avoidance response (i.e., no evidence of thermal discomfort). Instead, the adult winter flounder exodus Figure 6. Winter flounder thermal egg mortality. 2006 R.J. O’Neill, T.L. Englert, and J.K. Ko 85 reflects an evolutionary adaptation. That is, a temperature of 15 °C acts as a signal that Mount Hope Bay has begun its normal seasonal warming—a signal to which adult winter flounder respond by moving into the cooler ocean waters.6 As shown in Figure 7, for juvenile winter flounder, the predicted maximum percentage of habitat avoided over the period of occurrence is 2.9%, 2.8%, and 1.9% under current Station operation, reduced-heat-load operation, and the no-plant scenario, respectively. Potential for blockage at the entrance of the Mount Hope Bay tributaries Winter flounder juveniles move into Mount Hope Bay’s tributaries (Fig. 1) during May and June, remain there until October, and then return to the Bay’s more saline water to overwinter in November–December. The assessment of blockage tested whether passage through the “door” at each tributary mouth is prevented by temperatures that trigger an avoidance response. The blockage evaluation focused on movement into and through the tributaries in the northern portion of Mount Hope Bay. Such patterns of movement generally require passage through Brayton Point Station’s thermal plume. The potential for winter flounder blockage was quantified by comparing the mean avoidance temperatures shown in Figure 2 with predicted hourly temperatures at the mouth of each tributary. If the potential for blockage was indicated that is, if the mean avoidance temperature was less than the predicted plume temperatures then the cross-sectional area of the plume isotherms that exceeded the avoidance temperature was determined. This area was then compared to the total cross-sectional area available for migration into a given tributary. Figure 7. Juvenile winter flounder thermal avoidance. 86 Northeastern Naturalist Vol. 13, Special Issue 4 Because the analysis of migratory blockage accounted for the movement of winter flounder through the Bay and into the tributaries, the acclimation temperature was defined as the average temperature at Spar Island (located at the Bay’s approximate midpoint) over the seven days prior to the day evaluated. It was assumed that juveniles reside in the open portions of Mount Hope Bay before they move into the tributaries. In assessing the potential for blockage, this assumption resulted in the use of conservative acclimation temperatures (i.e., colder temperatures than might apply if juveniles reside closer to Brayton Point prior to moving into the tributaries). The acclimation temperatures at Spar Island and the exposure temperatures in the tributary mouths were determined using the hourly bottom-layer temperatures calculated by the ASA hydrothermal model for those locations. The results of this hourly analysis showed that no avoidance was predicted over any portion of any tributary mouth cross-section. Thus, passage for winter flounder was assured for all of the tributaries and operating scenarios considered. Potential for chronic thermal mortality due to a prolonged exposure The assessment tested whether sedentary (i.e., burrowed) winter flounder might be exposed to temperatures that would be lethal after prolonged (72- hour) exposure. If the temperatures are below the lethal threshold, then no effect is predicted.7 The assessment of winter flounder thermal mortality due to a prolonged exposure followed the same analytical approach used for avoidance. That is, the population response around the mean (i.e., TL5072 hr) was determined as shown in Figure 4. Several assumptions were made in assessing winter flounder chronic thermal mortality: Populations are static. The net movement of fish populations into and out of each grid cell was assumed to be zero. This assumption is conservative, because it ignores movement out of the cell to avoid elevated temperatures. Throughout the 72-hour exposure, each grid cell’s maximum daily temperature applies. This approach is conservative because the average daily temperature over a 3-day period would always be less than the highest daily value (assuming the daily values are not identical) and thus less likely to cause mortality. The results showed no chronic thermal mortality for adult winter flounder. Figures 8 through 10 present the average predicted chronic mortality for juveniles under the three scenarios examined. In general, while these results show an isolated pocket of elevated mortality in the Lee River (where the shallow water depths cause the plume temperatures to be higher than at the point of Station discharge), overall area-weighted mortality is small (e.g., the difference in area-weighted predicted mortality between current Station operation and the no-plant scenario is a modest 3.1 percentage points [4.6% - 1.5%]). 2006 R.J. O’Neill, T.L. Englert, and J.K. Ko 87 Figure 8. Juvenile winter flounder 72-hour chronic mortality under current Station operation. Figure 9. Juvenile winter flounder 72-hour chronic mortality under reduced-heat Station operation. 88 Northeastern Naturalist Vol. 13, Special Issue 4 Figure 10. Juvenile winter flounder 72-hour chronic mortality under the no-plant scenario. Summary and Conclusions Table 2 summarizes the winter flounder biothermal effects found. No effects were found for adults for any of the biological functions examined. For eggs and juveniles, effects were found for certain biological functions, but they were less than 5% in all cases except for thermal egg mortality. For that case, the effect was 5.2% under current operation, but the difference between that operating scenario and the no-plant scenario was only 2.7%. Thus, it is concluded, based on these model predictions, that the thermal plume occurring under either the current or reduced-heat operating scenario will not cause any appreciable harm to the Mount Hope Bay winter flounder population. This conclusion is consistent with findings for the other nine species evaluated in the biothermal assessment (USGenNE 2001). Importance of a conservative modeling approach Because the biothermal model accounts for direct thermal effects only, it was deemed important to take a conservative approach—that is, to model “worst-case” effects. This was done to ensure that the direct impact of the Station’s thermal plume would not be underestimated. As a result, it is likely that biothermal effects found are to a degree overestimates of the plume’s direct impact on the species and life stages examined. Key conservative aspects of the model include the following: 1. Effects for a very warm year were modeled. The year 1999 was chosen because it had the second highest water temperatures in a 40-year period. 2006 R.J. O’Neill, T.L. Englert, and J.K. Ko 89 2. It was assumed that winter flounder populations are static within grid cells. The assumption of net-zero movement in and out of grid cells in assessing chronic thermal mortality ignores the ability of fish to leave areas of elevated temperatures and thus avoid any thermal effects. 3. In the chronic mortality assessment, each grid cell’s highest daily temperature was modeled rather than an average of the daily temperatures within the 72-hour analytical period. Assuming that daily values are not equal, the average daily temperature would always be less than the temperature evaluated. 4. The chronic mortality assessment used the longest time frame (72 hours) found for such assessments in the literature. Chronic thermal assessments of 1, 3, 6, 24, and 48 hours were also found in the literature. 5. In assessing migratory blockage, the temperature near Spar Island (mid- Bay) was used in calculating acclimation temperature. If juveniles reside closer to Brayton Point prior to moving into the tributaries, calculated acclimated temperatures would be warmer values, which in turn would result in a lower avoidance response than calculated. Viewing results within their environmental context In examining the output of biothermal modeling, it is important to be mindful of the environmental context, as other direct effects may be operating within the sphere of study. In the case of Mount Hope Bay, documented fishing pressure appears to greatly exceed the modeled effects of the Station’s thermal plume on winter flounder. For example, the 1990–1998 average fishing exploitation rate for Narragansett Bay and adjacent coastal waters, deduced from the Report of the 36th Northeast Regional Stock Assessment Workshop (36th SAW)’s Stock Assessment Review Committee (SARC 2004), is 55 percent;8 Gibson (2000) reports a similar rate for this period—53 percent. It should be noted, however, that recent implementation of tightened commercial and recreational fishing regulations for Narragansett Bay and Rhode Island coastal waters has resulted in some success, as shown in SARC (2004), which puts average exploitation rates for 1999, 2000, and 2001 at 40, 39, and 37 percent, respectively. It should also be noted that the modeled effects from the Station’s thermal discharge occur only during the first year of life for each cohort. Fishing impacts, however, affect the same cohort during each year that is susceptible to exploitation by the fishery (i.e., from age 2), with a potential to compound effects over multiple years. Addressing model limitations The authors acknowledge two limitations inherent in biothermal assessment tools: the ability to readily mimic or simulate field conditions with models populated solely or primarily with laboratory data, and the ability to address the impacts and interactions of non-thermal environmental influences potentially affected by thermal inputs, such as competition, predation, and changes in water quality, food availability, and physiological condition 90 Northeastern Naturalist Vol. 13, Special Issue 4 (Wismer and Christie 1987). This assessment measures only direct thermal effects. Quantification of indirect thermal effects (e.g., increased predation or diminished predator avoidance) are currently beyond the reach of biothermal modeling, however, largely because insufficient data is available to properly quantify the associated variables and assess their complex interactions within a large natural ecosystem. In the future, the ability to improve the precision of in situ effects and behaviors will likely depend on technologies not yet available or currently in early stages of development. One possible enhancement to make more field data available to the modeling process could be the use of data storage tags (DSTs). DSTs attached directly to the species of concern can record such parameters as temperature, conductivity, and pressure. DSTs have “recently emerged as a new area of science that could combine the rigor of controlled conditions with the acquisition of detailed data measured in situ” (Ropert- Coudert and Wilson 2004). However, more time is likely needed for further advancement in miniaturization and sensor resolution in this new field of data acquisition before it can be applied—at least to the juvenile winter flounder population of Mount Hope Bay, which could suffer physical impairment or behavior alterations due to the current size of the archival units. More work is also needed to enhance the ability to predict the effect of multiple stimuli acting alone or in concert on the biota of a water body. For example, in addition to the thermal impact, egg mortality can also be caused by such environmental factors as predation, changes in salinity and dissolved oxygen levels, wave action, solar and UV radiation, and pollution. Inherited genetic flaws or poor parental condition also can influence egg mortality rates (Bunn et al. 2000). Nonetheless, the Mount Hope Bay biothermal assessment demonstrates that information currently available can be used to develop a sophisticated model of the direct impacts of thermal discharges on species resident in a water body. When conservative assumptions are incorporated throughout the analysis, they help to ensure that required approximations do not cause the analysis to underestimate the potential for thermal effects. Literature Cited Able, K.W., and M.P. Fahay. 1998. The First Year in the Life of Estuarine Fishes in the Middle Atlantic Bight. Rutgers University Press, New Brunswick, NJ. Beitinger, T.L., and W.A. Bennett. 2000. Quantification of the role of acclimation temperature in temperature tolerance of fishes. Environmental Biology of Fishes 58:277–288. Brett, J.R. 1956. Some principles in the thermal requirements of fishes. Quarterly Review of Biology 31(2):72–87. Buckley, L.J., A.S. Smigielski, T.A. Halavik, and G.C. Laurence. 1990. Effects of water temperature on size and biochemical composition of winter flounder Pseudopleuronectes americanus at hatching and feeding initiation. US National Marine Fisheries Service Fishery Bulletin 88(3):419–428. 2006 R.J. O’Neill, T.L. Englert, and J.K. Ko 91 Bunn, N.A., C.J. Fox, and T. Webb. 2000. Literature review of studies on fish egg mortality. Published on the Web in 2000 by the Centre for Environment, Fisheries, and Aquaculture Science (CEFAS). Available at: http://www.cefas.co.uk/ publications/techrep/tech111.pdf. Accessed July 2002. Casterlin, M.E., and W.W. Reynolds. 1982. Thermoregulatory behavior and diel activity of yearling winter flounder, Pseudopleuronectes americanus (Waldbaum). Environmental Biology of Fishes 7(2):177–180. Collie, J.S., and A.K. DeLong. 2002. Examining the decline of Narragansett Bay winter flounder. Final Report to RI DEM Division of Fish and Wildlife, Wickford, RI. July 23, 2002. Coutant, C.C. 1972. Biological aspects of thermal pollution. Vol. I: Entrainment and discharge canal effects. Critical Review in Environmental Control 3:341–381. Coutant, C.C. 1977. Physiological considerations of future thermal additions for aquatic life. Pp. 251–266, In M. Marois (Ed.). World Conference Toward a Plan of Actions for Mankind, Vol. 3: Biological Balance and Thermal Modifications. Pergamon Press, Oxford, UK. Fry, F.E.J. 1971. The effect of environmental factors on the physiology of fish. Pp. 1–98, In W.S. Hoar and D.J. Randall (Eds.). Fish Physiology, Vol. VI. Academic Press, New York, NY. Gibson, M.R. 2000. Recent trends in abundance, recruitment, and fishing mortality for winter flounder in Narragansett Bay and Rhode Island coastal waters. Report to the RI Marine Fisheries Council., RI Division of Fish and Wildlife, Wickford, RI. Gift, J.J., and J.R. Westman. 1971. Responses of some estuarine fishes to increasing thermal gradients. Ichthyological Associates, Inc., Ithaca, NY 154 pp. Hoff, J.G., and J.R. Westman. 1966. The temperature tolerance of three species of marine fishes. Journal of Marine Research 24(2):131–140. Hutchison, V.H., and J.D. Maness. 1979. The role of behavior in temperature acclimation and tolerance in ectotherms. American Zoologist 19:367–384. Ichthyological Associates, Inc. (IA). 1978. Predictive biological information to demonstrate the passage and maintenance of representative important species. Type III 316(a) Demonstration of Federal Water Pollution Control Act Amendments of 1972, PL92500 for Linden Generating Station. Report prepared for Public Service Electric and Gas Company, Newark, NJ. Kilgour, D.M., R.W. McCauley, and W. Kwain. 1985. Modeling the lethal effects of high temperature on fish. Canadian Journal of Fisheries and Aquatic Sciences 42(5):947–951. Klein-MacPhee, G. 1978. Synopsis of biological data for the winder flounder, Pseudopleuronectes americanus (Walbaum). US Department of Commerce, NOAA Technical Report, National Marine Fisheries Service Circular 414. FAO Fisheries Synopsis 117:1-43. Lawler, Matusky, and Skelly Engineers LLP. 2002. Comments on Draft NPDES Permit No. MA0003654 for Brayton Point Station, Somerset, MA. Lynch, T.R. 2000. Assessment of recreationally important finfish stocks in Rhode Island waters: Coastal fishery resource assessment trawl survey. Rhode Island Division of Fish and Wildlife, Providence, RI. Report # F-61-R-8. Mathur, D., R.M. Schutsky, E.J. Purdy, Jr., and C.A. Silver. 1983. Similarities in avoidance temperatures of freshwater fishes. Canadian Journal of Fisheries and Aquatic Sciences 40:2144–2152. 92 Northeastern Naturalist Vol. 13, Special Issue 4 McCracken, F.D. 1963. Seasonal movements of the winter flounder, Pseudopleuronectes americanus (Waldbaum), on the Atlantic coast. Journal of the Fisheries and Research Board of Canada 20(2):551–586. McCullough, D.A. 1999. A review and synthesis of effects of alterations to the water temperature regime on freshwater life stages of salmonids, with special reference to Chinook salmon. Columbia River Inter-Tribal Fish Commission, Portland OR. Meldrim, J.W., J.J. Gift, and B.R. Petrosky. 1974. The effect of temperature and chemical pollutants on the behavior of several estuarine organisms. Bulletin 11. Ichthyological Associates, Inc., Ithaca, NY. Marine Research, Inc. (MRI). 2000. Brayton Point Station annual report (draft), January–December 1999. Prepared for USGen New England, Inc. MRI, Falmouth, MA. September 15, 2000. National Academy of Sciences/National Academy of Engineering (NAS/NAE). 1973. Water quality criteria. 1972. Prepared for the USEPA. NAS/NAE, Washington, DC. Neill, W.H. 1979. Mechanisms of fish distribution in heterothermal environments. American Society of Zoologists 19:305–317. Olla, B.L., R. Wicklund and S. Wilk. 1969. Behavior of winter flounder in a natural habitat. Transactions. American Fisheries Society 98:717-720. Pearcy, W.G. 1962. Ecology of an estuarine population of winter flounder, Pseudopleuronectes americanus (Walbaum). II. Distribution, abundance, growth, and production of juveniles and survival of larvae and juveniles. Bull.etin of the Bingham Oceanographic Collection 18:39–64. Radle, E.W. 1971. A partial life history of the winter flounder (Pseudopleuronectes americanus) exposed to thermal addition in an estuary, Indian River Bay, Delaware. M.Sc. Thesis. University of Delaware, Lewes, DE. 74 pp. Ricker, W.E. 1975. Computation and interpretation of biological statistics of fish populations. Bulletin of the Fisheries Research Board of Canada 191:1–382. Rogers, C.A. 1976. Effects of temperature and salinity on the survival of winter flounder embryos. Fishery Bulletin 74(1):52–58. Ropert-Coudert, Y., and R.P. Wilson. 2004. Subjectivity in bio-logging: Do logged data mislead. Memoirs of the National Institute of Polar Research, Special Issue 58:23–33. SARC 2004. A report of the 38th Northeast Regional Stock Assessment Workshop. US Department of Commerce, National Oceanic and Atmospheric Administration, National Marine Fisheries Service, Northeast Fisheries Science Center, Woods Hole, MA. January 2004. Saucerman, S., and L.A. Deegan. 1991. Lateral and cross-channel movement of young-of-the-year winter flounder (Pseudopleuronectes americanus) in Waquoit Bay, Massachusetts. Estuaries 14(4):440–446. Spaulding, M.L., D. Mendelsohn, and J.C. Swanson. 1999. WQMAP: An integrated three-dimensional hydrodynamic and water quality model system for estuarine and coastal applications. Marine Technology Society Journal 33(3):38–54. Stoner, A.W., J.P. Manderson, and J.P. Pessutti. 2001. Spatially explicit analysis of estuarine habitat for juvenile winter flounder: Combining generalized additive models and geographic information systems. Marine Ecology Progress Series 213:253–271. Swanson, J.C., D. Mendelsohn, H. Rines, and H. Schuttenberg. 1998. Mount Hope Bay hydrodynamic model calibration and confirmation. Report to NEPCo, Westborough, MA. Applied Science Associates, Inc., Narragansett, RI. May 1998. 2006 R.J. O’Neill, T.L. Englert, and J.K. Ko 93 Terpin, K.M., M.C. Wyllie, and E.R. Holmstrom. 1977. Temperature preference, avoidance, shock, and swim speed studies with marine and estuarine organisms from New Jersey. Ichythyological Associates, Inc., Ithaca, NY. Bulletin #17. May 1977. US Environmental Protection Agency (USEPA). 1976. Quality criteria for water. Office of Water and Hazardous Materials. EPA-440/9-76-023. USEPA, Washington, DC. US Environmental Protection Agency (USEPA). 2002. Clean Water Act NPDES permitting determinations for thermal discharge and cooling water intake from Brayton Point Station in Somerset, MA. MA0003654 Determinations Document. USEPA, Region I - New England, Boston, MA. USGen New England, Inc. (USGenNE). 2001. Brayton Point Station 316(a) and 316(b) Demonstration. USGenNE, Somerset, MA. November 2001. Williams, G.C. 1975. Viable embryogenesis of the winter flounder Pseudopleuronectes americanus from -1.8 ° to 15 °C. Marine Biology (Berl.) 33:71–74. Wismer, D.A., and A.E. Christie. 1987. Temperature Relationships of Great Lakes Fishes: A Data Compilation. Great Lakes Fishery Commission Special Publication 87-3. 165 pp. Notes 1Under current operation of the plant, cooling water for the Station’s four generating units is withdrawn from and returned to Mount Hope Bay in a once-through cooling configuration during the warmer months; from October through May the Station operates in a “piggyback” configuration, in which cooling water from Units 1, 2, and 3 is recirculated to cool Unit 4. In this second set alternative cooling-water configuration, some of the heated water would be circulated to a mechanical-draft cooling tower, where it would be cooled with ambient air and then recirculated to condenser cooling-water inlets. This hypothetical configuration would reduce Station heat load (and Station flow) by 33% compared to current Station operation. 2The basic reason acclimation occurs is that fish lack the physiological mechanisms to control tissue temperature, and thus their peripheral body temperature is essentially the same as the surrounding water. Therefore, as water temperature and thus fish body temperature change, corresponding changes occur in thermal preference, avoidance, and mortality thresholds. 3The NAS/NEA 1973 critical growth limit equation was used for 9 of the 10 species evaluated. For winter flounder, however, an acclimation-preference data set was not found in the scientific literature. Thus, for this species, the upper limit of the growth zone was approximated by the field observations in Radle 1971. For example, field collection temperatures associated with the greater-than-half-full stomach observations for juvenile winter flounder were conservatively designated as the exposure “tipping point” at which elevated temperatures yield zero net growth. 4The decision to assess impacts in single cells, rather than in groups of cells, had a large impact on model run times and post-processing requirements. Nonetheless, it was pursued so that the precise location in the Bay of any predicted biothermal effect could be pinpointed. 94 Northeastern Naturalist Vol. 13, Special Issue 4 5For growth and the other biological functions examined, an “effect” is seen for the no-plant scenario because of natural seasonal variations in Mount Hope Bay ambient water temperatures in 1999. The no-plant effect demonstrates the model’s sensitivity to temperature variations. 6As stated in Neill (1979), “predictive thermoregulation comprises directed movements to a subset of habitat within which the fish “expects,” on the basis of individual or evolutionary experience, to find acceptable temperatures.” Predictive thermoregulation, a seasonal evolutionary adaptation, was outside the scope of the biothermal model. 7The potential for burrowing is supported by winter flounder physiological tolerance data, which reflect several instances of “no response” when the fish were subjected to elevated temperatures (Ichthyological Associates, Inc. 1978). 8SARC reports instantaneous fishing mortality rates for the Southern New England– Mid-Atlantic winter flounder stock complex, which includes Narragansett Bay. The SARC rates were converted to exploitation rates using the Type II fishery equation from Ricker (1975) and an instantaneous mortality rate of 0.2.