Massachusetts Estuaries Project

Linked Watershed-Embayment Model to Determine Critical Nitrogen Loading Thresholds for the Apponagansett Bay Estuary, Dartmouth, MA

University of Massachusetts Dartmouth Massachusetts Department of School of Marine Science and Technology Environmental Protection

DRAFT REPORT – JUNE 2015 Massachusetts Estuaries Project

Linked Watershed-Embayment Model to Determine Critical Nitrogen Loading Thresholds for the Apponagansett Bay Estuary, Dartmouth, MA

DRAFT REPORT – JUNE 2015

Brian Howes Tony Millham Ed Eichner Roland Samimy David Schlezinger

John Ramsey Trey Ruthven

Contributors:

US Geological Survey Don Walters and John Masterson

Applied Coastal Research and Engineering, Inc. Elizabeth Hunt and Sean Kelley

Massachusetts Department of Environmental Protection Charles Costello and Brian Dudley

SMAST Coastal Systems Program Jennifer Benson, Michael Bartlett, Sara Sampieri, and Elizabeth White

Lloyd Center for the Environment Mark Mello

MAE, Inc. Maisy McDarby Stanovich

ACKNOWLEDGMENTS

The Massachusetts Estuaries Project Technical Team would like to acknowledge the contributions of the many individuals who have worked tirelessly for the restoration and protection of the critical coastal resources of the Slocum's and Little River Estuaries and supported the application of the Linked Watershed-Embayment Model to Determine Critical Nitrogen Loading Thresholds for these systems. Without these stewards and their efforts, this project would not have been possible.

First and foremost we would like to recognize and applaud the significant time and effort in data collection and discussion spent by members of the Coalition for Buzzards Bay's BayWatcher, Water Quality Monitoring Program. These individuals gave of their time to consistent and sound nutrient related water quality from this system for over a decade. These data were enhanced by short-term intensive water quality monitoring efforts by the Lloyd Center Staff and interns participating in the Turn the Tide Project. These multiple efforts have resulted in a solid water quality database for this estuary, without which the present analysis would not have been possible. Of particular note has been the effort of the Coalition for Buzzards Bay's Monitoring Coordinator, Tony Williams, who has spent countless hours ensuring a scientifically defensible monitoring program and reviewing data and information with MEP Technical Team members. The present MEP Technical Report was developed as part of the on-going partnership project to restore the Town of Dartmouth's estuaries, known as Turn-The-Tide. This program is a partnership between the Town of Dartmouth, the Coalition for Buzzards Bay, the Lloyd Center for the Environment and the Coastal Systems Program at the School for Marine Science and Technology - University of Massachusetts Dartmouth. The goal is to restore water quality and living resources in Dartmouth's three major estuarine systems. Much of the present effort was conducted under this partnership, which greatly enhanced information and data transfers with the Town and its Departments. The Town of Dartmouth and its citizens have expended "extra" effort in the stewardship of these critical coastal systems and are to be commended. Departments and staff from the Town of Dartmouth have provided essential insights toward this effort, particularly the Department of Public Works and its commissioners who oversaw the Turn-The-Tide Project and specifically the help provided by David Hickox; Water and Sewer Division, Steve Sullivan; Animal Control, Sandra Gosselin and Wendy Henderson Board of Health. The technical team would like to specifically acknowledge the efforts of Mike O'Reilly, Dartmouth's Environmental Affairs Coordinator, who worked tirelessly as the Town interface with the Turn-The-Tide Project and was key in facilitating the overall effort. Other individuals of particular note include Mark Rasmussen, Executive Director CBB and Darcy McMahon Director of the Lloyd Center, who helped to guide and disseminate the information from the Turn-The-Tide effort. In addition to local contributions, technical, policy and regulatory support has been freely and graciously provided by Bill Napolitano and Karen Porter from SRPEDD; MaryJo Feurbach and Art Clark of the USEPA; and our MassDEP colleagues: Rick Dunn, Dave DeLorenzo and Geri Lambert. We are also thankful for the long hours in the field and laboratory spent by the technical staff, interns and students within the Coastal Systems Program at SMAST-UMD.

Support for this project was provided by the Town of Dartmouth through the Turn the Tide Project, private donations from citizens of the Town of Dartmouth, the MassDEP, and the USEPA.

This Nutrient Threshold Technical Report to be used by the Town of Dartmouth in the restoration of this key estuarine system is dedicated to Don Tucker former Director of the Lloyd Center, who helped to establish the Turn the Tide partnership project but passed away suddenly before its completion.

PROPER CITATION

Howes B.L., N.P. Millham, S.W. Kelley, J. S. Ramsey, R.I. Samimy, D.R. Schlezinger, E.M. Eichner (2007). Linked Watershed-Embayment Model to Determine Critical Nitrogen Loading Thresholds for the Apponagansett Bay Estuary, Dartmouth, Massachusetts. SMAST/DEP Massachusetts Estuaries Project, Massachusetts Department of Environmental Protection. Boston, MA.

© [2015] University of Massachusetts & Massachusetts Department of Environmental Protection All Rights Reserved No permission required for non-commercial use MASSACHUSETTS ESTUARIES PROJECT

TABLE OF CONTENTS I. INTRODUCTION ...... 1 I.1 THE MASSACHUSETTS ESTUARIES PROJECT APPROACH ...... 4 I.2 SITE DESCRIPTION ...... 7 I.3 WATER QUALITY MODELING ...... 14 I.4 REPORT DESCRIPTION ...... 14 II. PREVIOUS STUDIES RELATED TO NITROGEN MANAGEMENT ...... 16 III. DELINEATION OF WATERSHEDS ...... 35 III.1 BACKGROUND ...... 35 III.2 WATERSHED DELINEATION APPROACH ...... 35 IV. WATERSHED NITROGEN LOADING TO EMBAYMENT: LAND USE, STREAM INPUTS, AND SEDIMENT NITROGEN RECYCLING...... 39 IV.1 WATERSHED LAND USE BASED NITROGEN LOADING ANALYSIS ...... 39 IV.1.1 Land Use, Water Use, and Sewered Properties Database Preparation ...... 41 IV.1.2 Nitrogen Loading Input Factors ...... 44 IV.1.3 Calculating Nitrogen Loads ...... 51 IV.2 ATTENUATION OF NITROGEN IN SURFACE WATER TRANSPORT ...... 59 IV.2.1 Background and Purpose ...... 59 IV.2.2 Surface water Discharge and Attenuation of Watershed Nitrogen: Stream Discharge from Buttonwood Brook to the head of Apponagansett Bay ...... 63 IV.2.3 Surface water Discharge and Attenuation of Watershed Nitrogen: Stream Discharge Apponagansett Brook (aka. Vincent Brook) to head of Apponagansett Bay ...... 67 IV.3 BENTHIC REGENERATION OF NITROGEN IN BOTTOM SEDIMENTS ...... 70 IV.3.1 Sediment-Watercolumn Exchange of Nitrogen ...... 70 IV.3.2 Method for determining sediment-watercolumn nitrogen exchange ...... 71 IV.3.3 Rates of Summer Nitrogen Regeneration from Sediments ...... 73 V. HYDRODYNAMIC MODELING ...... 78 V.1 INTRODUCTION...... 78 V.2 FIELD DATA COLLECTION AND ANALYSIS ...... 80 V.2.1. Bathymetry ...... 80 V.2.2 Tide Data Collection and Analysis ...... 82 V.3 HYDRODYNAMIC MODELING ...... 89 V.3.1 Model Theory ...... 89 V.3.2 Model Setup ...... 89 V.3.2.1 Grid Generation ...... 90 V.3.2.2 Boundary Condition Specification ...... 90 V.3.3 Calibration ...... 92 V.3.4 Model Verification ...... 98 V.3.4.1 Model Circulation Characteristics ...... 100 V.4 FLUSHING CHARACTERISTICS ...... 101

i

MASSACHUSETTS ESTUARIES PROJECT

VI. WATER QUALITY MODELING ...... 105 VI.1 DATA SOURCES FOR THE MODEL ...... 105 VI.1.1 Hydrodynamics and Tidal Flushing in the Embayment ...... 105 VI.1.2 Nitrogen Loading to the Embayment ...... 105 VI.1.3 Measured Nitrogen Concentrations in the Embayment ...... 105 VI.2 MODEL DESCRIPTION AND APPLICATION ...... 105 VI.2.1 Model Formulation ...... 108 VI.2.2 Water Quality Model Setup ...... 108 VI.2.3 Boundary Condition Specification ...... 109 VI.2.4 Model Calibration ...... 110 VI.2.5 Model Salinity Verification ...... 112 VI.2.6 Build-Out and No Anthropogenic Load Scenarios ...... 114 VI.2.6.1 Build-Out ...... 115 VI.2.6.2 No Anthropogenic Load ...... 118 VII. ASSESSMENT OF EMBAYMENT NUTRIENT RELATED ECOLOGICAL HEALTH ...... 120 VII.1 OVERVIEW OF BIOLOGICAL HEALTH INDICATORS ...... 120 VII.2 BOTTOM WATER DISSOLVED OXYGEN ...... 121 VII.3 EELGRASS DISTRIBUTION - TEMPORAL ANALYSIS ...... 131 VII.4 BENTHIC INFAUNA ANALYSIS ...... 135 VIII. CRITICAL NUTRIENT THRESHOLD DETERMINATION AND DEVELOPMENT OF WATER QUALITY TARGETS ...... 144 VIII.1. ASSESSMENT OF NITROGEN RELATED HABITAT QUALITY ...... 144 VIII.2 THRESHOLD NITROGEN CONCENTRATIONS ...... 149 VIII.3. DEVELOPMENT OF TARGET NITROGEN LOADS ...... 152 IX. MANAGEMENT SCENARIO ...... 157 IX.1 BACKGROUND ...... 157 X. LIST OF REFERENCES ...... 160 APPENDIX 1 ...... A-1

ii

MASSACHUSETTS ESTUARIES PROJECT

LIST OF FIGURES

Figure I-1. Study region proximal to the Apponagansett embayment system for the Massachusetts Estuaries Project nutrient analysis. Tidal waters enter the outer bay in the lower right of the frame from Buzzards Bay. Freshwaters enter primarily from two surface water streams at the northern head of the inner bay...... 2 Figure I-2. Massachusetts Estuaries Project Critical Nutrient Threshold Analytical Approach. Section numbers refer to sections in this MEP report where the specified information is provided...... 7 Figure I-3. Apponagansett Bay with the embayment watershed and sub-watersheds ...... 9 Figure I-4. Apponagansett Bay subdivisions for water quality monitoring...... 12 Figure II-1. BayWatcher Water Quality Monitoring Program of the Apponagansett Bay Estuary, Town of Dartmouth. Estuarine water quality monitoring stations sampled by the Coalition for Buzzards Bay and the Lloyd Center / Turn the Tide Program. Stream water quality stations depicted in Section IV were sampled weekly by the Lloyd Center staff...... 18 Figure II-2a. Mouth of Coastal Rivers designation for Apponagansett Bay as determined by – MassDEP Wetlands Program...... 29 Figure II-2b. Mouth of Coastal Rivers designation for Apponagansett Bay as determined by – MassDEP Wetlands Program...... 30 Figure II-2c. Mouth of Coastal Rivers designation for Apponagansett Bay as determined by – MassDEP Wetlands Program...... 31 Figure II-3. Location of designated shellfish growing areas and their status relative to shellfish harvesting as determined by Mass Division of Marine Fisheries. Closures are generally related to bacterial contamination or "activities", such as the location of marinas...... 32 Figure II-4. Location of shellfish suitability areas within the Apponagansett Bay Estuary as determined by Mass Division of Marine Fisheries. Suitability does not necessarily mean "presence"...... 33 Figure II-5. Estimated Habitats for Rare Wildlife and State Protected Rare Species within the Apponagansett Bay Estuary as determined by - NHESP...... 34 Figure III-1. Watershed and sub-watershed delineations for the Apponagansett Bay estuary system. Outer watershed boundary is based on USGS/BBP (1991) topographic delineation with adjustments in the lower eastern boundary to reflect the stormwater system collection area of the City of New Bedford. Interior subwatershed delineations were completed by MEP staff using the same topographic examination techniques; interior subwatershed delineations were completed to match natural watershed or estuary features (e.g., basins on either side of Gulf Road crossing) or key measurement points (e.g., MEP stream gauges at Buttonwood Brook and Apponagansett Brook). The watershed is mostly in the Town of Dartmouth, but is also shared with the City of New Bedford...... 37 Figure IV-1. Land-use in the Apponagansett Bay watershed. The watershed is split among Town of Dartmouth and the City of New Bedford. Land use classifications are based on municipal assessors’ records: 2014 for Dartmouth and 2009 for New Bedford...... 42

iii

MASSACHUSETTS ESTUARIES PROJECT

Figure IV-2. Distribution of land-uses within the major sub-watersheds and whole watershed to the Apponagansett Bay estuary system. Only percentages greater than or equal to 3% are shown...... 43 Figure IV-3. Wetland areas in the Apponagansett Bay watersheds. All areas colored in green are wetlands areas delineated by MassGIS/MassDEP 1:12,000K coverage. Most of these areas are associated with freshwater streams that discharge into the Bay. All these areas were assigned a surface water nitrogen load in the MEP watershed nitrogen loading model...... 50 Figure IV-4 (a-b). Land use-specific unattenuated nitrogen load (by percent) to the (a) overall Apponagansett Bay Estuary System watershed and (b) Buttonwood Brook. “Overall Load” is the total nitrogen input within the watershed, while the “Local Control Load” represents only those nitrogen sources that could potentially be under local regulatory control...... 55 Figure IV-5. Developable Parcels in the Apponagansett Bay watershed. Parcels that are shown are either parcels with no existing development but classified by the respective town assessors as developable or parcels with existing development, but potential for additional development based on minimum lot sizes specified in respective town zoning regulations. Buildout assessments include changes based on review by municipal staff...... 58 Figure IV-6a. Locations of the two stream gaging sites in the Apponagansett Bay watershed. Red line is the watershed boundaries to Buttonwood Brook and Apponagansett Brook. The blue dashed line is the overall watershed boundary for the Apponagansett Bay system...... 61 Figure IV-6b. Location of stream gauges (red symbols) in the Apponagansett Bay embayment system...... 62 Figure IV-7. Buttonwood Brook discharge (solid blue line), nitrate+nitrite (yellow square) and total nitrogen (blue squares) concentrations for determination of annual volumetric discharge and nitrogen load from the upper watershed to Apponagansett Bay (Table IV-3)...... 66 Figure IV-9. Apponagansett Bay embayment system sediment sampling sites (red symbols) for determination of nitrogen regeneration rates (all sites sampled in both 2004 and 2005). Numbers are for reference to station identifications listed above...... 73 Figure IV-10. Conceptual diagram showing the seasonal variation in sediment N flux, with maximum positive flux (sediment output) occurring in the summer months, and maximum negative flux (sediment up-take) during the winter months...... 75 Figure V-1. Map of the Apponagansett Bay estuary system (from United States Geological Survey topographic maps)...... 79 Figure V-2. Map of the study region identifying locations of the tide gauges used to measure water level variations throughout the system. Four (4) gauges were deployed for the 36-day period between June18, and July 24, 2003. Each yellow dot represents the approximate locations of the tide gauges: (Apg-1) represents the gage in Buzzards Bay (Offshore), (Apg-2) inside the bay entrance, (Apg-3) in Dike Meadow Creek, (Apg-4) in Apponagansett River...... 81 Figure V-3. Bathymetric data interpolated to the finite element mesh of hydrodynamic model July 17th & 18th 2003...... 82 Figure V-4. Water elevation variations as measured at the seven locations within the Apponagansett Bay system, between June18 and July 23, 2003...... 83

iv

MASSACHUSETTS ESTUARIES PROJECT

Figure V-5 Plot showing two tide cycles tides at four stations in the Apponagansett Bay estuary system plotted together. Demonstrated in this plot is the phase delay effect caused by the propagation of the tide through the estuary...... 84 Figure V-6. Example of observed astronomical tide as the sum of its primary constituents. In this example the observed tide signal is the sum of individual constituents (M2, M4, K1, N2), with varying amplitude and frequency...... 86 Figure V-7. Results of the harmonic analysis and the separation of the tidal from the non-tidal, or residual, signal measured in Buzzards Bay (Apg-1)...... 88 Figure V-8. The model finite element mesh developed for Apponagansett Bay estuary system...... 91 Figure V-9. Depth contours of the completed Apponagansett Bay finite element mesh...... 92 Figure V-10. Material types assigned to the finite element mesh for the Apponagansett Bay estuary...... 93 Figure V-11. Comparison of water surface elevation at the Bridge Street Bridge (App-2) predicted by the model (dots) to those measured in the field (solid line) for the calibration time period...... 96 Figure V-12. Comparison of water surface elevation in Dike Meadow Creek (App-3) predicted by the model (dots) to those measured in the field (solid line) for the calibration time period...... 96 Figure V-13. Comparison of water surface elevation in Apponagansett River (App-4) predicted by the model (dots to those measured in the field (solid line) for the calibration time period...... 97 Figure V-14. Comparison of water surface elevation at the Bridge Street Bridge (App-2) predicted by the model (dots) to those measured in the field (solid line) for the verification time period...... 98 Figure V-15. Comparison of water surface elevation in Dike Meadow Creek (App-3) predicted by the model (dots) to those measured in the field (solid line) for the verification time period...... 99 Figure V-16. Comparison of water surface elevation in Apponagansett River (App-4) predicted by the model (dots to those measured in the field (solid line) for the verification time period...... 99 Figure V-17. Example of hydrodynamic model output at the entrance to Dike Meadow Creek for a single time step where maximum ebb velocities occur for this tide cycle. Color contours indicate flow speed, and vectors indicate the direction and magnitude of flow...... 101 Figure VI-1. Estuarine water quality monitoring station locations in the Apponagansett Bay System. Station labels correspond to those provided in Table VI-1...... 106 Figure VI-2. Map of Apponagansett Bay water quality model longitudinal dispersion coefficients. Color patterns designate the different areas used to vary model dispersion coefficient values...... 111 Figure VI-3. Comparison of measured total nitrogen concentrations and calibrated model output at stations in Apponagansett Bay System. For the left plot, station labels correspond with those provided in Table VI-1. Model output is presented as a range of values from minimum to maximum values computed during the simulation period (triangle markers), along with the average computed concentration for the same period (square markers). Measured data are presented as the total yearly mean at each station (circle markers), together with ranges that indicate ± one standard deviation of the entire dataset. For the plots to the right, model calibration

v

MASSACHUSETTS ESTUARIES PROJECT

target values are plotted against measured concentrations, together with the unity line...... 112 Figure VI-4. Contour plots of average total nitrogen concentrations from results of the present conditions loading scenario, for Apponagansett Bay System. The approximate location of the sentinel threshold station for Apponagansett Bay System (AB-4) is shown...... 113 Figure VI-5. Comparison of measured and calibrated model output at stations in Apponagansett Bay System. For the left plots, stations labels correspond with those provided in Table VI-1. Model output is presented as a range of values from minimum to maximum values computed during the simulation period (triangle markers), along with the average computed salinity for the same period (square markers). Measured data are presented as the total yearly mean at each station (circle markers), together with ranges that indicate ± one standard deviation of the entire dataset. For the plots to the right, model calibration target values are plotted against measured concentrations, together with the unity line...... 114 Figure VI-6. Contour plots of modeled salinity (ppt) in Apponagansett Bay System...... 115 Figure VI-7. Contour plots of modeled total nitrogen concentrations (mg/L) in Apponagansett Bay System, for projected build-out loading conditions, and bathymetry. The approximate location of the sentinel threshold station for Apponagansett Bay System (AB-4) is shown...... 117 Figure VI-8. Contour plots of modeled total nitrogen concentrations (mg/L) in Apponagansett Bay System, for no anthropogenic loading conditions, and bathymetry. The approximate location of the sentinel threshold station for Apponagansett Bay System (AB-4) is shown...... 119 Figure VII-1. Average water column respiration rates (micro-Molar/day) from water collected throughout the Popponesset Bay System (Schlezinger and Howes, unpublished data). Rates vary ~7 fold from winter to summer as a result of variations in temperature and organic matter availability...... 122 Figure VII-2. Aerial photograph of Apponagansett Bay system in Dartmouth showing locations of Dissolved Oxygen mooring deployments conducted in the Summer of 2003-2004...... 124 Figure VII-3. Bottom water record of dissolved oxygen at Apponagansett Bay Upper station, Summer 2003. Calibration samples represented as red dots...... 126 Figure VII-4. Bottom water record of dissolved oxygen in Apponagansett Bay Mid station, Summer 2004. Calibration samples represented as red dots...... 126 Figure VII-5. Bottom water record of dissolved oxygen in Apponagansett Bay Lower station, Summer 2004. Calibration samples represented as red dots...... 127 Figure VII-6. Bottom water record of Chlorophyll-a at Apponagansett Bay Upper station, Summer 2003. Calibration samples represented as red dots...... 127 Figure VII-7. Bottom water record of Chlorophyll-a in Apponagansett Bay Mid station, Summer 2004. Calibration samples represented as red dots...... 128 VII-8. Bottom water record of Chlorophyll-a in Apponagansett Bay Lower station, Summer 2004. Calibration samples represented as red dots...... 128 Figure VII-9. Eelgrass bed distribution within Apponagansett Bay system. The 1995 coverage is depicted by the green outline inside of which circumscribes the eelgrass beds. The yellow (2001) areas were mapped by DEP. All data was provided by the DEP Eelgrass Mapping Program...... 132 Figure VII-10. Eelgrass bed distribution within Apponagansett Bay system. The 1951 coverage is depicted by the dark green outline (hatched area) inside of which circumscribes the eelgrass beds. In the composite photograph, the

vi

MASSACHUSETTS ESTUARIES PROJECT

light green outline depicts the 1995 eelgrass coverage and the yellow outlined areas circumscribe the eelgrass coverage in 2001. The 1995 and 2001 areas were mapped by DEP. All data was provided by the DEP Eelgrass Mapping Program...... 133 Figure VII-12. Aerial photograph of Apponagansett Bay system showing location of benthic infaunal sampling stations (blue symbol)...... 139 Figure VII-13. Shellfishing closures in Apponagansett Bay. Areas BB-12.4, 12.1 and 12.2 are prohibited for shellfishing. The locations of the closed areas are related to surface water sources of stormwater runoff from Padanaram village areas and the upper Apponagansett Bay watershed. (Map source MA Division of Marine Fisheries)...... 141 Figure VII-14. Location of shellfish growing areas and their status relative to shellfish harvesting as determined by Mass Division of Marine Fisheries. Closures are generally related to bacterial contamination or "activities", such as the location of marinas...... 142 Figure VII-15. Location of shellfish suitability areas within the Farm Pond estuary as determined by Mass Division of Marine Fisheries. Suitability does not necessarily mean "presence"...... 143 Figure VIII-1. Contour plot of modeled average total nitrogen concentrations (mg/L) in Apponagansett Bay System under Build-out, for threshold conditions (0.50 mg/L at water quality monitoring station AB-4). The approximate location of the sentinel threshold station for Apponagansett Bay System (AB-4) is shown...... 153

vii

MASSACHUSETTS ESTUARIES PROJECT

LIST OF TABLES

Table III-1. Apponagansett Bay MEP Subwatershed Areas and Estimated Long-Term Freshwater Recharge...... 38 Table IV-1. Primary Nitrogen Loading Factors used in the Apponagansett Bay MEP analyses. General factors are from MEP modeling evaluation (Howes & Ramsey 2001). Site-specific factors are derived from watershed-specific data...... 52 Table IV-2. Apponagansett Bay Watershed Nitrogen Loads. Attenuation of system nitrogen loads occurs within Buttonwood Brook and Apponagansett Brook during transport to the estuary and attenuation rates are assigned based on MEP measured loads. All values are kg N yr-1...... 54 Table IV-3. Comparison of water flow and nitrogen discharges from streams (freshwater) discharging to the head of Apponagansett Bay. The “Stream” data is from the MEP stream gauging effort. Watershed data is based upon historic watershed delineations developed by the USGS and confirmed by the MEP Technical Team...... 65 Figure IV-8. Apponagansett Brook discharge (solid blue line), nitrate+nitrite (yellow square) and total nitrogen (blue triangles) concentrations for determination of annual volumetric discharge and nitrogen load from the upper watershed to Apponagansett Bay (Table IV-3)...... 68 Table IV-4. Summary of annual volumetric discharge and nitrogen load from Buttonwood Brook and Apponagansett Brook (freshwater) discharging to the Apponagansett Bay system based upon the data presented in Figures IV-7-8 and Table 3...... 69 Table IV-5. Rates of net nitrogen return from sediments to the overlying waters of the Apponagansett Bay Estuary. These values are combined with the basin areas to determine total nitrogen mass in the water quality model (see Chapter VI). Measurements represent July -August rates (average of 2004 and 2005)...... 77 Table V-1. Tide datums computed from records collected in the Apponagansett system June 18 - July 23, 2003. Datum elevations are given in feet relative to NGVD 29...... 85 Table V-2. Tidal Constituents, Apponagansett Bay System June18 - July 23, 2003 ...... 86 Table V-3. M2 Tidal Attenuation, Apponagansett Bay, June 18 - July 23, 2003 ...... 87 Table V-4. Percentages of Tidal versus Non-Tidal Energy, Apponagansett Bay ...... 88 Table V-5. Manning’s Roughness coefficients used in simulations of modeled embayments...... 94 Table V-6. Turbulence exchange coefficients (D) used in simulations of modeled embayment system...... 95 Table V-7. Comparison of Tidal Constituents calibrated RMA2 model versus measured tidal data for the period June 21 to June 28, 2003...... 97 Table V-8. Comparison of Tidal Constituents calibrated RMA2 model versus measured tidal data for the model verification period July 8 to July 15, 2003...... 100 Table V-9. Embayment mean volumes and average tidal prism of the Apponagansett Bay system during simulation period...... 103 Table V-10. Computed System and Local residence times for sub-embayments of the Apponagansett Bay estuary system...... 103

viii

MASSACHUSETTS ESTUARIES PROJECT

Table VI-1. Water quality monitoring data and modeled Nitrogen concentrations for the Apponagansett Bay System used in the model calibration plots of Figure VI-2. All concentrations are given in mg/L N. “Data mean” values are calculated as the average of the separate yearly means...... 107 Table VI-2. Sub-embayment loads used for total nitrogen modeling of the Apponagansett Bay System, with total watershed N loads, atmospheric N loads, and benthic flux. These loads represent present loading conditions...... 109 Table VI-3. Values of longitudinal dispersion coefficient, E, used in calibrated RMA4 model runs of salinity and nitrogen concentration for Apponagansett Bay System...... 112 Table VI-4. Comparison of sub-embayment watershed loads used for modeling of present, build-out, and no-anthropogenic (“no-load”) loading scenarios of the Apponagansett Bay System. These loads do not include direct atmospheric deposition (onto the sub-embayment surface) or benthic flux loading terms...... 114 Table VI-5. Build-out sub-embayment and surface water loads used for total nitrogen modeling of the Apponagansett Bay System, with total watershed N loads, atmospheric N loads, and benthic flux...... 116 Table VI-6. Comparison of model average total N concentrations from present loading and the build-out scenario, with percent change, for the Apponagansett Bay System. Sentinel threshold station is in bold print...... 117 Table VI-7. “No anthropogenic loading” (“no load”) sub-embayment and surface water loads used for total nitrogen modeling of Apponagansett Bay System, with total watershed N loads, atmospheric N loads, and benthic flux ...... 118 Table VI-8. Comparison of model average total N concentrations from present loading and the no anthropogenic (“no load”) scenario, with percent change, for the Apponagansett Bay System. Loads are based on atmospheric deposition and a scaled N benthic flux (scaled from present conditions). Sentinel threshold station is in bold print...... 118 Table VII-1. Duration, in days that in situ sensors were operated and the time periods that bottom water oxygen levels were below various benchmark oxygen levels. Even short-term oxygen declines below 3 mg/l result in a high level of stress to benthic and fish communities. Data collected by the Coastal Systems Program, SMAST...... 129 Table VII-2. Duration (% of deployment time ) that chlorophyll a levels exceed various benchmark levels within the embayment system. “Mean” represents the average duration of each event over the benchmark level and “S.D.” its standard deviation. Data collected by the Coastal Systems Program, SMAST...... 130 Table VII-3. Changes in eelgrass coverage in Apponagansett Bay Embayment System within the Town of Dartmouth over the past half century (C. Costello). Note that the 1995 and 2001 areas do not represent significant eelgrass beds, but rather fringing zones of sparsely distributed surviving eelgrass...... 135 Table VII-4. Benthic infaunal community data for Apponagansett Bay embayment system. Estimates of the number of species adjusted to the number of individuals and diversity (H’) and Evenness (E) of the community allow comparison between locations (Samples represent surface area of 0.018 m2)...... 138 Table VIII-1. Summary of nutrient related habitat quality within the Apponagansett Bay Estuary within the Town of Dartmouth, MA. and based upon assessments

ix

MASSACHUSETTS ESTUARIES PROJECT

described in Section VII. WQMP indicates BayWatcher Water Quality Monitoring Program...... 150 Table VIII-2. Comparison of sub-embayment watershed septic loads (attenuated) used for modeling of build-out and threshold loading scenarios of the Apponagansett Bay System. These loads do not include direct atmospheric deposition (onto the sub-embayment surface), benthic flux, runoff, or fertilizer loading terms...... 154 Table VIII-3. Comparison of sub-embayment total attenuated watershed loads (including septic, runoff, and fertilizer) used for modeling of build-out and threshold loading scenarios of the Apponagansett Bay system. These loads do not include direct atmospheric deposition (onto the sub- embayment surface) or benthic flux loading terms...... 154 Table VIII-4. Threshold sub-embayment loads and attenuated surface water loads used for total nitrogen modeling of the Apponagansett Bay system, with total watershed N loads, atmospheric N loads, and benthic flux ...... 155 Table VIII-5. Comparison of model average total N concentrations from present loading and the modeled threshold scenario, with percent change, for the Apponagansett Bay system. Sentinel threshold station is in bold print...... 155 Table IX-1. Apponagansett Bay Watershed Nitrogen Loads incorporating current (2014) Dartmouth Sewer Connections. Sewer connections are based on parcels currently receiving sewer bills. Slight land use changes are incorporated based on reconciling the sewer billing database and parcel classifications. Attenuation of system nitrogen loads within Buttonwood Brook and Apponagansett Brook are based on measurements discussed in Chapter 4. All values are kg N yr-1...... 158 Table IX-2. Comparison of sub-embayment watershed septic loads (attenuated) used for modeling loading conditions for 2014 Scenario. These loads do not include direct atmospheric deposition (onto the sub-embayment surface), benthic flux, runoff, or fertilizer loading terms...... 158 Table IX-3. Comparison of sub-embayment total attenuated watershed loads (including septic, runoff, and fertilizer) used for modeling of conditions for 2014 Scenario. These loads do not include direct atmospheric deposition (onto the sub-embayment surface) or benthic flux loading terms...... 159 Table IX-4. Sub-embayment loads used for total nitrogen modeling of the Apponagansett Bay system for present loading scenario with loading conditions for 2014 Scenario, with total watershed N loads, atmospheric N loads, and benthic flux...... 159 Table IX-5. Comparison of model average total N concentrations from 2014 scenario, with percent change, for the Apponagansett Bay System. The threshold station is shown in bold print...... 159

x

MASSACHUSETTS ESTUARIES PROJECT

I. INTRODUCTION

The Apponagansett Bay embayment is located within the Town of Dartmouth Massachusetts. The system has a southern shore bounded by waters of Buzzards Bay. About 90 percent of the watershed of Apponagansett Bay is located within the Town of Dartmouth, with about ten percent of the watershed located within the City of New Bedford.

The configuration of the Apponagansett Bay embayment results from the drowning by rising sea level of a shallow stream valley formed by long-term erosion of the underlying bedrock fabric that was more recently modified by repeated continental glaciations ending about 16,000 years before the present. The construction of the Padanaram causeway and bridge in 1830 brought a physical change to the bay by dividing the bay in two, separating Inner Apponagansett Bay from outer Apponagansett Bay (Glennon 2001). The causeway, which is a stone embankment built upon a pre-existing, but shorter sand bar, has limited tidal flow in and out of the inner bay to the width of the channel between the bridge abutments. The mouth of the Apponagansett Bay embayment is to the southeast and is open to Buzzards Bay so that the outer bay communicates freely with Buzzards Bay. The two principal streams of the Apponagansett Bay watershed discharge along the northeastern margin of the bay. Dike Marsh, a large tidal marsh, is located at the southwest corner of the inner bay and communicates with the inner bay through a channel that is restricted by the abutments of the bridge carrying Gulf Road. Dike Marsh has also gone through a series of changes from human manipulation and it is likely that it was originally a tidal salt marsh. At some time in the colonial period it was impounded by a stone spillway located about 750 m southwest of the Gulf Road bridge and most of the marsh became a fresh water wetland. In the first half of the 20th century the dike was breached and the marsh is now a tidal salt marsh.

Like a few other shallow coastal embayments in the region, inner Apponagansett Bay is a mesotrophic (moderately nutrient impacted) estuary. The watershed boundaries are defined primarily by a bedrock morphology of low ridges and a valley running roughly in a northward and northeastward direction (Zen, 1983). The soils and sediments of the watershed are of recent glacial derivation and consist of a variable fabric of low permeability basal till and ablation till mantling the bedrock ridges, ridge slopes, and valleys. The sediment in the shallow valleys consist of more permeable glacial deposits: outwash, glacial lake sediments and most recent alluvial and lake bottom deposits that include sand, silt, clay and peat (Williams and Tasker, 1978; Larson, 1982). Due to the generally low permeability of the uplands, the Apponagansett Bay watershed hydrology is characterized as a surface water-dominated system with a well- developed stream network, though groundwater flow is important in providing baseflow to much of the stream network (Bent, 1995). The Inner Apponagansett Bay portion of the overall estuary is a mixing zone for terrestrial fresh water and saline tidal flow from Buzzards Bay (Figure I-1). With a moderate-sized watershed of 20.7 km2 (8 mi2; ratio of land area to water area ratio is about 15 to 1) the inner bay subsequently has a moderate level of fresh surface water inputs, such that the inner Apponagansett Bay estuary is a mixing zone that has a salinity gradient that is influenced primarily by large rainfall events.

Apponagansett Bay and its watershed constitute an important natural and cultural component of Dartmouth and the City of New Bedford. Buttonwood Brook flows from the northern-most limit of the watershed to empty into inner Apponagansett Bay. Though degraded in many parts of its northern drainage due to the degree of urbanization, Buttonwood Brook provides a riparian corridor along the western side of New Bedford and then southward through

1 MASSACHUSETTS ESTUARIES PROJECT

Figure I-1. Study region proximal to the Apponagansett embayment system for the Massachusetts Estuaries Project nutrient analysis. Tidal waters enter the outer bay in the lower right of the frame from Buzzards Bay. Freshwaters enter primarily from two surface water streams at the northern head of the inner bay.

2 MASSACHUSETTS ESTUARIES PROJECT the Town of Dartmouth. The inner bay watershed to the east and north of Apponagansett Bay is relatively urbanized, with the highest intensity of use in the Route 6 commercial corridor in the north, a high percentage of residential use in the northern two-thirds of the watershed and in Padanaram village fringing the length of the eastern shore. The smaller remaining watershed area to the west and south of the bay is much less intensively developed. Both inner and outer Apponagansett Bay have high levels of recreational and commercial marine use for swimming, recreational boating, and commercial fishing and shellfishing, while substantial shoreline areas in Padanaram village are devoted to marine-related activities. Both the inner and outer bay waters are sources of substantial catches of fish and shellfish.

As an estuary in a populous region Apponagansett Bay is the meeting place for two opposing elements: as a protected marine bay it is the focus of boating, recreation and land- development; as a relatively enclosed body of water, the inner bay may not be readily flushed of the pollutants that it receives from the watershed. With 15 acres of land area contributing nutrients to each acre of estuary, the inner bay is at risk of eutrophication from nitrogen from stream and storm water runoff from the watershed. With substantial nitrogen inputs, one major ecological threat to the inner bay is degradation resulting from nutrient over-enrichment. Loading of the critical nutrient (nitrogen) to the embayment has increased over the last five decades and with continued development further increases are likely unless nitrogen management is implemented. The nitrogen loading to Apponagansett Bay derives from three principal sources: onsite disposal of wastewater in areas not served by the municipal wastewater systems of Dartmouth and New Bedford; storm water runoff containing nitrogen; residential and agricultural fertilizer use; and atmospheric deposition of nitrogen compounds on the land and water surface.

In the late 1960’s the rate and intensity of development in Dartmouth prompted state and federal regulatory agencies to require the construction of a municipal wastewater system to treat wastewater from Padanaram village and other areas of intensive land use in the town. Secondary-treated effluent from the municipal treatment plant is discharged to Buzzards Bay offshore of Smith Neck, thus exporting much wastewater-based nitrogen outside Apponagansett Bay. Despite the likely positive effects of sewers upon the inner bay water quality, inner Apponagansett Bay has exhibited poor water quality symptoms for the last couple of decades. Regular measurement of a decline in ecological health of Apponagansett Bay began in 1993 with the start of the Baywatchers Program in Dartmouth and other Buzzards Bay communities’ coastal waters. The monitoring program was initially conducted through a collaboration between the Coalition for Buzzards Bay and the Coastal Systems Program within the UMASS-D School for Marine Science and Technology. Baywatchers data from 1993 to 2001 has shown inner Apponagansett Bay to be the sixth poorest estuary segment out of more than 65 embayment segments, ranking slightly better than the Acushnet River/ New Bedford Harbor (Coalition for Buzzards Bay, 2003). Outer Apponagansett Bay water health ranks in the middle of all Buzzard Bay waters due to its improved flushing with cleaner Buzzards Bay waters (Costa and Howes, 1996). The documented poor inner and outer bay water quality is consistent with land-use analyses performed by the Buzzards Bay Project National Estuary Program beginning in 1991, which suggested that these embayments are receiving nitrogen pollution inputs one and one-half to two times higher than the threshold at which habitat decline is expected to begin (BBP CCMP, 1991; Costa et al 1999). The measured water quality data, the disappearance of the remaining inner bay eelgrass beds and reduced shellfish populations all underscore the level of impaired habitat quality within inner Apponagansett Bay.

The consistently poor record of water quality in the inner Apponagansett Bay and the other Dartmouth estuaries mobilized watershed residents, Town officials and local advocates to

3 MASSACHUSETTS ESTUARIES PROJECT organize to restore the health of Apponagansett Bay, the Slocums River, and Little River. In 2002, that effort became Turn the Tide: Restore Dartmouth’s Estuaries, a partnership among the Town of Dartmouth, the Coastal Systems Program within the University of Massachusetts School for Marine Science and Technology (SMAST) as the Scientific Lead in Massachusetts Estuaries Project (MEP), the Coalition for Buzzards Bay and the Lloyd Center for the Environment.

As noted above, the Baywatchers monitoring has documented that Apponagansett Bay along with the other Dartmouth embayments, is currently showing water quality declines. Based on the information obtained over the past 13 years of Baywatchers monitoring of these coastal embayments, Apponagansett Bay was included in the Massachusetts Estuaries Project to receive state-of-the-art analysis and modeling.

The MEP linked watershed-embayment modeling approach has been utilized in each of Dartmouth's estuaries to develop nitrogen loads, establish circulation and flushing characteristics and create linkages to specific ecological criteria that together form the basis for the nitrogen threshold and subsequent Total Maximum Daily Load (TMDL) for nitrogen to be developed by the MassDEP. The nitrogen threshold developed by the MEP is ultimately necessary as a guide to develop nitrogen management alternatives needed by the Town of Dartmouth to reverse nutrient impairment in the estuaries of the Town. The modeling tools developed as part of the MEP provide the quantitative information necessary for the Town Dartmouth to develop and evaluate the most cost effective nitrogen management alternatives to restore these valuable coastal resources which are currently being degraded by nitrogen overloading.

I.1 THE MASSACHUSETTS ESTUARIES PROJECT APPROACH Coastal embayments throughout the Commonwealth of Massachusetts (and along the U.S. eastern seaboard) are becoming nutrient enriched. Nutrients are primarily related to changes in watershed land-use associated with increasing population within the coastal zone over the past half century. Many of Massachusetts’ embayments have nutrient levels that are approaching or are currently over this assimilative capacity, which begins to cause declining ecological health. The result is the loss of fisheries habitat, eelgrass beds, and a general disruption of benthic communities. At its higher levels, enhanced loading from surrounding watersheds causes aesthetic degradation and inhibits even recreational uses of coastal waters. In addition to nutrient related ecological declines, an increasing number of embayments are being closed to swimming, shellfishing and other activities as a result of bacterial contamination. While bacterial contamination does not generally degrade the habitat, it restricts human use. Similar to nutrients, bacterial contamination is related to changes in land-use as watersheds become more developed. Regional effects of both nutrient loading and bacterial contamination span the spectrum from environmental to socio-economic impacts and have direct consequences to culture, economy, and tax base of Massachusetts’s coastal communities.

The primary nutrient causing the increasing impairment of the Commonwealth’s coastal embayments is nitrogen and the primary sources of this nitrogen are wastewater disposal, fertilizers, and changes in the freshwater hydrology associated with development. At present there is a critical need for state-of-the-art approaches for evaluating and restoring nitrogen sensitive and impaired embayments. Within Southeastern Massachusetts alone, almost all of the municipalities (as is the case with the Town of Dartmouth) are grappling with Comprehensive Wastewater Planning and/or environmental management issues related to the declining health of their estuaries.

4 MASSACHUSETTS ESTUARIES PROJECT

Municipalities are seeking guidance on the assessment of nitrogen sensitive embayments, as well as available options for meeting nitrogen goals and approaches for restoring impaired systems. Many of the communities have encountered problems with “first generation” watershed based approaches, which do not incorporate estuarine processes. The appropriate method must be quantitative and directly link watershed and embayment nitrogen conditions. This “Linked” Modeling approach must also be readily calibrated, validated, and implemented to support planning. Although it may be technically complex to implement, results must be understandable to the regulatory community, town officials, and the general public.

The Massachusetts Estuaries Project represents the newest generation of watershed based nitrogen management approaches. The Massachusetts Department of Environmental Protection (MA DEP), the University of Massachusetts – Dartmouth School of Marine Science and Technology (SMAST), and others including the Cape Cod Commission (CCC) have undertaken the task of providing a quantitative tool for watershed-embayment management for communities throughout Southeastern Massachusetts.

The Massachusetts Estuary Project is founded upon science-based management. The Project is using a consistent, state-of-the-art approach throughout the region’s coastal waters and providing technical expertise and guidance to the municipalities and regulatory agencies tasked with their management, protection, and restoration. The overall goal of the Massachusetts Estuaries Project is to provide the DEP with technical guidance to support policies on nitrogen loading to embayments. In addition, the technical reports prepared for each embayment system will serve as the basis for the development of Total Maximum Daily Loads (TMDLs). Development of TMDLs is required pursuant to Section 303(d) of the Federal Clean Water Act. TMDLs must identify sources of the pollutant of concern (in this case nitrogen) from both point and non-point sources, the allowable load to meet the state water quality standards and then allocate that load to all sources taking into consideration a margin of safety, seasonal variations, and several other factors. In addition, each TMDL must contain an implementation plan. That plan must identify, among other things, the required activities to achieve the allowable load to meet the allowable loading target, the time line for those activities to take place, and reasonable assurances that the actions will be taken.

In appropriate estuaries, TMDLs for bacterial contamination will also be conducted in concert with the nutrient effort (particularly if there is a 303d listing). However, the goal of the bacterial program is to provide information to guide targeted sampling for specific source identification and remediation. As part of the overall effort, the evaluation and modeling approach will be used to assess available options for meeting selected nitrogen goals that are protective of embayment health.

The major Project goals are to:

 develop a coastal TMDL working group for coordination and rapid transfer of results,  determine the nutrient sensitivity of each of the 89 embayments in Southeastern MA  provide necessary data collection and analysis required for quantitative modeling,  conduct quantitative TMDL analysis, outreach, and planning,  keep each embayment’s model “alive” to address future regulatory needs.

The core of the Massachusetts Estuaries Project analytical method is the Linked Watershed-Embayment Management Modeling Approach. This approach represents the “next generation” of nitrogen management strategies. It fully links watershed inputs with embayment

5 MASSACHUSETTS ESTUARIES PROJECT circulation and nitrogen characteristics. The Linked Model builds on and refines well accepted basic watershed nitrogen loading approaches such as those used in the Buzzards Bay Project, the CCC models, and other relevant models. However, the Linked Model differs from other nitrogen management models in that it:

 requires site specific measurements within each watershed and embayment;  uses realistic “best-estimates” of nitrogen loads from each land-use (as opposed to loads with built-in “safety factors” like Title 5 design loads);  spatially distributes the watershed nitrogen loading to the embayment;  accounts for nitrogen attenuation during transport to the embayment;  includes a 2D or 3D embayment circulation model depending on embayment structure;  accounts for basin structure, tidal variations, and dispersion within the embayment;  includes nitrogen regenerated within the embayment;  is validated by both independent hydrodynamic, nitrogen concentration, and ecological data;  is calibrated and validated with field data prior to generation of “what if” scenarios.

The Linked Model has been applied for watershed nitrogen management in ca. 15???? embayments throughout Southeastern Massachusetts. In these applications it has become clear that the Linked Model Approach’s greatest assets are its ability to be clearly calibrated and validated, and its utility as a management tool for testing “what if” scenarios for evaluating watershed nitrogen management options.

The Linked Watershed-Embayment Model when properly parameterized, calibrated and validated for a given embayment becomes a nitrogen management planning tool, which fully supports TMDL analysis. The Model suggests “solutions” for the protection or restoration of nutrient related water quality and allows testing of “what if” management scenarios to support evaluation of resulting water quality impact versus cost (i.e., “biggest ecological bang for the buck”). In addition, once a model is fully functional it can be “kept alive” and corrected for continuing changes in land-use or embayment characteristics (at minimal cost). In addition, since the Model uses a holistic approach (the entire watershed, embayment and tidal source waters), it can be used to evaluate all projects as they relate directly or indirectly to water quality conditions within its geographic boundaries.

Linked Watershed-Embayment Model Overview: The Model provides a quantitative approach for determining an embayment’s: (1) nitrogen sensitivity, (2) nitrogen threshold loading levels (TMDL) and (3) response to changes in loading rate. The approach is fully field validated and unlike many approaches, accounts for nutrient sources, attenuation, and recycling and variations in tidal hydrodynamics (Figure I-2). This methodology integrates a variety of field data and models, specifically:

 Monitoring - multi-year embayment nutrient sampling  Hydrodynamics - - embayment bathymetry - site specific tidal record - current records (in complex systems only) - hydrodynamic model  Watershed Nitrogen Loading - watershed delineation - stream flow (Q) and nitrogen load - land-use analysis (GIS)

6 MASSACHUSETTS ESTUARIES PROJECT

- watershed N model  Embayment TMDL - Synthesis - linked Watershed-Embayment N Model - salinity surveys (for linked model validation) - rate of N recycling within embayment - D.O record - Macrophyte survey - Infaunal survey

Nitrogen Thresholds Analysis

Tide, Benthic Flux and Water Column Watershed Bathymetry, and Measurements Delineation & N Load Current Section IV Section III and IV Measurements

Hydrodynamic Total Nitrogen Modeling Modeling Section V Section VI

D.O., Eelgrass Infauna Surveys Section VII

Thresholds Development Section IX

Figure I-2. Massachusetts Estuaries Project Critical Nutrient Threshold Analytical Approach. Section numbers refer to sections in this MEP report where the specified information is provided.

I.2 SITE DESCRIPTION Apponagansett Bay is oriented roughly north to south with a sub-embayment to the southwest that is a salt marsh and tidal creek system, Dike Marsh. The inner main embayment is open to Buzzards Bay through the swing bridge opening on the Padanaram causeway and

7 MASSACHUSETTS ESTUARIES PROJECT the outer embayment communicates directly to Buzzards Bay on the southern end of the estuary. The configuration of the embayment results from the drowning of a shallow valley due to post-glacial sea-level rise. The shallow river valley formed by both long-term and short-term processes. Long-term processes include erosion of the bedrock by chemical and physical weathering of the bedrock over the past 30 million years or more, which maintained a drainage network in the region. The bedrock underlying the soils of the region is primarily crystalline bedrock of about 600 million years ago (Zen, 1983). The bedrock types include granite, gneiss, schist and other rocks (Murray 1990). Over the past 600,000 years a series of repeated continental glaciations have advanced and retreated across the region, with the most recent ending about 16,000 years ago. The glaciations lowered the local bedrock erosional surfaces by small amounts, leaving smoothed upland bedrock surfaces and a variety of glacial sediments covering much of the underlying bedrock structure. Bedrock outcrops occur throughout the watershed and the exposed and buried bedrock topography is the primary control upon the local topography.

The Apponagansett Bay watershed boundary is defined on the west and northwest by a series of northward trending, low bedrock ridges, beginning along Bakerville Road and then continuing to the north along Slocums Road (Figure I-3). To the north and northeast, a series of low bedrock saddles and higher ridges separate the watershed from that of adjacent Slocums River drainage. On the east side, a series of low rises with a northeastward trend define the eastern boundary with the watersheds of New Bedford Harbor, Clarks Cove and Buzzards Bay. The embayment watershed is about 10.2 km (6.36 mi) long from south to north and about 3.0 km (1.9 mi) wide from east to west. The Apponagansett Bay embayment is about 3.5 km (2.2 mi) long from the breakwater in the south to the northern terminus of the embayment. The outer embayment is about 1.5 km (1.25 mi) from the breakwater to the Padanaram causeway which divides the inner from the outer embayment. Outer and inner embayment widths are 0.86km (0.53 mi) and 1.15 km (0.72 mi), respectively. Dike Marsh is about 2.1 km (1.31 mi) from the Gulf Road bridge to the marsh southern margin. Approximately two-thirds of the watershed lies north of the Padanaram causeway.

The soils and sediments of the Apponagansett Bay watershed differ by both elevation and location. Northward from the north end of the estuary at Russells Mills Road the watershed sediments are dominated by ablation and basal tills (Williams and Tasker, 1978). The two principal streams of the watershed, Buttonwood Brook and Apponagansett Bay Brook flow from the north, discharging to the uppermost end of the Bay (Figure I-3). The floors of the two adjacent stream valleys also contain substantial amounts of ablation and basal till. Glacial fluvial, lacustrine and ice contact sediment deposits which typically have much higher permeabilities than tills, appear to be a minor factor in the hydrologic character of the two streams, comprising less than 1% of the surficial sediments (Bent, 1995; Williams and Tasker, 1978). To the east of the estuary, across Padanaram village and extending to the eastern watershed boundary, ablation and basal till also predominate (Williams and Tasker, 1978). More permeable glacial fluvial sediments fringe the western estuary salt marshes and extend to the western watershed boundary at the junction of Russells Mills and Tucker Roads (Williams and Tasker, 1978). In the southwestern portion of the watershed, between Gulf Road and Rock o’ Dundee Road, glacial fluvial sand and gravel deposits are located within the lowland that also contains Dike Marsh. The lowland continues southward to form the basin of the watershed of Little River. Northward-trending bedrock ridges, along Bakerville Road and on the east, along Smith Neck Road define the Apponagansett Bay watershed in this area. In this southern watershed area, these ridges and upland slopes are also covered with both ablation and basal tills (Williams and Tasker, 1978).

8 MASSACHUSETTS ESTUARIES PROJECT

Figure I-3. Apponagansett Bay with the embayment watershed and sub-watersheds (red line) and principal streams.

The regional bedrock has a very low permeability, while the basal tills (locally known as hardpan) and ablation tills have low to moderately low permeability. The combination of exposed and shallowly buried bedrock and generally low permeability glacial tills strongly affects

9 MASSACHUSETTS ESTUARIES PROJECT the hydrology of the watershed. In the northern and eastern parts of the Apponagansett Bay watershed, the buried bedrock and low permeability upland soils form a surface-water dominated regime where rainfall and snow melt tend to flow over the ground surface in a greater proportion than they percolate into the soils to become groundwater (Bent 1995). The stream network is therefore well-developed in the northern two-thirds of the bay’s watershed and most upland groundwater flow is localized, towards the nearest stream. Of the total amount of freshwater falling across the northern two-thirds of the watershed and entering the estuary, the stream network transports a high proportion into the embayment while groundwater transports a relatively small amount to the shore. Groundwater underflow beneath the major stream, Buttonwood Brook is very limited by the bedrock surrounding the streambed near the mouth of the brook. Very low flows in Buttonwood Brook during warm months also tend to confirm the low percentage of permeable sediments in the upper basin and the consequent small role that groundwater plays in the hydrologic character of the upper watershed. Other streams in the watershed are short due to the limited distances from the estuary shore to the watershed boundary. In the eastern portion of the watershed two small streams that drained the Padanaram village area have been severely altered: one stream discharging to the Bay near Fort Street has been partially buried; and the other stream drained areas in the center of the village and is now completely incorporated into the storm drain network. As such, these streams did not represent significant surface water features and therefore were not monitored by the MEP stream gauging program.

In the area contributing to the Dike Marsh sub-watershed to the south, the upland areas are similar to those in the north with low bedrock ridges mantled with till, with lowlands containing fresh and saltwater wetland peat that are both above and alongside glacial-fluvial deposits and ice contact deposits with relatively high permeability (Williams and Tasker, 1978).

Because of the layout of the Apponagansett Bay embayment, it is a mixing zone for watershed freshwater and marine waters. At the northern embayment terminus there is a variable salinity regime that is dependent upon the amount of freshwater discharging from Buttonwood Brook and Apponagansett Bay. Both brooks display a “flashy” discharge character: during and immediately following rainfall events, the two streams levels rise and fall rapidly and stream flow between events (baseflow) also decreases relatively rapidly, so that the amounts discharged to the embayment come in pulses with low flow rates in between. The periodic nature of high flows for the two streams means that embayment salinities in the northernmost area are highly variable from brackish (<10 ppt) during high stream discharge to approaching Buzzards Bay salinities (20-30 ppt) between rain events. Despite this, the embayment as a whole could be characterized as tidally dominated for much of the year due to the relatively rapid flushing with Buzzards Bay waters. Southward in the inner embayment salinities are relatively high, about 25-30 ppt. The outer embayment has salinities that are similar to Buzzards Bay levels (30 ppt) due to the higher flushing rate for this part of the embayment. The Dike Marsh sub-embayment has a limited upland watershed surrounding it due to the fact that much of the watershed is salt marsh and water. It is also tidally dominated with salinities between 20 and 30 ppt, varying seasonally.

Apponagansett Bay is a relatively recent (<16,000 years) ecological system due to the effects of recent glacial erosion “planing” off older soils and plant cover. After the wasting of the last glacier about 16,000 years ago, the present day watershed was occupied by a freshwater stream system that continued southward into Buzzards Bay. Drainage of the local waters of the Buzzards Bay basin continued southwestward to the edge of the continental shelf where the shoreline of the lowered Atlantic Ocean was at that time (O’Hara and Oldale, 1980). As sea level rose rapidly after deglaciation, the shoreline migrated shoreward and entered Buzzards

10 MASSACHUSETTS ESTUARIES PROJECT

Bay. As sea level continued to rise more slowly over the past 6,000 years shoreline processes of the advancing sea altered the glacial sediments of the area just south of the present day mouth of the embayment eroding the till from along Smith Neck and from Ricketson’s Point. Beginning about 2,500 years ago, shoreline process began to modify the outer embayment leaving a boulder-strewn shoreline along the outer embayment and a relatively unmodified inner embayment. Due to the predominance of till and bedrock along the shores, the configuration of the outer embayment shoreline and the slope of the outer embayment bottom, there was a limited supply of sand within the system to be available for building shoals and barriers into the embayment. What sand that has been mobile in the system has been collecting as a short spit near the opening to inner Apponagansett Bay which was exploited in 1830 for the construction the Padanaram causeway and bridge (Glennon 2001).

Anthropogenic influences of the past couple of centuries has significantly modified the morphology and thus the ecology of the Apponagansett Bay embayment system. The Padanaram causeway and bridge created a barrier that now gives structural definition to the inner and outer embayment. The causeway was constructed between 1827 and 1830 and effectively reduced the width of the embayment from about 518m (1700ft) to about 125m (410ft) or a reduction of about 76% of the former width at that site. The outer embayment jetty was constructed in the 19th century and reduced the outer harbor communication with Buzzards Bay from 862 m (2831 ft) to about 610 m (2002 ft), a reduction of about 29%. In the Dike Marsh sub- embayment the approaches and bridge abutments for the Gulf Road bridge reduced the communication between the marsh and the inner embayment. The effects of this restriction at this location today can be observed an hour or two after the ebb begins during spring tide events.

Given the rate of post-glacial sea level rise, it is likely that Apponagansett Bay had estuarine conditions for at least the past 4,500 years and to have been always been a tidally dominated estuary. The tidal forcing for Apponagansett Bay is from Buzzards Bay, which has mean tide range of nearly 1.8 m.

Apponagansett Bay is a shallow mesotrophic (moderately nutrient impacted) to eutrophic (nutrient-rich) embayment in Dartmouth. As noted above, Apponagansett Bay is an estuary with significant fresh surface water inputs at the estuary northern headwaters (Buttonwood Brook and Apponagansett Bay Brook) and tidal exchange of marine waters from Buzzards Bay (tide range of approximately 1.8 m) at its southern mouth. For the MEP analysis, Apponagansett Bay was divided into inner and outer Apponagansett Bay and Dike Marsh systems (Figure I-4). Inner Apponagansett Bay receives discharge from the two major streams and is of locally variable salinity; Dike Marsh receives relatively small amounts of freshwater from its small upland watershed and is dominated by tidal exchange with inner Apponagansett Bay. Outer Apponagansett Bay reflects the mixing of terrestrial nutrient inputs from the inner bay and the rapid flushing of Buzzards Bay waters.

Given the present hydrodynamic characteristics of the Apponagansett embayment system, it appears that estuarine habitat quality is dependent on both the level of nutrient loading to embayment waters and the tidal characteristics. In inner and outer Apponagansett Bay, some enhancements to tidal flushing may be achieved via modification of the bridge causeway resulting in some mediation of the nutrient loading impacts from the Apponagansett Bay watershed. The details of such are a part of the MEP analysis (Section V) described in this report.

11 MASSACHUSETTS ESTUARIES PROJECT

Buttonwood Brook

Apponagansett Bay Brook

Padanaram Village

Inner Apponagansett Bay

Causeway and bridge

Outer Apponagansett Bay Dike Marsh

Buzzards Bay

Figure I-4. Apponagansett Bay subdivisions for water quality monitoring.

The watershed of Apponagansett Bay is located in Dartmouth and the City of New Bedford. About 89% is within the Town of Dartmouth and about 11% of the watershed lies in the City of New Bedford and located in the northern and northeastern part of the watershed. Based upon land-use and much of the watershed being predominantly within Dartmouth, it appears that substantial nitrogen management for Apponagansett Bay restoration may formulated and implemented through Town of Dartmouth actions. Cooperation with the City of New Bedford on

12 MASSACHUSETTS ESTUARIES PROJECT planning and management will still be critical to the long-term success of a restoration plan due to the nature of the intensity of development in the watershed: substantial amounts of the commercial and high density residential development lie within New Bedford. As management alternatives are being developed and evaluated, it is important to note that relatively strong gradients define the nutrient characteristics the Apponagansett Bay and these also control the associated habitat impacts. There is a clear nutrient gradient between inner and outer Apponagansett Bay with highest nitrogen and lowest environmental health in the innermost portions of the embayment and particularly in the northern pat of the inner system and lowest nitrogen and greatest health in the outer embayment near Buzzards Bay. The inner embayment of Apponagansett Bay is presently showing poor water quality and “Eutrophic” conditions. Large areas of macroalgae cover the bottom in the middle portion and northern parts of the inner bay, eelgrass is absent now but was present within the past 50 years (not field verified). Outer Apponagansett Bay is demonstrating better water quality and maintains eelgrass beds.

Surface and groundwater flows are pathways for the transfer of land-sourced nutrients to coastal waters. Fluxes of the primary ecosystem structuring nutrients, nitrogen and phosphorus, differ as a result of their hydrologic transport pathway (i.e. streams versus groundwater). In a glacial till watershed with small amounts of mixed stratified drift such as in the watershed to Apponagansett Bay embayment watershed system, phosphorus is generally strongly retained Nitrogen, primarily as plant available nitrate, is also readily transported through the watershed both during surface flow and through groundwater systems. Coastal estuaries tend to have algal growth limited by nitrogen availability, rather than by the availability of phosphorus due to the relatively large amounts of phosphorous in coastal system waters and due to the embayment’s flooding with low nitrogen coastal waters (Ryther and Dunstan 1971). Tidal reaches within Apponagansett Bay follow this general pattern, where the primary nutrient of eutrophication in these systems is nitrogen. However due to the nature of the hydrology and the soils in the watershed, it is possible that phosphorous is locally important to the growth of algae in the inner Apponagansett Bay embayment (Ullman, 2001). The roles of both nitrogen and phosphorous were considered during the MEP analysis of Apponagansett Bay.

Nutrient related water quality decline represents one of the most serious threats to the ecological health of the nearshore coastal waters. Coastal embayments, because of their enclosed basins, shallow waters and large shoreline area, are generally the first indicators of nutrient pollution from terrestrial sources. By nature, these systems are highly productive environments, but nutrient over-enrichment of these systems worldwide is resulting in the loss of their aesthetic, economic and commercially valuable attributes.

Each embayment system maintains a capacity to assimilate watershed nitrogen inputs without degradation. However, as loading increases a point is reached at which the capacity (termed assimilative capacity) is exceeded and nutrient related water quality degradation occurs. As nearshore coastal salt ponds and embayments are the primary recipients of nutrients carried via surface and groundwater transport from terrestrial sources, it is clear that activities within the watershed, often miles from the water body itself, can have chronic and long lasting impacts on these fragile coastal environments.

Protection and restoration of coastal embayments from nitrogen overloading has resulted in a focus on determining the assimilative capacity of these aquatic systems for nitrogen. While this effort is ongoing (e.g. USEPA TMDL studies), southeastern Massachusetts has been the site of intensive efforts in this area (Eichner et al., 1998, Costa et al., 1992 and in press, Ramsey et al., 1995, Howes and Taylor, 1990). While each approach may be different, they all focus on changes in nitrogen loading from watershed to embayment and aim at projecting the

13 MASSACHUSETTS ESTUARIES PROJECT level of increase in nitrogen concentration within the receiving waters. Each approach depends upon estimates of circulation within the embayment; however, few directly link the watershed and hydrodynamic models, and virtually none include internal recycling of nitrogen (as was done in the present effort). However, determination of the “allowable N concentration increase” or “threshold nitrogen concentration” used in previous studies had a significant uncertainty due to the need for direct linkage of watershed and embayment models and site-specific data. In the present effort we have integrated site-specific data on nitrogen levels and the gradient in N concentration throughout the Apponagansett Bay system monitored by the Coalition for Buzzards Bay and CSP/SMAST staff with site-specific habitat quality data (D.O., eelgrass, phytoplankton blooms, benthic animals; macroalgae; and finfish) to establish an embayment specific nitrogen threshold that would be restorative of healthy estuarine habitat.

I.3 WATER QUALITY MODELING Evaluation of upland nitrogen loading provides important “boundary conditions” for water quality modeling of the Apponagansett Bay system; however, a thorough understanding of estuarine circulation is required to accurately determine nitrogen concentrations within the system. Therefore, water quality modeling of tidally influenced estuaries must include a thorough evaluation of the hydrodynamics of the estuarine system. Estuarine hydrodynamics control a variety of coastal processes including tidal flushing, pollutant dispersion, tidal currents, sedimentation, erosion, and water levels. Numerical models provide a cost-effective method for evaluating tidal hydrodynamics since they require limited data collection and may be utilized to numerically assess a range of management alternatives. Once the hydrodynamics of an estuary system are understood, computations regarding the related coastal processes become relatively straightforward extensions to the hydrodynamic modeling. The spread of pollutants may be analyzed from tidal current information developed by the numerical models.

The MEP water quality evaluation examined the potential impacts of nitrogen loading into Apponagansett Bay. A two-dimensional depth-averaged hydrodynamic model based upon the tidal currents and water elevations was employed for the system. Once the hydrodynamic properties of the estuarine system were computed, two-dimensional water quality model simulations were used to predict the dispersion of the nitrogen at current loading rates.

Using standard dispersion relationships for estuarine systems of this type, the water quality model and the hydrodynamic models were then integrated in order to generate estimates regarding the spread of total nitrogen from the site-specific hydrodynamic properties. The distributions of nitrogen loads from watershed sources were determined from land-use analysis, based upon watershed delineations by SMAST and Lloyd Center staff for watershed and sub- watershed areas designated by MEP. Almost all nitrogen entering the Apponagansett Bay embayment is transported by surface water (streams). Concentrations of total nitrogen and salinity of Buzzards Bay source waters and throughout the Apponagansett Bay system were measured over two summers during 2004-2005. Measurements of current, salinity and nitrogen and salinity distributions throughout estuarine waters of the system were used to calibrate and validate the water quality model (under existing loading conditions).

I.4 REPORT DESCRIPTION This report presents the results generated from the implementation of the Massachusetts Estuaries Project linked watershed-embayment approach to the Apponagansett Bay system for the Town of Dartmouth. A review of existing water quality studies is provided (Section II). The development of the watershed delineations and associated detailed land use analysis for watershed based nitrogen loading to the coastal system is described in Sections III

14 MASSACHUSETTS ESTUARIES PROJECT and IV. In addition, nitrogen input parameters to the water quality model are described. Since benthic flux of nitrogen from bottom sediments is a critical (but often overlooked) component of nitrogen loading to shallow estuarine systems, determination of the site-specific magnitude of this component also was performed (Section IV). Nitrogen loads from the watershed and sub- watershed surrounding the estuary were derived from Southeastern Regional Planning and Economic Development District (SRPEDD) data and offshore water column nitrogen values were derived from an analysis of a monitoring station in Buzzards Bay (Section IV). Intrinsic to the calibration and validation of the linked-watershed embayment modeling approach is the collection of background water quality monitoring data (conducted by municipalities) as discussed in Section IV. Results of hydrodynamic modeling of embayment circulation are discussed in Section V and nitrogen (water quality) modeling, as well as an analysis of how the measured nitrogen levels correlate to observed estuarine water quality are described in Section VI. This analysis includes modeling of current conditions, conditions at watershed build-out, and with removal of anthropogenic nitrogen sources. In addition, an ecological assessment of each embayment was performed that included a review of existing water quality information, temporal changes in eelgrass distribution, dissolved oxygen records and the results of a benthic infaunal animal analysis and other bioassays (Section VII). The modeling and assessment information is synthesized and nitrogen threshold levels developed for restoration of each embayment in Section VIII. Additional modeling is conducted to produce an example of the type of watershed nitrogen reduction required to meet the determined threshold for restoration in a given embayment. This latter assessment represents only one of many solutions and is produced to assist the Town in developing a variety of alternative nitrogen management options for the Apponagansett Bay system. Finally, analyses of the Apponagansett Bay system was relative to potential alterations to loading from the watershed to improve nitrogen related water quality. The results of the nitrogen modeling for the alternative loading scenario is presented in Section IX.

15 MASSACHUSETTS ESTUARIES PROJECT

II. PREVIOUS STUDIES RELATED TO NITROGEN MANAGEMENT

In most marine and estuarine systems such as Apponagansett Bay, the limiting nutrient and thus the nutrient of primary concern is nitrogen. In large part, if nitrogen addition is controlled then eutrophication is controlled. This approach has been formalized through the development of tools for predicting nitrogen loads from watersheds and the concentrations of water column nitrogen that may result. Additional development of the eutrophication management approaches via the reduction of nitrogen loads has also generated specific guidelines as to what is to be considered acceptable water column nitrogen concentrations to achieve desired water quality goals (e.g., see Cape Cod Commission 1991, 1998; Howes et al. 2003).

Until recently, these tools for predicting loads and concentrations tended to be generic in nature, and overlooked some of the site-specific characteristics associated with a given water body. The present Massachusetts Estuaries Project (MEP) study focuses on linking water quality model predictions, based upon watershed nitrogen loading and embayment recycling and system hydrodynamics, to actual measured values for specific nutrient species. The linked watershed-embayment model is built using embayment specific measurements, thus enabling calibration of the prediction process for specific conditions in each of the coastal embayments of southeastern Massachusetts, including Apponagansett Bay.

Beginning in 1990, nutrient loading evaluations for all Dartmouth embayments were included in the initial and then subsequently updated nitrogen loading and management strategy plans for all Buzzards Bay embayments (Buzzards Bay Comprehensive Conservation and Management Plan (CCMP) issued by the Buzzards Bay National Estuary Program (BBNEP 1991). The 1991 CCMP used embayment–specific hydrodynamic data and available land-use data to characterize an embayment’s Total Maximum Annual Nitrogen Load or TMAL. The 1991 tiered nitrogen loading model assigned values for nitrogen generation and transport within a watershed using: watershed delineations; land usage characterizations (e.g., forest, water, cropland and pasture, commercial, residential, industrial, marsh, transportation, etc.); their respective land-use area measurements using GIS (Geographical Information Systems); and the hydrodynamic characteristics of the embayment (bathymetry, volume, estuary turn-over time) to calculate the nitrogen loading to each embayment. The TMAL is a measure of an embayments ability to meet one of several regulatory water quality classifications. For Apponagansett Bay, a classification of SA, the second highest marine waters classification, has been set by MA Department of Environmental Protection. The 1991 CCMP loading model calculated a recommended load limit of 35,700 kg of nitrogen per year while the calculated 1991 existing nitrogen load to Apponagansett Bay was 52,000 kg nitrogen per year, or about 46% more than the recommended limit.

The CCMP Nitrogen loading model data was used as a starting point for the nitrogen management portion of the CCMP, the Buzzards Bay Action Plan which outlined measures that Towns could adopt to manage nitrogen inputs to the regions embayments. The Action Plan recommendation for municipalities included conducting parcel by parcel build-out analysis of the watersheds; sewering when appropriate; adoption of nitrogen–loading bylaws for sensitive embayments; reduction of agricultural fertilizer use by cranberry growers- there are two bogs in the Slocums River watershed; and the implementation of agricultural best-management practices for fertilizers and manure.

16 MASSACHUSETTS ESTUARIES PROJECT

Since the late 1980’s the Town of Dartmouth has instituted many of the recommendations of the CCMP in order to improve ground and surface water quality in the Town. However, the maturity of development and small amount of developable land available in the northern two- thirds of the Apponagansett watershed may have limited opportunities to apply management tools. Nevertheless, the Planning Board has adopted stormwater regulations that require constructed wetlands to treat parking lot and roof runoff for new commercial and residential projects. When older commercial developments are remodeled or increased in scale, the Planning Board negotiates improvements in stormwater management infrastructure that tend to improve stream water quality. The Department of Public Works has worked to improve stormwater runoff by installing several stormwater treatment units in the Apponagansett basin and recently purchased storm drain maintenance equipment to improve the performance of existing infrastructure. The Department of Public Works is planning to extend the sewer network to serve the Bayview neighborhood on Smith Neck Road, and when completed this project will reduce the nitrogen inputs to Outer Apponagansett Bay. It is likely that these combined efforts have contributed to and will continue to contribute to a slowing in the rate of increase of nitrogen loading to the Town embayments.

Nutrient data collected prior to 1992 relative to nutrient levels in inner Apponagansett Bay is limited. In 1986 MA DEP did a synoptic sampling of the fresh and marine waters in western Buzzards Bay watersheds including Apponagansett Bay (MA DEOE 1986). Total Kjeldahl nitrogen and ammonium were measured; nitrate was not sampled at the marine sites. Two sites were sampled twice daily on two successive days in July 1986. The mean of four total kjeldahl nitrogen samples from just north of the bridge was 1.75 mg/l (125 uM nitrogen) and in the mean from an outer harbor site were 1.58 mg/l (113 uM nitrogen) perhaps indicating a gradient between the inner and outer harbor. These levels are relatively high but are similar to nitrogen measurements made during this survey in other western Buzzards Bay embayments. Since these samples were collected on two successive days, they do not give much insight into the range of TKN in the harbor at that time and are of limited use in characterizing the health of the bay at that time.

Apponagansett Bay Water Quality Monitoring Program - Beginning in 1993, summer measurement of nutrient levels (dissolved and particulate nitrogen; phosphorus); and other water quality indicators, (chlorophyll; secchi depth, dissolved oxygen and temperature) was begun in the Apponagansett Bay embayment (inner and outer bay) by the BayWatchers program instituted by the Coalition for Buzzards Bay, the Buzzards Bay Project and scientists now at SMAST-UMassD (then at WHOI). The Coalition’s BayWatcher Program, with staff support from the Lloyd Center and the Turn the Tide Program, has collected the baseline water quality data necessary to support ecological management of the Apponagansett Bay Embayment System. The BayWatcher monitoring was undertaken was a collaborative effort, with CBB (Tony Williams) coordinating the field effort and chemical assays being completed by the SMAST Coastal Systems Analytical Facility until 2008. The Coastal Systems Analytical Facility is located in the School for Marine Science and Technology UMASS-Dartmouth, 706 S. Rodney French Blvd, New Bedford, MA, and the laboratory Points of Contact are Sara Sampieri 508-910-6325 ([email protected]) or Mike Bartlett ([email protected]). Use of the SMAST Analytical Facility ensured sufficient sensitivity and accuracy of the analytical protocols and that proper QA/QC procedures were followed to allow incorporation of the data into the MEP analysis. Baseline water quality data were a prerequisite to entry into the MEP. Implementation of the MEP Linked Watershed-Embayment Approach necessarily incorporates the quantitative water column nitrogen data (2003-2011) gathered by the Monitoring Program and watershed and embayment data collected by MEP staff.

17 MASSACHUSETTS ESTUARIES PROJECT

The focus of the Coalition for Buzzards Bay BayWatcher Water Quality Monitoring Program effort has been to gather site-specific data on the current nitrogen related water quality throughout all the embayments tributary to Buzzards Bay. The program was tailored to the gathering of data specifically to support evaluations relating observed water quality to habitat health. The BayWatcher Water Quality Monitoring Program in the Apponagansett Bay Embayment System developed a data set that elucidated the long-term water quality of this system (Figure II-1). The BayWatcher Program provided the quantitative water column nitrogen data (1992-2010) for validating the MEP’s Linked Watershed-Embayment Approach (see Section VI).

Figure II-1. BayWatcher Water Quality Monitoring Program of the Apponagansett Bay Estuary, Town of Dartmouth. Estuarine water quality monitoring stations sampled by the Coalition for Buzzards Bay and the Lloyd Center / Turn the Tide Program. Stream water quality stations depicted in Section IV were sampled weekly by the Lloyd Center staff.

18 MASSACHUSETTS ESTUARIES PROJECT

In 1999 the first 7 years of embayment data from 30 embayments monitored by the Baywatchers Program was compared with the 1991 BBNEP Buzzards Bay nitrogen loading model results (Costa et al, 1999). The 1999 study undertook a comparative analysis of previous studies of nitrogen loading and ecosystem responses and compared those analyses with the Baywatchers results using the Baywatchers data as a yardstick for CCMP model evaluation. Costa et al. found that the revised loading methodology yielded somewhat lower existing loading levels to Apponagansett Bay from 52,000 kg/y to 31,314 kg/y of nitrogen and at the same time revising the TMAL recommended threshold downward from 35,700 kg/y to 15,000 kg/y of nitrogen. The 1999 change in nitrogen load recommendation shows an excess of nitrogen of more than 200% entering the embayment, underscoring an imbalance that exists between the Apponagansett Bay recommended nitrogen “carrying capacity” and the 1999 estimate of nitrogen load (Costa et al. 1999).

After the first ten years of monitoring, the Baywatchers data set was reviewed in 2001 by the Coalition for Buzzards Bay and summarized in a report for the first period, 1992-2001 (CBB 2002). For the stations included in Apponagansett Bay Inner which is the upper half of inner Apponagansett Bay the Baywatchers data found poor/eutrophic water quality for five summers between 1992 and 1998 and fair water quality for the years 1999-2001. For Apponagansett Bay- Mid Harbor which includes the area south of Little Island to the causeway and bridge, the water quality was fair during the same period. At the Apponagansett Bay – Outer stations, water quality was in the upper Fair range for all eight years of measurement up to 2001. More recent 5-year running averages of the health indexes for the three Apponagansett Bay through the summer of 2005 show variable results with Apponagansett Bay Inner showing a decline in the 2005 data, Apponagansett Bay Mid Harbor showing no change and Apponagansett Bay –Outer showing a small decline in the 2005 5-year average. On balance the water quality for Inner Apponagansett Bay (north of the causeway and bridge) changes from poor to fair in a gradient southward from the mouths of the two main streams, Buttonwood Brook and Apponagansett Bay Brook, probably reflecting the effects of dilution with lower nitrogen concentrations in Buzzards Bay waters. In comparison to other Buzzards Bay embayments sampled by the Baywatchers program, Apponagansett Bay Inner “Health Index” of about 30 was consistently in the lowest third of all 29 Buzzards Bay embayments, while water quality improved near the bridge to about 57 and improved still more in the Outer Bay to about 63 south of the bridge in the Outer harbor area. For comparative purposes, several of the healthiest embayments, including Aucoot Cove Outer, Quisset Harbor Outer, and Penikese Island had Health Indices above 84, while nearby Clarks Cove had Health Index numbers from 76 to 83 for the same period.

Sampling of Apponagansett Bay waters was conducted as part of a study of coastal embayments in the Southeastern New England area with the purpose of calibration of a watershed and coastal embayment model by Kremer (unpublished data 1998). Surface water samples taken during four visits to the sites showed a seasonal increase in chlorophyll a in the inner bay from about 1 ug/L in the mid-June 1998 samples to 35 ug/L in August 1998. Mean values of about 20ug/l with a maximum of 57 ug/L were found in 26 samples for the four dates. While the average values for the mid June samples were relatively low (1 ug/l) the values from late June (7.7 ug/l), July (25.5 ug/L) and August (35 ug/L) showed a pattern of increasing chlorophyll a levels and the July and August means would generally be characterized as indicating eutrophic conditions for waters of this salinity (polyhaline:18-30 ppt; US EPA 2003).

In summary, both loading model estimates of BBNEP in 1991 and the revised loading levels of 1999 are confirmed by the past and ongoing Baywatchers nitrogen and other water quality indicator measurements and agree that Inner Apponagansett Bay is well past its carrying

19 MASSACHUSETTS ESTUARIES PROJECT capacity for nitrogen. In comparison with neighboring Slocums River, the estimated excess nitrogen loading rate is substantially lower and reflects the relationship between the embayment morphology and tidal flushing rate and the relative terrestrial watershed size and land use. For Outer Apponagansett Bay, the data from previous studies indicates the current level nutrient loading is resulting in moderately healthy water quality which is likely due to the good mixing with cleaner Buzzards Bay water.

The MEP effort has integrated and built upon previous watershed delineation and land- use analyses, river transport, embayment water quality and eelgrass surveys in the present effort. This information is integrated with MEP collected higher order biogeochemical analyses and water quality modeling necessary to develop critical nitrogen targets for the Apponagansett Bay Embayment System. The incorporation of appropriate data from previous studies to which the MEP Technical Team had access both enhance the determination of nitrogen thresholds for the Apponagansett Bay System and has reduced analysis costs for the Town of Dartmouth.

The City of New Bedford CZM –Non Point Source Pollution Program Buttonwood Brook Stormwater Sampling (Fall 2003) This study was undertaken in 2003 as a partnership between the City of New Bedford, the Lloyd Center for Environmental Studies and the Coastal Systems Program at the University of Massachusetts-Dartmouth, School for Marine Science and Technology. It was motivated by the well documented reduction in the water quality of the three principal estuaries in the Town of Dartmouth, Apponagansett Bay, Slocums River and Little River due to bacterial contamination and excess nutrients. In ten years of water quality assessment studies conducted by The Coalition for Buzzards Bay’s “Baywatchers Program“ (1992-present), water quality in Apponagansett Bay has ranked in the bottom 10% of all harbors and coves monitored in Buzzards Bay. Many of the water quality problems of this estuary are related to excessive nutrient inputs. In addition, the Apponagansett Bay and its principal tributary Buttonwood Brook are on the state 303(d) list for pathogens. Due to the chronic bacterial contamination in these waters, valuable soft and hard shell clam and scallop fisheries in Apponagansett Bay are subject to permanent and seasonal closures. This project focused on a previously identified area of concern along the upper reaches of Buttonwood Brook, the principal source of fresh surface water to Apponagansett Bay. Approximately forty percent of the upper stream channel of Buttonwood Brook including Buttonwood Pond lies within the City of New Bedford. This project consisted of sampling stormwater inputs in the New Bedford portion of the stream system. The major focus of the study was to locate the regions of Buttonwood Brook which are recipient of watershed contaminants, primarily bacterial indicators (fecal coliform, E. coli, Enterococcus) and nutrients with secondary focus on organics (TPH) and metals. The goal was to determine regions of high source loads in order to restore/manage the shellfish resources within Apponagansett Bay. As a result data was analyzed within the context of the Massachusetts State Water Quality Classification for Apponagansett Bay, SA, and for Buttonwood Brook which has not been classified. The shellfish resource target for the study was Apponagansett Bay.

The goal of this project was to assess upper Buttonwood Brook in greater detail to identify the sources of and extent of nutrient, pathogen and sediment problems at Route 6 and along adjacent reaches of the New Bedford portion of Buttonwood Brook.

This project of water quality assessment was conducted to provide the data necessary to identify source areas of contamination along the upper reaches of Buttonwood Brook within the City of New Bedford. The results were intended to lead to a subsequent proposal for design and implementation of best management practices (BMP) for upper Buttonwood Brook. Though submitted by the Conservation Commission of the City of New Bedford, the proposal was

20 MASSACHUSETTS ESTUARIES PROJECT intended to be a complement to the watershed-based assessment and management project for Apponagansett Bay, the receiving waters of Buttonwood Brook. That project was called, Turn the Tide-Restore Dartmouth’s Estuaries, an integrated, watershed-based, $1.8 million, four-year plan to assess and restore Dartmouth’s’ estuaries. The Turn the Tide partnership was formed in 2001 among the Coalition for Buzzards Bay, the Town of Dartmouth, the UMass Dartmouth School of Marine Science and Technology (SMAST), and The Lloyd Center for Environmental Studies (LC), and was modeled upon and integrated with the methodology of the Massachusetts Estuaries Project.

Specific Project Goals for the Buttonwood Brook Storm Sampling

The primary goals of the Upper Buttonwood Brook Storm Water Input Assessment included:

 Identifying non-storm (dry weather) inputs of bacteria to upper Buttonwood Brook  Identifying storm-related inputs of bacteria to upper Buttonwood Brook  Identifying organic pollutant inputs to upper Buttonwood Brook  Identifying hotspots of metals contamination through sediment sampling  Evaluating nitrogen loading inputs during wet and dry weather,  Characterizing sources of pollution  Prioritizing pollutant sources as targets for Best Management Practices or other management strategies to reduce bacterial and nutrient loading to Buttonwood Brook and Apponagansett Bay

Bacterial results were separated into discussions of the Buttonwood Zoo, Buttonwood Pond and the Route 140 Main channel of Buttonwood Brook, including the intersection of Kempton St and Route 140.

Buttonwood Zoo The results of water samples taken downstream of Buttonwood Zoo at Site 1 (located at the intersection of Hawthorne and Brownell) indicated that at the time of the study the zoo was a source of bacterial contamination to Buttonwood Brook. Results from the wet weather events of September 16 and October 15, 2003 yielded concentrations greater than 230,000/100ml for Fecal coliform, E. Coli and Enterococci, and increases below the Zoo were two orders of magnitude higher than levels just above the Zoo. High mass loadings were also evident during the first two wet weather events for all three assays.

Buttonwood Pond The results of sampling the outlet of Buttonwood Pond (south of Court St) were variable. The graphical depictions for mass loading Below and Above the pond illustrated that during the first storm event, bacterial loads were lower Below the Pond (Site 2) than Above the Pond (Site 5) for all three analytes. However, during the second and third storm sampling events, bacterial loads Below the Pond exceeded loads Above the Pond for E. coli but not for Enterococcus.

Stream north of Buttonwood Pond/Intersection of Route 6/Kempton St The sampling conducted north of the pond (Sites 3 through 9) consisted of sampling either the main stem of Buttonwood Brook (Sites 5, 8 and 9), feeder channels to Buttonwood Brook (Sites 4, 6, and 10) or pipe ends (Sites 3 and 7). During August of 2003, prior to the onset of sampling, Dr. Tony Millham observed solid sewage waste in the channel at Site 6. The discharge occurred following a significant rain event on August 11, 2003 (2.6” according to readings taken at the

21 MASSACHUSETTS ESTUARIES PROJECT

New Bedford Regional Airport). The sewage was not observed prior to August 6 when various project personnel were conducting preliminary surveys of sample sites.

The sampling results for this portion of the study area revealed sources of bacterial contamination. Samples from pipe outlets (Sites 3 and 7), and dry feeder channels (Site 4 and 6) during wet weather events, contained varying loads of bacteria for all analytes as shown in Appendix 2 (wet-dry weather graphs) of the report. Due to the flashy flow conditions during wet weather, it was not possible to obtain samples from every dry Site during every wet weather event and some channels and pipes were dry during dry weather sampling. However, when sampled, high concentrations and significant mass loads were observed. Site 6, where the sewage discharge was observed prior to the sampling program, contained significant bacterial concentrations. Additionally, sampling of the main channel of Buttonwood Brook, Sites 5, 8 & 9, and the perennial tributary parallel to Route 140 (Site 10) provided evidence that bacteria is a problem at these locations too.

Total dissolved nitrogen and total nitrogen levels measured at 5 stations along the main stem of the stream were moderate for an urbanized stream with most of the nitrogen in the dissolved fraction. There was also an increase in the mean concentration of both nitrogen assays with distance downstream.

Orthophosphate levels were low to moderate for the urban setting and variable with no trend for downstream increase.

Total suspended solids were highly variable with very high storm event values in the upper channelized streambed at site CZ8. The maximum value for TSS occurred at CZ8 during the 10/15/03 rain event of 1.46”.

Particulate carbon, like TSS was highly variable and dependent upon rainfall amounts and stream channel setting and exhibited no downstream pattern of increase or decrease.

Nutrients in Zoo Region Nutrient values above and below the zoo Station 2 (above zoo) and Station 1 (below zoo) were variable depending upon the analyte (presented in Appendix III of study). For example, both orthophosphate and nitrate + nitrite were higher downstream of the zoo than above the zoo, while dissolved organic nitrogen and ammonium were variable. Particulates, PON, POC and TSS were higher above the zoo. Of these analytes, only the orthophosphate was seen as relevant to animal husbandry at the zoo.

Nutrients at CZ6 The culvert channel that had a sewage overflow on 8/8/03 during an intense rainstorm of more than 2.5”, only had flow twice during the study period. The average values for TDN and TN were similar to those of the other stream sites (TDN: 0.88 mg/l and TN: 1.22 mg/l). POC and TSS were also similar ( TSS: 9.97 mg/l and POC: 3.79 mg/l) while PO4 values were about twice the mean of other sites (0.12 mg/l).

Recommendations to Remediate Pollutant Sources in Buttonwood Brook The purpose of the grant was to identify sources of pollutants to Buttonwood Brook within the upper reaches of the watershed within the City of New Bedford. The successful completion of the water quality sampling in 2003 led to the identification of pollution sources. Bacterial contamination was the most significant pollutant source identified. As such, the study

22 MASSACHUSETTS ESTUARIES PROJECT recommendations concentrated on addressing the bacterial contamination sources in the upper reaches of the Buttonwood Brook Watershed.

The recommendations were broken down by location as follows:

Buttonwood Zoo The City of New Bedford was to proceed to investigate the bacterial contamination emanating from Buttonwood Park Zoo. A preliminary meeting was been held with the Zoo Director and a course of action for addressing the problem included the following components:

1. A more detailed bacterial sampling program was to be developed for the zoo. This would include wet weather sampling of Buttonwood Brook during 2004, at selected sample sites within the zoo boundaries. The purpose of the additional sampling was to accurately pinpoint the sources of bacterial contamination.

2. Concurrently with the sampling effort, the Conservation Commission was to undertake in partnership with the Buttonwood Zoo, an evaluation of all Best Management Practices currently utilized by the zoo to prevent non point source runoff of animal waste into Buttonwood Brook.

Following completion of the sampling and evaluation of Best Management Practices, the next steps towards remediation would be identified. At the time of the study it seemed likely a combination of techniques would have to be employed to reduce bacterial loads. Techniques could include development of new Best Management Techniques to control non point source runoff, and developing methods to remove bacteria from the Buttonwood Brook prior to it exiting the zoo boundaries. As bacteria can be a difficult pollutant to remove from the water column, it was suggested that methods to address remediation might have to include some type of treatment facility.

Buttonwood Brook north of Buttonwood Pond The results of the sampling in this area provided significant evidence of bacterial contamination of the perennial stream channel. It was thought that the sources of the contamination could be illegal cross connections of sewer pipes into the storm drain systems, or in the case of Site 6 where sewage sludge was observed, it could be some sort of an old overflow from a pump station, possibly located in Dartmouth. To address the bacterial contamination of Buttonwood Brook in this portion of the study site a coordinated effort would required between the New Bedford Conservation Commission, the City of New Bedford Department of Infrastructure, who is responsible for the maintenance of the City sewer and storm drain systems, and the Town of Dartmouth Public Works (to address Site 6) and begin development of a workplan for addressing the observed sewage situation. The stated goal for 2004 was to form a partnership with the concerned parties and develop a program for identifying the sources and eliminating them.

Buttonwood Pond The results of the sampling upstream and downstream of Buttonwood Pond were variable. At times, bacterial mass loads were substantively lower below the pond (Site 2) than above the pond (Site 5) but these results were not consistent as the reverse was also true. Therefore, recommendations for the future included additional sampling within the pond itself and outlet to determine more accurately what, if any, pollutant contributions Buttonwood Pond is making to the downstream watershed. As a valuable recreational resource for the City of New Bedford, a

23 MASSACHUSETTS ESTUARIES PROJECT concerted effort was suggested to develop a management plan for the pond to restore the pond banks, control waterfowl densities and address eutrophication.

Turn the Tide Project: Natural Resources Assessment of the Slocums River, Little River and Apponagansett Bay Watersheds (April 2009) - This report presented the findings of a series of surveys undertaken to determine the current state of several important natural resources within the three main estuaries in the Town of Dartmouth: the Slocums River, Little River and Apponagansett Bay, and the watersheds associated with those estuaries. The surveys focused specifically on shell fish, birds, freshwater macro-invertebrates, state listed rare and endangered species, and fin fish. These surveys were conducted by the Lloyd Center for the Environment as part of the Turn the Tide Project, a joint effort by the Town of Dartmouth, the Coalition for Buzzards Bay, the Lloyd Center for the Environment and the Coastal Systems Program at the School for Marine Science and Technology (SMAST), University of Massachusetts Dartmouth.

Since its inception in 2002, the goals of the Turn the Tide Project have been to address the problem of the long-term decline of water quality and loss of native species and suitable habitat in these three (3) Dartmouth estuaries and their surrounding watersheds. The results of the surveys were completed to serve as both an assessment of the current state of these natural resources and a data base with which to investigate potential effective management practices to improve existing conditions. Some of the results of the natural resource surveys that are pertinent to the Apponagansett Bay estuarine system are summarized herein.

Shellfish The value of the landings of commercially important shellfish species has fallen in the Town of Dartmouth for most years since the peak landing year of 1985. The amounts of individual species harvested have varied widely from year to year and are influenced by many factors including: availability, market price, fishing effort; cycles of reproduction, loss of habitat, predation, disease, closures due to toxic algal blooms, oil spills, hurricanes and rainfall intensity; and poaching. In an effort to quantify the current status of shellfish abundance and distribution, a survey was conducted in Apponagansett Bay, as well as the Slocums River and Little River, between 2003 and 2005.

In Apponagansett Bay Quahogs were most abundant in the shallower depths below MLW but were scarce in deeper subtidal areas. Soft shelled clams were found in low numbers in shallow areas near MLW and were absent from subtidal areas. As was the case in Slocums and Little Rivers, recruitment of Quahogs in Appongansett Bay has been poor. Additionally, very few live but many dead Oysters were found in Apponagansett Bay as well as the other two estuaries of the Town. Determining the reason(s) for such high mortality was beyond the scope of the survey and could not be determined.

Birds In all 3 estuaries of Dartmouth, inclusive of Apponagansett Bay, geese, ducks and gulls were the most dominant bird groups documented during the surveys conducted from 2001-2005.

In Apponagansett Bay, Ring-billed Gulls, Buffleheads (winter), Herring Gulls, Mallards and Canada Geese were dominant. There are also Red Breasted Mergansers and Scaup which, with the noted abundance of Buffleheads, indicate a more open estuary than either Slocums or Little River. Buffleheads have shown a decline over the past 20 years as have some less abundant species such as Red Breasted Mergansers and Scaup. Buttonwood Pond (within the Apponagansett Bay watershed) was also surveyed where Canada Geese, Mallards, Ring-billed

24 MASSACHUSETTS ESTUARIES PROJECT

Gulls Rock Doves and Herring Gulls were most abundant. It is important to note that while overall densities of birds were generally found to be low, local densities of populations in the vicinity of popular feeding areas (Apponagansett Bay Park, Padanarum Bridge, Buttonwood Pond) may be high enough to cause localized nutrient- or bacteria-related problems. Moreover, as nutrient management is undertaken in the Apponagansett Bay watershed it is important to consider that this watershed support habitat for State Listed species and Priority Species including Piping Plover, Common Tern and Least Tern.

Freshwater Macro-invertebrates A Sensitivity Index was created using the data from the 2009 survey and data from earlier surveys in the Apponagansett Bay watershed in 1993 and 2002. This index is an indicator of stream health based on the type and abundance of taxa present. In the Apponagansett Bay watershed, many of the survey sites were in areas that were semi-permanent. Apponagansett Bay Brook had the lowest Index while many of the stations in Buttonwood Brook were higher and comparable to stations in the Paskamansett River in the adjacent Slocums River watershed. Data from most sites indicated that water quality is not poor enough to have a detrimental effect on macro-invertebrate populations in these two main streams discharging to the head of Apponagansett Bay. The 2009 surveys constitute a baseline for future surveys.

Rare Species within the Apponagansett Bay Watershed The following species that are listed as rare in the Massachusetts Endangered Species Act have been recorded within wetland habitats associated with the Slocums, Little and Paskamansett Rivers, Destruction Brook, Turners Pond, and/or the Apponagansett Bay watershed. The species identified in the 2009 survey will work will need to be considered as nutrient management in the Apponagansett Bay watershed is undertaken in the future, particularly N-management involving the implementation of soft solutions oriented towards enhancing natural attenuation of bogs, ponds, wetlands, stream riparian areas or even abandoned cranberry bogs as they may exist in the watershed.

Slocums River Piping Plover (Charadrius melodus) – Threatened (State and Federal) Least Tern (Sterna antillarum) – Special Concern Common Tern (Sterna hirundo) – Special Concern Northern Harrier (Circus cyaneus) – Threatened Diamondback Terrapin (Malaclemys terrapin) – Threatened Spotted Turtle (Clemmys guttata) – Delisted in 2006. Box Turtle (Terrapene carolina) – Special Concern (Likely but Not Confirmed) Marbled Salamander (Ambystoma opacum) – Threatened Four-toed Salamander (Hemidactylium scutatum) – Special Concern Attenuated Bluet (Enallagma daeckii) – Special Concern Scarlet Bluet (Enallagma pictum) – Threatened Hessel’s Hairstreak (Callophrys hesseli) – Special Concern Spartina Borer (Spartiniphaga inops) – Special Concern Pale Green Pinion Moth (Lithophane viridipallens) – Special Concern Coastal Swamp Amphipod (Synurella chamberlaini) – Special Concern Sea Pink (Sabatia stellaris) – Endangered (Historical Records only)

Little River Common Tern (Sterna hirundo) – Special Concern Northern Harrier (Circus cyaneus) – Threatened

25 MASSACHUSETTS ESTUARIES PROJECT

Box Turtle (Terrapene carolina) – Special Concern (Likely but Not Confirmed) Four-toed Salamander (Hemidactylium scutatum) – Special Concern (Likely but Not Confirmed) Spartina Borer (Spartiniphaga inops) – Special Concern (Likely but Not Confirmed)

Apponagansett Bay Common Tern (Sterna hirundo) – Special Concern Diamondback Terrapin (Malaclemys terrapin) – Threatened (Likely but Not Confirmed) Spotted Turtle (Clemmys guttata) – Delisted in 2006 (Likely but Not Confirmed) Box Turtle (Terrapene carolina) – Special Concern (Likely but Not Confirmed) Four-toed Salamander (Hemidactylium scutatum) – Special Concern (Likely but Not Confirmed)

Fin Fish Estuaries have long been recognized as essential nurseries for finfish. Any degradation in water quality such as influx of nutrients negatively impacts an estuary’s ability to function in this role. Historical systematic finfish surveys of southeastern Massachusetts estuaries are scant. In order to assess the condition of the current finfish communities in the Apponagansett Bay estuary, the system was sampled on a monthly basis in the lower, mid-, and upper reaches of the Bay. In the 2009 finfish survey conducted in Apponagansett Bay, 7,103 individuals were captured representing 25 species. Silversides and Herring were the most abundant species captured in Apponagansett Bay. Killifish, Tautog and Black Sea Bass were also present in large numbers. Species diversity indices (H’) were calculated at each sampling site for the 2009 finfish survey in Apponagansett Bay and it was found that H’ varied from 1.010-2.389. The 2009 finfish surveying in Apponagansett Bay serves as a current baseline against which future conditions can be compared as nutrient management is undertaken to improve water quality and habitat quality in Apponagansett Bay.

Basis for Development of Total Maximum Daily Load for Bacteria in Apponagansett Bay: This report (October 2007) was undertaken by technical staff from the Coastal Systems Program within the University of Massachusetts-Dartmouth School for Marine Science and Technology and represented the basis for the development of a TMDL for bacteria in the Apponagansett Bay watershed including Buttonwood Brook, in the Town of Dartmouth and the City of New Bedford, Massachusetts. The goal was the development of a TMDL for Fecal Coliform, E. coli and Enterococcus. Fecal Coliform is a general classification of bacteria that are typically associated with animal and human waste. E coli are typically found in the intestines of animals and humans. Some strains are known to be toxic to humans. Enterococcus, a bacterium toxic to humans, is thought to be a better indicator of human health risk than Fecal Coliforms. At present, Fecal Coliform is the bacterial indicator used to manage shell fish resources and Enterococcus the indicator to manage primary (i.e. bathing beaches) and secondary contact recreation

The intent of the report was for the TMDL to use the data provided in order to establish the bacterial limits for the water resource and outline corrective actions to achieve the restoration goal. Based on the format and content of Bacterial Technical Reports previously developed by the Coastal Systems Program for the Massachusetts Estuaries Project (MEP) and to be utilized by the MassDEP, this report was not meant to direct the reader to specific bacterial sources (point or non-point), nor was it intended to produce bacteria Waste Load Allocations (WLAs) or Load Allocations by bacteria source for the Apponagansett Bay system. This report was developed to point to likely geographic sections of the system that are the most likely sources of the highest bacterial concentrations recorded to date. Historical data were compiled from multiple agencies and were synthesized along with new data (2003) on flow and bacteria in Buttonwood brook and Apponagansett Brook that were collected by the SMAST/Turn

26 MASSACHUSETTS ESTUARIES PROJECT the Tide Team. In order to identify likely sections of Apponagansett Bay and Buttonwood Brook responsible for highest bacterial contamination, geometric means and percent exceedences were developed for current and historical data obtained for this report.

In mass load terms, Buttonwood Brook is the largest contributor of bacteria to the inner bay. Based on the data reviewed, there are three areas of bacteria sources that cause high bacterial levels in Buttonwood Brook. First, in the upper watershed upstream of BWB 9, there is a strong source for wet weather contamination. Second, it appears likely that as Buttonwood Brook travels through the Buttonwood Zoo grounds, conditions at the Zoo cause a large increase in bacterial levels in the stream and contribute disproportionately under both wet and dry conditions to the bacterial load at the mouth of Buttonwood Brook (based on 2003 flow and bacteria sample collection). Third, the southwestern tributary sampled at station D-4 is a factor in raising bacterial levels in the stream during wet weather. Identification and management of the three problem areas described above should reduce the sources of bacteria in the Buttonwood watershed and substantially lower the bacterial load at the mouth of the brook.

Reductions will likely come from multiple and diverse management tools that concentrate in the headwaters and tributaries of the stream. These tools likely will include: extending sewers into some small subwatersheds in the most northern reaches of the drainage (City of New Bedford); identifying sources of intermittent raw sewage releases to the stream at Rt. 6 (underway now by the Town of Dartmouth and the City of New Bedford); identifying and abating high bacterial sources in storm drain networks in upper Buttonwood drainage; improvements to the storm water retention capacity of Buttonwood Pond; and use of peak storm flow diversion and short-term storage in the riparian zones on the middle section of the stream drainage. An effort to locate sources in the problem areas in the upper watershed (West Rockdale neighborhood) has been underway since 2004 by City of New Bedford environmental officials, though no sources have been pinpointed thus far.

Regarding Apponagansett Bay Brook, from the sanitary survey that was conducted in 2005, there was a significant source of bacterial contamination along the entire length of the stream and especially in the Utica Lane section of lower Apponagansett Bay Brook. Storm drains did not appear to be the source of the extremely high bacteria counts in the stream itself. Since this area is sewered, the most likely source of bacteria was determined to be wildlife and domesticated animals. The north storm drain outfall pipe at Russells Mills Rd. (RMR N) was identified as a major source of bacteria contamination in the area of the mouth of the brook. Identifying and abating the source of the high bacteria levels would help in an overall mass load reduction to the inner bay..

Focusing on the East Shore of Inner Apponagansett Bay, the Fort St. (AP 17), Bridge Street. and Tradewinds Lane outfalls were identified as the most significant sources of bacteria contamination to the east bay. Most of the smaller outfalls along the eastern shore showed both relatively low bacterial levels and low discharges, and contributed relatively small amounts to the inner bay. The Fort Street/Ap-17 drainage system was pointed out as a candidate for one or more sanitary surveys to characterize the source of the bacteria contamination in this area. The Bridge St. storm drain was also identified as a candidate for more intensive investigation via one or more sanitary surveys to identify the sources of bacterial contamination.

Regulatory Assessments of Apponagansett Bay Resources -

The Apponagansett Bay Estuary contains a variety of natural resources of value to the citizens of Dartmouth as well as to the Commonwealth. As such, over the years surveys have

27 MASSACHUSETTS ESTUARIES PROJECT been conducted to support protection and management of these resources. The MEP gathers the available information on these resources as part of its assessment, and presents them here (Figures II-2 through II-6) for reference by those providing stewardship for this estuary. For the Apponagansett Bay Estuary these include:

 Mouth of River designation - MassDEP (Figure II-2a,b,c)  Designated Shellfish Growing Area – MassDMF (Figure II-3)  Shellfish Suitability Areas - MassDMF (Figure II-4)  Estimated Habitats for Rare Wildlife and State Protected Rare Species – NHESP (Figure II-5)  Anadromous Fish Runs - MassDMF (None Present in Apponagansett Bay system)

28 MASSACHUSETTS ESTUARIES PROJECT

Figure II-2a. Mouth of Coastal Rivers designation for Apponagansett Bay as determined by – MassDEP Wetlands Program.

29 MASSACHUSETTS ESTUARIES PROJECT

Figure II-2b. Mouth of Coastal Rivers designation for Apponagansett Bay as determined by – MassDEP Wetlands Program.

30 MASSACHUSETTS ESTUARIES PROJECT

Figure II-2c. Mouth of Coastal Rivers designation for Apponagansett Bay as determined by – MassDEP Wetlands Program.

31 MASSACHUSETTS ESTUARIES PROJECT

Figure II-3. Location of designated shellfish growing areas and their status relative to shellfish harvesting as determined by Mass Division of Marine Fisheries. Closures are generally related to bacterial contamination or "activities", such as the location of marinas.

32 MASSACHUSETTS ESTUARIES PROJECT

Figure II-4. Location of shellfish suitability areas within the Apponagansett Bay Estuary as determined by Mass Division of Marine Fisheries. Suitability does not necessarily mean "presence".

33 MASSACHUSETTS ESTUARIES PROJECT

Figure II-5. Estimated Habitats for Rare Wildlife and State Protected Rare Species within the Apponagansett Bay Estuary as determined by - NHESP.

34 MASSACHUSETTS ESTUARIES PROJECT

III. DELINEATION OF WATERSHEDS

III.1 BACKGROUND Apponagansett Bay watershed is located along the northern edge of the Buzzards Bay watershed basin. The Buzzards Bay Basin is the result of glacial processes that defined the surficial geology of the region during the retreat of the Cape Cod Lobe of the Laurentide Ice sheet approximately 18,000 years ago. The underlying granitic and metamorphic bedrock is located at depths ranging from surface outcrops to approximately 100 to 200 feet below land surface depending on the location in the basin (Bent, 1995). Most of the surficial deposits in the Buzzards Bay Basin were deposited during the retreat of the glaciers during the last glacial period and are primarily composed of till and stratified drift deposits. The till is generally overlain by the stratified drift deposits, but is found at the land surface more frequently in the western portion of the basin. The Apponagansett Bay watershed area is mostly composed of till and bedrock, with drift materials along the western shoreline of the Bay. As described by Melvin and others (1992), the till deposits in southern New England are relatively sandy. In areas not overlain by stratified drift deposits, the thickness of the till layer can be as much as 30 feet. Unlike till, stratified drift deposits are composed of sorted and layered glaciofluvial and glacial lacustrine deposits of all grain sizes ranging from cobbles to clay (inclusive of silts, sands and gravels). The glaciofluvial deposits were generated mainly by glacial meltwater streams in outwash plains and river valleys (Stone and Peper, 1982), while glaciolacustrine deposits were generated during the presence of glacial lakes. The fluvial deposits tend to have coarser materials (e.g., sands and gravels), while the lacustrine deposits tend to be finer materials (e.g., silts and clays). Stratified drift deposits in the Mattapoisett River valley and westward along Buzzards Bay tend to follow north-south river valleys, but this is not the case in the till- dominated upper portions of the Apponagansett Bay watershed area. As these stratified drift, till, bedrock materials are heterogeneous throughout the northern Buzzards Bay Basin and are characterized by varying permeabilities and hydraulic conductivities, direct rainwater run-off is typically higher than for the sandy outwash sediments along the eastern shore of Buzzards Bay (i.e., Cape Cod). Therefore, freshwater inflow from northern Buzzards Bay rivers leading to the estuarine systems tends to be a significant transport mechanism along with usual direct groundwater discharge to the estuarine receiving water.

III.2 WATERSHED DELINEATION APPROACH A watershed divide or boundary can be described as the line from which rainwater or snowmelt flows on the surface and through groundwater towards one stream, river or estuary, while rainfall and groundwater on the other side of the divide flows away to another water body. The underlying water table, or the surface of the saturated sediments (aquifer), in these areas tends to reflect the changes in surface elevation within bedrock and till dominated landscapes, but can be modified by layers of low hydraulic conductivity sediments within the aquifer. Delineating watershed divides in areas like Apponagansett Bay typically is completed by a technique called topographic inspection. Topographic inspection begins with developing an understanding of the watershed stratigraphy and hydrogeology to determine the validity of this method of watershed delineation. In the case of the Apponagansett Bay, the surficial till on high elevation areas and the dominance of bedrock in forming the watershed supports the use of this method. Analysis focuses on determining the pattern of lines of local maximum elevation upon a US Geological Survey 1:25,000 topographic map and draws watershed divides based upon the tendency of surface water and groundwater to flow downhill perpendicularly to the topographic contour lines. Divides drawn upon topographic maps can be confirmed by observing general patterns of groundwater flow and surface water flow during rainfall or snow melt or by measuring the flow of water in streams over a hydrologic cycle.

35

MASSACHUSETTS ESTUARIES PROJECT

The initial watershed delineation for Apponagansett Bay was conducted in 1991 by the US Geological Survey as part of determining the watersheds for all the sub-embayments to Buzzard Bay for the Buzzards Bay Project, now the Buzzards Bay National Estuary Program (BBP, 1991). The boundaries were determined by the method of topographic inspection and focused on the outer boundary of each sub-embayment. MEP staff reviewed the delineation for Apponagansett Bay and generally found it to be sufficient for advancing a land-use analysis of this system. In order to complete the MEP assessment, however, subwatershed delineations were developed to address the major freshwater features in the estuary watershed (e.g., Buttonwood Brook and Apponagansett Brook) and to provide nitrogen loadings at spatial scales matching the sub-embayment segmentation of the MEP tidal hydrodynamic model (e.g., loading north and south of the Gulf Road crossing of the estuary).

There are seven (7) subwatersheds to Apponagansett Bay (Figure III-1). Six of the subwatersheds were delineated based mostly on topographic inspection, while the subwatershed to Buttonwood Brook includes an adjustment to match the City of New Bedford stormwater collection system; this adjusted watershed boundary matches the boundary to the Acushnet River Estuary/New Bedford Inner Harbor estuary system (Howes et al., 2013). The delineations allow proper distribution of watershed nitrogen loads in the MEP water quality modeling. The subwatersheds include contributing areas to the two MEP stream gauges on Buttonwood Brook and Apponagansett Brook. Delineation of these subwatersheds allows direct comparison between the expected discharge flows and nitrogen loads from the delineated areas and measured data from the gauges. This effort also supported quantification of nitrogen attenuation prior to discharge to estuarine waters. Attenuation is a critical element in the development of the inputs to the estuary water quality model (see section IV.2).

Based upon the delineated sub-watersheds and annual average recharge, freshwater streamflow and direct groundwater input were determined for the Apponagansett Bay estuary system (Table III-1). The streamflow estimate determined by this method was compared to measured streamflows collected by the MEP (see Section IV). Annual recharge was based on a review of available precipitation data for the region. The National Oceanic and Atmospheric Administration (NOAA) maintains a long-term precipitation gauge in New Bedford. Annual average precipitation at this site between 1961 and 2000 is 47.8 inches (CDM, 2006), while the average between 1971 and 2000 is 50.8 inches (NOAA, 2004). This more-recent long term average was used as the annual precipitation rate for further analysis.

A portion of precipitation is utilized by plants on the land surface (transpiration) and a portion is evaporated back into the atmosphere. USGS recharge rates used in groundwater modeling on Cape Cod are approximately 60% of long-term precipitation rates (e.g., Walter and Whealan, 2005). USGS modeling of recharge in the Charles River basin, which is more similar to the geology of the Apponagansett Bay watershed, has found recharge variations of 43 to 56% of precipitation with a strong reliance on measured streamflows for the development of the model (DeSimone, et al., 2002). Given the uncertainty in many of the factors for developing the percentage of recharge, MEP staff conservatively assumed 60% of precipitation or 30.5 inches per year is an appropriate recharge rate in the Apponagansett Bay watershed. This is the same recharge rate that was used for the MEP assessments of the two adjacent estuary systems: Acushnet River Estuary/New Bedford Inner Harbor (Howes and others, 2013) and the Slocums River (Howes and others, 2012). It should be noted that this rate provided reasonable agreement between measured and estimated streamflows in other stream watersheds reviewed by the MEP. This recharge rate is used to develop the long-term freshwater inflows in Table III- 1 and is also used in the watershed nitrogen loading estimates (see Chapter IV). It should also be noted that this recharge analysis is used as an independent comparison of measured and

36

MASSACHUSETTS ESTUARIES PROJECT modeled annual stream flow, but measured streamflow and loads are used in the water quality modeling for the two gauged watersheds: Buttonwood Brook and Apponagansett Brook.

Figure III-1. Watershed and sub-watershed delineations for the Apponagansett Bay estuary system. Outer watershed boundary is based on USGS/BBP (1991) topographic delineation with adjustments in the lower eastern boundary to reflect the stormwater system collection area of the City of New Bedford. Interior subwatershed delineations were completed by MEP staff using the same topographic examination techniques; interior subwatershed delineations were completed to match natural watershed or estuary features (e.g., basins on either side of Gulf Road crossing) or key measurement points (e.g., MEP stream gauges at Buttonwood Brook and Apponagansett Brook). The watershed is mostly in the Town of Dartmouth, but is also shared with the City of New Bedford.

37

MASSACHUSETTS ESTUARIES PROJECT

Table III-1. Apponagansett Bay MEP Subwatershed Areas and Estimated Long-Term Freshwater Recharge.

Watershed Discharge Watershed Name # Area (acres) 3 3 m /day ft /day Buttonwood Brook 1 1,931 16,562 584,877 Apponagansett Brook Gaged 2 398 3,411 120,464 Apponagansett Bay Inner E 3 645 5,535 195,469 Apponagansett Bay Inner W 4 1,043 8,945 315,880 Dike Marsh 5 1,178 10,103 356,777 Apponagansett Bay Outer W 6 133 1,143 40,376 Apponagansett Bay Outer E 7 74 631 22,289 Apponagansett Bay System Total 5,401 46,330 1,636,132 Notes: 1) discharge volumes are based on 30.46 inches of annual recharge over the watershed; 2) recharge is based on 60% of annual precipitation of 50.77 inches (1971-2000 average at New Bedford NOAA gauge); 3) areas do not include the surface of the estuary 4) flows do not include precipitation on the surface of the estuary; 5) totals may not match due to rounding.

38

MASSACHUSETTS ESTUARIES PROJECT

IV. WATERSHED NITROGEN LOADING TO EMBAYMENT: LAND USE, STREAM INPUTS, AND SEDIMENT NITROGEN RECYCLING

IV.1 WATERSHED LAND USE BASED NITROGEN LOADING ANALYSIS Management of nutrient related water quality and habitat health in coastal waters requires determination of the amount of nitrogen transported by freshwaters (surface water flow, groundwater flow) from the surrounding watershed to the receiving embayment of interest. In southeastern Massachusetts, the nutrient of management concern for estuarine systems is nitrogen and this is true for the Apponagansett Bay system as well. Determination of watershed nitrogen inputs to these embayment systems requires the (a) identification and quantification of the nutrient sources and their loading rates to the land or aquifer, (b) confirmation that the transported load has reached the embayment at the time of analysis, and (c) quantification of nitrogen attenuation that can occur during travel through lakes, ponds, streams and marshes. This latter natural attenuation process results from biological processes that naturally occur within ecosystems. Failure to account for attenuation of nitrogen during transport results in an over-estimate of nitrogen inputs to an estuary and an underestimate of the sensitivity of a system to new inputs (or removals). In addition to the nitrogen transport from land to sea, the amount of direct atmospheric deposition on each embayment surface must be determined as well as the amount of nitrogen recycling within the embayment, specifically nitrogen regeneration from sediments. Sediment nitrogen recycling results primarily from the settling and decay of phytoplankton and macroalgae (and eelgrass when present). During decay, organic nitrogen is transformed to inorganic forms, which may be released to the overlying waters or lost to denitrification within the sediments. Permanent burial of nitrogen is generally small relative to the amount cycled. Sediment nitrogen regeneration can be a seasonally important source of nitrogen to embayment waters or in some cases a sink for nitrogen reaching the bottom. Failure to include the nitrogen balance of estuarine sediments generally leads to errors in predicting water quality, particularly in determination of summertime nitrogen load to embayment waters.

In order to determine watershed nitrogen loading inputs to the Apponagansett Bay estuary system, the MEP Technical Team developed nitrogen loading rates (Section IV.1) to each component of the estuary and its watersheds (Section III). This effort was coordinated with staff from the Town of Dartmouth and the City of New Bedford, as well as the Buzzards Bay National Estuary Program (BBNEP). The Apponagansett Bay sub-watersheds were delineated to define contributing areas to the two main freshwater features (i.e., Buttonwood Brook and Apponagansett Brook) and to each major portion of the estuary. A total of seven sub-watershed areas were delineated within the Apponagansett Bay study area (see Section III). Freshwater inflow from the two gauged brooks accounts for approximately 40% of the watershed inputs to the estuary (see Chapter III).

The initial task in the MEP land use analysis is to gauge whether or not nitrogen discharges to the watershed have reached the estuary. This generally involves a temporal review of land use changes, review of data at natural collection points, such as streams and ponds, and, in groundwater dominated systems, the time of groundwater travel provided by a USGS groundwater model. The Apponagansett Bay watershed system is a stream-dominated system because of its underlying geology, so this portion of the review focused heavily on land use development and data from stream gauges. Comparison of subwatershed nitrogen loads to the overall system estimates show that 57% of the unattenuated load is in the lower portion of the watershed, which was evident once the land use map was reviewed. In addition, the distance from the edges of the watershed to Buttonwood Brook and Apponagansett Brook, their tributaries, or adjacent wetlands is generally less than 0.8 km (~2,600 ft) often with wetlands or

39 MASSACHUSETTS ESTUARIES PROJECT feeder streams extending to within hundreds of meters of the watershed boundaries. If it is assumed that groundwater discharge is the only mechanism for transfer of nitrogen and water to the main stem of each stream and that groundwater travels at approximately 1 ft/d (a common assumption in porous outwash or till materials), the areas furthest from the primary stream channels (close to the sub-watershed boundaries) would take approximately 7 years to reach the main stem. Since the groundwater system is constrained by underlying bedrock and USGS quadrangles show extensive tributaries feeding into the two brooks, flow to the Buttonwood Brook and Apponagansett Brook in the northern portions of the watershed must be 10 years or less from the outer edges of the watershed. In the southern portions of the watershed, most of the dense development is close to the Bay on the eastern side, while the western subwatersheds generally include extensive wetland systems that extend close to the watershed divides. These features suggest that most, if not all, of the watershed nitrogen load would reach the Bay in less than 10 years. Given that other MEP reviews in groundwater-dominated systems have shown that if most development is within 10 years or less, then the watershed and nitrogen loads are in relative balance with the estuary nitrogen concentrations, the MEP team has a high level of confidence that the present watershed nitrogen load appears to accurately reflect the present nitrogen sources to the estuaries (after accounting for natural attenuation discussed below).

In order to determine nitrogen loads from the watersheds, the MEP team utilizes detailed individual lot-by-lot data to develop nitrogen loads from most areas, while information developed from other detailed site-specific studies is applied to the remaining portions of the watershed. The Linked Watershed-Embayment Management Model (Howes and Ramsey, 2001) uses a land-use Nitrogen Loading Sub-Model based upon sub-watershed specific land uses and pre- determined nitrogen loading rates based on detailed source studies in southeastern Massachusetts. For the Apponagansett Bay Embayment System, the model uses land-use data from the City of New Bedford and the Town of Dartmouth. This land-use data is transformed to nitrogen loads using both regional nitrogen loading factors and local watershed- specific data (such as parcel-by-parcel water use). Determination of the nitrogen loads required obtaining watershed-specific information regarding wastewater (including municipal sewer connections), fertilizers, runoff from impervious surfaces and atmospheric deposition. The primary regional factors were derived for southeastern Massachusetts from direct measurements. The resulting nitrogen loads represent the “potential” or unattenuated nitrogen load to each receiving embayment, since attenuation during transport is included at a later stage.

Natural attenuation of nitrogen during transport from land-to-sea within the Apponagansett Bay watershed was determined based upon site-specific measurements of stream flow at gauges on Buttonwood Brook and Apponagansett Brook. The subwatersheds to these stream discharge points allowed comparison between field-collected data from the brooks and estimates from the nitrogen-loading sub-model. Stream flow and associated surface water attenuation is included in the MEP nitrogen attenuation and freshwater flow investigation, presented in Section IV.2. If smaller aquatic features that have not been included in this MEP analysis were providing additional attenuation of nitrogen, nitrogen loading to the estuary would only be slightly (~10%) overestimated given the distribution of nitrogen sources within the watershed.

Based upon the evaluation of the watershed system, the MEP Technical Team used the watershed Nitrogen Loading Model to estimate nitrogen loads for the sub-watersheds that directly discharge groundwater to the estuary without flowing through one of these interim stream measuring points. Internal nitrogen recycling was also determined throughout the tidal

40 MASSACHUSETTS ESTUARIES PROJECT reaches of the Apponagansett Bay Estuarine System; measurements were made to capture the spatial distribution of sediment nitrogen regeneration from the sediments to the overlying water- column. Nitrogen regeneration focused on summer months, the critical nitrogen management interval and the focal season of the MEP approach and application of the Linked Watershed- Embayment Management Model (Section IV.3).

IV.1.1 Land Use, Water Use, and Sewered Properties Database Preparation Since the watershed to the Apponagansett Bay includes portions of the City of New Bedford and the Town of Dartmouth, Estuaries Project staff obtained digital parcel and tax assessor’s data from these municipalities to serve as a base for the watershed nitrogen loading model. Digital parcels and land use/assessors data for New Bedford are from 2010, while Dartmouth data were also originally from 2010. These databases were linked with available water use and sewer account databases from the respective municipalities. The land use databases contain traditional information regarding land use classification based on MassDOR (2015) land use codes. Significant effort was made to reconcile and link all of the databases, including QA/QC by MEP staff to review incomplete entries in the datasets.

Figure IV-1 shows the land uses within the Apponagansett Bay estuary watershed areas based on the municipal assessor codes. Land uses in the study area are grouped into ten land use categories: 1) residential, 2) commercial, 3) industrial, 4) mixed use, 5) undeveloped, 6) agricultural, 7) recreational, 8) public service/government, including road rights-of-way, 9) parcels that are not classified by the assessors, and 10) freshwater features (e.g. ponds and streams). These land use categories, except for freshwater features and those parcels that are unclassified, are aggregations derived from the major categories in the land uses classification guidance to Massachusetts Assessors (MADOR, 2015). These categories are common to each municipality in the watershed. “Public service” properties in the MassDOR coding system are tax-exempt properties, including lands owned by government (e.g., wellfields, schools, open space, roads) and private non-profit groups like churches and colleges.

In the overall Apponagansett Bay System watershed, the predominant land use based on area is residential, which accounts for 49% of the overall watershed area (Figure IV-2). Residential land uses occupy most of the area in all of the subwatersheds too, which range from 43% (Buttonwood Brook) to 65% (Apponagansett Brook). Public service/ROW areas account for the second largest area (28% of the overall watershed area), followed by parcels classified as undeveloped (16%). Other land use categories are less than 5% of the total area. Public service/ROW areas range from 16% to 39% of the subwatershed land areas.

Parcel counts present a different perspective; residential parcels are the majority of parcels in almost all of the sub-watersheds and in all the groupings in Figure IV-2. Residential parcels are 83% of all parcels in the overall watershed and range between 42% and 89% of the total parcel counts in the seven subwatersheds. Undeveloped parcels are the second highest percentage of the watershed parcel counts, accounting for 11% of the parcel count for the entire watershed. This type of information provides a sense of how many potential land owners exist in each subwatershed portion and how information on watershed management strategies might be tailored to address predominant land uses and concerns. Review of average undeveloped lot size shows that the average undeveloped watershed parcel is 1.3 acres, but average parcels are much smaller in the Buttonwood Brook subwatershed (0.6 acres) and much larger in the Apponagansett Brook subwatershed (2.6 acres). Most of the undeveloped properties are in the Dike Marsh subwatershed, where they average 9.2 acres in area. These findings suggest that

41 MASSACHUSETTS ESTUARIES PROJECT

NEW BEDFORD DARTMOUTH 1

2

3 4

Land Use 7 (Municipal Assessors) Residential Commercial 5 6 Industrial Recreational Agricultural Undeveloped

Figure IV-1. Land-use in the Apponagansett Bay watershed. The watershed is split among Town of Dartmouth and the City of New Bedford. Land use classifications are based on municipal assessors’ records: 2014 for Dartmouth and 2009 for New Bedford.

42 MASSACHUSETTS ESTUARIES PROJECT

100%

16% 90% 23% 22% 21% Freshwater 28% Unclassified 80% 39% 19% Public Service/ROW 70% 14% 20% Undeveloped 25% 16% 60% Recreational 10% 4% 13% 4% Agricultural 50% 3% 4% Mixed Use 40% Industrial

65% Commercial 30% 53% 53% Residential 47% 49% 20% 43%

10%

0% Buttonwood Apponagansett Apponagansett Dike Marsh Apponagansett Whole System Brook Brook Bay Inner Bay Outer

Figure IV-2. Distribution of land-uses within the major sub-watersheds and whole watershed to the Apponagansett Bay estuary system. Only percentages greater than or equal to 3% are shown. 43

MASSACHUSETTS ESTUARIES PROJECT the majority of the undeveloped parcels in the Buttonwood Brook subwatershed are likely already subdivided and surrounded by developed parcels, while undeveloped parcels in the Dike Marsh subwatershed could be subdivided further. Further review of the residential parcels shows that single-family residences (MassDOR land use code 101) are 93% of residential parcels in the overall watershed and are 90% of the residential area in watershed.

MEP analyses generally use water use as a proxy for wastewater flows and these loads are adjusted for any sewer collection systems. In the Apponagansett Bay watershed, the Town of Dartmouth and the City of New Bedford provided municipal water use databases for individual parcels and sewer billing records were used to identify parcels with sewer connections. No private wastewater treatment facilities were identified in MassDEP databases. Wastewater-based nitrogen loading from the individual parcels using on-site septic systems is based upon the average water-use, nitrogen concentration, and consumptive loss of water before the remainder is treated in a septic system (see Section IV.1.2). The wastewater treatment facilities connected to the sewered parcels discharge their treated effluent outside of the Bay watershed, so these wastewater nitrogen loads are removed from the watershed.

One significant issue that was resolved during the development of the watershed nitrogen loading model was differences between the measured nitrogen loads at the two gauged brooks and the loads that were initially estimated based on the land use and sewer connections. Initial municipal sewer connections within the Dartmouth portion of the watershed were identified using the town’s sewer billing database. This database includes dates of sewer betterments and these dates were identified as a reasonable proxy for when connections to the sewer system occurred or, in other words, when septic system nitrogen loads were removed from the watershed. MEP measurements on Buttonwood Brook and Apponagansett Brook were collected in 2003-2004 (see Section IV.2). In order match this time period, MEP staff selected 2001 as the cutoff year for sewer connections; this year would account for a conservative estimated average two-year travel time to the brooks. This estimate was reasonable based on the distance from the edges of the subwatersheds to the brooks or their tributaries. However, the initial review of the sewer betterment database found that more than half of the ~7,200 sewered parcels had a placeholder betterment date of January 1, 2002 and that more than 2,500 of these within the Apponagansett Bay watershed. With the help of Town of Dartmouth staff, MEP staff reviewed individual parcel paper records maintained by the Department of Public Works and determined the actual date of sewer connection for 44% of the placeholder watershed parcels. This information was incorporated into the watershed nitrogen loading model. The resulting percentage of sewered parcels among those with confirmed betterment dates was then used for those parcels that still had unresolved placeholder betterment dates. MEP staff used these combined confirmed and estimated dates to determine the percentage of sewered properties within each subwatershed that had sewer connections in 2001 or before. Any parcels without a sewer bill are assumed to utilize a septic system for wastewater treatment and any developed parcel without a municipal water account is assumed to utilize a private well for drinking water. The resulting parcel/landuse/sewer database was then used to develop the MEP watershed nitrogen loading model.

IV.1.2 Nitrogen Loading Input Factors Wastewater/Water Use

The Massachusetts Estuaries Project septic system nitrogen loading rate is fundamentally based upon a per capita nitrogen load to the receiving aquatic system. Specifically, the MEP septic system wastewater nitrogen loading is based upon a number of studies and additional

44 MASSACHUSETTS ESTUARIES PROJECT information that directly measured septic system and per capita loads on Cape Cod or in similar geologic settings (Nelson et al.1990, Weiskel & Howes 1991, 1992, Koppelman 1978, Frimpter et al. 1990, Brawley et al. 2000, Howes and Ramsey 2000, Costa et al. 2001). Variation in per capita nitrogen load has been found to be relatively small, with average annual per capita nitrogen loads generally between 1.9 to 2.3 kg person-yr-1.

However, given the seasonal shifts in occupancy and rapid population growth throughout southeastern Massachusetts, decennial census data yields accurate estimates of total population only in selected watersheds. To correct for this uncertainty and more accurately assess current nitrogen loads, the MEP Technical Team employs a water-use approach. The water-use approach is applied on a parcel-by-parcel basis within a watershed, where annual water meter data is linked to assessor’s parcel information using GIS techniques. The parcel specific water use data is converted to septic system nitrogen discharges (to the receiving aquatic systems) by adjusting for consumptive use (e.g., irrigation) and applying a wastewater nitrogen concentration. The water use approach focuses on the nitrogen load, which reaches the aquatic receptors downgradient in the aquifer.

All nitrogen losses within a septic system are incorporated into the MEP analysis. For example, information developed on Title 5 septic systems at the MassDEP Massachusetts Alternative Septic System Test Center at the Massachusetts Military Reservation have shown nitrogen removals between 21% and 25%. Multi-year monitoring from the Test Center has revealed that nitrogen removal within the septic tank was small (1% to 3%), with most (20 to 22%) of the removal occurring within five feet of the soil adsorption system (Costa et al. 2001). Downgradient studies of septic system plumes indicate that further nitrogen loss during aquifer transport is negligible (Robertson et al. 1991, DeSimone and Howes 1996).

In its application of the water-use approach to septic system nitrogen loads, the MEP Technical Team has ascertained for the Estuaries Project region that while the per capita septic load is well constrained by direct studies, the consumptive use and nitrogen concentration data are less certain. As a result, the Technical Team has derived a combined term for an effective N Loading Coefficient (consumptive use times N concentration) of 23.63, to convert water (per volume) to nitrogen load (N mass). This coefficient uses a per capita nitrogen load of 2.1 kg N person-yr-1 and is based upon direct measurements and corrects for changes in concentration that result from per capita shifts in water-use (e.g. due to installing low plumbing fixtures or high versus low irrigation usage).

The nitrogen loads developed using this approach have been validated in a number of long and short term field studies where integrated measurements of nitrogen discharge from watersheds could be directly measured. Weiskel and Howes (1991, 1992) conducted a detailed watershed/stream tube study that monitored septic systems, leaching fields and the transport of the nitrogen in groundwater to adjacent Buttermilk Bay. This monitoring resulted in estimated annual per capita nitrogen loads of 2.17 kg (as published) to 2.04 kg (if new attenuation information is included). Further, modeled and measured nitrogen loads were determined for a small subwatershed to Mashapaquit Creek in West Falmouth Harbor (Smith and Howes, manuscript in review) where measured nitrogen discharge from the aquifer was within 5% of the modeled N load. Another evaluation was conducted by surveying nitrogen discharge to the Mashpee River in reaches with swept sand channels and in winter when nitrogen attenuation is minimal. The modeled and observed loads showed a difference of less than 8%, easily attributable to the low rate of attenuation expected at that time of year in this type of ecological situation (Samimy and Howes, unpublished data).

45 MASSACHUSETTS ESTUARIES PROJECT

While census based population data has limitations in the highly seasonal MEP region, part of the regular MEP analysis is to compare expected water used based on average residential occupancy to measured average water uses. This is performed as a quality assurance check to increase certainty in the final results. This comparison has shown that the larger the watershed the better the match between average water use and occupancy. For example, in the cases of the combined Great Pond, Green Pond and Bournes Pond watershed in the Town of Falmouth and the Popponesset Bay/Eastern Waquoit Bay watershed, which covers large areas and have significant year-round populations, the septic nitrogen loading based upon the census data is within 5% of that from the water use approach. This comparison matches some of the variability seen in census data itself. Census blocks, which are generally smaller areas of any given town, have shown up to a 13% difference in average occupancy from town-wide occupancy rates. These analyses provide additional support for the use of the water use approach in the MEP study region.

Overall, the MEP water use approach for determining septic system nitrogen loads has been both calibrated and validated in a variety of watershed settings. The approach: (a) is consistent with a suite of studies on per capita nitrogen loads from septic systems in sandy outwash aquifers; (b) has been validated in studies of the MEP Watershed “Module”, where there has been excellent agreement between the nitrogen load predicted and that observed in direct field measurements corrected to other MEP Nitrogen Loading Coefficients (e.g., stormwater, lawn fertilization); (c) the MEP septic nitrogen loading coefficient agrees in specific studies of consumptive water use and nitrogen attenuation between the septic tank and the discharge site; and (d) the watershed module provides estimates of nitrogen attenuation by freshwater systems that are consistent with a variety of ecological studies. It should be noted that while points b-d support the use of the MEP Septic N Coefficient, they were not used in its development. The MEP Technical Team has developed the septic system nitrogen load over many years, and the general agreement among the number of supporting studies has greatly enhanced the certainty of this critical watershed nitrogen loading term.

The independent validation of the water quality model (Section VI) and the reasonableness of the freshwater attenuation (Section IV.2) add additional weight to the nitrogen loading coefficients used in MEP analyses and a variety of other MEP embayments. While the MEP septic system nitrogen load is the best estimate possible, to the extent that it may underestimate the nitrogen load from this source reaching receiving waters provides a safety factor relative to other higher loads that are generally used in regulatory situations. The lower concentration results in slightly higher amounts of nitrogen mitigation (estimated at 1% to 5%)) needed to lower embayment nitrogen levels to a nitrogen target (e.g. nitrogen threshold, cf. Section VIII). The additional nitrogen removal is not proportional to the septic system nitrogen level, but is related to the how the septic system nitrogen mass compares to the nitrogen loads from all other sources that reach the estuary (i.e. attenuated loads).

In order to provide an independent validation of the average residential water use within the Apponagansett Bay watershed, MEP staff reviewed US Census population values for the City of New Bedford and the Town of Dartmouth. Previous evaluations of watersheds with available water use and sewer service areas have shown that water use provides a reasonable estimate of wastewater generation. The state on-site wastewater design regulations (i.e., 310 CMR 15, Title 5) assume that two people occupy each bedroom and each bedroom has a wastewater flow of 110 gallons per day (gpd), so for the purposes of Title 5 each person generates 55 gpd of wastewater. Based on data collected during the 2010 US Census, average occupancy within New Bedford and Dartmouth are 2.45 and 3.03 people per housing unit, respectively and these averages are approximately the same as during the 2000 US Census

46 MASSACHUSETTS ESTUARIES PROJECT

(2.46 and 2.91, respectively). Seasonal properties are a small component of the housing stock in each municipality, so potential variability associated with seasonal fluctuations should not be a significant concern. Year-round occupancies in the municipalities were both 90% in 2010 and 92% and 94%, respectively in 2000. Based on available water use data within the watershed, the average single family residence water use is 153 gpd. If this flow is then divided by 55 gpd, the average estimated occupancy within the watershed based on the water use is 2.77 people per household. This occupancy is reasonably within the range of the town-wide averages. These comparisons provide confidence in the water use information and show that water use is an appropriate basis for determining septic system wastewater nitrogen loads within Apponagansett Bay watershed.

Water use information exists for 99% of the 5,011 developed parcels in the Apponagansett Bay watershed. Developed parcels without water use accounts are assumed to utilize private wells for drinking water. These are properties classified with land use codes that should be developed (e.g., 101 or 325) and have been confirmed as having buildings on them through a review of aerial photographs, but do not have a listed account in the water use databases. Of the 68 developed parcels without water use accounts, 50 (74%) are classified as single-family residences (land use code 101). These parcels are assumed to utilize private wells and are assigned the MEP Apponagansett Bay watershed average water use of 153 gpd in the watershed nitrogen loading modules. Another 11 developed parcels without water use are parcels classified as other types of residential properties (e.g., multi-family or condominiums). These parcels are assumed to utilize private wells and are assigned the watershed average water use of 224 gpd for residential properties that are not single family residences.

Nitrogen Loading Input Factors: Fertilized Areas

The second largest source of estuary watershed nitrogen loading is usually fertilized areas: lawns, golf courses, and agricultural land uses (e.g., cranberry bogs, row crops, etc.). Residential lawns are usually the predominant source within this category. In order to add this source to the watershed nitrogen loading model for Apponagansett Bay, MEP staff reviewed available information about residential lawn fertilizing practices and incorporated site-specific information to determine nitrogen loading from other fertilization applications in the watershed. Within the watershed, MEP staff reviewed available regional information about residential lawn fertilizing practices. The primary site-specific information in this watershed is for crop nitrogen loads, which were determined based on previous studies conducted in southeastern Massachusetts. Other site-specific fertilized areas information includes turf for the Country Club of New Bedford, the only golf course within the watershed, and playing fields at Bishop Stang High School.

Residential lawn fertilizer use has rarely been directly measured in watershed-based nitrogen loading investigations. Instead, lawn fertilizer nitrogen loads have been estimated based upon a number of assumptions: a) each household applies fertilizer, b) cumulative annual applications are 3 pounds per 1,000 sq. ft., c) each lawn is 5000 sq. ft., and d) only 25% of the nitrogen applied reaches the groundwater (leaching rate). Because many of these assumptions had not been rigorously reviewed in over a decade, the MEP Technical Staff undertook an assessment of lawn fertilizer application rates and a review of leaching rates for inclusion among the standard factors used in the Watershed Nitrogen Loading Sub-Model.

The initial effort in this assessment was to determine nitrogen fertilization rates for residential lawns in the Towns of Falmouth, Mashpee and Barnstable (White, 2003). This

47 MASSACHUSETTS ESTUARIES PROJECT assessment, which was completed prior to the start of the MEP, accounted for proximity to fresh ponds and embayments. Based upon ~300 interviews and over 2,000 site surveys, a number of findings emerged: 1) average residential lawn area is ~5000 sq. ft., 2) half of the residences did not apply lawn fertilizer, and 3) the weighted average application rate was 1.44 applications per year, rather than the 4 applications per year recommended on standard fertilizer bags. Integrating the average residential fertilizer application rate with a leaching rate of 20% results in a fertilizer contribution of N to groundwater of 1.08 lb N per residential lawn; these factors are used in the MEP nitrogen loading calculations. A similar survey within the Town of Orleans in 2003 found similar residential lawn fertilizer practices (Howes and White, 2005). It should also be noted that a recent data review of lawn fertilizer leaching in settings similar to those on Cape Cod confirmed that the 20% leaching rate is reasonable (HWG, 2009). It is likely that these rates still represent a conservative estimate of nitrogen load from residential lawns. It should also be noted that professionally maintained lawns in the three town survey were found to have the higher rate of fertilizer application and hence higher estimated annual contribution to groundwater of 3 lb/lawn/yr.

Project staff also determined the areas of fertilized agricultural fields and golf courses and consulted with municipal staff on appropriate nitrogen loading rates. Agricultural fields were generally assigned crop nitrogen application rates based on rates determined for various Massachusetts land use codes (previously used other MEP assessments). Staff were unsuccessful in contacting a staff person at the Country Club of New Bedford who could provide course-specific fertilizer application rates, so the watershed nitrogen loading model utilized averages developed for tees, greens, fairways, and roughs from 19 other courses of contacted through a number of previous MEP watershed analyses. Turf areas within the golf course were determined based on review of aerial photographs. Approximately 1 acre of fairway and tee is located within the Buttonwood Brook subwatershed.

Nitrogen Loading Input Factors: Freshwater Wetlands

The data collected at the MEP gauge sites to Buttonwood Brook and Apponagansett Brook watersheds generally produced measured nitrogen loads that were higher than what the preliminary MEP watershed nitrogen loading model indicated. Since the MEP assessment approach is data-driven, MEP staff began the process of exploring the cause of these higher nitrogen loads by re-reviewed all of the data leading to the preliminary watershed loads, including the watershed delineations, the nitrogen loading inputs, and re-reviewing the streamflow and concentration data (see Section IV.2). These steps confirmed the evaluations and suggested that there was another nitrogen source in the Apponagansett Bay watershed that was not included in the preliminary model. A similar approach was taken in the Slocums River, Westport River, and Nasketucket Bay systems, where MEP staff identified extensive wetland and swamp lands surrounding most of the streams and rivers feeding into the estuaries as the most likely cause of the high nitrogen loads.

The nitrogen load assigned to freshwater wetlands bordering the freshwater portions of the Buttonwood Brook and Apponagansett Brook and their tributaries is consistent with the nitrogen loading assigned to the freshwater wetlands bordering the Slocums/Paskamansett River (Howes, et al., 2008) and the Westport River (Howes, et al., 2012). It was clear from both the Westport River and Paskamansett River measurements that attenuation of nitrogen in their riverine wetlands was relatively low, with added nitrogen being transformed, but not removed. Specifically, the indication of wetland “N saturation” is based up atmospheric N deposition being only partially removed. This determination for the Apponagansett Bay brooks is reasonable based on the available similarly determined measurements for the two larger rivers, as well as

48 MASSACHUSETTS ESTUARIES PROJECT observation developed during the assessment of other smaller freshwater systems in similar settings.

The Westport River, Paskamansett River, and Acushnet River have similar geology, are structurally similar, have significant associated freshwater wetlands (especially the Acushnet River stream), and are highly nitrogen-enriched (TN’s of 1.3 mg/L, 1.2 mg/L and 1.1 mg/L, respectively). The geology of these systems (particularly the Westport and Paskamansett Rivers) differs from those on Cape Cod and the Islands in that the watersheds tend to be topographically defined and have exposed bedrock and till, whereas the systems from the Wareham River and to the east are primarily glacial sands and moraines with watersheds defined by differences in water table elevations within the porous matrix. As such, based on site-specific analysis that is the foundation of the MEP, some of the factors associated with surface water systems in the Westport River, Paskamansett River, Acushnet River, and Buttonwood Brook and Apponagansett Brook watersheds are different than those developed for the Cape Cod and the Islands watersheds.

In most of the Cape Cod streams, application of the MEP N loading approach has produced very good agreement with measured stream nitrogen loads. These sandy aquifer- dominated systems typically support limited freshwater wetland areas and much lower stream flows than found in western Buzzards Bay streams. In these Cape Cod systems, nitrogen attenuation rates of 20% to 30% are typical, possibly due to higher watershed retention times. In contrast, the rivers along the northwestern edge of the Buzzards Bay watershed are underlain by bedrock and till, have comparatively high stream flows, and extensive bordering freshwater wetlands. In these western Buzzards Bay systems, nitrogen contact time in the wetlands will be shorter and, like freshwater ponds with short residence times, they should attenuate less nitrogen. In addition, reviews of river wetlands have indicated that they have threshold effects like those seen in estuaries and ponds. This means that these freshwater wetlands can become nearly completely loaded with nitrogen and once in that condition act as transformers of nitrogen (changing nitrate+nitrite to organic forms), but not attenuators of nitrogen (e.g., USDA, 2011). This change appears to be related to the amount of nitrogen received, as well as inter- related factors such as hydraulic residence time, temperature, plant surface coverage, and plant density (e.g., Hagg et al., 2011; Kröger, et al., 2009; Alexander, et al., 2008). It is important to note that the wetlands are not actually a nitrogen source, but they merely have a lower rate of nitrogen removal of the nitrogen deposited upon them, than in the smaller, lower flow wetlands.

The MEP results are also consistent with studies by other researchers that found the ability of river wetlands to attenuate nitrogen is directly related to their hydraulic residence time with longer residence times resulting in greater nitrogen reduction (e.g., Jansson, et al., 1994; Perez, et al., 2011; Toet, et al., 2005). Direct data in the overall MEP study area generally confirms this relationship with lower flow/longer residence time streams on the eastern portion of the overall MEP study area having greater nitrogen attenuation, as well as attenuation in ponds and lakes, which have even longer residence times, having nitrogen attenuation rates of 50% or higher (e.g., Howes, et al., 2006).

In order to incorporate the nitrogen loading from the wetland areas in the Apponagansett Bay watershed, MEP staff assigned the water surface nitrogen loading factor to the wetland areas identified in a MassGIS/MassDEP wetland coverage (Figure IV-3). The wetlands are interpreted from 1:12,000 scale, stereo color-infrared photography captured over a series of years between 1990 and 2000 (MassGIS, 2009). For the purposes of the MEP assessment, the treatment of these wetlands as water surfaces is appropriately conservative without further data to refine the spatial differences in residence times, plant communities/densities and the role of

49 MASSACHUSETTS ESTUARIES PROJECT seasonal impacts along the various streams and rivers in the Apponagansett Bay watershed system.

NEW BEDFORD DARTMOUTH

Buzzards Bay

Figure IV-3. Wetland areas in the Apponagansett Bay watersheds. All areas colored in green are wetlands areas delineated by MassGIS/MassDEP 1:12,000K coverage. Most of these areas are associated with freshwater streams that discharge into the Bay. All these areas were assigned a surface water nitrogen load in the MEP watershed nitrogen loading model.

50 MASSACHUSETTS ESTUARIES PROJECT

Nitrogen Loading Input Factors: Other

MEP staff also assigned nitrogen loads based on the number of farm animals within the watersheds. Farm animals on individual parcels were provided by Town of Dartmouth Board of Health (personal communication, Wendy Henderson, 11/12). The provided list identified over 1,200 animals on 35 parcels. Details of the animal counts are available in the MEP Data Disk that accompanies this report.

Staff also contacted staff at the Buttonwood Park Zoo to estimate potential nitrogen loads from animal manure (personal communication, L. Garabaldi, Zoo Director, 9/07). No fertilizer was used on the zoo site and animal manure was transported to a large dumpster on a daily basis. The dumpster was not covered prior to 2007 and MEP staff estimated a nitrogen load based on the watershed recharge rate from the dumpster area and average nitrogen in runoff released from a manure pile at the Woodland Park (WA) zoo (E&A Environmental Consultants, 2000). This load is removed in the buildout scenario based on current practices to regularly cover the manure dumpster.

The nitrogen loading factors for atmospheric deposition, impervious surfaces and natural areas in the Apponagansett Bay assessment are from the MEP Embayment Modeling Evaluation and Sensitivity Report (Howes and Ramsey 2001). The factors are similar to those utilized by the Cape Cod Commission Nitrogen Loading Technical Bulletin (Eichner and Cambareri, 1992) and the MassDEP Nitrogen Loading Computer Model Guidance Document (1999). The recharge rate for natural areas and lawn areas is the same as utilized in the watershed modeling effort (Section III). Factors used in the MEP nitrogen loading analysis for the Apponagansett Bay watershed are summarized in Table IV-1.

For impervious surfaces, MEP reviewed a number of different sources and selected the most reliable sources. Road areas are based on MassHighway GIS information, which provides road width for various road segments. MEP staff utilized the GIS to sum these segments and their various widths by sub-watershed. Project staff also checked this information against parcel-based rights-of-way. Town assessor’s databases for both watershed towns also include parcel-specific building footprint information, which was used for roof areas. MEP impervious surface nitrogen loading factors were applied to these road and roof areas.

IV.1.3 Calculating Nitrogen Loads Once all the land and water use information was linked to the parcel coverages, parcels were assigned to various watersheds based initially on whether at least 50% or more of the land area of each parcel was located within a respective watershed. Following the initial assigning of boundary parcels, all large parcels were examined individually and were split (as appropriate) in order to obtain less than a 2% difference between the total land area of each sub-watershed based on the watershed delineations and the sum of the area of the parcels within each sub- watershed.

The review of individual parcels straddling watershed boundaries included corresponding reviews and individualized assignment of nitrogen loads associated with lawn areas, septic systems, and impervious surfaces. Individualized information for parcels with atypical nitrogen loading (condominiums, golf courses, etc.) was also assigned at this stage. It should be noted that small shifts in nitrogen loading due to the above assignment procedure generally have a negligible effect on the total nitrogen loading to the Apponagansett Bay estuary. The

51 MASSACHUSETTS ESTUARIES PROJECT assignment effort was undertaken to better define sub-estuary loads and enhance the use of the Linked Watershed-Embayment Model for the analysis of management alternatives.

Following the assignment of all parcels, sub-watershed modules were generated for each of the seven sub-watersheds summarizing water use, parcel area, frequency, sewer connections, private wells, and road area. The individual sub-watershed modules were then integrated to create an Apponagansett Bay Watershed Nitrogen Loading module with summaries for each of the individual sub-embayments and sub-estuaries. The sub- embayments represent the functional embayment units for the Linked Watershed-Embayment Model’s estuary water quality component.

Table IV-1. Primary Nitrogen Loading Factors used in the Apponagansett Bay MEP analyses. General factors are from MEP modeling evaluation (Howes & Ramsey 2001). Site-specific factors are derived from watershed-specific data. Nitrogen Concentrations: mg/l Recharge Rates:2 in/yr Road Run-off 1.5 Impervious Surfaces 45.7 Roof Run-off 0.75 Natural and Lawn Areas 30.46 Direct Precipitation on 1.09 Water Use/Wastewater: Embayments and Ponds Natural Area Recharge 0.072 Existing developed parcels Wastewater Coefficient 23.63 wo/water accounts and buildout 153 gpd3 Fertilizers: single family residence parcels

Average Residential Lawn Size 4 5,000 Multi-family residential parcels 224 gpd (sq ft)1 Residential Watershed Nitrogen Existing developed parcels Measured annual 1.08 Rate (lbs/lawn)1 w/water accounts: water use Nitrogen leaching rate 20% Commercial and Industrial buildout additions5 Crops kg/ha/yr Commercial Wastewater flow Hay, Pasture 5 95 (gpd/1,000 ft2 of building): Hay, Pasture leaching rate 100%7 Building lot coverage: 18% Corn, Vegetables, Vineyard, Fruit 34 Crop N leaching rate 30% Industrial Wastewater flow Farm Animals kg/yr/animal 67 (gpd/1,000 ft2 of building): Horse 32.4 Building lot coverage: 17% Cow/Steer 55.8 Goats/Sheep 7.3 Hogs 14.5 Chickens 0.4 Animal N leaching rate 40% Notes: 1) Data from MEP lawn studies in Falmouth, Mashpee & Barnstable 2001 and Orleans 2005. 2) Based on precipitation rate of 50.77 inches per year (1971-2000 NOAA average for closest long-term precipitation gauge (New Bedford)) 3) Apponagansett Bay watershed average from single-family residences with water use accounts 4) Apponagansett Bay watershed average from other residential properties with water use accounts 5) Based on characteristics of respective land uses within two watershed municipalities 6) Hay, Pasture leaching rate is 100% because the assigned nitrogen loading rate already incorporates a leaching rate

52 MASSACHUSETTS ESTUARIES PROJECT

For management purposes, the aggregated watershed nitrogen loads are partitioned by the major types of nitrogen sources in order to focus development of nitrogen management alternatives. Within the Apponagansett Bay study area, the major types of nitrogen loads are: wastewater (e.g., septic systems), fertilizers (and including the Buttonwood Zoo loading), farm animals, freshwater wetlands, impervious surfaces, direct atmospheric deposition to water surfaces, and recharge within natural areas (Table IV-2). The output of the watershed nitrogen- loading model is the annual mass (kilograms) of nitrogen added to the contributing area of component sub-embayments, by each source category (Figure IV-4 a-d). The annual watershed nitrogen input is then reduced by natural nitrogen attenuation in Buttonwood Brook and Apponagansett Brook during transport and the estuary receives this reduced load. The nitrogen loads used in the MEP embayment water quality sub-model are a combination of the estimated loads in Table IV-2 and the measured loads from the brooks discussed in Section IV.2.

Buildout

Part of the regular MEP watershed nitrogen loading modeling is to prepare a buildout assessment (or scenario) of potential development and accompanying nitrogen loads within the study area watersheds. The MEP buildout is relatively straightforward and is generally completed in four steps: 1) each residential parcel classified by the town assessor as developable is identified and divided by minimum lot sizes specified in town zoning and the resulting number of new residential units is rounded down, 2) parcels classified as developable commercial and industrial parcels by the town assessor are identified, 3) residential, commercial and industrial parcels with existing development and areas greater than twice zoning’s minimum lot size are identified, divided by the minimum lot size and the resulting number of new units is rounded down, and 4) results are discussed with local staff and/or planning board members and the analysis results are modified based on local knowledge. The final step can include discussions with municipal staff, regional government entities, and/or non-governmental advocacy groups.

It should be noted that the initial MEP buildout approach is relatively simple and does not include any modifications/refinements for lot line setbacks, wetlands, road construction, frontage requirements, parcel shape requirements, or other more detailed zoning provisions. The MEP buildout approach also does not include potential impacts associated with the higher densities usually associated with 40B affordable housing projects. The discussions with local planners in the fourth step usually leads to additional insights on developments that are planned, especially developments planned on government or public service parcels, and updates to assessor classifications, including lands purchased by the town as open space. This final step may lead to removal and/or additions to the number of parcels initially identified as developable and may include application of more detailed zoning provisions.

As an example of how the MEP buildout approach might apply, assume an 81,000 square foot lot is classified by the town assessor as a developable residential lot (MassDOR land use code 130). This lot is divided by the 40,000 square foot minimum lot size specified in municipal zoning and the result is rounded down to two. As a result, two additional residential lots would be added to the subwatershed in the MEP buildout scenario. This addition could then be modified during discussion of municipal staff.

53 MASSACHUSETTS ESTUARIES PROJECT

Table IV-2. Apponagansett Bay Watershed Nitrogen Loads. Attenuation of system nitrogen loads occurs within Buttonwood Brook and Apponagansett Brook during transport to the estuary and attenuation rates are assigned based on MEP measured loads. All values are kg N yr-1. Apponagansett Bay N Loads by Input (kg/y): Present N Loads Buildout N Loads % of Water Body Watershed Farm Impervious River "Natural" Pond UnAtten N Atten Atten N UnAtten N Atten Atten N Wastewater Fertilizers Surface Buildout ID# Animals Surfaces Wetlands Surfaces Outflow Load % Load Load % Load Area Name Apponagansett Harbor System 11291 4094 566 4477 3217 3219 686 2560 27550 25755 27719 25851 Buttonwood Brook 1 4334 1999 136 2691 1029 8 190 1069 10387 11% 9245 10672 11% 9498 Apponagansett Brook Gaged 2 1058 200 1 291 152 0 62 287 1764 37% 1111 1877 37% 1182 Apponagansett Bay Inner E 3 2747 573 125 765 227 0 76 286 4514 4514 4355 4355 Apponagansett Bay Inner W 4 1997 874 163 435 738 37 137 253 4379 4379 4194 4194 Dike Marsh 5 580 336 140 152 1068 0 186 600 2461 2461 2907 2907 Apponagansett Bay Outer W 6 372 92 0 88 3 0 22 5 576 576 316 316 Apponagansett Bay Outer E 7 204 21 0 56 0 0 14 61 294 294 225 225 Dike Marsh Estuary Surface 213 213 213 213 213 Apponagansett Bay Outer W Estuary Surface 26 26 26 26 26 Apponagansett Bay Inner Estuary Surface Area 1633 1633 1633 1633 1633 Apponagansett Bay Outer Estuary Surface Area 1301 1301 1301 1301 1301 54

MASSACHUSETTS ESTUARIES PROJECT

Wastewater 2% Fertilizers

15% Farm Animals 16% 20%

Impervious 41% 3% Surfaces 55% 12%

River Wetlands 22%

12% Water Body Surface Area 2% "Natural" Overall Load Surfaces Local Control Load a. Apponagansett Harbor Whole Estuary System

Wastewater

1% Fertilizers

19% Farm Animals 22% 2%

Impervious 42% 26% Surfaces 47% 29% River Wetlands

10% Water Body Surface Area 2% 0% "Natural" Overall Load Local Control Load Surfaces b. Buttonwood Brook Figure IV-4 (a-b). Land use-specific unattenuated nitrogen load (by percent) to the (a) overall Apponagansett Bay Estuary System watershed and (b) Buttonwood Brook. “Overall Load” is the total nitrogen input within the watershed, while the “Local Control Load” represents only those nitrogen sources that could potentially be under local regulatory control.

55 MASSACHUSETTS ESTUARIES PROJECT

Wastewater

Fertilizers

Farm Animals 11% 0% 60% 68% Impervious 13% Surfaces 17% 0%

River Wetlands 19% 9% Water Body Surface Area 3% 0% "Natural" Overall Load Local Control Load Surfaces c. Apponagansett Brook

Wastewater

Fertilizers 4% 15% Farm Animals

20% Impervious 12% 61% Surfaces 48% 4% River Wetlands 17% 15%

Water Body Surface Area 0% 4% "Natural" Overall Load Local Control Load Surfaces d. Apponagansett Bay Lower (subwatersheds 3-7)

Figure IV-4 (c-d). Land use-specific unattenuated nitrogen load (by percent) to the (c) Apponagansett Brook and (d) lower Bay subwatersheds (watersheds 3-7). “Overall Load” is the total nitrogen input within the watershed, while the “Local Control Load” represents only those nitrogen sources that could potentially be under local regulatory control.

56 MASSACHUSETTS ESTUARIES PROJECT

Other provisions of the MEP buildout assessment include town assessor classification of undevelopable lots, standard treatment of commercial and industrial properties, and assumptions for lots less than the minimum areas specified by zoning. Properties classified by the Town of Dartmouth or City of New Bedford assessors as “undevelopable” (e.g., MassDOR codes 132, 392, and 442) are not assigned any development at buildout (unless revised by the municipal review). Commercial and industrial properties classified as developable are not subdivided; the area of each parcel and the factors in Table IV-1 are used to determine an estimated building size and wastewater flow for these properties. Pre-existing lots classified by the municipal assessor as developable are also treated as developable even if they are less than the minimum lot size specified in zoning; so, for example, a 10,000 square foot lot classified by the town assessor as a developable residential property (MassDOR 130 land use code) and located in a zoning area with a 40,000 square feet minimum lot size will be assigned an additional residential dwelling in the MEP buildout scenario. Most municipal zoning bylaws have a lower minimum lot size for pre-existing lots (usually 5,000 square feet) that will minimize instances of regulatory takings. Existing developed residential properties that are larger than zoning’s minimum lot sizes are also assigned additional development potential only if enough area is available to accommodate at least one additional lot as specified by the zoning minimum. Also in MEP buildouts, agricultural lands, Chapter 61 open space, town preserved open space, recreational areas and most other land uses other than residential, commercial and industrial properties are assumed to remain under their current use unless this is modified through discussions with municipal staff.

Discussions with municipal planners, boards, and/or wastewater consultants can generate some additional insights on planned development, and often include discussion of developments planned for government or public service parcels, and updates to assessor classifications, including lands purchased by the town as open space. Refinements of the MEP buildout can continue as municipalities within the watersheds begin nitrogen management planning and could include updates on parcels initially identified as developable or undevelopable and application of more detailed zoning provisions. As planning proceeds municipalities may request additional refined buildout scenarios to account for specific land-use shifts or projects that may be deemed likely within the watershed.

Following the completion of the initial buildout assessment for the Apponagansett Bay watersheds, MEP staff contacted and shared the results with Town of Dartmouth and City of New Bedford officials. Dartmouth Environmental Affairs Coordinator and Town Planner reviewed and adjusted the preliminary watershed buildout results in April 2012. All adjustments provided by the municipalities were incorporated into the MEP buildout.

All the parcels with additional buildout potential within the Apponagansett Bay watershed under the MEP buildout scenario are shown in Figure IV-5 and details for individual parcels are included in the MEP Data Disk that accompanies this report. The MEP buildout scenario for the Apponagansett Bay watershed includes 864 additional residential units (531 with sewer connections), 79,701 square feet of commercial buildings (83% without sewer), and 96,812 square feet of industrial buildings (all sewered). Each additional residential, commercial, or industrial property added at buildout is assigned respective nitrogen loads that depend on the impervious surfaces (roof and driveway), wastewater treatment (whether or not the parcel is connected to a municipal sewer system), and lawn fertilizer additions (only for residential unit additions). All properties not connected to the sewers are assumed to utilize Title 5 on-site septic systems for wastewater treatment. The buildout scenario also utilizes Dartmouth sewer connections as of 2014 and New Bedford sewer connections as of 2009. Cumulative

57 MASSACHUSETTS ESTUARIES PROJECT

Buildout Potential Developed Residential with Additional Development NEW Potential BEDFORD Developable Residential 1 Developed DARTMOUTH Commercial with Additional Development Potential Developable Commercial 2 Developed Industrial with Additional Development Potential Developable Industrial

3 4

7 5 6

Figure IV-5. Developable Parcels in the Apponagansett Bay watershed. Parcels that are shown are either parcels with no existing development but classified by the respective town assessors as developable or parcels with existing development, but potential for additional development based on minimum lot sizes specified in respective town zoning regulations. Buildout assessments include changes based on review by municipal staff.

58 MASSACHUSETTS ESTUARIES PROJECT unattenuated and attenuated buildout loads are indicated in separate columns in Table IV-2. As a result of sewer connections that have occurred between 2001 (the baseline “existing” conditions) and 2014, the buildout additions within the Apponagansett Bay system watershed only increase the unattenuated nitrogen loading rate by 1%.

IV.2 ATTENUATION OF NITROGEN IN SURFACE WATER TRANSPORT

IV.2.1 Background and Purpose Modeling and predicting changes in coastal embayment nitrogen related water quality is based, in part, on determination of the inputs of nitrogen from the surrounding contributing land or watershed. This watershed nitrogen input parameter is the primary term used to relate present and future loads (build-out, sewering analysis, enhanced flushing, pond/wetland restoration for natural attenuation, etc.) to changes in water quality and habitat health. Therefore, nitrogen loading is the primary threshold parameter for protection and restoration of estuarine systems. Rates of nitrogen loading to the sub-watersheds of the Apponagansett Bay system being investigated under this nutrient threshold analysis was based upon the delineated watersheds (Section III) and their land-use coverages (Section IV.1). If all of the nitrogen applied or discharged within a watershed reaches an embayment the watershed land-use loading rate represents the nitrogen load to the receiving waters. This condition exists in watersheds where nitrogen transport from source to estuarine waters is through groundwater flow in sandy outwash aquifers as well as stratified drift deposits (such as the developed region of the watersheds to this Town of Dartmouth system). The lack of nitrogen attenuation in these aquifer systems results from the lack of biogeochemical conditions needed for supporting nitrogen sorption and denitrification. However, in most watersheds in southeastern Massachusetts, nitrogen passes through a surface water ecosystem (pond, wetland, stream) on its path to the adjacent embayment. Surface water systems, unlike sandy aquifers, do support the needed conditions for nitrogen retention and denitrification. The result is that the mass of nitrogen passing through lakes, ponds, streams and marshes (fresh and salt) is diminished by natural biological processes that represent removal (not just temporary storage). However, this natural attenuation of nitrogen load is not uniformly distributed within the watershed, but is associated with ponds, streams and marshes. In the case of the Dartmouth embayment system watershed for Apponagansett Bay, a portion of the freshwater flow and transported nitrogen passes through several surface water systems (Buttonwood Brook and Apponagansett Brook) prior to entering estuarine system, producing the opportunity for significant nitrogen attenuation.

Failure to determine the attenuation of watershed derived nitrogen overestimates the nitrogen load to receiving estuarine waters. If nitrogen attenuation is significant in one portion of a watershed and insignificant in another the result is that nitrogen management would likely be more effective in achieving water quality improvements if focused on the watershed region having unattenuated nitrogen transport (other factors being equal). In addition to attenuation by freshwater ponds (see Section IV.1.3, above), attenuation in surface water flows is also important. An example of the significance of surface water nitrogen attenuation relating to embayment nitrogen management was seen in the Agawam River, where >50% of nitrogen originating within the upper watershed was attenuated prior to discharge to the Wareham River Estuary (CDM 2000). Similarly, MEP analysis of the Quashnet River indicates that in the upland watershed, which has natural attenuation predominantly associated with riverine processes, the integrated attenuation was 39% (Howes et al. 2004). In addition, a preliminary study of Great, Green and Bournes Ponds in Falmouth, measurements indicated a 30% attenuation of nitrogen during stream transport (Howes and Ramsey 2001). An example where natural attenuation played a significant role in nitrogen management can be seen relative to West Falmouth Harbor

59 MASSACHUSETTS ESTUARIES PROJECT

(Falmouth, MA), where ~40% of the nitrogen discharge to the Harbor originating from the groundwater effluent plume emanating from the WWTF was attenuated by a small salt marsh prior to reaching Harbor waters. Clearly, proper development and evaluation of nitrogen management options requires determination of the nitrogen loads reaching an embayment, not just loaded to the watershed.

Given the importance of determining accurate nitrogen loads to embayments for developing effective management alternatives and the potentially large errors associated with ignoring natural attenuation, direct integrated measurements of upper watershed attenuation were undertaken as part of the MEP Approach. MEP conducted long-term measurements of natural attenuation relating to surface water discharges to the head of the Apponagansett Bay embayment system considered in this report in addition to the natural attenuation measures by fresh kettle ponds, addressed above (Section IV.1). These additional site-specific studies were conducted in the 2 major surface water flow systems in the watersheds to the head of Apponagansett Bay, 1) Buttonwood Brook and 2) Apponagansett Brook (sometimes referred to as Vincent Brook) as depicted in Figure IV-6a,b.

Quantification of watershed based nitrogen attenuation is contingent upon being able to compare nitrogen load to the embayment system directly measured in freshwater stream flow (or in tidal marshes, net tidal outflow) to nitrogen load as derived from the detailed land use analysis (Section IV.1). Measurement of the flow and nutrient load associated with the freshwater streams discharging to the estuary provides a direct integrated measure of all of the processes presently attenuating nitrogen in the contributing area up-gradient from the various gaging sites. Flow and nitrogen load were measured at the gauges in each freshwater stream site for between 16 and 24 months of record depending on the stream gauging location (Figures 6a-6b, Figures 7-8). During each study period, velocity profiles were completed on each creek every month to two months and sometimes more frequently. The summation of the products of stream subsection areas of the stream cross-section and the respective measured velocities represent the computation of instantaneous stream flow (Q).

Determination of stream flow at each gauge was calculated and based on the measured values obtained for stream cross sectional area and velocity. Stream discharge was represented by the summation of individual discharge calculations for each stream subsection for which a cross sectional area and velocity measurement were obtained. Velocity measurements across the entire stream cross section were not averaged and then applied to the total stream cross sectional area.

The formula that was used for calculation of stream flow (discharge) is as follows:

Q = (A * V) where by:

Q = Stream discharge (m3/s) A = Stream subsection cross sectional area (m2) V = Stream subsection velocity (m/s)

Thus, each stream subsection will have a calculated stream discharge value and the summation of all the sub-sectional stream discharge values will be the total calculated discharge for the stream.

60 MASSACHUSETTS ESTUARIES PROJECT

Buttonwood Brook

Apponagansett Brook

Figure IV-6a. Locations of the two stream gaging sites in the Apponagansett Bay watershed. Red line is the watershed boundaries to Buttonwood Brook and Apponagansett Brook. The blue dashed line is the overall watershed boundary for the Apponagansett Bay system.

Periodic measurement of flows over the entire stream gauge deployment period allowed for the development of a stage-discharge relationship (rating curve) that could be used to obtain flow volumes from the detailed record of stage measured by the continuously recording stream gauges. Water level data obtained every 10-minutes was averaged to obtain hourly stages for a given river. These hourly stages values where then entered into the stage-discharge relation to compute hourly flow. Hourly flows were summed over a period of 24 hours to obtain daily flow and further, daily flows summed to obtain annual flow. In the case of tidal influence on stream stage, the diurnal low tide stage value was extracted on a day-by-day basis in order to resolve the stage value indicative of strictly freshwater flow. The two low tide stage values for any given day were averaged and the average stage value for a given day was then entered into the stage – discharge relation in order to compute daily flow. A complete annual record of stream flow (365 days) was generated for the surface water discharges flowing into the Slocums River system.

61 MASSACHUSETTS ESTUARIES PROJECT

Figure IV-6b. Location of stream gauges (red symbols) in the Apponagansett Bay embayment system.

The annual flow record for the surface water flow at each gauge was merged with the nutrient data set generated through the weekly water quality sampling performed at the gauge locations to determine nitrogen loading rates to the head of the Apponagansett Bay system. Nitrogen discharge from the streams was calculated using the paired daily discharge and daily nitrogen concentration data to determine the mass flux of nitrogen through a specific gauging site. For each of the stream gauge locations, water samples were generally collected weekly (at low tide for a tidally influenced stage) in order to determine nutrient concentrations from which nutrient load was calculated. Grab samples were collected at weekly intervals from November 1

62 MASSACHUSETTS ESTUARIES PROJECT through May 1 and at twice-weekly intervals between May 1 and October 30. In order to pair daily flows with daily nutrient concentrations, interpolation between weekly nutrient data points was necessary. These data are expressed as nitrogen mass per unit time (kg/d) and can be summed in order to obtain weekly, monthly, or annual nutrient load to the embayment system as appropriate. Comparing these measured nitrogen loads based on stream flow and water quality sampling to predicted loads based on the land use analysis allowed for the determination of the degree to which natural biological processes within the watershed to each pond currently reduces (percent attenuation) nitrogen loading to the embayment system.

IV.2.2 Surface water Discharge and Attenuation of Watershed Nitrogen: Stream Discharge from Buttonwood Brook to the head of Apponagansett Bay At the Buttonwood Brook gauge site (gauge located at Russells Mills Road crossing), a continuously recording vented calibrated water level gauge was installed to yield the level of water in the discharge that carries nitrogen load from the wetland to the head of the Apponagansett Bay estuary. To confirm that freshwater was being measured, salinity measurements were conducted on the weekly water quality samples collected from the gauge site. Average salinity was determined to be no greater than 0.1 ppt. Based on the salinity, the gauge location was deemed acceptable for making freshwater flow measurements at low tide. Calibration of the gauge was checked monthly. The gauge on Buttonwood Brook to Apponagansett Bay was installed on May 22, 2003 and was set to operate continuously for 16 months such that two summer seasons would be captured in the flow record. Stage data collection continued until March 21, 2005 for a total deployment of 22 months. The 12-month uninterrupted record used in this analysis encompasses the summer 2004 field season. Despite the relatively long deployment, only one complete hydrologic year was captured due to instrument failures. The period of record ultimately used in this analysis was September 2003 to August 2004.

River flow (volumetric discharge) was measured every 4 to 6 weeks using a Marsh- McBirney electromagnetic flow meter. A rating curve was developed for the Buttonwood Brook site based upon these flow measurements and measured water levels at the gauge site. The rating curve was then used for conversion of the continuously measured stage data to obtain daily freshwater flow volume. Water samples were collected weekly for nitrogen analysis. Integrating the flow and nitrogen concentration datasets allowed for the determination of nitrogen mass discharge to the head of Apponagansett Bay (Figure IV-7 and Table IV-3,4). In addition, a water balance was constructed based upon the US Geological Survey watershed delineations to determine long-term average freshwater discharge expected at the gauge site.

The annual freshwater flow record for Buttonwood Brook measured by the MEP was compared to the long-term average flows determined by the watershed area/recharge rate approach. The measured freshwater discharge from Buttonwood Brook to Apponagansett Bay was initially 23% below the long-term average modeled flows. Measured flow in Buttonwood Brook was obtained for one hydrologic year (September 2003 to August 2004). The average daily flow based on the MEP measured flow data was 12,822 m3/day compared to the long term average flows based on recharge rate (16,562 m3/day). The difference between the long-term average flow based on recharge rates over the watershed area and the MEP measured flow in Buttonwood Brook is due to below average annual precipitation before and during the stream gauge deployment period, based on rainfall records obtained from a rain gauge in the City of New Bedford. Twelve years of rainfall data (1993-2005) indicate that the average rainfall in the vicinity of Buttonwood Brook was 48.8 inches. By comparison, rainfall for the hydrologic period 2001-2002 was 34% below (31.85 in.) long term average, for the period 2002-2003 it was 4%

63 MASSACHUSETTS ESTUARIES PROJECT above average (50.80 in.) and for the gauge deployment period 2003-2004 rainfall was 21% below average (38.54 in.). It should be recognized that for the period 2001 to 2004 rainfall was below average with only one year (Sept. 2002-Aug. 2003) above average by only 4%, thus the water table is likely to have been lower than usual due to the 2 years of lower rainfall. This is significant relative to measured flow in the Buttonwood Brook surface water system as it is essentially a groundwater fed feature. Adjusting for the 21% lower annual precipitation during the 2003-2004 gauge deployment period, stream measured flow (15,898 m3/d) compared to long term flow calculated based on average recharge over the watershed area (16,562 m3/d) was only 4% different. This suggests that the stream is capturing the up-gradient recharge (and loads) accurately.

Total nitrogen concentrations within the Buttonwood Brook outflow to Apponagansett Bay were high, 1.97 mg N L-1, yielding an average daily total nitrogen discharge to the estuary of 25.26 kg/day and a measured total annual TN load of 9,220 kg/yr. In the Buttonwood Brook surface water system, nitrate + nitrite (NOx) was the predominant form of nitrogen (69%), while DON dissolved organic matter accounted for about 24% of TN. This pattern may reflect the influx of DIN rich water from a specific source or the predominance of DIN in Buttonwood Brook watershed may also reflect the higher permeability of the watershed sediments, which have a high relative amount of sand and gravel and thus groundwater transport plays a more dominant role in the watershed hydrologic regime. In this case, nitrate from septic systems, fertilizers and other watershed sources such as stormwater may have less opportunity for uptake and transformation in the soils before entering the groundwater flow system. The nutrient characteristics of the Buttonwood Brook flow indicates that groundwater nitrogen (typically dominated by nitrate) has little opportunity for uptake prior to reaching the Brook and was not completely taken up by plants within the stream ecosystem. The high concentration of inorganic nitrogen in the out-flowing Buttonwood Brook waters also suggests that plant production within the limited up-gradient freshwater ecosystems is not nitrogen limited. In addition, the high nitrate level suggests the possibility for additional uptake by freshwater systems might be accomplished in this system should a suitable location be identified.

From the measured nitrogen load discharged by Buttonwood Brook to the estuary and the nitrogen load determined from the watershed based land use analysis, it appears that there is nitrogen attenuation of upper watershed derived nitrogen during transport to the estuary. Based upon the slightly lower nitrogen load (9,220 kg yr-1) discharged from the stream compared to that added by the various land-uses to the associated watershed (10,387 kg yr-1), the integrated attenuation in passage through the wetland prior to discharge to the estuary is 11% (i.e. 11% of nitrogen input to watershed does not reach the estuary). This low level of attenuation compared to other streams evaluated under the MEP is expected given the limited number of up-gradient ponds or wetlands available to naturally attenuate nitrogen in the stream flow. The directly measured nitrogen loads from the stream was used in the Linked Watershed- Embayment Modeling of water quality (see Section VI, below)

64 MASSACHUSETTS ESTUARIES PROJECT

Table IV-3. Comparison of water flow and nitrogen discharges from streams (freshwater) discharging to the head of Apponagansett Bay. The “Stream” data is from the MEP stream gauging effort. Watershed data is based upon historic watershed delineations developed by the USGS and confirmed by the MEP Technical Team.

Stream Discharge Parameter Buttonwood Brook Apponagansett Brook Data Discharge(a) Discharge(a) Source Apponagansett Bay Apponagansett Bay Total Days of Record 365(b) 365(b) (1)

Flow Characteristics Stream Average Discharge (m3/day) ** 15,898 2,729 (1) Contributing Area Average Discharge (m3/day) 16,562 3,411 (2) Discharge Stream 2004-05 vs. Long-term Discharge 4% 20%

Nitrogen Characteristics Stream Average Nitrate + Nitrite Concentration (mg N/L) 1.369 1.015 (1) Stream Average Total N Concentration (mg N/L) 1.97 1.374 (1) Nitrate + Nitrite as Percent of Total N (%) 69% 74% (1)

Total Nitrogen (TN) Average Measured Stream Discharge (kg/day) 25.26 3.02 (1) TN Average Contributing UN-attenuated Load (kg/day) 28.46 4.83 (3) Attenuation of Nitrogen in Pond/Stream (%) 11.24% 37.47% (4)

(a) Flow and N load to streams discharging to Apponagansett Bay includes apportionments of Pond contributing areas. (b) September 1, 2003 to August 31, 2004. ** Flow is an average of annual flow for 2003-2004 adjusted upwards to reflect average precip conditions

(1) MEP gage site data (2) Calculated from MEP watershed delineations to ponds upgradient of specific gages; the fractional flow path from each sub-watershed which contribute to the flow in the streams to Apponagansett Bay; and the annual recharge rate. (3) As in footnote (2), with the addition of pond and stream conservative attentuation rates. (4) Calculated based upon the measured TN discharge from the rivers vs. the unattenuated watershed load.

65

MASSACHUSETTS ESTUARIES PROJECT

Massachusetts Estuaries Project Town of Dartmouth - Buttonwood Brook to Apponagansett Bay Predicted Flows relative to Stream Nutrient Concentrations (Sept. 2003 - Sept. 2004)

180000 4000

160000 3500

140000 3000

120000 2500 100000 2000 80000

Flow (m3/day) 1500 60000 Concentration (mg/m3) Concentration 1000 40000

20000 500

0 0

0 0 0 0 :00 :00 :00 :00 0 0:00 0 0 0:0 0 0:00 3 4 4 4 4 03 0:0 03 0:0 04 0 0 003 0 00 00 /200 2 /2 /2 /2 /200 7 /6/2 2 0 /28/2 1 0 /14/2004 0:0 3/4 /1 8/1/200 /2 6 8/ 1 1 4/23/2004 0:00 6 9 11/9 11/25/ Date / Time

Predicted Flow Nox Concentration TN Concentration

Figure IV-7. Buttonwood Brook discharge (solid blue line), nitrate+nitrite (yellow square) and total nitrogen (blue squares) concentrations for determination of annual volumetric discharge and nitrogen load from the upper watershed to Apponagansett Bay (Table IV-3). 66

MASSACHUSETTS ESTUARIES PROJECT

IV.2.3 Surface water Discharge and Attenuation of Watershed Nitrogen: Stream Discharge Apponagansett Brook (aka. Vincent Brook) to head of Apponagansett Bay At the Apponagansett Brook gauge site, a continuously recording vented calibrated water level gauge was installed to yield the level of water in the discharge that carries nitrogen load from the up-gradient sub-watershed (relatively devoid of aquatic resources such as ponds, bogs, wetlands) to the upper portion of the Apponagansett Bay system close to the discharge site for Buttonwood Brook. To confirm that freshwater was being measured, salinity measurements were conducted on the weekly water quality samples collected from the gauge site. Average salinity was determined to be no greater than 0.4 ppt. Based on the low salinity, the gauge location was deemed acceptable for making freshwater flow measurements. Calibration of the gauge was checked monthly. The gauge on the Apponagansett Brook outflow to the head of the Apponagansett Bay system was installed on June 11, 2003 and was set to operate continuously for 16 months such that two summer seasons would be captured in the flow record. Stage data collection continued until March 2, 2005 for a total deployment of 21 months. The 12-month uninterrupted record used in this analysis covered the period of September 1, 2003 to August 31,2004, a complete hydrologic year from one low flow period to the next.

River flow (volumetric discharge) was measured every 4 to 6 weeks using a Marsh- McBirney electromagnetic flow meter. A rating curve was developed for the Apponagansett Brook gauge site based upon these flow measurements and measured water levels at the gauge site. The rating curve was then used for conversion of the continuously measured stage data to obtain daily freshwater flow volume. Water samples were collected weekly for nitrogen analysis. Integrating the flow and nitrogen concentration datasets allowed for the determination of nitrogen mass discharge to the head of the Apponagansett Bay system (Figure IV-8 and Table IV-3,4). In addition, a water balance was constructed based upon the US Geological Survey watershed delineations to determine long-term average freshwater discharge expected at the gauge site based on recharge rates appropriate to the region.

The annual freshwater flow record for Apponagansett Brook measured by the MEP was compared to the long-term average flows determined by the watershed area/recharge rate approach. The measured freshwater discharge from Apponagansett Brook to Apponagansett Bay was initially 35% below the long-term average modeled flows. Measured flow in Apponagansett Brook was obtained for one hydrologic year (September 2003 to August 2004). The average daily flow based on the MEP measured flow data was 2,201 m3/day compared to the long term average flows based on recharge rate (3,411 m3/day). The difference between the long-term average flow based on recharge rates over the watershed area and the MEP measured flow in Apponagansett Brook is due to below average annual precipitation before and during the stream gauge deployment period, based on rainfall records obtained from a rain gauge in the City of New Bedford. Twelve years of rainfall data (1993-2005) indicate that the average rainfall in the vicinity of Buttonwood Brook was 48.8 inches. By comparison, rainfall for the hydrologic period 2001-2002 was 34% below (31.85 in.) long term average, for the period 2002-2003 it was 4% above average (50.80 in.) and for the gauge deployment period 2003- 2004 rainfall was 21% below average (38.54 in.). It should be recognized that for the period 2001 to 2004 rainfall was below average with only one year (Sept. 2002-Aug. 2003) above average by only 4%, thus the water table is likely to have been lower than usual due to the 2 years of lower rainfall. This is significant relative to measured flow in the Apponagansett Brook surface water system as it is essentially a groundwater fed feature. Adjusting for the 21% lower

67 MASSACHUSETTS ESTUARIES PROJECT

Massachusetts Estuaries Project Town of Dartmouth - Apponagansett (Vincent) Brook to Apponagansett Bay Predicted Flow relative to Nutrient Concentrations (Sept. 2003 to March 2005)

40000 2500

35000

2000 30000

25000 1500

20000

1000 Flow (m3/day) 15000 Concentration (mg/m3) Concentration 10000 500

5000

0 0

0 0 0 0 0 :0 :0 0 0 0:0 0:0 /03 /03 /04 /04 0:0 /04 8 6 4 3 9 8/1/04 0:00 6/2 10/ 1/1 4/2 11/ 2/17/05 0:00 5/28/05 0:00 Date / Time

Predicted Flow Nox Concentration TN Concentration

Figure IV-8. Apponagansett Brook discharge (solid blue line), nitrate+nitrite (yellow square) and total nitrogen (blue triangles) concentrations for determination of annual volumetric discharge and nitrogen load from the upper watershed to Apponagansett Bay (Table IV-3). 68

MASSACHUSETTS ESTUARIES PROJECT

Table IV-4. Summary of annual volumetric discharge and nitrogen load from Buttonwood Brook and Apponagansett Brook (freshwater) discharging to the Apponagansett Bay system based upon the data presented in Figures IV-7-8 and Table 3.

Stream Annual Flow Annual Load Annual Load Watershed Area Nox TN (m3/yr) (kg/yr) (kg/yr) (Km2)

Apponagansett Bay 18.52 Buttonwood Brook 5802770 6407 9220 8.08 Vincent Brook 996085 815 1104 1.91 (other subwatershed areas) 8.53

Stream Watershed Area Annual Unit Flow Annual Unit Load Annual Unit Load Nox TN (Km2) (m3/km2/yr) (kg/km2/yr) (kg/km2/yr)

Apponagansett Bay 18.52 Buttonwood Brook 8.08 718165 793 1141 Vincent Brook 1.91 521510 427 578 (other subwatershed areas) 8.53

DISCHARGE ATTENUATED LOAD (Kg/yr) EMBAYMENT SYSTEM PERIOD OF RECORD (m3/year) Nox TN

Apponagansett Bay Buttonwood Brook MEP September 1, 2004 to August 31, 2005 5,802,770 6407 9220

Apponagansett Bay Buttonwood Brook CCC Based on Watershed Area and Recharge 6,039,290 -- --

Apponagansett Bay Apponagansett (aka Vincent) Brook MEP September 1, 2004 to August 31, 2005 996,085 815 1104

Apponagansett Bay Apponagansett (aka Vincent) Brook CCC Based on Watershed Area and Recharge 1,243,920 -- --

69

MASSACHUSETTS ESTUARIES PROJECT annual precipitation during the 2003-2004 gauge deployment period, stream measured flow (2,729 m3/d) compared to long term flow calculated based on average recharge over the watershed area (3,411 m3/d) was 20% different.

Total nitrogen concentrations within the Apponagansett Brook outflow were high, 1.37 mg N L-1, yielding an average daily total nitrogen discharge to the estuary of 3.02 kg/day and a measured total annual TN load of 1,104 kg/yr. In the Apponagansett Brook surface water system, nitrate+nitrite (NOx) was well more than half of the total nitrogen load (74%), indicating that groundwater nitrogen (typically dominated by nitrate) discharging to the freshwater brook was not completely taken up by plants within the watershed or stream ecosystems. In the Apponagansett Brook discharge to the estuary, dissolved organic nitrogen (DON) was a small fraction of the total nitrogen pool (~21%). Similar to Buttonwood Brook, this balance of nitrogen species indicates that about 2/3 of the nitrogen is available for immediate uptake by primary producers (algae) in the Apponagansett Bay embayment and less than one-third of the total nitrogen in the river discharge is not immediately available for uptake in the estuary, but would need to be further processed by estuarine biota before being available to primary production.

From the measured nitrogen load discharged by the Apponagansett Brook to the estuary and the nitrogen load determined from the watershed based land use analysis, it appears that there is practically no nitrogen attenuation of upper watershed derived nitrogen during transport to the estuary. Based upon the measured nitrogen load (1,104 kg yr-1) discharged from the stream compared to that added by the various land-uses to the associated watershed (1,764 kg yr-1), the integrated attenuation in passage through the wetland prior to discharge to the estuary is 37% (i.e. 37% of nitrogen input to watershed does not reach the estuary). This moderate level of attenuation compared to other streams evaluated under the MEP is expected given the up-gradient conditions available to naturally attenuate nitrogen in the stream flow. The directly measured nitrogen loads from the stream was used in the Linked Watershed-Embayment Modeling of water quality (see Chapter VI, below).

IV.3 BENTHIC REGENERATION OF NITROGEN IN BOTTOM SEDIMENTS The overall objective of the benthic nutrient flux surveys was to quantify the summertime exchange of nitrogen, between the sediments and overlying waters throughout the Apponagansett Bay Estuarine System. The mass exchange of nitrogen between water column and sediments is a fundamental factor in controlling nitrogen levels within coastal waters. These fluxes and their associated biogeochemical pools relate directly to carbon, nutrient and oxygen dynamics and the nutrient related ecological health of these shallow marine ecosystems. In addition, these data are required for the proper modeling of nitrogen in shallow aquatic systems, both fresh and salt water.

IV.3.1 Sediment-Watercolumn Exchange of Nitrogen As stated in above sections, nitrogen loading and resulting levels within coastal embayments are the critical factors controlling the nutrient related ecological health and habitat quality within a system. Nitrogen enters the complex Apponagansett Bay Estuarine System predominantly in highly bioavailable forms from the surrounding upland watershed and more refractory forms in the inflowing tidal waters. If all of the nitrogen remained within the water column (once it entered) then predicting water column nitrogen levels would be simply a matter of determining the watershed loads, dispersion, and hydrodynamic flushing. However, as nitrogen enters the embayment from the surrounding watersheds it is predominantly in the bioavailable form nitrate. This nitrate and other bioavailable forms are rapidly taken up by phytoplankton for growth, i.e. it is converted from dissolved forms into phytoplankton “particles”.

70 MASSACHUSETTS ESTUARIES PROJECT

Most of these “particles” remain in the water column for sufficient time to be flushed out to a down gradient larger water body (like Buzzards Bay). However, some of these phytoplankton particles are grazed by zooplankton or filtered from the water by shellfish and other benthic animals and deposited on the bottom. Also, in longer residence time systems (greater than 8 days) these nitrogen rich particles may die and settle to the bottom. In both cases (grazing or senescence), a fraction of the phytoplankton with their associated nitrogen “load” become incorporated into the surficial sediments of the bays.

In general the fraction of the phytoplankton population which enters the surficial sediments of a shallow embayment: (1) increases with decreased hydrodynamic flushing, (2) increases in low velocity settings, (3) increases within enclosed tributary basins, particularly if they are deeper than the adjacent embayment. To some extent, the settling characteristics can be evaluated by observation of the grain-size and organic content of sediments within an estuary.

Once organic particles become incorporated into surface sediments they are decomposed by the natural animal and microbial community. This process can take place both under oxic (oxygenated) or anoxic (no oxygen present) conditions. It is through the decay of the organic matter with its nitrogen content that bioavailable nitrogen is returned to the embayment water column for another round of uptake by phytoplankton. This recycled nitrogen adds directly to the eutrophication of the estuarine waters in the same fashion as watershed inputs. In some systems that have been investigated by SMAST and the MEP, recycled nitrogen can account for about one-third to one-half of the nitrogen supply to phytoplankton blooms during the warmer summer months. It is during these warmer months that estuarine waters are most sensitive to nitrogen loadings. In contrast in some systems with salt marsh tidal creeks and basins, like Dike Marsh adjacent Apponagansett Bay, the sediments can be a net sink for nitrogen even during summer (e.g. Mashapaquit Creek Salt Marsh, West Falmouth Harbor; Centerville River Salt Marsh; Namskaket and Little Namskaket Salt Marshes). Embayment basins can also be net sinks for nitrogen to the extent that they support relatively oxidized surficial sediments, such as was found within the lower reach of the Slocum's River. In contrast, regions of enhanced deposition typically support moderate levels of nitrogen release during summer months.

Failure to account for the site-specific nitrogen balance of the sediments and its spatial variation from the tidal creeks and basins will result in significant errors in determination of the threshold nitrogen loading to the Apponagansett Bay Estuary. In addition, since the sites of recycling can be different from the sites of nitrogen entry from the watershed, both recycling and watershed data are needed to determine the best approaches for nitrogen mitigation.

IV.3.2 Method for determining sediment-watercolumn nitrogen exchange For the Apponagansett Bay Estuary, in order to determine the contribution of sediment regeneration to nutrient levels during the most sensitive summer interval (July-August), sediment samples were collected and incubated under in situ conditions. Sediment samples were collected from 22 sites, 6 within the outer portion of the Apponagansett Bay system and 2 within the Dike Marsh sub-embayment tributary to the inner portion of Apponagansett Bay (Figure IV-9), in July-August in 2 years, all in 2004 and all but 3 in 2005. Measurements of total dissolved nitrogen, nitrate + nitrite, ammonium were made in time-series on each incubated core sample.

Rates of nitrogen release were determined using undisturbed sediment cores incubated for 24 hours in temperature-controlled baths. Sediment cores (15 cm inside diameter) were collected by SCUBA divers and then transported by small boat to a shore side field lab. Cores

71 MASSACHUSETTS ESTUARIES PROJECT were maintained from collection through incubation at in situ temperatures. Bottom water was collected and filtered from each core site to replace the headspace water of the flux cores prior to incubation. The number of core samples from each site (Figure IV-9) per incubation are as follows:

Apponagansett Bay Estuary Benthic Nutrient Regeneration Cores (2004, 2005)  APB-1 2 cores (Outer Basin)  APB-2 2 cores (Outer Basin)  APB-3 1 cores (Outer Basin)  APB-4 2 cores (Outer Basin)  APB-5 2 cores (Outer Basin)  APB-6 1 cores (Outer Basin)  APB-7 2 cores (Inner Basin)  APB-8 2 cores (Inner Basin)  APB-9 2 cores (Inner Basin)  APB-10 2 cores (Inner Basin)  APB-11 2 cores (Inner Basin)  APB-12 2 cores (Inner Basin)  APB-13 2 cores (Inner Basin)  APB-14 2 cores (Inner Basin)  APB-15/16 2 cores (Inner Basin)  APB-17 2 cores (Inner Basin)  APB-18 1 cores (inner Basin)  APB-19 2 cores (Inner Basin)  APB-20 2 core (Dike Marsh)  APB-21 2 core (Dike Marsh)  APB-23 2 cores (Inner Basin)

Sampling was distributed throughout the primary embayment sub-basins of this system: the outer Apponagansett Bay, inner Apponagansett Bay and within the Dike Marsh salt marsh tributary the inner Apponagansett Bay. The results for each site were then combined for calculating the net nitrogen regeneration rates for the water quality modeling effort.

Sediment-water column exchange follows the methods of Jorgensen (1977), Klump and Martens (1983), and Howes et al. (1998) for nutrients and metabolism. Upon return to the field laboratory, the location having generously been provided by the White family, local Dartmouth residents, the cores were transferred to pre-equilibrated temperature baths. The headspace water overlying the sediment was replaced, magnetic stirrers emplaced, and the headspace enclosed. Periodic 60 ml water samples were withdrawn (volume replaced with filtered water), filtered into acid leached polyethylene bottles and held on ice for nutrient analysis. Ammonium (Scheiner 1976) and ortho-phosphate (Murphy and Reilly 1962) assays were conducted within 24 hours and the remaining samples frozen (-20oC) for assay of nitrate + nitrite (Cd reduction: Lachat Autoanalysis), and DON (D'Elia et al. 1977). Rates were determined from linear regression of analyte concentrations through time.

Chemical analyses were performed by the Coastal Systems Analytical Facility at the School for Marine Science and Technology (SMAST) at the University of Massachusetts in New Bedford, MA. The laboratory follows standard methods for saltwater analysis and sediment geochemistry (S.Sampieri, Manager, [email protected]).

72 MASSACHUSETTS ESTUARIES PROJECT

APB14

APB13

APB12

APB18 APB11 APB9 APB19 APB15/16 APB10 APB7 APB17 APB8 APB20 APB23

APB6

APB5

APB21 APB3 APB4

APB2

APB22

APB1

Figure IV-9. Apponagansett Bay embayment system sediment sampling sites (red symbols) for determination of nitrogen regeneration rates (all sites sampled in both 2004 and 2005). Numbers are for reference to station identifications listed above.

IV.3.3 Rates of Summer Nitrogen Regeneration from Sediments Water column nitrogen levels are the balance of inputs from direct sources (land, rain etc), losses (denitrification, burial), regeneration (water column and benthic), and uptake (e.g. photosynthesis). As stated above, during the warmer summer months the sediments of shallow embayments typically act as a net source of nitrogen to the overlying waters and help to stimulate eutrophication in organic rich systems. However, some sediments may be net sinks

73 MASSACHUSETTS ESTUARIES PROJECT for nitrogen and some may be in “balance” (organic N particle settling = nitrogen release). Sediments may also take up dissolved nitrate directly from the water column and convert it to dinitrogen gas (termed “denitrification”), hence effectively removing it from the ecosystem. This process is typically a small component of sediment denitrification in embayment sediments, since the water column nitrogen pool is typically dominated by organic forms of nitrogen, with very low nitrate concentrations. However, this process can be very effective in removing nitrogen loads in some systems, particularly in streams, ponds and salt marshes, where overlying waters support high nitrate levels.

In addition to nitrogen cycling, there are ecological consequences to habitat quality of organic matter settling and mineralization within sediments, these relate primarily to sediment and water column oxygen status. However, for the modeling of nitrogen within an embayment it is the relative balance of nitrogen input from water column to sediment versus regeneration which is critical. Similarly, it is the net balance of nitrogen fluxes between water column and sediments during the modeling period that must be quantified. For example, a net input to the sediments represents an effective lowering of the nitrogen loading to down-gradient systems and net output from the sediments represents an additional load.

The relative balance of nitrogen fluxes (“in” versus “out” of sediments) is dominated by the rate of particulate settling (in), the rate of denitrification of nitrate from overlying water (in), and regeneration (out). The rate of denitrification is controlled by the organic levels within the sediment (oxic/anoxic) and the concentration of nitrate in the overlying water. Organic rich sediment systems with high overlying nitrate frequently show large net nitrogen uptake throughout the summer months, even though organic nitrogen is being mineralized and released to the overlying water as well. The rate of nitrate uptake, simply dominates the overall sediment nitrogen cycle. Similarly, organic enriched sediments with high oxygen in overlying water and high mineralization rates (organic N --> ammonium) can also support high rates of nitrogen removal by coupled nitrification-denitrification. Sediment systems like those in the open water basins of Apponagansett Bay are candidates for this nitrogen removal pathway.

In order to model the nitrogen distribution within an embayment it is important to be able to account for the net nitrogen flux from the sediments within each part of each system. This requires that an estimate of the particulate input and nitrate uptake be obtained for comparison to the rate of nitrogen release. Only sediments with a net release of nitrogen contribute a true additional nitrogen load to the overlying waters, while those with a net input to the sediments serve as an “in embayment” attenuation mechanism for nitrogen.

Overall, coastal sediments are not overlain by nitrate rich waters and the major nitrogen input is via phytoplankton grazing or direct settling. In these systems, on an annual basis, the amount of nitrogen input to sediments is generally higher than the amount of nitrogen release. This net sink results from the burial of reworked refractory organic compounds, sorption of inorganic nitrogen and some denitrification of produced inorganic nitrogen before it can “escape” to the overlying waters. However, this net sink evaluation of coastal sediments is based upon annual fluxes. If seasonality is taken into account, it is clear that sediments undergo periods of net input and net output. The net output is generally during warmer periods and the net input is during colder periods. The result can be an accumulation of nitrogen within late fall, winter, and early spring and a net release during summer. The conceptual model of this seasonality has the sediments acting as a battery with the flux balance controlled by temperature (Figure IV- 10).

74 MASSACHUSETTS ESTUARIES PROJECT

100

80

60

40

20

0

-20

-40

-60 JFMAMJJASOND

Figure IV-10. Conceptual diagram showing the seasonal variation in sediment N flux, with maximum positive flux (sediment output) occurring in the summer months, and maximum negative flux (sediment up-take) during the winter months.

Unfortunately, the tendency for net release of nitrogen during warmer periods coincides with the periods of lowest nutrient related water quality within temperate embayments. This sediment nitrogen release is in part responsible for poor summer nutrient related health. Other major factors causing the seasonal water quality decline are the lower solubility of oxygen during summer, the higher oxygen demand by marine communities, and environmental conditions supportive of high phytoplankton growth rates.

In order to determine the net nitrogen flux between water column and sediments, all of the above factors were taken into account. The net input or release of nitrogen within a specific embayment was determined based upon the measured total dissolved nitrogen uptake or release, and estimate of particulate nitrogen input.

Sediment sampling was conducted throughout the primary embayment sub-basins of this system: the outer basin of Apponagansett Bay, the inner basin of Apponagansett Bay and throughout the Dike Marsh tributary salt marsh basin adjacent the inner basin of Apponagansett Bay. Sediment sampling and analysis was completed in order to obtain the nitrogen regeneration rates required for parameterization of the water quality model. The distribution of cores was established to cover gradients in sediment type, flow field and phytoplankton density. For each core the nitrogen flux rates (described in the section above) were evaluated relative to measured sediment organic carbon and nitrogen content and sediment type and an analysis of each site’s tidal flow velocities. The maximum bottom water flow velocity at each coring site was determined from the hydrodynamic model. These data were then used to determine the nitrogen balance within each sub-embayment.

The magnitude of the settling of particulate organic carbon and nitrogen into the sediments was accomplished by determining the average depth of water within each sediment site, the average summer particulate carbon and nitrogen concentration within the overlying water and the tidal velocities from the hydrodynamic model (Chapter V). Two levels of settling

75 MASSACHUSETTS ESTUARIES PROJECT were used. If the sediments were organic rich and fine grained, and the hydrodynamic data showed low tidal velocities, then a water column particle residence time of 8 days was used (based upon phytoplankton and particulate carbon studies of poorly flushed basins). If the sediments indicated coarse-grained sediments and low organic content and high velocities, then half this settling rate was used. Adjusting the measured sediment releases was essential in order not to over-estimate the sediment nitrogen source and to account for those sediment areas which are net nitrogen sinks for the aquatic system. This approach has been previously validated in outer Cape Cod embayments (Town of Chatham embayments) by examining the relative fraction of the sediment carbon turnover (total sediment metabolism), which would be accounted for by daily particulate carbon settling. This analysis indicated that sediment metabolism in the highly organic rich sediments of the wetlands and depositional basins is driven primarily by stored organic matter (ca. 90%). Also, in the more open lower portions of larger embayments, storage appears to be low and a large proportion of the daily carbon requirement in summer is met by particle settling (approximately 33% to 67%). This range of values and their distribution is consistent with ecological theory and field data from shallow embayments. Additional, validation has been conducted on deep enclosed basins (with little freshwater inflow), where the fluxes can be determined by multiple methods. In this case the rate of sediment regeneration determined from incubations was comparable to that determined from whole system balance.

Net nitrogen release or uptake from the sediments within the Apponagansett Bay open water basins and Dike Marsh were comparable to other similar embayment / salt marsh systems with similar configuration and flushing rates. The spatial distribution of nitrogen release/uptake by the sediments of the open water basins (upper and lower basins) of Apponaganett Bay were similar, showing negligible release to low rates of net nitrogen uptake with a relatively narrow range, 0.5 to -23.8 mg N m-2 d-1. Only the area directly up-gradient of the bridge, which is a depositional area of soft mud, showed a moderate rate of nitrogen release to the overlying waters, 19.8 mg L-1. Dike Marsh showed a negligible net nitrogen release, likely related to its high tidal velocities and sandy sediments in many areas, 0.4 mg L-1. The observed rates agree well with those from other open water embayments and salt marsh dominated systems in southeastern Massachusetts. For example Swan Pond and Seine Pond (-8.0 and -16.9 mg N m-2 d-1, respectively) the main basins of Swan River and Parkers River estuaries, also the lower basin of the Slocum's River(-13.2 mg N m-2 d-1), main basin of Waquoit Bay (-6.4 to -31.9 mg N m-2 d-1) and associated basin of Eel Pond (-27.1 mg N m-2 d-1) all previously assessed by the MEP.

Finally, sediment-watercolumn nitrogen exchange within the salt marsh dominated tidal creek of Dike Marsh (0.4 mg L-1), was also similar to other nearby Buzzards Bay tidal marsh basins. For example the creeks of the salt marsh dominated portions of the Back River (Bourne) to the north and the Slocum's and Little River Estuaries (Dartmouth) support similarly small net release rates of 6.5 mg N m-2 d-1 and 4.6-9.0 mg N m-2 d-1, respectively. The observed rates of net nitrogen release in the Dike Marsh main channel uptake were also similar to other oxidized sediments in salt marsh dominated systems on Cape Cod such as the sandy main creeks of Nauset Marsh in the region of the tidal inlet, 8.4 mg N m-2 d-1. It appears that the sediment nitrogen release rates within the component basins to the Apponagansett Bay Estuarine System are very comparable to analogous basins in other estuaries in the region and specifically in Buzzards Bay.

Net nitrogen release rates for use in the water quality modeling effort for the component reaches of the Apponagansett Bay Estuarine System (Section VI) are presented in Table IV-5. There was a clear spatial pattern of sediment nitrogen flux, with highest rates of uptake in the

76 MASSACHUSETTS ESTUARIES PROJECT deeper basins. The sediments within the Apponagansett Bay Estuary showed nitrogen fluxes typical of similarly structured systems within the region and appear to be in balance with the overlying waters and the nitrogen flux rates consistent with the level of nitrogen loading to this system and its relatively high flushing rate.

Table IV-5. Rates of net nitrogen return from sediments to the overlying waters of the Apponagansett Bay Estuary. These values are combined with the basin areas to determine total nitrogen mass in the water quality model (see Chapter VI). Measurements represent July -August rates (average of 2004 and 2005). Sediment Nitrogen Flux (mg N m-2 d-1) Station Location Mean S.E. # sites i.d. * Apponagansett Bay - Inner Basin Upper Basin 0.5 7.6 6 12, 13, 14 Main Basin -12.2 3.1 15 3, 11, 15-20, 23 Lowest Reach 19.8 18.8 5 7, 8, 9 Dike Marsh 0.4 5.3 4 21, 22 Apponagansett Bay - Outer Basin (below bridge) Upper (Bridge) -23.8 8.3 3 5, 6 Lower (Breakwater) -9.6 7.7 7 1, 2, 3, 4 * Station numbers refer to Figure IV-9.

77 MASSACHUSETTS ESTUARIES PROJECT

V. HYDRODYNAMIC MODELING

V.1 INTRODUCTION This section summarizes field data collection effort and the development of hydrodynamic models for the Apponagansett Bay estuary system (Figure V-1). For this system, the final calibrated model offers an understanding of water movement through the estuary, and provides the first step towards evaluating water quality, as well as tool for later determining nitrogen loading “thresholds”. Tidal flushing information is utilized as the basis for a quantitative evaluation of water quality. Nutrient loading data combined with measured environmental parameters within the various sub-embayments become the basis for an advanced water quality model based on total nitrogen concentrations. This type of model provides a tool for evaluating existing estuarine water quality, as well as determining the likely positive impacts of various alternatives for improving overall estuarine health, enabling the bordering residence to understand how pollutant loadings into the estuary will affect the biochemical environment and its ability to sustain a healthy marine habitat.

In general, water quality studies of tidally influenced estuaries must include a thorough evaluation of the hydrodynamics of the estuarine system. Estuarine hydrodynamics control a variety of coastal processes including tidal flushing, pollutant dispersion, tidal currents, sedimentation, erosion, and water levels. Numerical models provide a cost-effective method for evaluating tidal hydrodynamics since they require limited data collection and may be utilized to numerically assess a range of management alternatives. Once the hydrodynamics of an estuary system are understood, computations regarding the related coastal processes become relatively straightforward extensions to the hydrodynamic modeling. For example, the spread of pollutants may be analyzed from tidal current information developed by the numerical models.

Estuarine water quality is dependent upon nutrient and pollutant loading and the processes that help flush nutrients and pollutants from the estuary (e.g., tides and biological processes). Relatively low nutrient and pollutant loading and efficient tidal flushing are indicators of high water quality. The ability of an estuary to flush nutrients and pollutants is proportional to the volume of water exchanged with a high quality water body (i.e. Buzzards Bay). Several embayment-specific parameters influence tidal flushing and the associated residence time of water within an estuary. For the Apponagansett Bay system, the most important parameters are the tide range along with the shape, length and depth of the estuary.

Shallow coastal embayments are the initial recipients of freshwater flows (i.e., groundwater and surfacewater) and the nutrients they carry. An embayment’s shape influences the time that nutrients are retained in them before being flushed out to adjacent open waters, and their shallow depths both decrease their ability to dilute nutrient (and pollutant) inputs and increase the secondary impacts of nutrients recycled from the sediments. Degradation of coastal waters and development are tied together through inputs of pollutants in runoff and groundwater flows, and to some extent through direct disturbance, i.e. boating, oil and chemical spills, and direct discharges from land and boats. Excess nutrients, especially nitrogen, promote phytoplankton blooms and the growth of epiphytes on eelgrass and attached algae, with adverse consequences including low oxygen, shading of submerged aquatic vegetation, and aesthetic problems.

The Apponagansett Bay estuary (Figure V-1) is a tidally dominated embayment system open to Buzzards Bay. The Apponagansett estuary has three main sub-systems. The Upper Bay consists of the areas north of the Bridge Street Bridge, including the Apponagansett River

78

MASSACHUSETTS ESTUARIES PROJECT while the Lower Bay is that area which runs south of the Bridge to the breakwater and Buzzards Bay. The third sub-system is Dike Meadow Creek, which consists of a marsh system with a coverage area of approximately 160 acres. The Lower Bay area serves as the transition region between Buzzards Bay and the inner parts of Apponagansett Bay. The average water depth is - 10 feet NGVD. The Upper Bay system is a shallow sub-embayment with a mean depth of -4 ft NGVD. Across the whole system, the greatest depths (-27 ft NGVD) occur within the channel that runs along the east side of the Bridge Street Bridge.

Figure V-1. Map of the Apponagansett Bay estuary system (from United States Geological Survey topographic maps).

79

MASSACHUSETTS ESTUARIES PROJECT

Since the water elevation difference between Buzzards Bay and the inland reaches of Apponagansett Bay are the primary driving force for tidal exchange, the local tide range naturally limits the volume of water flushed during a tidal cycle. Tidal damping (reduction in tidal amplitude) along the length of the bay is negligible, indicating a system that flushed efficiently. Any issues with water quality, therefore, would likely be due other factors including nutrient loading conditions from the system’s watersheds, and the tide range in Buzzards Bay.

Circulation in Apponagansett Bay system was simulated using the RMA-2 numerical hydrodynamic model. To calibrate the model, field measurements of water elevations and bathymetry were required. Tide data were acquired within Buzzards Bay at a gage station installed offshore of the mouth of Apponagansett Bay, and at 3 stations located within the estuary (Figure V-2). All temperature-depth recorders (TDRs or tide gages) were installed for a 36-day period to measure tidal variations through one spring-neap tidal cycle. In this manner, attenuation of the tidal signal as it propagates through the various sub-embayments was evaluated accurately.

V.2 FIELD DATA COLLECTION AND ANALYSIS Accurate modeling of system hydrodynamics is dependent upon measured conditions within the estuary for two important reasons:

 To accurately define the system geometry and boundary conditions for the numerical model  To provide ‘real’ observations of hydrodynamic behavior to calibrate and verify the model results

System geometry is defined by the shoreline of the system, including all coves, creeks, and marshes, as well as accompanying depth (or bathymetric) information. The three- dimensional surface of the estuary is mapped as accurately as possible, since the resulting hydrodynamic behavior is strongly dependent upon features such as channel widths and depths, sills, marsh elevations, and inter-tidal flats. Hence, this study included an effort to collect bathymetric information in the field.

Boundary conditions for the numerical model consist of variations of water surface elevations measured in Buzzards Bay. These variations result principally from tides, and provide the dominant hydraulic forcing for the system, and are the principal forcing function applied to the model. Additional pressure sensors were installed at selected interior locations to measure variations of water surface elevation along the length of the system (gauging locations are shown in Figure V-2). These measurements were used to calibrate and verify the model results, and to assure that the dynamic of the physical system were properly simulated.

V.2.1. Bathymetry Bathymetry data (i.e., depth measurements) for the hydrodynamic model of the Apponagansett Bay system was assembled from a recent hydrographic survey performed specifically for this study. Historical NOS survey data, where available, were used for areas in Buzzards Bay that were not covered by these more recent surveys.

80

MASSACHUSETTS ESTUARIES PROJECT

Figure V-2. Map of the study region identifying locations of the tide gauges used to measure water level variations throughout the system. Four (4) gauges were deployed for the 36-day period between June18, and July 24, 2003. Each yellow dot represents the approximate locations of the tide gauges: (Apg-1) represents the gage in Buzzards Bay (Offshore), (Apg-2) inside the bay entrance, (Apg-3) in Dike Meadow Creek, (Apg-4) in Apponagansett River.

The hydrographic survey of August, 2003 (CRE, 2003) was designed to cover the entire Apponagansett Bay estuary system. The survey was conducted from an outboard motorboat with an installed precision fathometer (with a depth resolution of approximately 0.1 foot), coupled together with a differential GPS to provide position measurements accurate to approximately 1-3 feet. Digital data output from both the echo sounder fathometer and GPS were logged to a laptop computer, which integrated the data to produce a single data set consisting of water depth as a function of geographic position (latitude/longitude).

The raw measured water depths were merged with water surface elevation measurements to determine bathymetric elevations relative to the NGVD 1929 vertical datum.

81

MASSACHUSETTS ESTUARIES PROJECT

Once rectified, the finished processed data were archived as ‘xyz’ files containing x-y horizontal position (in Massachusetts State Plan 1983 coordinates) and vertical elevation of the bottom (z). These xyz files were then interpolated into the finite element mesh used for the hydrodynamic simulations. The final processed bathymetric data from the survey are presented in Figure V-3.

Figure V-3. Bathymetric data interpolated to the finite element mesh of hydrodynamic model July 17th & 18th 2003.

V.2.2 Tide Data Collection and Analysis Variations in water surface elevation were measured at a station in at three locations in the Apponagansett Bay, and at a single station in Buzzards Bay. TDRs were deployed at each gauging station in at the end of June 2003, and recovered in late July, 2003. The duration of

82

MASSACHUSETTS ESTUARIES PROJECT the TDR deployment allowed time to conduct the bathymetric surveys (CRE July 2003), as well as sufficient data to perform a thorough analysis of the tides in the system.

The tide records from Apponagansett Bay were corrected for atmospheric pressure variations and then rectified to the NGVD 29 vertical datum. Atmospheric pressure data, available in one-hour intervals from the NDBC Buzzards Bay C-MAN platform, were used to pressure correct the raw tide data. Final processed tide data from stations used for this study are presented in Figure V-4, for the complete 36-day period of the TDR deployment.

Buzzards Bay 4

2

0 feet (NGVD29) -2 06/18 06/23 06/28 07/03 07/08 07/13 07/18 07/23

Padanaram Bridge 4

2

0 feet (NGVD29) -2 06/18 06/23 06/28 07/03 07/08 07/13 07/18 07/23

Apponagansett River 4

2

0 feet (NGVD29) -2 06/18 06/23 06/28 07/03 07/08 07/13 07/18 07/23

Dike Meadow Creek 4

2

0 feet (NGVD29) -2 06/18 06/23 06/28 07/03 07/08 07/13 07/18 07/23 Date Figure V-4. Water elevation variations as measured at the seven locations within the Apponagansett Bay system, between June18 and July 23, 2003.

Tide records longer than 29 days are necessary for a complete evaluation of tidal dynamics within the estuarine system. Although a one-month record likely does not include extreme high or low tides, it does provide an accurate basis for typical tidal conditions governed by both lunar and solar motion. For numerical modeling of hydrodynamics, the typical tide conditions associated with a one-month record are appropriate for driving tidal flows within the estuarine system.

The loss of amplitude together with increasing phase delay with increasing distance from the mouth of an estuary is called tidal attenuation. Tide attenuation can be a useful indicator of flushing efficiency in an estuary. Attenuation of the tidal signal is caused by the geometry of the nearshore region, where channel restrictions (e.g., bridge abutments) and also the depth of an

83

MASSACHUSETTS ESTUARIES PROJECT estuary are the primary factors which influence tidal damping in estuaries. For Apponagansett Bay, a visual comparison in Figure V-5 between tide elevations at the three stations along the system demonstrates how little change there is between the tide range and timing from Buzzards Bay to the farthest inland reaches of the system. This provides an initial indication that flushing conditions in the Apponagansett Bay are ideal, with minimal loss of tidal energy along the length of the system.

July 12 - 13, 2003 4 Buzzards Bay Padanaram Bridge Apponagansett River Dike Meadow Creek 3

2

1 Water Elevation, feet (NGVD) 0

-1

-2 00:00 04:00 08:00 12:00 16:00 20:00 00:00 04:00 08:00 12:00 16:00 20:00 00:00 hours:min

Figure V-5 Plot showing two tide cycles tides at four stations in the Apponagansett Bay estuary system plotted together. Demonstrated in this plot is the phase delay effect caused by the propagation of the tide through the estuary.

To better quantify the changes to the tide from the inlet to inside the system, the standard tide datums were computed from the 36-day records. These datums are presented in Table V- 1. The Mean Higher High Water (MHHW) and Mean Lower Low Water (MLLW) levels represent the mean of the daily highest and lowest water levels. The Mean High Water (MHW) and Mean Low Water (MLW) levels represent the mean of all the high and low tides of a record, respectively. The Mean Tide Level (MTL) is simply the mean of MHW and MLW. The tides in Buzzards Bay are semi-diurnal, meaning that there are typically two tide cycles in a day. There is usually a small variation in the level of the two daily tides. This variation can be seen in the differences between the MHHW and MHW, as well as the MLLW and MLW levels.

84

MASSACHUSETTS ESTUARIES PROJECT

Table V-1. Tide datums computed from records collected in the Apponagansett system June 18 - July 23, 2003. Datum elevations are given in feet relative to NGVD 29. Bridge Dike Buzzards Apponagansett Tide Datum Street Meadow Bay River Bridge Creek Maximum Tide 4.00 4.00 4.00 3.95 MHHW 2.84 2.86 2.87 2.87 MHW 2.60 2.62 2.64 2.64 MTL 0.92 0.92 0.92 0.92 MLW -0.77 -0.79 -0.81 -0.80 MLLW -0.85 -0.87 -0.89 -0.88 Minimum Tide -1.50 -1.55 -1.52 -1.56

For most NOAA tide stations, these datums are computed using 19 years of tide data, the duration of a tidal epoch. For this study, a significantly shorter time span of data was available; however, these datums still provide a useful comparison of tidal dynamics within the system. From the computed datums, it further apparent that there is little tide damping throughout the system. Again, the absence of tide damping exhibited in the Apponagansett Bay system suggests that it flushes efficiently.

A more thorough harmonic analysis was also performed on the time series data from each gauging station in an effort to separate the various component signals which make up the observed tide. The analysis allows an understanding of the relative contribution that diverse physical processes (i.e. tides, winds, etc.) have on water level variations within the estuary. Harmonic analysis is a mathematical procedure that fits sinusoidal functions of known frequency to the measured signal. The amplitudes and phase of 23 tidal constituents, with periods between 4 hours and 2 weeks, result from this procedure. The observed tide is therefore the sum of an astronomical tide component and a residual atmospheric component. The astronomical tide in turn is the sum of several individual tidal constituents, with a particular amplitude and frequency. For demonstration purposes a graphical example of how these constituents add together is shown in Figure V-6.

Table V-2 presents the amplitudes of eight significant tidal constituents. The M2, or the familiar twice-a-day lunar semi-diurnal, tide is the strongest contributor to the signal with an amplitude of 1.62 feet in Buzzards Bay. The tidal range of the M2 constituent is twice the amplitude, or about 3.24 feet. The diurnal (once daily) tide constituents, K1 (solar) and O1 (lunar), possess amplitudes of approximately 0.31 and 0.18 feet respectively and account for the semi-diurnal variance between high tides and low tides seen in figure V-5. The N2 tide, another semi-diurnal lunar constituent that together with L2 it modifies the amplitude and frequency of M2 for the effects of variation in the Moon’s orbital speed due to its elliptical orbit, is the next largest tidal constituent at and is 4.2 times smaller then the main semi-diurnal constituent (M2). The M4 tide, a higher frequency harmonic of the M2 lunar tide (twice the frequency of the M2), results from frictional dissipation of the M2 tide in shallow water and therefore have more influence on the shape of the tide signal than the other tidal constituents. The effect of the comparatively large amplitude can be seen most clearly in Figure V-5 as the sharply rising and falling semi-diurnal high and low tides.

85

MASSACHUSETTS ESTUARIES PROJECT

Table V-2. Tidal Constituents, Apponagansett Bay System June18 - July 23, 2003 Constituent Amplitude (feet) M2 M4 M6 S2 N2 K1 O1 Msf Period (hours) 12.42 6.21 4.14 12 12.66 23.93 25.82 354.61 Buzzards Bay (Apg-1) 1.62 0.21 0.01 0.22 0.38 0.31 0.18 0.12 Bridge Street Bridge (Apg-2) 1.64 0.21 0.02 0.23 0.39 0.30 0.18 0.12 Apponagansett River (Apg-4) 1.64 0.21 0.02 0.23 0.39 0.31 0.18 0.11 Dike Meadow Creek (Apg-3) 1.63 0.22 0.01 0.23 0.39 0.31 0.18 0.11

Figure V-6. Example of observed astronomical tide as the sum of its primary constituents. In this example the observed tide signal is the sum of individual constituents (M2, M4, K1, N2), with varying amplitude and frequency.

Table V-3 presents the phase delay (in other words, the travel time required for the tidal wave to propagate throughout the system) of the M2 tide at all tide gauge locations inside the Bay. The greatest delay occurs between the Buzzards Bay and the Dike Meadow Creek gauging stations. There is virtually no phase delay between the north an south ends of Apponagansett Bay. This suggests that the only noticeable tidal attenuation occurs at the narrow opening to Dike Meadow Creek. However, the degree of attenuation is not significant relative to the hydraulic efficiency of the system because the effects of attenuation are observed only in the phase delay across the inlet, and not as a reduction in the amplitude of the tide.

86

MASSACHUSETTS ESTUARIES PROJECT

Table V-3. M2 Tidal Attenuation, Apponagansett Bay, June 18 - July 23, 2003 Delay in minutes relative to Buzzards Bay Location Delay (minutes) Bridge Street Bridge 3.3 Apponagansett River 3.7 Dike Meadow Creek 11.9

The tide data were further evaluated to determine the importance of tidal versus non- tidal processes to changes in water surface elevation. Non-tidal processes include wind forcing (set-up or set-down) within the estuary, as well as sub-tidal oscillations of the sea surface. Variations in water surface elevation can also be affected by freshwater discharge into the system, if these volumes are relatively large compared to tidal flow. The energy content of the astronomical tidal signal alone is determined by summing the contributions from the 23 constituents determined by the harmonic analysis. Subtracting this “pure” tidal signal from the original elevation time series resulted with the non-tidal, or residual, portion of the water elevation changes. The energy of this non-tidal signal is compared to the tidal signal, and yields a quantitative measure of how important these non-tidal physical processes are relative to hydrodynamic circulation within the estuary. Figure V-7 shows the comparison of the measured tide from Buzzards Bay, with the computed astronomical tide resulting from the harmonic analysis, and the resulting non-tidal residual.

Table V-4 shows that the percentage contribution of tidal energy was essentially equal in all parts of the system, which indicates that local effects due to winds and other non-tidal processes are minimal throughout the systems. The analysis also shows that tides are responsible for approximately 99% of the water level changes in all parts of the Apponagansett Bay system. The remaining 1% was the result of atmospheric forcing, due to winds, or barometric pressure gradients acting upon the collective water surface of Buzzards Bay and Apponagansett Bay. The total energy content of the tide signal from each gauging station does not change significantly, nor does the relative contribution of tidal vs. non-tidal forces along the estuary basin. This is further indication that tide attenuation through the system is negligible. It is also an indication that the source of the non-tidal component of the tide signal is generated completely offshore, with no additional non-tidal energy input inside the system (e.g., from wind set-up of the Bay).

The results from Table V-4 indicate that hydrodynamic circulation throughout Apponagansett Bay is dependent primarily upon tidal processes while wind and other non-tidal effects have a minimal contribution. For completeness however, the actual tide signal from Buzzards Bay was used to force the hydrodynamic model so that the effects of non-tidal energy are included in the analysis.

87

MASSACHUSETTS ESTUARIES PROJECT

Table V-4. Percentages of Tidal versus Non-Tidal Energy, Apponagansett Bay

Total Variance Total Tidal Non-tidal Location (ft2) (%) (%) (%) Buzzards Bay 1.49 100 98.7 1.3 Bridge Street Bridge 1.49 100 98.6 1.4 Apponagansett River 1.49 100 98.6 1.4 Dike Meadow Creek 1.49 100 98.5 1.5

Original Series 3

2

1

0

feet (NGVD29) -1

-2 06/18 06/23 06/28 07/03 07/08 07/13 07/18 07/23

Predicted Tide 3

2

1

0

feet (NGVD29) -1

-2 06/18 06/23 06/28 07/03 07/08 07/13 07/18 07/23

Residual Tide

3

2

1

0

feet (NGVD29) -1

-2 06/18 06/23 06/28 07/03 07/08 07/13 07/18 07/23 Date

Figure V-7. Results of the harmonic analysis and the separation of the tidal from the non-tidal, or residual, signal measured in Buzzards Bay (Apg-1).

88

MASSACHUSETTS ESTUARIES PROJECT

V.3 HYDRODYNAMIC MODELING The focus of this study was the development of a numerical model capable of accurately simulating hydrodynamic circulation within the Apponagansett Bay estuary system. Once calibrated, the model was used to calculate water volumes for selected sub-embayments (e.g., the upper reach of Apponagansett River) as well as determine the volumes of water exchanged during each tidal cycle. These parameters are used to calculate system residence times, or flushing rates. The ultimate utility of the hydrodynamic model is to supply required input data for the water quality modeling effort described in Chapter VI.

V.3.1 Model Theory This study of Apponagansett Bay utilized a state-of-the-art computer model to evaluate tidal circulation and flushing. The particular model employed was the RMA-2 model developed by Resource Management Associates (King, 1990). It is a two-dimensional, depth-averaged finite element model, capable of simulating transient hydrodynamics. The model is widely accepted and tested for analyses of estuaries or rivers. Applied Coastal staff members have utilized RMA-2 for numerous flushing studies for estuary systems in southeast Massachusetts, including systems in Chatham, Falmouth’s ‘finger’ ponds, and Popponesset Bay.

In its original form, RMA-2 was developed by William Norton and Ian King under contract with the U.S. Army Corps of Engineers (Norton et al., 1973). Further development included the introduction of one-dimensional elements, state-of-the-art pre- and post-processing data programs, and the use of elements with curved borders. Recently, the graphic pre- and post- processing routines were updated by Brigham Young University through a package called the Surfacewater Modeling System or SMS (BYU, 1998). SMS is a front- and back-end software package that allows the user to easily modify model parameters (such as geometry, element coefficients, and boundary conditions), as well as view the model results and download specific data types. While the RMA model is essentially used without cost or constraint, the SMS software package requires site licensing for use.

RMA-2 is a finite element model designed for simulating one- and two-dimensional depth- averaged hydrodynamic systems. The dependent variables are velocity and water depth, and the equations solved are the depth-averaged Navier-Stokes equations. Reynolds assumptions are incorporated as an eddy viscosity effect to represent turbulent energy losses. Other terms in the governing equations permit friction losses (approximated either by a Chezy or Manning formulation), Coriolis effects, and surface wind stresses. All the coefficients associated with these terms may vary from element to element. The model utilizes quadrilaterals and triangles to represent the prototype system. Element boundaries may either be curved or straight.

The time dependence of the governing equations is incorporated within the solution technique needed to solve the set of simultaneous equations. This technique is implicit; therefore, unconditionally stable. Once the equations are solved, corrections to the initial estimate of velocity and water elevation are employed, and the equations are re-solved until the convergence criterion is met.

V.3.2 Model Setup There are three main steps required to implement RMA-2V: • Grid generation • Boundary condition specification • Calibration

89

MASSACHUSETTS ESTUARIES PROJECT

The extent of the finite element grid was generated using digital aerial photographs from the MassGIS online orthophoto database. A time-varying water surface elevation boundary condition (measured tide) was specified at the entrance of the system based on the tide gauge data collected in Buzzards Bay. Once the grid and boundary conditions were set, the model was calibrated to ensure accurate predictions of tidal flow. Various friction and eddy viscosity coefficients were adjusted, through several model calibration simulations to obtain agreement between measured and modeled tides. The calibrated model provides the requisite information for future detailed water quality modeling.

V.3.2.1 Grid Generation The grid generation process for the model was assisted through the use of the SMS package. The digital shoreline and bathymetry data were imported to SMS, and a finite element grid was generated to represent the estuary with 1649 elements and 4194 nodes (Figure V-8). All regions in the system were represented by two-dimensional (depth-averaged) elements. The finite element grid for the system provided the detail necessary to evaluate accurately the variation in hydrodynamic properties within the estuary. Grid resolution is governed by two factors: 1) expected flow patterns, and 2) the bathymetric variability in each region. Smaller cross channel node spacing in the river channels was designed to provide a more detailed analysis in these regions of rapidly varying velocities and bathymetry. Widely spaced nodes were utilized in areas where velocity gradients were likely to be small; for example, on marsh plains, the broad, deep channel sections south of the Bridge Street Bridge and the flat areas in the central portion of the Upper Bay. The completed grid is made up of quadrilateral and triangular two-dimensional elements. Reference water depths at each node of the model were interpreted from bathymetry data obtained in the recent field surveys and the NOS data archive. The final interpolated grid bathymetry is shown in Figure V-9.

In addition to assigning water depths for all areas of the grid, material types are also assigned to each region of the study area (Figure V-10). By dividing the model domain into separate materials, each area of the system can be assigned unique properties such as individual friction coefficients (discussed in Section V.3.3.1) and turbulent exchange coefficients (discussed in Section V.3.3.2). With the spatial extent, water depth and physical properties of each model element defined, the boundary conditions are applied.

V.3.2.2 Boundary Condition Specification Two types of boundary conditions were employed for the RMA-2 model: 1) "slip" boundaries, and 2) tidal elevation boundaries. All of the elements with land borders have "slip" boundary conditions, where the direction of flow was constrained shore-parallel. The model generated all internal boundary conditions from the governing conservation equations.

The model was forced at the open boundary using water elevations measurements obtained in Buzzards Bay (described in section V.2.2). This measured time series consists of all physical processes affecting variations of water level: tides, winds, and other non-tidal oscillations of the sea surface. The rise and fall of the tide in Buzzards Bay is the primary driving force for estuarine circulation. Dynamic (time-varying) model simulations specified a new water surface elevation at the offshore boundary every 10 minutes. The model specifies the water elevation at the offshore boundary, and uses this value to calculate water elevations at every nodal point within the system, adjusting each value according to solutions of the model equations. Changing water levels in Buzzards Bay produce variations in surface slopes within

90

MASSACHUSETTS ESTUARIES PROJECT the estuary; these slopes drive water either into the system (if water is higher offshore) or out of the system (when the water level falls in Buzzards Bay).

Figure V-8. The model finite element mesh developed for Apponagansett Bay estuary system.

91

MASSACHUSETTS ESTUARIES PROJECT

Figure V-9. Depth contours of the completed Apponagansett Bay finite element mesh.

V.3.3 Calibration After developing the finite element grid and specifying boundary conditions, the model was calibrated. Calibration ensured the model predicts accurately what was observed during the field measurement program. Numerous model simulations were required to calibrate the model, with each run varying specific parameters such as friction coefficients, turbulent exchange coefficients, fresh water inflow, and subtle modifications to the system bathymetry to achieve a best fit to the data. Calibration of the flushing model required a close match between the modeled and measured tides in each of the sub-embayments where tides were measured. Initially, the model was calibrated by the visual agreement between modeled and measured tides. To refine the calibration procedure, water elevations were output from the model at the same locations in the estuary where tide gauges were installed, and the data were processed to calculate standard

92

MASSACHUSETTS ESTUARIES PROJECT error as well harmonic constituents (of both measured and modeled data) over the seven-day calibration period. The amplitude and phase of four constituents (M2, M4, M6, and K1) were compared and the corresponding errors for each were calculated. The intent of the calibration procedure is to minimize the error in amplitude and phase of the individual constituents. In general, minimization of the M2 amplitude and phase becomes the highest priority, since this is the dominant constituent. Emphasis is also placed on the M4 constituent, as this constituent has the greatest impact on the degree of tidal distortion within the system, and provides the unique shape of the modified tide wave at various points in the system.

Figure V-10. Material types assigned to the finite element mesh for the Apponagansett Bay estuary.

The calibration was performed for an approximate seven-day period, beginning June 21, 2003 and ending June 28, 2003. This time period was selected because it includes tidal conditions where the wind-induced portion of the signals (i.e. the residual) was minimal, hence

93

MASSACHUSETTS ESTUARIES PROJECT more typical of tidal circulation within the estuary. The selected time period also spans the transition from neap to spring tide ranges, which is representative of average tidal conditions in the embayment system. The ability to model a range of flow conditions is a primary advantage of a numerical tidal flushing model. Modeled tides were evaluated for time (phase) lag and height damping of dominant tidal constituents. The calibrated model was used to analyze existing detailed flow patterns and compute residence times.

V.3.3.1 Friction Coefficients Friction inhibits flow along the bottom of estuary channels or other flow regions where water depths can become shallow and velocities relatively high. Friction is a measure of the channel roughness, and can cause both significant amplitude attenuation and phase delay of the tidal signal. Friction is approximated in RMA-2 as a Manning coefficient. First, Manning's friction coefficient values of 0.025 were specified for all elements. These values correspond to typical Manning's coefficients determined experimentally in smooth earth-lined channels with no weeds (low friction) to winding channels with pools and shoals with higher friction (Henderson, 1966). On the marsh plains of Dike Meadow Creek, damping of flow velocities typically is controlled more by “form drag” associated with marsh plants than the bottom friction described above. However, simulation of this drag is performed using Manning’s coefficients as well, with values ranging from 2-to-10 times friction coefficients used in sandy channels.

Final calibrated friction coefficients (listed in Table V-5) were largest for marsh plain area, where values were set at 0.5. Small changes in these values did not change the accuracy of the calibration.

Table V-5. Manning’s Roughness coefficients used in simulations of modeled embayments. Embayment Bottom Friction Offshore 0.025 Lower Bay 0.025 Upper Bay 0.025 Apponagansett River 0.025 Bridge 0.1 Marsh Channel 0.03 Marsh Plain 0.5

V.3.3.2 Turbulent Exchange Coefficients Turbulent exchange coefficients approximate energy losses due to internal friction between fluid particles. The significance of turbulent energy losses increases where flow is swift, such as inlets and bridge constrictions. According to King (1990), these values are proportional to element dimensions (numerical effects) and flow velocities (physics). The model was mildly sensitive to turbulent exchange coefficients, with areas of marsh plain being most sensitive. In other regions where the flow gradients were not as strong, the model was much less sensitive to changes in the turbulent exchange coefficients. Typically, model turbulence coefficients (D) are set between 20 and 100 lb-sec/ft2 (as listed in Table V-6). Higher values (up to 500 lb-sec/ft2) can be used on the marsh plains, to ensure solution stability.

94

MASSACHUSETTS ESTUARIES PROJECT

Table V-6. Turbulence exchange coefficients (D) used in simulations of modeled embayment system. Embayment D (lb-sec/ft2) Offshore 20 Lower Bay 20 Upper Bay 20 Apponagansett River 20 Bridge 50 Marsh Channel 40 Marsh Plain 100

V.3.3.3 Wetting and Drying/Marsh Porosity Processes Modeled hydrodynamics were complicated by wetting/drying cycles on the marsh plain included in the model as part of the Dike Meadow Creek. Cyclically wet/dry areas of the marsh will tend to store waters as the tide begins to ebb and then slowly release water as the water level drops within the creeks and channels. This store-and-release characteristic of these marsh regions was partially responsible for the distortion of the tidal signal, and the elongation of the ebb phase of the tide. On the flood phase, water rises within the channels and creeks initially until water surface elevation reaches the marsh plain, when at this point the water level remains nearly constant as water fans out over the marsh surface. The rapid flooding of the marsh surface corresponds to a flattening out of the tide curve approaching high water. Marsh porosity is a feature of the RMA-2 model that permits the modeling of hydrodynamics in marshes. This model feature essentially simulates the store-and-release capability of the marsh plain by allowing grid elements to transition gradually between wet and dry states. This technique allows RMA-2 to gradually change the amount of water retained in each element.

V.3.3.4 Comparison of Modeled Tides and Measured Tide Data Several calibration model runs were performed to determine how changes to various parameters (e.g. friction and turbulent exchange coefficients) affected the model results. These trial runs achieved excellent agreement between the model simulations and the field data. Comparison plots of modeled versus measured water levels at the three interior gauge locations are presented in Figures V-11 through V-13. Although visual calibration reveal that the modeled tidal hydrodynamics are reasonable, tidal constituent calibration is required to quantify the accuracy of the models. Measured tidal constituent amplitudes and time lags (lag) for the calibration time period are shown in Table V-7. The constituent values in for the calibration time period differ from those in Tables V-2 because constituents were computed for only 7 days, rather than the entire 32-day period represented in Table V-2. Errors associated with tidal constituent height were on the order of hundredths of feet, which was an order of magnitude better than the accuracy of the tide gage gauges (0.12 ft). Time lag errors were less than the time increment used in the model and measured tide data (10 minutes), indicating good agreement between the model and data.

95

MASSACHUSETTS ESTUARIES PROJECT

Figure V-11. Comparison of water surface elevation at the Bridge Street Bridge (App-2) predicted by the model (dots) to those measured in the field (solid line) for the calibration time period.

Figure V-12. Comparison of water surface elevation in Dike Meadow Creek (App-3) predicted by the model (dots) to those measured in the field (solid line) for the calibration time period.

96

MASSACHUSETTS ESTUARIES PROJECT

Figure V-13. Comparison of water surface elevation in Apponagansett River (App-4) predicted by the model (dots to those measured in the field (solid line) for the calibration time period.

Table V-7. Comparison of Tidal Constituents calibrated RMA2 model versus measured tidal data for the period June 21 to June 28, 2003. Model Data Constituent Amplitude (ft) Phase (degrees) Location M2 M4 M6 K1 ΦM2 ΦM4 Bridge Street Bridge 1.26 0.19 0.02 0.33 -57.3 -73.8 Dike Meadow Creek 1.26 0.19 0.02 0.33 -56.6 -72.2 Apponagansett River 1.26 0.19 0.02 0.33 -57.2 -73.7 Measured Tidal Data Constituent Amplitude (ft) Phase (degrees) Location M2 M4 M6 K1 ΦM2 ΦM4 Bridge Street Bridge 1.26 0.19 0.02 0.32 -55.9 -72.0 Dike Meadow Creek 1.26 0.20 0.01 0.33 -52.8 -61.7 Apponagansett River 1.26 0.19 0.02 0.33 -55.9 -72.2 Error Constituent Amplitude (ft) Phase (minutes) Location M2 M4 M6 K1 ΦM2 ΦM4 Bridge Street Bridge 0.00 0.00 0.00 -0.01 2.9 1.9 Dike Meadow Creek 0.00 0.01 -0.01 0.00 7.8 10.9 Apponagansett River 0.00 0.00 0.00 0.00 2.8 1.5

97

MASSACHUSETTS ESTUARIES PROJECT

V.3.4 Model Verification Typically, the calibrated model is verified by comparing flow rates predicted by the model to those recorded in the field at certain sections of the system. The Apponagansett Bay study however had no current data to compare to and so the verification was performed through an additional model run using tide data during a spring tide. In this case, the water levels in Buzzards Bay from July 8, 2003 - July 15, 2003 were used. This time represents the largest tide range during the instrument deployment.

Modeled water elevations were compared to field data as means to verify that the model parameters could accurately predict the hydrodynamics of this spring tide condition. The plots of the model/field comparisons are shown in Figures V-14 to V-16 below.

Figure V-14. Comparison of water surface elevation at the Bridge Street Bridge (App-2) predicted by the model (dots) to those measured in the field (solid line) for the verification time period.

98

MASSACHUSETTS ESTUARIES PROJECT

Figure V-15. Comparison of water surface elevation in Dike Meadow Creek (App-3) predicted by the model (dots) to those measured in the field (solid line) for the verification time period.

Figure V-16. Comparison of water surface elevation in Apponagansett River (App-4) predicted by the model (dots to those measured in the field (solid line) for the verification time period.

99

MASSACHUSETTS ESTUARIES PROJECT

As with the analysis of the model results during the calibration stage, the model verification also included a closer look at the match between model and field data by considering the major tidal constituents for each data set. As Table V-8 below shows, there is excellent agreement between the amplitudes and phases of the major constituents. The amplitudes are within one-one hundredth of a foot while the phase differences are mostly smaller than the time step of the model (10 minutes). The only exception to this is apparent in the results from Dike Meadow Creek where there phase of the M2 constituent is off by one time step and the smaller M4 constituent is off by 3 time steps. These are small discrepancies and as a whole the model was judged to have been verified as accurate. It is also worth noting that the actual water quality modeling based on this hydrodynamic model will be using the calibration time period discussed is Section V.3.3.4, where the results are in excellent agreement.

Table V-8. Comparison of Tidal Constituents calibrated RMA2 model versus measured tidal data for the model verification period July 8 to July 15, 2003. Model Data Constituent Amplitude (ft) Phase (degrees) Location M2 M4 M6 K1 ΦM2 ΦM4 Bridge Street Bridge 2.01 0.25 0.03 0.45 -7.6 4.5 Dike Meadow Creek 2.01 0.24 0.03 0.45 -6.7 6.3 Apponagansett River 2.02 0.25 0.03 0.45 -7.4 4.3 Measured Tidal Data Constituent Amplitude (ft) Phase (degrees) Location M2 M4 M6 K1 ΦM2 ΦM4 Bridge Street Bridge 2.02 0.25 0.03 0.44 -6.1 9.1 Dike Meadow Creek 2.01 0.23 0.02 0.46 -0.7 35.4 Apponagansett River 2.02 0.25 0.03 0.45 -5.9 7.9 Error Constituent Amplitude (ft) Phase (minutes) Location M2 M4 M6 K1 ΦM2 ΦM4 Bridge Street Bridge 0.01 0.00 0.00 -0.01 3.0 4.8 Dike Meadow Creek 0.00 -0.01 -0.01 0.01 12.5 30.1 Apponagansett River 0.00 0.00 0.00 0.00 3.2 3.7

V.3.4.1 Model Circulation Characteristics The final calibrated and validated model serves as a useful tool for investigating the circulation characteristics of the Apponagansett Bay system. Using model inputs of bathymetry and tide data, current velocities and flow rates can be determined at any point in the model domain. This is a very useful feature of a hydrodynamic model, where a limited amount of collected data can be expanded to determine the physical attributes of the system in areas where no physical data record exists.

A close-up of the model output is presented in Figure V-17, which shows contours of flow velocity, along with velocity vectors which indicate the direction and magnitude of flow, for a single model time-step, at the portion of the tide where maximum ebb velocities occur at the entrance to Dike Meadow Creek.

In addition to depth averaged velocities, the total flow rate of water flowing through a channel can be computed with the hydrodynamic model. During the simulation time period,

100

MASSACHUSETTS ESTUARIES PROJECT maximum modeled flow rates in the vicinity of the Bridge Street Bridge were 3000 ft3/sec on both flood and ebb tides.

Figure V-17. Example of hydrodynamic model output at the entrance to Dike Meadow Creek for a single time step where maximum ebb velocities occur for this tide cycle. Color contours indicate flow speed, and vectors indicate the direction and magnitude of flow.

V.4 FLUSHING CHARACTERISTICS The primary mechanism controlling estuarine water quality within Apponagansett Bay is tidal exchange. A rising tide offshore in Buzzards Bay creates a slope in water surface from the bay into the estuary. Consequently, water flows into (floods) the system. Similarly, the estuary drains into the open waters of Buzzards Bay on an ebbing tide. This exchange of water between the estuary and the larger water body of Buzzards Bay is defined as tidal flushing. The calibrated hydrodynamic model is a tool to evaluate quantitatively tidal flushing of each system, and was used to compute flushing rates (residence times) and tidal circulation patterns.

Flushing rate, or residence time, is defined as the average time required for a parcel of water to migrate out of an estuary from points within the system. For this study, system residence times were computed as the average time required for a water parcel to migrate from a point within the each embayment to the entrance of the system. System residence times are computed as follows:

101

MASSACHUSETTS ESTUARIES PROJECT

V T  system t system P cycle where Tsystem denotes the residence time for the system, Vsystem represents volume of the (entire) system at mean tide level, P equals the tidal prism (or volume entering the system through a single tidal cycle), and tcycle the period of the tidal cycle, typically 12.42 hours (or 0.52 days). To compute system residence time for a sub-embayment, the tidal prism of the sub-embayment replaces the total system tidal prism value in the above equation.

In addition to system residence times, a second residence, the local residence time, was defined as the average time required for a water parcel to migrate from a location within a sub- embayment to a point outside the sub-embayment. Using the Upper Bay as an example, the system residence time is the average time required for water to migrate from the Upper Bay, into the Lower Bay, and finally into Buzzards Bay, where the local residence time is the average time required for water to migrate from the Upper Bay to the Lower Bay (not all the way out of the system). Local residence times for each sub-embayment are computed as:

V T  local t local P cycle where Tlocal denotes the residence time for the local sub-embayment, Vlocal represents the volume of the sub-embayment at mean tide level, P equals the tidal prism (or volume entering the local sub-embayment through a single tidal cycle), and tcycle the period of the tidal cycle (again, 0.52 days).

Residence times are provided as a first order evaluation of estuarine water quality. Lower residence times generally correspond to higher water quality; however, residence times may be misleading depending upon pollutant/nutrient loading rates and the overall quality of the receiving waters. As a qualitative guide, system residence times are applicable for systems where the water quality within the entire estuary is degraded and higher quality waters provide the only means of reducing the high nutrient levels. For the modeled system, this approach is applicable, since it assumes the main system has relatively low quality water relative to Buzzards Bay.

The rate of pollutant/nutrient loading and the quality of water outside the estuary both must be evaluated in conjunction with residence times to obtain a clear picture of water quality. Efficient tidal flushing (low residence time) is not an indication of high water quality if pollutants and nutrients are loaded into the estuary faster than the tidal circulation can flush the system. Neither are low residence times an indicator of high water quality if the water flushed into the estuary is of poor quality. Advanced understanding of water quality will be obtained from the calibrated hydrodynamic model by extending the model to include a total nitrogen dispersion model (Section VI). The water quality model will provide a valuable tool to evaluate the complex mechanisms governing estuarine water quality in the Apponagansett Bay and its component sub-embayments.

The volume of the each sub-embayment, as well as their respective tidal prisms, were computed in cubic feet (Table V-9). Model divisions used to define the system sub- embayments for the two systems include 1) the whole of Apponagansett Bay, 2) the Lower Bay, 3) the Upper Bay, and 4) Dike Meadow Creek and the surrounding marsh. The model computed total volume of each material type (using the divisions shown in Figure V-10), at

102

MASSACHUSETTS ESTUARIES PROJECT every time step, and this output was used to calculate mean sub-embayment volume and average tide prism. Since the 7 day period used to compute the flushing rates of the system represent average tidal conditions, the measurements provide the most appropriate method for determining mean flushing rates for the system sub-embayments.

Residence times were averaged for the tidal cycles comprising a representative 7 day period, and are listed in Table V-10. Residence times were computed for the entire estuary, as well selected sub-embayments within the system. In addition, system and local residence times were computed to indicate the range of conditions possible for the system. Residence times were calculated as the volume of water (based on mean volumes computed for the simulation period) in the entire system divided by the average volume of water exchanged with each sub-embayment over a flood tidal cycle (tidal prism).

Table V-9. Embayment mean volumes and average tidal prism of the Apponagansett Bay system during simulation period. Embayment Mean Volume (ft3) Tide Prism Volume (ft3) Apponagansett Bay 203,179,000 68,511,000 Lower Bay 126,656,000 29,358,000 Upper Bay 69,433,000 35,934,000 Dike Meadow Creek 7,141,000 3,437,000

Table V-10. Computed System and Local residence times for sub- embayments of the Apponagansett Bay estuary system. System Residence Local Residence Embayment Time (days) Time (days) Apponagansett Bay 1.5 1.5 Lower Bay 3.6 2.2 Upper Bay 2.9 1.0 Dike Meadow Creek 30.6 1.1

The relatively low local residence time for the whole Apponagansett Bay system (1.5 days) shows that it has good flushing conditions, and therefore, water quality within the system is moderately sensitive to the combined nutrient load input from the system watersheds, benthic sediments and direct atmospheric deposition. The limit to the flushing ability of Apponagansett Bay is not due to hydrodynamic constrictions, but rather the small tide range in Buzzards Bay itself. The relatively long residence time for Dike Meadow Creek reveal the inadequacy of using system residence time alone to evaluate water quality. By definition, smaller sub- embayments have longer residence times; therefore, residence times may be misleading for small, remote parts of the estuary. In each areas of Apponagansett Bay, local flushing rates are similar to the average for the entire system. This indicates that the health of each of the sub- systems within the estuary is dependent upon the health of the remainder of the estuary. If water quality conditions are good in the main portions of the estuary, then Apponagansett River and Dike Meadow Creek would likely have good conditions as well.

Based on our knowledge of estuarine processes, the combined errors associated with the method applied to compute residence times are within 10% to 15% of “true” residence

103

MASSACHUSETTS ESTUARIES PROJECT times, for the Apponagansett Bay estuary system. Possible errors in computed residence times can be linked to two sources: the bathymetry information and simplifications employed to calculate residence time. In this study, the most significant errors associated with the bathymetry data result from the process of interpolating the data to the finite element mesh, which was the basis for all the flushing volumes used in the analysis. In addition, limited topographic measurements were available in some of the smaller sub-embayments of the system.

Minor errors may be introduced in residence time calculations by simplifying assumptions. Flushing rate calculations assume that water exiting an estuary or sub-embayment does not return on the following tidal cycle. For regions where a strong littoral drift exists, this assumption is valid. However, water exiting a small sub-embayment on a relatively calm day may not completely mix with estuarine waters. In this region of Buzzards Bay, the littoral drift is moderate and the assumption of full mixing of the water exiting the estuary should be fairly accurate.

104

MASSACHUSETTS ESTUARIES PROJECT

VI. WATER QUALITY MODELING

VI.1 DATA SOURCES FOR THE MODEL Several different data types and calculations are required to support the water quality modeling effort for the Apponagansett Bay System. These include the output from the hydrodynamics model, calculations of external nitrogen loads from the watersheds, measurements of internal nitrogen loads from the sediment (benthic flux), and measurements of nitrogen in the water column.

VI.1.1 Hydrodynamics and Tidal Flushing in the Embayment Extensive field measurements and hydrodynamic modeling of the embayment were an essential preparatory step to the development of the water quality model. The result of this work, among other things, was a calibrated model output representing the transport of water within the system embayment. Files of node locations and node connectivity for the RMA-2 model grid were transferred to the RMA-4 water quality model; therefore, the computational grid for the hydrodynamic model also was the computational grid for the water quality model. The period of hydrodynamic output for the water quality model calibration was an 10-tidal cycle period in June 2003. Each modeled scenario (e.g., present conditions, build-out) required the model be run for a 28-day spin-up period, to allow the model to reach a dynamic “steady state”, and ensure that model spin-up would not affect the final model output.

VI.1.2 Nitrogen Loading to the Embayment Three primary nitrogen loads to an embayment are recognized in this modeling study: external loads from the watersheds, nitrogen load from direct rainfall on the embayment surface, and internal loads from the sediments. Additionally, there is a fourth load to the Apponagansett Bay System, consisting of the background concentrations of total nitrogen in the waters entering from Buzzards Bay. This load is represented as a constant concentration along the seaward boundary of the model grid.

VI.1.3 Measured Nitrogen Concentrations in the Embayment In order to create a model that realistically simulates the total nitrogen concentrations in a system in response to the existing flushing conditions and loadings, it is necessary to calibrate the model to actual measurements of water column nitrogen concentrations. The refined and approved data for each monitoring station used in the water quality modeling effort are presented in Table VI-1. Station locations are indicated in Figure VI-1. The multi-year averages present the “best” comparison to the water quality model output, since factors of tide, temperature and rainfall may exert short-term influences on the individual sampling dates and even cause inter-annual differences. Three years of baseline field data is the minimum required to provide a baseline for MEP analysis. Ten years of data (collected between 1999 and 2008) were available for stations monitored by SMAST in the Apponagansett Harbor and the outer Bay.

VI.2 MODEL DESCRIPTION AND APPLICATION A two-dimensional finite element water quality model, RMA-4 (King, 1990), was employed to study the effects of nitrogen loading in the Apponagansett Bay System. The RMA-4 model has the capability for the simulation of advection-diffusion processes in aquatic environments. It is the constituent transport model counterpart of the RMA-2 hydrodynamic model used to simulate the fluid dynamics of the Apponagansett Harbor. Like RMA-2 numerical code, RMA-4 is a two-dimensional depth averaged finite element model capable of simulating time-dependent

105 MASSACHUSETTS ESTUARIES PROJECT constituent transport. The RMA-4 model was developed with support from the US Army Corps of Engineers (USACE) Waterways Experiment Station (WES), and is widely accepted and tested. Applied Coastal staff have utilized this model in numerous water quality studies of other embayments along the south coast of Massachusetts.

Figure VI-1. Estuarine water quality monitoring station locations in the Apponagansett Bay System. Station labels correspond to those provided in Table VI-1.

The overall approach involves modeling total nitrogen as a non-conservative constituent, where bottom sediments act as a source or sink of nitrogen, based on local biochemical characteristics. This modeling represents summertime conditions, when algal growth is at its maximum. Total nitrogen modeling is based upon various data collection efforts and analyses presented in previous sections of this report. Nitrogen loading information was derived from the SMAST and Town of Dartmouth watershed loading analysis (based on the USGS watersheds), as well as the measured bottom sediment nitrogen fluxes. Water column nitrogen measurements were utilized as model boundaries and as calibration data. Hydrodynamic model output (discussed in Section V) provided the remaining information (tides, currents, and bathymetry) needed to parameterize the water quality model of the system.

106 MASSACHUSETTS ESTUARIES PROJECT

Table VI-1. Water quality monitoring data and modeled Nitrogen concentrations for the Apponagansett Bay System used in the model calibration plots of Figure VI-2. All concentrations are given in mg/L N. “Data mean” values are calculated as the average of the separate yearly means.

Sub- Monitor 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 s.d. all model model model Embaym ing mean N mean mean mean mean mean mean mean mean mean mean data min max avg ent station

Head of AB-1A 0.876 1.562 1.519 2.030 1.512 1.613 1.393 1.315 1.602 1.544 1.477 0.349 35 1.18 1.66 1.34 Bay North Little AB-4 0.512 0.463 0.456 0.574 0.492 0.419 0.464 0.597 0.544 0.600 0.518 0.090 36 0.48 0.58 0.53 Island Upper Basin- AB-3 0.454 0.441 0.446 0.408 0.484 0.373 0.414 0.501 0.490 0.489 0.452 0.081 74 0.40 0.48 0.45 Lower Lower Basin- AB-2 0.422 0.431 0.461 0.454 0.477 0.376 0.524 0.457 0.525 0.421 0.456 0.074 39 0.33 0.43 0.38 Upper Lower AB-6 0.387 -- 0.379 0.399 0.411 0.297 0.462 0.436 0.468 0.375 0.402 0.065 35 0.32 0.37 0.34 Basin 107

MASSACHUSETTS ESTUARIES PROJECT

VI.2.1 Model Formulation The formulation of the model is for two-dimensional depth-averaged systems in which concentration in the vertical direction is assumed uniform. The depth-averaged assumption is justified since vertical mixing by wind and tidal processes prevent significant stratification in the modeled sub-embayments. The governing equation of the RMA-4 constituent model can be most simply expressed as a form of the transport equation, in two dimensions:  c c c    c  c 

  u  v    Dx  Dy     t x    xy  yx y  where c in the water quality constituent concentration; t is time; u and v are the velocities in the x and y directions, respectively; Dx and Dy are the model dispersion coefficients in the x and y directions; and  is the constituent source/sink term. Since the model utilizes input from the RMA-2 model, a similar implicit solution technique is employed for the RMA-4 model.

The model is therefore used to compute spatially and temporally varying concentrations c of the modeled constituent (i.e., total nitrogen), based on model inputs of 1) water depth and velocity computed using the RMA-2 hydrodynamic model; 2) mass loading input of the modeled constituent; and 3) user selected values of the model dispersion coefficients. Dispersion coefficients used for each system sub-embayment were developed during the calibration process. During the calibration procedure, the dispersion coefficients were incrementally changed until model concentration outputs matched measured data.

The RMA-4 model can be utilized to predict both spatial and temporal variations in total for a given embayment system. At each time step, the model computes constituent concentrations over the entire finite element grid and utilizes a continuity of mass equation to check these results. Similar to the hydrodynamic model, the water quality model evaluates model parameters at every element at 10-minute time intervals throughout the grid system. For this application, the RMA-4 model was used to predict tidally averaged total nitrogen concentrations throughout Apponagansett Bay System.

VI.2.2 Water Quality Model Setup Required inputs to the RMA-4 model include a computational mesh, computed water elevations and velocities at all nodes of the mesh, constituent mass loading, and spatially varying values of the dispersion coefficient. Because the RMA-4 model is part of a suite of integrated computer models, the finite-element meshes and the resulting hydrodynamic simulations previously developed for the Apponagansett Bay System was used for the water quality constituent modeling portion of this study.

Based on groundwater recharge rates from the USGS, the hydrodynamic model was set- up to include ground water flowing into the system from the watersheds. Apponagansett Bay has seven watersheds contributing to the groundwater and stream flow, the combined flow rate into the system is 16.91 ft3/sec (41,380 m3/day). Watersheds 1 (Buttonwood Brook) and 2 (Apponagansett Brook) were entered as freshwater streams, Watershed 3 to 7 (Apponagansett Bay Inner East, Apponagansett Bay Inner West, Dike Marsh, Apponagansett Bay Outer West, and Apponagansett Bay Outer East) were represented as groundwater flow. The groundwater flow rates for each of the watershed are: Watershed 3, 2.26 ft3/sec (5,535 m3/day); Watershed 4, 3.66 ft3/sec (8,945 m3/day); Watershed 5, 4.13 ft3/sec (10,103 m3/day); Watershed 6, 0.47 ft3/sec (1,143 m3/day); and Watershed 7, 0.26 ft3/sec (631 m3/day). The stream flow rates for

108 MASSACHUSETTS ESTUARIES PROJECT watersheds 1 and 2 are: Watershed 1, 5.24 ft3/sec (12,822 m3/day) and Watershed 2, 0.90 ft3/sec (2,201 m3/day).

For the model, an initial total N concentration equal to the concentration at the open boundary was applied to the entire model domain. The model was then run for a simulated month-long (28 day) spin-up period. At the end of the spin-up period, the model was run for an additional 5 tidal-day (125 hour) period. Model results were recorded only after the initial spin- up period. The time step used for the water quality computations was 10 minutes, which corresponds to the time step of the hydrodynamics input for the Apponagansett Bay System.

VI.2.3 Boundary Condition Specification Mass loading of nitrogen into each model included 1) sources developed from the results of the watershed analysis, 2) estimates of direct atmospheric deposition, and 3) summer benthic regeneration. Nitrogen loads from each separate sub-embayment watershed were distributed across the sub-embayment. For example, the combined watershed direct atmospheric deposition load for Dike Marsh was evenly distributed across the grid cells that formed the marsh embayment. Benthic regeneration load was distributed among another sub-set of grid cells which are in the interior portion of each basin.

The loadings used to model present conditions in Apponagansett Bay System are given in Table VI-2. Watershed and depositional loads were taken from the results of the analysis of Section IV. Summertime benthic flux loads were computed based on the analysis of sediment cores in Section IV. The area rate (g/sec/m2) of nitrogen flux from that analysis was applied to the surface area coverage computed for each sub-embayment (excluding marsh coverage, when present), resulting in a total flux for each embayment (as listed in Table VI-2). Due to the highly variable nature of bottom sediments and other estuarine characteristics of coastal embayments in general, the measured benthic flux for existing conditions also is variable. For present conditions, the benthic flux is negative within Apponagansett Harbor and Bay and positive benthic flux exists within Dike Marsh.

Table VI-2. Sub-embayment loads used for total nitrogen modeling of the Apponagansett Bay System, with total watershed N loads, atmospheric N loads, and benthic flux. These loads represent present loading conditions. direct watershed benthic flux atmospheric sub-embayment load net deposition (kg/day) (kg/day) (kg/day) Apponagansett Harbor 24.36 4.48 -4.05 Apponagansett Bay 2.38 3.64 -14.38 Dike Marsh 6.74 0.58 0.04 Surface Water Sources Buttonwood Brook 25.33 -- -- Apponagansett Brook 3.04 -- --

In addition to mass loading boundary conditions set within the model domain, a concentration along the model open boundary was specified. The model uses the specified concentration at the open boundary during the flooding tide periods of the model simulations. TN concentration of the incoming water is set at the value designated for the open boundary.

109 MASSACHUSETTS ESTUARIES PROJECT

The boundary concentration in Buzzards Bay was set at 0.295 mg/L, based on data from Padanaram Breakwater water quality measurement station. The open boundary total nitrogen concentration represents long-term average summer concentrations found within Buzzards Bay.

VI.2.4 Model Calibration Calibration of the total nitrogen model proceeded by changing model dispersion coefficients so that model output of nitrogen concentrations matched measured data. Generally, several model runs of each system were required to match the water column measurements. Dispersion coefficient (E) values were varied through the modeled system by setting different values of E for each grid material type, as designated in Figure VI-2. Observed values of E (Fischer, et al., 1979) vary between order 10 and order 1000 m2/sec for riverine estuary systems characterized by relatively wide channels (compared to channel depth) with moderate currents (from tides or atmospheric forcing). Generally, the relatively quiescent areas of Apponagansett Bay (Dike Marsh) require values of E that are lower compared to the riverine estuary systems evaluated by Fischer, et al., (1979). Observed values of E in these calmer areas typically range between order 10 and order 0.001 m2/sec (USACE, 2001). The final values of E used in each sub-embayment of the modeled systems are presented in Table VI-3. These values were used to develop the “best-fit” total nitrogen model calibration. For the case of TN modeling, “best fit” can be defined as minimizing the error between the model and data at all sampling locations, utilizing reasonable ranges of dispersion coefficients within each sub- embayment.

Comparisons between model output and measured nitrogen concentrations are shown in plots presented in Figure VI-3. In these plots, means of the water column data and a range of two standard deviations of the annual means at each individual station are plotted against the modeled maximum, mean, and minimum concentrations output from the model at locations which corresponds to the water quality monitoring stations.

For model calibration, the mid-point between maximum modeled TN and average modeled TN was compared to mean measured TN data values, at each water-quality monitoring station. The calibration target would fall between the modeled mean and maximum TN because the monitoring data are collected, as a rule, during mid ebb tide.

Also presented in this figure are unity plot comparisons of measured data verses modeled target values for the system. The model fit is exceptional for the Apponagansett Bay System, with rms error of 0.07 mg/L and an R2 correlation coefficient of 0.98.

A contour plot of calibrated model output is shown in Figure VI-4 for Apponagansett Bay System. In the figure, color contours indicate nitrogen concentrations throughout the model domain. The output in the figure show average total nitrogen concentrations, computed using the full 5-tidal-day model simulation output period.

110 MASSACHUSETTS ESTUARIES PROJECT

Figure VI-2. Map of Apponagansett Bay water quality model longitudinal dispersion coefficients. Color patterns designate the different areas used to vary model dispersion coefficient values.

111 MASSACHUSETTS ESTUARIES PROJECT

Table VI-3. Values of longitudinal dispersion coefficient, E, used in calibrated RMA4 model runs of salinity and nitrogen concentration for Apponagansett Bay System. E Embayment Division m2/sec Buzzards Bay 10.0 Dike Marsh Channels 4.0 Dike Marsh 3.0 Inner Harbor 4.5 Outer Harbor 6.5 River 4.0 Bridge 5.0 Apponagansett Brook 7.0 Buttonwood Brook 7.0

Figure VI-3. Comparison of measured total nitrogen concentrations and calibrated model output at stations in Apponagansett Bay System. For the left plot, station labels correspond with those provided in Table VI-1. Model output is presented as a range of values from minimum to maximum values computed during the simulation period (triangle markers), along with the average computed concentration for the same period (square markers). Measured data are presented as the total yearly mean at each station (circle markers), together with ranges that indicate ± one standard deviation of the entire dataset. For the plots to the right, model calibration target values are plotted against measured concentrations, together with the unity line.

VI.2.5 Model Salinity Verification In addition to the model calibration based on nitrogen loading and water column measurements, numerical water quality model performance is typically verified by modeling salinity. This step was performed for the Apponagansett Bay System using salinity data collected at the same stations as the nitrogen data. The only required inputs into the RMA4 salinity model of each system, in addition to the RMA2 hydrodynamic model output, were salinities at the model open boundary, and groundwater inputs. The open boundary salinity was set at 31.68 ppt. The total groundwater input used for the model was 10.77 ft3/sec (26,357

112 MASSACHUSETTS ESTUARIES PROJECT m3/day) distributed amongst the watersheds. Groundwater flows were distributed evenly within each watershed through grid cells within each watershed’s boundary.

Figure VI-4. Contour plots of average total nitrogen concentrations from results of the present conditions loading scenario, for Apponagansett Bay System. The approximate location of the sentinel threshold station for Apponagansett Bay System (AB-4) is shown.

Comparisons of modeled and measured salinities are presented in Figure VI-5, with contour plots of model output shown in Figure VI-6. Though model dispersion coefficients were not changed from those values selected through the nitrogen model calibration process, the model capably represents salinity gradients in Apponagansett Bay System. The rms error of the models was 1.92 ppt, and correlation coefficient was 0.99. The salinity verification provides a further independent confirmation that model dispersion coefficients and represented freshwater inputs to the model correctly simulate the real physical systems.

113 MASSACHUSETTS ESTUARIES PROJECT

Figure VI-5. Comparison of measured and calibrated model output at stations in Apponagansett Bay System. For the left plots, stations labels correspond with those provided in Table VI-1. Model output is presented as a range of values from minimum to maximum values computed during the simulation period (triangle markers), along with the average computed salinity for the same period (square markers). Measured data are presented as the total yearly mean at each station (circle markers), together with ranges that indicate ± one standard deviation of the entire dataset. For the plots to the right, model calibration target values are plotted against measured concentrations, together with the unity line.

VI.2.6 Build-Out and No Anthropogenic Load Scenarios To assess the influence of nitrogen loading on total nitrogen concentrations within the embayment system, two standard water quality modeling scenarios were run: a “build-out” scenario based on potential development (described in more detail in Section IV) and a “no anthropogenic load” or “no load” scenario assuming only atmospheric deposition on the watershed and sub-embayment, as well as a natural forest within each watershed. Comparisons of the alternate watershed loading analyses are shown in Table VI-4. Loads are presented in kilograms per day (kg/day) in this Section, since it is inappropriate to show benthic flux loads in kilograms per year due to seasonal variability.

Table VI-4. Comparison of sub-embayment watershed loads used for modeling of present, build-out, and no-anthropogenic (“no-load”) loading scenarios of the Apponagansett Bay System. These loads do not include direct atmospheric deposition (onto the sub-embayment surface) or benthic flux loading terms. present build out no load build out no load sub-embayment load % % (kg/day) (kg/day) (kg/day) change change Apponagansett Harbor 24.36 23.42 -0.04 2.65 -89.13% Apponagansett Bay 2.38 1.48 -0.38 0.13 -94.60% Dike Marsh 6.74 7.96 +18.08% 2.53 -62.45% Surface Water Sources Buttonwood Brook 25.33 26.02 +2.74% 3.23 -87.24% Apponagansett Brook 3.04 3.24 +6.39% 0.39 -87.22%

114 MASSACHUSETTS ESTUARIES PROJECT

Figure VI-6. Contour plots of modeled salinity (ppt) in Apponagansett Bay System.

VI.2.6.1 Build-Out In general, certain sub-embayments would be impacted more than others. The build-out scenario indicates that there would be a increase in watershed nitrogen load to the Buttonwood Brook and Apponagansett Brook as a result of potential future development, and minor reductions within the Apponagansett Harbor. Specific watershed areas would experience large load increases, for example the loads to Dike Marsh would increase 18% from the present day loading levels. For the no load scenarios, a majority of the load entering the watershed is

115 MASSACHUSETTS ESTUARIES PROJECT removed; therefore, the load is significantly lower than existing conditions by over 80% overall for a majority of the sub-embayments.

For the build-out scenario, a breakdown of the total nitrogen load entering the Apponagansett Bay System sub-embayments is shown in Table VI-5. The benthic flux for the build-out scenarios is assumed to vary proportional to the watershed load, where an increase in watershed load will result in an increase in benthic flux (i.e., a positive change in the absolute value of the flux), and vise versa.

Projected benthic fluxes (for both the build-out and no load scenarios) are based upon projected PON concentrations and watershed loads, determined as:

(Projected N flux) = (Present N flux) * [PONprojected]/[PONpresent] where the projected PON concentration is calculated by,

[PONprojected] = Rload * ∆PON + [PON(present offshore)], using the watershed load ratio,

Rload = (Projected N load) / (Present N load), and the present PON concentration above background,

∆PON = [PON(present flux core)] – [PON(present offshore)].

Table VI-5. Build-out sub-embayment and surface water loads used for total nitrogen modeling of the Apponagansett Bay System, with total watershed N loads, atmospheric N loads, and benthic flux. direct watershed benthic flux atmospheric sub-embayment load net deposition (kg/day) (kg/day) (kg/day) Apponagansett Harbor 23.42 4.48 -5.34 Apponagansett Bay 1.48 3.64 -14.40 Dike Marsh 7.96 0.58 0.04 Surface Water Sources Buttonwood Brook 26.02 -- -- Apponagansett Brook 3.24 -- --

Following development of the nitrogen loading estimates for the build-out scenario, the water quality model of Apponagansett Bay System was run to determine nitrogen concentrations within each sub-embayment (Table VI-6). Total nitrogen concentrations in the receiving waters (i.e., Buzzards Bay) remained identical to the existing conditions modeling scenarios. The stations in Apponagansett Harbor show slight increase in nitrogen from head of the system to the inlet, with the largest increase within at the head of the bay due to the load increases in Buttonwood Brook and Apponagansett Brook. Color contours of model output for the build-out scenario are present in Figure VI-7. The range of nitrogen concentrations shown are the same as for the plot of present conditions in Figure VI-4, which allows direct comparison of nitrogen concentrations between loading scenarios.

116 MASSACHUSETTS ESTUARIES PROJECT

Table VI-6. Comparison of model average total N concentrations from present loading and the build-out scenario, with percent change, for the Apponagansett Bay System. Sentinel threshold station is in bold print. monitoring present build-out Sub-Embayment % change station (mg/L) (mg/L) Head of Bay AB-1A 1.34 1.38 +3.6% North Little Island AB-4 0.53 0.54 +1.6% Upper Basin-Lower AB-3 0.45 0.45 +1.1% Lower Basin-Upper AB-2 0.38 0.38 +0.6% Lower Basin AB-6 0.34 0.34 +0.3%

Figure VI-7. Contour plots of modeled total nitrogen concentrations (mg/L) in Apponagansett Bay System, for projected build-out loading conditions, and bathymetry. The approximate location of the sentinel threshold station for Apponagansett Bay System (AB-4) is shown.

117 MASSACHUSETTS ESTUARIES PROJECT

VI.2.6.2 No Anthropogenic Load A breakdown of the total nitrogen load entering each sub-embayment for the no anthropogenic load (“no load”) scenario is shown in Table VI-7. The benthic flux input to each embayment was reduced (toward zero) based on the reduction in the watershed load (as discussed in §VI.2.6.1). Compared to the modeled present conditions and build-out scenario, atmospheric deposition directly to each sub-embayment becomes a greater percentage of the total nitrogen load as the watershed load and related benthic flux decrease.

Table VI-7. “No anthropogenic loading” (“no load”) sub-embayment and surface water loads used for total nitrogen modeling of Apponagansett Bay System, with total watershed N loads, atmospheric N loads, and benthic flux direct benthic flux watershed load atmospheric sub-embayment net (kg/day) deposition (kg/day) (kg/day) Apponagansett Harbor 2.65 4.48 -2.70 Apponagansett Bay 0.13 3.64 -10.07 Dike Marsh 2.53 0.58 0.03 Surface Water Sources Buttonwood Brook 3.23 -- -- Apponagansett Brook 0.39 -- --

Following development of the nitrogen loading estimates for the no load scenario, the water quality model was run to determine nitrogen concentrations within each sub-embayment. Again, total nitrogen concentrations in the receiving waters (i.e., Buzzards Bay) remained identical to the existing conditions modeling scenarios. The relative change in total nitrogen concentrations resulting from “no load” was significant as shown in Table VI-8, with reductions ranging from 14% in the lower Apponagansett Bay with greater than 81% reduction in total nitrogen at the head of Apponagansett Harbor. Results for each system are shown pictorially in Figure VI-8.

Table VI-8. Comparison of model average total N concentrations from present loading and the no anthropogenic (“no load”) scenario, with percent change, for the Apponagansett Bay System. Loads are based on atmospheric deposition and a scaled N benthic flux (scaled from present conditions). Sentinel threshold station is in bold print. monitoring present no-load Sub-Embayment % change station (mg/L) (mg/L) Head of Bay AB-1A 1.34 0.25 -81.6% North Little Island AB-4 0.53 0.29 -44.7% Upper Basin-Lower AB-3 0.45 0.30 -33.8% Lower Basin-Upper AB-2 0.38 0.29 -22.8% Lower Basin AB-6 0.34 0.29 -14.0%

118 MASSACHUSETTS ESTUARIES PROJECT

Figure VI-8. Contour plots of modeled total nitrogen concentrations (mg/L) in Apponagansett Bay System, for no anthropogenic loading conditions, and bathymetry. The approximate location of the sentinel threshold station for Apponagansett Bay System (AB-4) is shown.

119 MASSACHUSETTS ESTUARIES PROJECT

VII. ASSESSMENT OF EMBAYMENT NUTRIENT RELATED ECOLOGICAL HEALTH

The nutrient related ecological health of an estuary can be gauged by the nutrient, chlorophyll, and oxygen levels of its waters and the plant (eelgrass, macroalgae) and animal communities (fish, shellfish, infauna) which it supports. For the Apponagansett Bay Estuary, the ecological assessment is based upon data from the water quality monitoring database (1999- 2008) and surveys of eelgrass distribution, macroalgae coverage, benthic animal communities and sediment characteristics, shell fish and water bird populations and dissolved oxygen records conducted during the period 2003-2006. These data form the basis of an assessment of this system’s present health, and when coupled with a full water quality synthesis and projections of future conditions based upon the water quality modeling effort, will support complete nitrogen threshold development for these systems (Section VIII).

VII.1 OVERVIEW OF BIOLOGICAL HEALTH INDICATORS There are a variety of indicators that can be used in concert with water quality monitoring data for evaluating the ecological health of embayment systems. The best biological indicators are those species which are non-mobile and which persist over relatively long periods, if environmental conditions remain constant. The concept is to use species which integrate environmental conditions over seasonal to annual intervals. The approach is particularly useful in environments where high-frequency variations in structuring parameters (e.g. light, nutrients, dissolved oxygen, etc.) are common, making adequate field sampling difficult.

As a basis for a nitrogen thresholds determination, MEP focused on major habitat quality indicators: (1) bottom water dissolved oxygen and chlorophyll a (Section VII.2), (2) eelgrass distribution over time (Section VII.3), and (3) benthic animal communities (Section VII.4). Additional surveys of macroalgae coverage (Section VII.5) and shellfish, finfish, and water bird surveys (see Appendix 1) were conducted by staff from the Lloyd Center as part of the Turn the Tide Project1. Dissolved oxygen depletion is frequently the proximate cause of habitat quality decline in coastal embayments (the ultimate cause being nitrogen loading). However, oxygen conditions can change rapidly and frequently show strong tidal and diurnal patterns. Even if severe levels of oxygen depletion occur only infrequently, they can have important effects on system health. To capture this variation, the MEP Technical Team deployed dissolved oxygen sensors at three sites around the Inner Basin of Apponagansett Bay to record the frequency and duration of low oxygen conditions during the critical summer period. The Inner Basin was selected based on results of oxygen grab samples (BayWatchers). The MEP habitat analysis uses eelgrass and benthic animal communities as sentinel indicators of nitrogen over-loading to coastal embayments. Eelgrass is a fundamentally important species in the ecology of shallow coastal systems, providing both habitat structure and sediment stabilization. Mapping of the eelgrass beds within Apponagansett Bay was conducted for comparison to historic records (DEP Eelgrass Mapping Program, C. Costello; Lloyd Center eelgrass and macroalgae surveys). Temporal trends in the distribution of eelgrass beds are used by the MEP to assess the stability of the habitat and to determine trends potentially related to water quality. Benthic animal communities are a critical component of estuarine ecosystems and provide food for commercial and recreational fish species, shorebirds and general ecosystem functioning. Impairment of this habitat has cascading negative impacts to both estuarine function and offshore fisheries. Eelgrass beds can decrease and benthic animal habitats can become degraded within

1 Turn-the-Tide was a project to support MEP efforts in the Town of Dartmouth and was supported by the Town, private donors and UMD.

120

MASSACHUSETTS ESTUARIES PROJECT embayments in response to a variety of causes, but throughout almost all of the embayments within southeastern Massachusetts, the primary cause appears to be related to increases in embayment nitrogen levels and for eelgrass, associated decreases in light penetration. Within Apponagansett Bay System, temporal changes in eelgrass distribution and spatial analysis of benthic animal habitat quality can provide a basis for evaluating increases in nitrogen loading and nutrient enrichment.

In areas that do not support eelgrass beds, benthic animal indicators were used to assess the level of habitat health from “healthy” (low organic matter loading, high D.O.) to “highly stressed” (high organic matter loading-low D.O.). The basic concept is that certain species or species assemblages reflect the quality of their habitat. Benthic animal species from sediment samples were identified and the environments ranked based upon the fraction of healthy, transitional, and stressed indicator species. The analysis is based upon life-history information on the species and a wide variety of field studies within southeastern Massachusetts waters, including the Wild Harbor oil spill (Hampson, 1978), benthic population studies in Buzzards Bay (e.g. Hampson, 1989) and New Bedford (SMAST, unpublished data), and more recently the Woods Hole Oceanographic Institution Nantucket Harbor Study (Howes et al. 1997). These data are coupled with the level of diversity (H’) and Evenness (E) of the benthic community and the total number of individuals to determine the infaunal habitat quality.

VII.2 BOTTOM WATER DISSOLVED OXYGEN Dissolved oxygen levels near atmospheric equilibration are important for maintaining healthy animal and plant communities. Short-duration oxygen depletions can significantly affect communities even if they are relatively rare on an annual basis. For example, for the Chesapeake Bay it was determined that restoration of nutrient degraded habitat requires that instantaneous oxygen levels not drop below 3.8 mg L-1. Massachusetts State Water Quality Classification indicates that SA (high quality) waters maintain oxygen levels above 6 mg L-1. The tidal waters of Apponagansett Bay are currently listed under this Classification as SA. It should be noted that the Classification system represents the water quality that the embayment should support, not the existing level of water quality. It is through the MEP and TMDL processes that management actions are developed and implemented to keep or bring the existing conditions in line with the Classification.

Dissolved oxygen levels in temperate embayments vary seasonally, due to changes in oxygen solubility, which varies inversely with temperature. In addition, biological processes that consume oxygen from the water column (water column respiration) vary directly with temperature, with several fold higher rates in summer than winter (Figure VII-1). It is not surprising that the largest levels of oxygen depletion (departure from atmospheric equilibrium) and lowest absolute levels (mg L-1) are found during the summer in southeastern Massachusetts embayments when water column respiration rates are greatest. Since oxygen levels can change rapidly, several mg L-1 in a few hours, traditional grab sampling programs typically underestimate the frequency and duration of low oxygen conditions within shallow embayments (Taylor and Howes, 1994). To more accurately capture the degree of bottom water dissolved oxygen depletion during the critical summer period, autonomously recording oxygen sensors were moored 30 cm above the embayment bottom within key regions of Apponagansett Bay embayments (Figure VII-2). The sensors (YSI 6600) were first calibrated in the laboratory and then checked with standard oxygen mixtures at the time of initial instrument mooring deployment. In addition periodic calibration samples were collected at the sensor depth and assayed by Winkler titration (potentiometric analysis, Radiometer) during each deployment. Each instrument mooring was serviced and calibration samples collected at least

121

MASSACHUSETTS ESTUARIES PROJECT biweekly and sometimes weekly during a minimum deployment of 30 days within the interval from July through mid-September. The mooring data from the Apponagansett Bay was collected in 2003 and 2004.

Watercolumn Respiration Rates

40 35 30 25 20 15

WCR (uM/d) 10 5 0 FAJ JSND Date

Figure VII-1. Average water column respiration rates (micro-Molar/day) from water collected throughout the Popponesset Bay System (Schlezinger and Howes, unpublished data). Rates vary ~7 fold from winter to summer as a result of variations in temperature and organic matter availability.

Similar to other embayments in southeastern Massachusetts, the Apponagansett Bay system evaluated in this assessment showed high frequency variation, apparently related to diurnal and sometimes tidal influences. Nitrogen enrichment of embayment waters generally manifests itself in the dissolved oxygen record, both through oxygen depletion and through the magnitude of the daily excursion. The high degree of temporal variation in bottom water dissolved oxygen concentration at each mooring site, underscores the need for continuous monitoring within these systems.

Dissolved oxygen and chlorophyll-a records were examined both for temporal trends and to determine the percent of the deployment periods that these parameters were below/above various benchmark concentrations (Tables VII-1, VII-2). These data indicate both the temporal pattern of minimum or maximum levels of these critical nutrient related constituents, as well as the intensity of the oxygen depletion events and phytoplankton blooms. However, it should be noted that the frequency of oxygen depletion needs to be integrated with the actual temporal pattern of oxygen levels, specifically as it relates to daily oxygen excursions. The use of only the duration of oxygen below, for example 4 mg L-1, can underestimate the level of habitat impairment in these locations. The effect of nitrogen enrichment is to cause oxygen depletion; however, with increased phytoplankton (or epibenthic algae) production, oxygen levels will rise in daylight to above atmospheric equilibration levels in shallow systems (generally ~7-8 mg L-1 at the mooring sites). The high oxygen excursions were not typical of the upper mooring site, but modest high excursions to 9-10 mg L-1 were observed infrequently at the mid and lower sites indicating only moderate eutrophication by that measure (Figures VII-3, 5, 7). Likewise, oxygen levels at Apponagansett Upper and the Lower sites fell below 3 mg/L for an insignificant period, less than 6 hours total over the total record, and were >4 mg L-1 for 94% and 91% of the record

122

MASSACHUSETTS ESTUARIES PROJECT respectively, with the mid site being >4 mg L-1. Daily dissolved oxygen excursions at the Upper embayment site are therefore moderate, usually ~3 mg/L or less from the diurnal maximum to minimum, lower than the same measure at the Mid and Lower sites (Figure VII-5 and 7). However, the percentage of time that the Upper site fell below 6 mg/L was high, 92%, indicating a chronic condition of somewhat depressed water column oxygen. It is not clear why this pattern of oxygen depletion occurs at the Upper site. The pattern may be due to the balance between two processes: first, shallow water and elevated temperatures and sediment oxygen demand (including macroalgae) that is fueled by high nitrogen loads transported by the two principal streams discharging to the Upper embayment; and second, relatively short water residence time in the Upper embayment whereby Upper embayment nutrient-enriched waters are rapidly mixed with and replaced with lower nutrient, more oxygen-rich Buzzard Bay water. Both the upper and mid sites generally maintained moderate oxygen levels (>4 mg L-1) with the lower site showing greater depletion. However, the oxygen dynamics at the lower site are likely influenced by low oxygen waters ebbing from the adjacent salt marsh, due to the salt marsh's naturally organic enrichment. Salt marshes typically have periodic oxygen depletions (even anoxia) that are not an impairment to the marsh, but would be an impairment to open water basins. Overall, the oxygen records are comparable to the long-term grab sample data where the uppermost tidal reach has oxygen declines to <4 mg L-1 in 4% of the samples (N=248) versus 6% of the mooring record but remained >4 mg L-1 in 99% of the samples adjacent Little Island.

Analysis of the level of oxygen excursion at the Mid and Lower mooring sites showed the dissolved oxygen levels were periodically in the 8-10 mg/l range indicating some moderate effects of eutrophication (Figure VII-4), and the larger daily oxygen excursion may also reflect the influence of the large amounts of macroalgae on the bottom in the Mid mooring area (See Figure VII-10 below). At the Lower embayment site adjacent the Dike Marsh outlet, the levels of dissolved oxygen are in excess of 7-8 mg/l with a frequency similar than at the Mid embayment site (Figure VII-5). Dissolved oxygen at the Mid and Lower sites remained below 6 mg/L for 51% and 45% of the deployment period but very low oxygen levels generally were not observed.

Generally low to moderate effects of nitrogen enrichment are demonstrated in the record of chlorophyll-a at the Mid and Lower embayment sites in 2004, with no events >20 µg/L occurring at the either site (Table VII-2). On the other hand in late July 2003 at the Apponagansett Upper site (Figure VII-6), concentrations of chlorophyll-a exceeded 20 µg/L for a short period and then for an extended period in mid August 2003, exceeding 20 µg/L during 25% of the 45 day record period and reaching 60 µg/L on two occasions. The August 2003 algal bloom may have been a response to a period of moderate rainstorms between August 4-13, 2003 which amounted to 3.08 inches of rainfall and likely flushed a supply of fresh nutrients into the embayment. As noted above, the Upper Apponagansett site is closest to the two main sources of stream runoff, Apponagansett Bay Brook and Buttonwood Brook, and both have high nitrogen loading. The mean chlorophyll-a value at the Upper embayment site (19.9 ug L-1) was about twice those recorded at the Mid (10.9 ug L-1) and Lower (8.38 ug L-1) Inner Basin sites (Table VII-2; Figs VII-7, 8). In contrast, the Outer Basin averaged 6.5 - 7.0 ug L-1 in the long- term monitoring record. Chlorophyll-a summer averages >10 ug L-1 have been used to indicate eutrophication.

The dissolved oxygen, chlorophyll and total nitrogen data vary spatially with an enriched zone (Upper) and low enrichment zone (Lower) site and an intermediate site (Mid). The low oxygen conditions periodically at the Lower site are almost certainly the result of ebbing waters from Dike Marsh while the Upper and Mid sites results from moderate nitrogen and organic matter enrichment, resulting in oxygen depletions to levels moderately stressful to benthic animal communities. In addition, analysis of benthic community habitat quality needs to also

123

MASSACHUSETTS ESTUARIES PROJECT include the distribution of macroalgal accumulations in the upper and western cove regions of the Inner Basin. It appears that major regions of the Inner Basin are currently showing low to moderate impairment from oxygen depletion and organic matter enrichment (chlorophyll-a) and that these observations are consistent with the level and spatial distribution of nitrogen enrichment. Further evidence of the nitrogen associated habitat quality (both eelgrass and benthic animals) in Apponagansett Bay is presented in the sections below.

Figure VII-2. Aerial photograph of Apponagansett Bay system in Dartmouth showing locations of Dissolved Oxygen mooring deployments conducted in the Summer of 2003-2004.

124

MASSACHUSETTS ESTUARIES PROJECT

125

MASSACHUSETTS ESTUARIES PROJECT

Dartmouth, Apponagansett - Upper

12

10

8

6

Dissolved Oxygen (mg/L) Oxygen Dissolved 4

2

0 7/10/03 7/15/03 7/20/03 7/25/03 7/30/03 8/4/03 8/9/03 8/14/03 8/19/03 8/24/03 Time

Figure VII-3. Bottom water record of dissolved oxygen at Apponagansett Bay Upper station, Summer 2003. Calibration samples represented as red dots.

Dartmouth, Apponagansett - Mid

12

10

8

6

Dissolved Oxygen (mg/L) Oxygen Dissolved 4

2

0 7/6/04 7/11/04 7/16/04 7/21/04 7/26/04 7/31/04 Time

Figure VII-4. Bottom water record of dissolved oxygen in Apponagansett Bay Mid station, Summer 2004. Calibration samples represented as red dots.

126

MASSACHUSETTS ESTUARIES PROJECT

Dartmouth, Apponagansett - Lower

12

10

8

6

Dissolved (mg/L) Oxygen 4

2

0 7/6/04 7/11/04 7/16/04 7/21/04 7/26/04 7/31/04 Time

Figure VII-5. Bottom water record of dissolved oxygen in Apponagansett Bay Lower station, Summer 2004. Calibration samples represented as red dots.

Dartmouth, Apponagansett - Upper

70

60

50

40

30

Total Chlorophyll Pigment (ug/L) Pigment Chlorophyll Total 20

10

0 7/10/03 7/15/03 7/20/03 7/25/03 7/30/03 8/4/03 8/9/03 8/14/03 8/19/03 8/24/03 Time

Figure VII-6. Bottom water record of Chlorophyll-a at Apponagansett Bay Upper station, Summer 2003. Calibration samples represented as red dots.

127

MASSACHUSETTS ESTUARIES PROJECT

Dartmouth, Apponagansett - Mid

25

20

15

10 Total Chlorophyll Pigment (ug/L)

5

0 7/6/04 7/11/04 7/16/04 7/21/04 7/26/04 7/31/04 Time

Figure VII-7. Bottom water record of Chlorophyll-a in Apponagansett Bay Mid station, Summer 2004. Calibration samples represented as red dots.

Dartmouth, Apponagansett - Lower

20

18

16

14

12

10

8

6 Total Chlorophyll Pigment (ug/L) Pigment Chlorophyll Total

4

2

0 7/6/04 7/11/04 7/16/04 7/21/04 7/26/04 7/31/04 Time

VII-8. Bottom water record of Chlorophyll-a in Apponagansett Bay Lower station, Summer 2004. Calibration samples represented as red dots.

128

MASSACHUSETTS ESTUARIES PROJECT

Table VII-1. Duration, in days that in situ sensors were operated and the time periods that bottom water oxygen levels were below various benchmark oxygen levels. Even short-term oxygen declines below 3 mg/l result in a high level of stress to benthic and fish communities. Data collected by the Coastal Systems Program, SMAST. Total <6 mg/L <5 mg/L <4 mg/L <3 mg/L Mooring ID. Deployment Duration Duration Duration Duration Start Date End Date (Days) (Days) (Days) (Days) (Days) Apponagansett Bay – Upper 7/10/2003 8/25/2003 46.0 42.29 14.21 2.91 0.19 Mean 0.76 0.27 0.14 0.05 Min 0.02 0.02 0.02 0.02 Max 1.85 0.86 0.34 0.13 S.D. 0.39 0.21 0.10 0.05 Apponagansett Bay - Mid 7/6/2004 8/2/2004 27.0 13.82 4.59 0.13 0.01 Mean 0.41 0.18 0.04 0.01 Min 0.01 0.01 0.01 0.01 Max 1.46 0.42 0.09 0.01 S.D. 0.32 0.13 0.05 N/A Apponagansett Bay - Lower 7/6/2004 8/2/2004 25.7 11.65 6.33 2.31 0.25 Mean 0.39 0.25 0.14 0.06 Min 0.01 0.01 0.01 0.01 Max 1.70 0.63 0.27 0.11 S.D. 0.34 0.16 0.07 0.05

129

MASSACHUSETTS ESTUARIES PROJECT

Table VII-2. Duration (% of deployment time ) that chlorophyll a levels exceed various benchmark levels within the embayment system. “Mean” represents the average duration of each event over the benchmark level and “S.D.” its standard deviation. Data collected by the Coastal Systems Program, SMAST. Total >5 ug/L >10 ug/L >15 ug/L >20 ug/L >25 ug/L Deployment Duration Duration Duration Duration Duration Embayment System Start Date End Date (Days) (Days) (Days) (Days) (Days) (Days)

Apponagansett Bay - Upper 7/10/2003 8/25/2003 45.3 96% 71% 43% 25% 13% Mean Chl Value = 19.90 ug/L Mean 2.06 0.60 0.35 0.25 0.15 S.D. 3.34 1.28 0.64 0.40 0.14 Apponagansett Bay - Mid 7/6/2004 8/2/2004 27.0 100% 60% 5% 0% 0% Mean Chl Value = 10.88 ug/L Mean 13.50 0.39 0.13 N/A N/A S.D. 19.03 0.35 0.10 N/A N/A Apponagansett Bay - Lower 7/6/2004 8/2/2004 25.7 85% 31% 2% 0% 0% Mean Chl Value = 8.38 ug/L Mean 1.22 0.26 0.08 N/A N/A S.D. 2.61 0.21 0.03 N/A N/A

130

MASSACHUSETTS ESTUARIES PROJECT

VII.3 EELGRASS DISTRIBUTION - TEMPORAL ANALYSIS Eelgrass surveys and analysis of historical data was conducted for the Apponagansett Bay System by the MassDEP Eelgrass Mapping Program as part of the MEP Technical Team with additional surveying done previously (Costa 1988). MassDEP Surveys were conducted in 1995 and 2001 as part of this program with additional data collected in 2004 by Lloyd Center staff. Additional analysis of available aerial photos from 1951 was used to reconstruct the eelgrass distribution when watershed development was at a low level. The 1951 data were only anecdotally validated and used only for comparison in the management analysis. In contrast, the 1995 and 2001 MassDEP eelgrass distribution maps and the maps based on 1984 and 1985 surveys (Costa 1988) were field validated. In addition, eelgrass observations were also made throughout Apponagansett Bay and by SMAST Staff during sampling for mooring calibrations, benthic regeneration and infaunal community analysis in 2003, 2004 and 2005. The primary use of the data is to indicate: (a) if eelgrass once or currently colonizes a basin and (b) if large-scale system-wide shifts have occurred. Integration of these data sets provides a view of temporal trends in eelgrass distribution from 1951 to 1995 to 2005 (Figure VII-9 and VII- 10); the period in which watershed nitrogen loading significantly increased to its present level. This temporal information can be used to determine the stability of the eelgrass community.

From the surveys completed in 1984/85, 1995, 2001 and confirmed in 2003-2005, eelgrass habitat was not found within inner Apponagansett Bay (north of the causeway and bridge). Based on the 2001 eelgrass survey, conducted by the DEP Eelgrass Mapping Program, the eelgrass appears to be distributed only in outer Apponagansett Bay south of the causeway and bridge (Figure VII-9). The 1951 map of the inner embayment indicates the "possibility" of small beds in the inner basin, but as stated by Costa for this upper basin, not possible to determine with certainty because "identification of photographs is difficult in some areas because of drift material". The upper basin supports drift macroalgae, so these 1951 interpretations (unvalidated) were not used in the temporal analysis of eelgrass coverage in this system. The 1984-85 surveys did identify eelgrass beds along the east and west shore in outer Apponagansett Bay, consistent with the 1951 photographic analysis and 1995 and 2001 surveys. These areas do not accumulate drift macro-algae. There have been reports of a small patch of eelgrass southeast of Little Island in the inner basin, but a targeted survey in 2006 failed to find eelgrass. Further, it is reported that eelgrass has not been reported south of Little Island over decades (BBP 2015)2.In the outer embayment eelgrass beds appear to have diminished in size between 1995 and 2001 along the eastern shore, though none have disappeared over that interval.

Other anecdotal evidence indicates that Apponagansett Bay has had eelgrass beds historically as evidenced by the bay scallop fishery it supported into the mid-1980’s. Since scallops were harvested in eelgrass beds throughout the Town waters and not always separated by water body in Town Annual Reports, it is not possible to parse the volume and timing of scallops taken in inner versus outer Apponagansett Bay. Town Report figures show that in 1946 5,000 bushels of scallops were harvested from Apponagansett Bay and substantial scallop harvests continued through the early 1970’s, with several annual catches of 6,000 to 7,000 bushels occurring in 1965 and 1971, respectively. A few years of limited harvests occurred in the mid-1980’s with the last substantial scallop harvest in Dartmouth of 1,975 bushels occurring in 1985 and between 1994 and 1996 small recreational catches of bay

2 http://buzzardsbay.org/historical-eelgrass-apponagansett-bay.htm

131

MASSACHUSETTS ESTUARIES PROJECT scallops were recorded, but the source areas of these last harvests were unlikely to have been in the inner bay.

Figure VII-9. Eelgrass bed distribution within Apponagansett Bay system. The 1995 coverage is depicted by the green outline inside of which circumscribes the eelgrass beds. The yellow (2001) areas were mapped by DEP. All data was provided by the DEP Eelgrass Mapping Program.

132

MASSACHUSETTS ESTUARIES PROJECT

Figure VII-10. Eelgrass bed distribution within Apponagansett Bay system. The 1951 coverage is depicted by the dark green outline (hatched area) inside of which circumscribes the eelgrass beds. In the composite photograph, the light green outline depicts the 1995 eelgrass coverage and the yellow outlined areas circumscribe the eelgrass coverage in 2001. The 1995 and 2001 areas were mapped by DEP. All data was provided by the DEP Eelgrass Mapping Program.

133

MASSACHUSETTS ESTUARIES PROJECT

The existing coverage data indicates a gradual decline of eelgrass within Apponagansett Bay, primarily in the lower basin. This is consistent with the moderate chlorophyll a and periodic low dissolved oxygen levels and high water column nitrogen concentrations within this system. It is not documented that eelgrass beds were significant in the Inner Bay in recent history (last 50 years). This is consistent with the moderate to high chlorophyll a levels (mooring average 8.4 - 19.9 ug L-1, long-term average 7.7 - 9.1 ug L-1, BayWatcher 1999-2008). The historic record and the persistence of a gradual decline of beds in the outer embayment suggests a system just beyond its ability to assimilate additional nitrogen inputs without impairment. This indicates that should nitrogen loading to the estuary be lowered, eelgrass coverage should expand. If loading to the inner basin is lowered the result would include a lowering of nitrogen levels in the lower basin (below the bridge) and associated improvements of eelgrass habitat and benthic animal habitat, as well.

In systems like Apponagansett Bay, the general pattern is for highest nitrogen levels to be found within the innermost basins, with concentrations declining moving toward the tidal inlet. In Apponagansett Bay the general gradient in nitrogen is influenced by the large salt marsh, Dike Marsh to the southwest with tidal flow into and out of this marsh influencing circulation within the inner bay. The location of the embayment relative to its watershed means that most of the nitrogen from the watershed enters the uppermost region of the inner basin via discharges from Buttonwood Brook and Apponagansett Bay Brook. The nitrogen gradient throughout Apponagansett Bay (0.51-0.30 mg L-1) is comparable to other estuaries in s.e. Massachusetts with similar nitrogen loading, basin geomorphology and tidal flushing. In the Megansett- Squeteague Estuary a similar nitrogen gradient is observed, and also a comparable eelgrass coverage, with no historic evidence of eelgrass habitat in the inner basin (Squeteague Harbor) but with fringing beds throughout the outer basin (Megansett Harbor). It is typical of these estuaries to have eelgrass habitat in the outer or lower basins near the tidal inlet and no eelgrass habitat in the inner or upper basins where watershed nitrogen inputs are focused. In Apponagansett Bay the gradual decline in eelgrass in the lower basin (below the bridge) follows the pattern of loss from the upper or deeper areas to the lower or shallower areas. The temporal pattern is a “retreat” of beds toward the region of the tidal inlet. The gradual decline of eelgrass coverage in the lower basin is similar to that found elsewhere (Green Pond and Quissett Harbor, Falmouth; Westport Rivers, Westport; Slocums River, Dartmouth).

Other factors which influence eelgrass bed loss in embayments were examined for the Apponagansett Bay system, though the decline/loss seems completely in-line with nitrogen enrichment. A brief listing of non-nitrogen related factors is useful. Eelgrass bed loss could be related to mooring density or docks/piers and to shellfishing. Apponagansett Bay has extensive recreational boating activities, mainly as a harbor and mooring area. Several docks mainly in the lower basin add to boating activities, but likely have only a small effect on eelgrass (note docks found mainly where eelgrass persists). Shellfishing along the eastern shore of the inner embayment has not been permitted for some time due to chronic excessive bacterial levels, so it was not a likely contributing factor in the loss of the last eelgrass bed. If water quality were to improve in inner Apponagansett Bay, or have been found in the inner basin, the area southeast of Little Island would be the likely place due to its access to the incoming waters from Buzzards Bay.

Based on the MEP analysis of chlorophyll, D.O., light penetration, nutrient concentrations, and historical patterns of bed loss, the Technical Team is confident that if nitrogen loading were to decrease, eelgrass could be restored in the lowest portion of the inner Apponagansett Bay embayment. Improvements in water clarity from nitrogen reductions would also help to protect the existing beds in the outer embayment.

134

MASSACHUSETTS ESTUARIES PROJECT

Table VII-3. Changes in eelgrass coverage in Apponagansett Bay Embayment System within the Town of Dartmouth over the past half century (C. Costello). Note that the 1995 and 2001 areas do not represent significant eelgrass beds, but rather fringing zones of sparsely distributed surviving eelgrass.

EMBAYMENT 1951 1995 2001 % Difference (acres) (acres) (acres) (1951 to 2001)

Apponagansett Bay 42.08 17.64 12.01 71%

VII.4 MACROALGAE SURVEYS

Apponagansett Bay A macroalgae survey was conducted in inner Apponagansett Bay in September and October of 2004 by Lloyd Center staff. Macroalgae accumulation was quite variable. Within Apponagansett Bay, the highest percentage of coverage occurred in the northern and northwestern areas with little accumulation in other places (Figure VII-11). These high accumulations were comprised mainly of mixed Gracillaria and Ulva, with much of the Ulva as drift algae (as opposed to attached). In the overall collection, the predominant species were, in decreasing order: Gracillaria, Ulva, Codium, Enteromorpha and Fucus (Fucus was on rocks along the shoreline). As with nitrogen enrichment, the highest accumulations of macroalgae were in the upper less well flushed regions will much lower (or no) accumulation in the lower ~3/4 of the inner basin. It is noteworthy that significant macroalgal accumulations were not observed in Dike Marsh, consistent with it generally being high quality salt marsh habitat. Significant macroalgal accumulations of drift algae were also not observed in the lower basin of Apponagansett Bay, likely due to its lower nitrogen inputs and water column concentrations, as well as its high tidal flushing.

VII.4 BENTHIC INFAUNA ANALYSIS Quantitative sediment sampling was conducted at 18 locations within the Apponagansett Bay Estuary (Figure VII-12), with replicate assays at most sites. In all areas and particularly those that do not support eelgrass beds, benthic animal indicators can be used to assess the level of habitat health from healthy (low organic matter loading, high D.O.) to highly stressed (high organic matter loading-low D.O.). The basic concept is that certain species or species assemblages reflect the quality of the habitat in which they live. Benthic animal species from sediment samples are identified and ranked as to their association with nutrient related stresses, such as organic matter loading, anoxia, and dissolved sulfide. The analysis is based upon life-history information and animal-sediment relationships (Rhoads and Germano 1986). Assemblages are classified as representative of healthy conditions, transitional, or stressed conditions. Both the distribution of species and the overall population density are taken into account, as well as the general diversity and evenness of the community. It should be noted that it is clear that eelgrass habitat within the Outer Basin of Apponagansett Bay has been relatively stable with some apparent loss at the margins of beds in the outer basin (below the bridge), with no evidence of significant eelgrass habitat within the Inner Basin, inclusive of Dike Marsh. This apparent loss is likely the result of the system reaching its assimilative capacity for nitrogen. To the extent that the overall system can support not only eelgrass but more diverse benthic infaunal communities, the benthic infauna analysis is important for determining the level of habitat impairment (healthymoderately impairedsignificantly impairedseverely

135

MASSACHUSETTS ESTUARIES PROJECT degraded) and what nutrient concentrations are supportive of healthy habitat. This assessment is also important for the establishment of site-specific nitrogen thresholds (Chapter VIII).

Buttonwood Brook Apponagansett Bay Brook

Padanaram

Figure VII-11. Macroalgae coverage in percent area covered in Apponagansett Bay 2004.

Analysis of the evenness and diversity of the benthic animal communities was also used to support the density data and the natural history information. The evenness statistic can range from 0-1 (one being most even), while the diversity index does not have a theoretical upper limit. The highest quality habitat areas, as shown by the oxygen and chlorophyll records and eelgrass coverage, have the highest diversity (generally >3) and evenness (~0.7). The converse is also true, with poorest habitat quality found where diversity is <1 and evenness is <0.5.

Overall, the infauna survey indicated a gradient in benthic animal habitat quality, with moderate impairment in the Inner Basin (open water above bridge) and high quality habitat in the Outer Basin, with Dike Marsh also having high quality habitat based upon its function as salt marsh containing naturally organic enriched sediments. It appears that organic deposition in the Inner Basin is in part due to macroalgal accumulations contributing to the cause of the

136

MASSACHUSETTS ESTUARIES PROJECT stress. The gradient in habitat quality is inversely related to the levels of watercolumn nitrogen and phytoplankton biomass levels and magnitude and frequency of oxygen depletion (Table VII- 4).

The Inner Basin supports areas of significant macroalgal accumulations (up to 80% cover), with high phytoplankton biomass, particularly in the inner (above Little Island) and western coves, averaging 19.9 and 10.9 ug L-1 respectively, consistent with the elevated TN levels in the upper portions of this basin (AB1a and AB4, 1.48-0.53 mg L-1). While periodic oxygen depletion occurs in these "cove regions" it appears that only moderate stress results as oxygen generally remains >4 mg L-1, in >97% of the samples from both the BayWatcher grab samples and the mooring record. The benthic animal community metrics reflect these stresses with a reduced number of species (13), individuals (214), Diversity (2.14) and Evenness (0.63). Habitat impairment is underscored by the general presence of significant numbers of stress indicator species (Capitellids and amphipods). Impairment in the open water Inner Basin declines approaching the bridge opening where tidal exchange occurs. The main basin and region near the bridge show moderate to high quality benthic habitat with moderate to high numbers of species (23-25), diversity (2.70-2.92), with high species numbers (604-1024) and the low numbers of stress indicator species.

In contrast, the Outer Basin is currently supporting high quality benthic habitat based upon the key community indices, the Weiner Diversity Index (H') and Evenness consistent with its lower TN, chlorophyll-a and generally high oxygen levels (>6 mg L-1, 96% of BayWatcher samples, AB-7). Only the in the uppermost reach nearest the bridge are there periodic oxygen declines with 4% of samples <4 mg L-1, most likely due to ebb tidal waters from the Inner Basin. The overall Outer Basin is currently supporting high quality benthic habitat with high numbers of species (27-32), high numbers of individuals (636-867) and moderate to high diversity (2.26- 2.74). Equally important the species dominating the communities were generally representative of non-stressed environments (diverse communities of polychaetes, mollusks and crustaceans).

The salt marsh creek of Dike Marsh is also supporting high quality benthic animal habitat associated with salt marshes as seen in the numbers of species (17), and numbers of individuals (260), high diversity (2.87), and Evenness (0.72), and the low numbers of stress indicator species.

Overall, the Outer Basin supported diverse communities of polychaetes, mollusks and crustaceans typically associated with high quality coastal environments. The levels of diversity and the species numbers present in Outer Basin of Apponagansett Bay are among the highest encountered throughout the MEP region. While the Inner Basin is supporting a productive benthic animal habitat, it is clearly showing moderately impaired habitat, particularly in the upper most and western cove regions. This spatial pattern of habitat quality parallels nitrogen and associated organic enrichment and is common to estuaries in general, with lower habitat quality in the inner versus outer regions.

The high quality benthic animal habitat regions of the Outer Basin of Apponagansett Bay are similar to other low nutrient estuarine basins. For example, the outer stations within Lewis Bay in Barnstable on Cape Cod currently support high quality benthic habitat as seen in the numbers of individuals (502 per sample) and number of species (32). Similarly, Quissett Harbor supports high quality habitat with similarly high numbers of individuals (412) and species (28). Equally important in all cases of observed high quality habitat, the benthic community is not consistent with nutrient enrichment and is composed of a variety of polychaete, crustacean and mollusk species, as opposed to stress tolerant small opportunistic oligochaete worms. All of the

137

MASSACHUSETTS ESTUARIES PROJECT benthic animal habitat metrics are significantly lower in the upper regions of the Inner Basin, particularly above Little Island, consistent with the lower water quality and higher nitrogen and organic enrichment. All of the habitat and environmental metrics are consistent with impaired benthic animal habitat.

Integration of all of the metrics clearly indicates that the Outer Basin is generally supporting high quality benthic animal habitat, while the Inner Basin is just beyond its capacity to assimilate nitrogen loads without impairment (i.e. it is near but above its nitrogen threshold). Since the Outer Basin of Apponagansett Bay is also just beyond its nitrogen threshold to support healthy eelgrass habitat as noted by the recent gradual losses, a slight reduction in nitrogen enrichment to enhance the benthic habitat in the Inner Basin will also lower nitrogen in the Outer Basin on the ebbing tides and should also restore the moderate impairment of the eelgrass habitat. The overall level of impairment of each basin (Table VIII-1) and the level of nitrogen needed for restoration of impaired areas is discussed in Section VIII.3, below.

Table VII-4. Benthic infaunal community data for Apponagansett Bay embayment system. Estimates of the number of species adjusted to the number of individuals and diversity (H’) and Evenness (E) of the community allow comparison between locations (Samples represent surface area of 0.018 m2). Total Total Species Weiner Actual Actual Calculated Diversity Evenness Station Location Species Individuals @75 Indiv. (H') (E) i.d. Apponagansett Bay - Inner Basin Upper Basin 13 214 5 2.06 0.63 12,13,14 Main Basin 23 604 12 2.92 0.67 11,15-19,23 Lowest Reach 25 1024 13 2.70 0.58 22 Dike Marsh 17 260 12 2.87 0.72 7,8,9 Apponagansett Bay - Outer Basin (below bridge) Upper 32 867 13 2.74 0.55 5,6 Lower 27 636 10 2.26 0.47 1.3.4 * Station numbers refer to Figure VII-12

138

MASSACHUSETTS ESTUARIES PROJECT

APB14

APB13

APB12

APB11 APB9 APB15/16 APB19 APB10 APB7 APB17 APB8 APB23

APB6

APB5

APB3 APB4

APB22

APB1

Figure VII-12. Aerial photograph of Apponagansett Bay system showing location of benthic infaunal sampling stations (blue symbol).

In addition to benthic infaunal community characterization undertaken as part of the MEP field data collection, other biological resources assessments were integrated into the habitat assessment portion of the MEP nutrient threshold development process as developed by the Commonwealth as well as CSP scientists from UMD-SMAST working with Lloyd Center for Environmental Studies staff working on the Turn the Tide Project (2009).

The Massachusetts Division of Marine Fisheries has an extensive library of shellfish resources maps which indicate the current status of shellfish areas closed to harvest (Figure VII-13) as well as the suitability of a system for the propagation of shellfish (Figure VII-14). As is the case with other systems in southeastern Massachusetts, all of the enclosed waters of Apponagansett Bay are classified as Conditionally Approved for the taking of shellfish during

139

MASSACHUSETTS ESTUARIES PROJECT any given season of the year, indicating the system is clearly impaired relative to the taking of shellfish. This is likely due to bacterial concerns which would be a result of both human activity (septic systems in the watershed) as well as natural fauna. Nevertheless, the Apponagansett Bay system has also been classified as supportive of specific shellfish communities (Figure VII- 15). The major shellfish species with potential habitat within the Farm Pond Estuary are soft shell clams (Mya) and quahogs (Mercenaria) extending essentially along the shallow waters at the edge of the pond forming a ring most of the water around the pond. It should be noted that the observed pattern of shellfish growing area is consistent with the observed organic rich sediments within the basins and the hypoxia in the bottom waters, the "ring" encompasses the shallows where hypoxia is less frequent and severe. Improving benthic animal habitat quality should also expand the shellfish growing area within this system.

140

MASSACHUSETTS ESTUARIES PROJECT

BB-12.4 BB-12.1

BB-12.20

BB-12.3 BB-12.2

BB-12.7

BB-12.5

Figure VII-13. Shellfishing closures in Apponagansett Bay. Areas BB-12.4, 12.1 and 12.2 are prohibited for shellfishing. The locations of the closed areas are related to surface water sources of stormwater runoff from Padanaram village areas and the upper Apponagansett Bay watershed. (Map source MA Division of Marine Fisheries).

141

MASSACHUSETTS ESTUARIES PROJECT

Figure VII-14. Location of shellfish growing areas and their status relative to shellfish harvesting as determined by Mass Division of Marine Fisheries. Closures are generally related to bacterial contamination or "activities", such as the location of marinas.

142

MASSACHUSETTS ESTUARIES PROJECT

Figure VII-15. Location of shellfish suitability areas within the Farm Pond estuary as determined by Mass Division of Marine Fisheries. Suitability does not necessarily mean "presence".

143

MASSACHUSETTS ESTUARIES PROJECT

VIII. CRITICAL NUTRIENT THRESHOLD DETERMINATION AND DEVELOPMENT OF WATER QUALITY TARGETS

VIII.1. ASSESSMENT OF NITROGEN RELATED HABITAT QUALITY Determination of site-specific nitrogen thresholds for an embayment requires integration of key habitat parameters (infauna and eelgrass), sediment characteristics, and nutrient related water quality information (particularly dissolved oxygen and chlorophyll). Additional information on temporal changes within each sub-embayment of an estuary, its associated watershed nitrogen load and geomorphological considerations of basin depth, stratification and functional type further strengthen the analysis. These data were collected to support threshold development for the Apponagansett Bay Estuarine System by the MEP and were discussed in Section VII. Nitrogen threshold development builds on this data and links habitat quality to summer water column nitrogen levels from the baseline Water Quality Monitoring Program conducted by the Coalition for Buzzards Bay BayWatchers with analytical support from the Coastal Systems Analytical Facility at SMAST-UMass Dartmouth (through 2008).

The Apponagansett Bay Estuarine System is a drown river valley estuary with a tributary salt marsh (Dike Marsh), a large Outer Basin and a shallow Inner Basin separated by a bridge/causeway. Tidal waters enter from Buzzards Bay and tidal flushing is relatively efficient throughout the component basins. While there is some fringing salt marsh in the Inner Basin and Dike Marsh, the open water basins are currently functioning as typical coastal embayment basins with free tidal exchange with the waters of Buzzards Bay. Each type of functional component to an estuary (salt marsh basin, embayment basin, tidal river, deep basin {sometimes drown kettles}, shallow basin, etc.) has a different natural sensitivity to nitrogen enrichment and organic matter loading. Evaluation of eelgrass and infaunal habitat quality must consider the natural structure of the specific basin and its ability to support eelgrass beds and infaunal communities. At present, the Apponagansett Bay system is just beyond its ability to assimilate nitrogen without impairment. It is presently showing a low-moderate level of nitrogen enrichment, with some moderate impairment of eelgrass (Outer basin) and appears to be just beyond its nitrogen loading limit relative to sustaining high quality infaunal habitats within the Inner Basin. This is particularly evident in the upper and western tidal reaches of the system which receive freshwater discharges from Buttonwood Brook and Apponagansett Brook (Table VIII-1). The evidence indicates that nitrogen management of this system will be for restoration rather than for protection or maintenance of an unimpaired system.

The measured levels of oxygen depletion and enhanced chlorophyll-a levels follow the spatial pattern of total nitrogen levels in this system (Section VI), and the parallel variation in these water quality parameters is consistent with watershed based nitrogen enrichment. The spatial pattern indicated that the magnitude of organic matter enrichment of sediments, enhancement of chlorophyll-a levels and total nitrogen concentrations increased from the offshore waters to the Lower Basin (below the bridge) and were highest within the uppermost reaches of the Inner Basin.

Oxygen records obtained from both the moorings (3) deployed throughout the Inner Basin of Apponagansett Bay and the long-term monitoring program indicate moderate organic matter enrichment resulting in periodic moderate oxygen depletion within the Inner Basin with generally only minor oxygen decline in the Outer Basin. Oxygen stress was indicated within the upper most tidal reaches of this system associated with organic matter deposition of both phytoplankton and macrophyte detritus and drift macroalgae accumulations.

144

MASSACHUSETTS ESTUARIES PROJECT

The effect of nitrogen enrichment is to cause oxygen depletion; however, with increased phytoplankton (or epibenthic algae) production, oxygen levels will rise in daylight to above atmospheric equilibration levels in shallow systems (generally ~7-8 mg L-1 at the mooring sites).

Analysis of the level of oxygen excursion at the Mid and Lower mooring sites within the Inner Basin, showed the dissolved oxygen levels were periodically in the 8-10 mg/l range indicating some moderate effects of eutrophication (Figure VII-4), and the larger daily oxygen excursion may also reflect the influence of the large amounts of macroalgae on the bottom in the Mid mooring area. At the Lower mooring in the Inner Basin, adjacent the Dike Marsh outlet, the levels of dissolved oxygen are in excess of 7-8 mg/l with a frequency similar to the Mid embayment site (Figure VII-5). Dissolved oxygen depletion was also evident at the Mid and Lower sites, with declines below 6 mg/L for 51% and 45% of the deployment period, but very low oxygen levels generally were not observed. High oxygen excursions were not typical of the upper mooring site, but modest excursions to 9-10 mg L-1 were observed infrequently at the Mid and Lower sites in the Upper Basin indicating only moderate eutrophication by that measure. Likewise, oxygen levels at Apponagansett Upper and Lower mooring sites fell below 3 mg/L for an insignificant period, less than 6 hours total over the total record, and were >4 mg L-1 for 94% and 91% of the record respectively, with the mid site being >4 mg L-1. However, the percentage of time that the Upper site fell below 6 mg/L was high, 92%, indicating a chronic condition of somewhat depressed water column oxygen.

The oxygen dynamics throughout the Inner Basin, particularly the uppermost and western cove areas may be due to the balance between two processes: first, a shallow water and elevated temperatures and sediment oxygen demand (including macroalgae) that is fueled by high nitrogen loads transported by the two principal streams and second, relatively short water residence time in the Upper embayment whereby Upper embayment nutrient-enriched waters are rapidly mixed with and replaced with lower nutrient, more oxygen-rich Buzzard Bay water (driving the D.O. excursions). Both the upper and mid sites generally maintained moderate oxygen levels (>4 mg L-1) with the lower site showing greater depletion. However, the oxygen dynamics at the lower site are likely influenced by low oxygen waters ebbing from the adjacent salt marsh, due to the salt marsh's naturally organic enrichment. Salt marshes typically have periodic oxygen depletions (even anoxia) that are not an impairment to the marsh, but would be an impairment to open water basins. Overall, the oxygen records are comparable to the long- term grab sample data where the uppermost tidal reach has oxygen declines to <4 mg L-1 in 4% of the samples (N=248) versus 6% of the mooring record but remained >4 mg L-1 in 99% of the samples adjacent Little Island.

Similarly, low to moderate effects of nitrogen enrichment are demonstrated in the record of chlorophyll-a within the Inner Basin, with no blooms >20 µg/L at either site (Table VII-2). On the other hand the upper reach of the Inner Basin had periodic blooms to >40 ug L-1, with a summer bloom in 2003 producing chlorophyll-a levels exceeding 20 µg/L during 25% of the 45 day record and reaching 50 µg/L for a short period. The mean chlorophyll-a value at the Upper mooring site (19.9 ug L-1) was about twice those recorded at the Mid (10.9 ug L-1) and Lower (8.38 ug L-1) sites within the Inner Basin. In contrast, the Outer Basin averaged 6.5 - 7.0 ug L-1 in the long-term monitoring record. Chlorophyll-a summer averages >10 ug L-1 have been used to indicate eutrophication.

The dissolved oxygen, chlorophyll and total nitrogen data vary spatially throughout the overall system with an enriched zone (Inner Basin) clearly observed. Within the Inner Basin

145

MASSACHUSETTS ESTUARIES PROJECT

oxygen depletion is consistent with moderate nitrogen and organic matter enrichment, resulting in oxygen depletions to levels moderately stressful to benthic animal communities. In addition, analysis of benthic community habitat quality needs to also include the distribution of macroalgal accumulations in the upper and western cove regions of the Inner Basin. It appears that major regions of the Inner Basin are currently showing low to moderate impairment from oxygen depletion and organic matter enrichment (chlorophyll-a) and that these observations are consistent with the level and spatial distribution of nitrogen enrichment.

Given the water depths, basin structure, nitrogen loading and circulation of the Inner Basin, macroalgae periodically accumulates in the upper tidal reaches. Within Apponagansett Bay, the highest percentage of coverage occurs in the northern and northwestern areas with little accumulation in other places. These high accumulations are comprised mainly of Gracillaria and Ulva, with much of the Ulva as drift algae (as opposed to attached). Overall, the predominant species comprising accumulations are, in decreasing order: Gracillaria, Ulva, Codium, Enteromorpha and Fucus (Fucus was attached to shoreline rocks). As with nitrogen enrichment, the highest accumulations of macroalgae were in the upper less well flushed regions with much lower (or no) accumulation in the lower ~3/4 of the Inner Basin. It is noteworthy that significant macroalgal accumulations were not observed in Dike Marsh, consistent with its generally being high quality salt marsh habitat. Significant macroalgal accumulations of drift algae also were not observed in the lower basin of Apponagansett Bay, likely due to its lower nitrogen inputs and watercolumn concentrations, as well as its high tidal flushing.

Eelgrass surveys and analysis of historical data for the Apponagansett Bay System were collected by the MassDEP Eelgrass Mapping Program, with additional surveying done in 1984- 85 (Costa 1988). MassDEP Surveys were conducted in 1995 and 2001 as part of this program with additional data collected in 2004 by Lloyd Center staff. Additional analysis of available aerial photos from 1951 was used to reconstruct the eelgrass distribution when watershed development was at a low level. The 1951 data were only anecdotally validated and used only for comparison in this management analysis. In contrast, the 1995 and 2001 MassDEP eelgrass distribution maps and the maps based on 1984 and 1985 surveys (Costa 1988) were field validated. In addition, eelgrass observations were also made throughout Apponagansett Bay and by SMAST Staff during sampling for mooring calibrations, benthic regeneration and infaunal community analysis in 2003, 2004 and 2005. All of these surveys by different groups showed comparable results.

In the surveys in 1984/85, 1995, 2001 and confirmed in 2003-2005, eelgrass habitat was not found within the Inner Basin of Apponagansett Bay (north of the causeway and bridge). Based on the 2001 eelgrass survey conducted by the DEP Eelgrass Mapping Program, the eelgrass appears to be distributed only in outer Apponagansett Bay south of the causeway and bridge. The 1951 map of the inner embayment indicates the "possibility of small beds in the inner basin", but as stated by Costa for this upper basin, the "identification of photographs is difficult in some areas because of drift material". The uppermost and western cove areas of the Inner basin support drift macroalgae. As a result the 1951 interpretations (unvalidated) have significant uncertainty and were therefore not used in the temporal analysis of eelgrass coverage in this system. The 1984-85 surveys did identify eelgrass beds along the east and west shore in the Outer Basin of Apponagansett Bay, consistent with the 1951 photographic analysis, and 1995 and 2001 surveys. These areas do not accumulate drift macro-algae and have low levels of nitrogen enrichment. However, the outer embayment eelgrass beds, while

146

MASSACHUSETTS ESTUARIES PROJECT

stable, appear to have diminished in size between 1995 and 2001 along the eastern shore, though none have disappeared over that interval.

In systems like Apponagansett Bay, the general pattern is for highest nitrogen levels to be found within the innermost tidal reaches, with concentrations declining moving toward the tidal inlet. In Apponagansett Bay the general gradient in nitrogen is influenced by the large salt marsh (Dike Marsh) to the southwest with tidal flow into and out of this marsh influencing circulation within the inner bay. The location of the embayment relative to its watershed means that most of the nitrogen from the watershed enters the uppermost region of the Inner Basin via discharges from Buttonwood Brook and Apponagansett Bay Brook. The nitrogen gradient throughout Apponagansett Bay (0.53 - 0.30 mg L-1) is comparable to other estuaries in southeastern Massachusetts with similar nitrogen loading, basin geomorphology and tidal flushing. A similar nitrogen gradient is observed in the Megansett-Squeteague Harbor Estuary and also a comparable eelgrass coverage, with no historic evidence of eelgrass habitat in the inner basin (Squeteague Harbor) but with fringing beds throughout the outer basin (Megansett Harbor). It is typical of these estuaries to have eelgrass habitat in the outer or lower basins near the tidal inlet and no eelgrass habitat in the inner or upper basins where watershed nitrogen inputs are focused. In Apponagansett Bay the gradual decline in eelgrass in the lower basin (below the bridge) follows the pattern of loss from the upper or deeper areas to the lower or shallower areas. The temporal pattern is a “retreat” of beds toward the region of the tidal inlet. The gradual decline of eelgrass coverage in the lower basin is similar to that found elsewhere (Green Pond and Quissett Harbor, Falmouth; Westport Rivers, Westport; Slocums River, Dartmouth).

The existing coverage data indicates a gradual decline of eelgrass within Apponagansett Bay, primarily in the Outer Basin. This is consistent with the moderate chlorophyll-a and moderate oxygen level depletion as well as low to moderate water column nitrogen concentrations within this basin. It is not documented that eelgrass beds were significant in the inner Bay in recent history (last 50 years). This is consistent with the moderate to high chlorophyll-a levels (mooring average 8.4 - 19.9 ug L-1, long-term average 7.7 - 9.1 ug L-1, BayWatcher 1999-2008). The historic record and the persistence of a gradual decline of beds in the outer embayment suggests a system just beyond its ability to assimilate additional nitrogen inputs without impairment. This indicates that should nitrogen loading to the estuary be lowered, eelgrass coverage should expand. Further, if loading to the Inner Basin is lowered the result would include a lowering of nitrogen levels in the Outer Basin (below the bridge) and associated improvements of eelgrass habitat and benthic animal habitat as well.

Overall, the infauna survey indicated a gradient in benthic animal habitat quality, with moderate impairment in the Inner Basin (open water above bridge) and high quality habitat in the Outer Basin, with Dike Marsh also having high quality habitat based upon its function as salt marsh, with naturally organic enriched sediments. It appears that organic deposition in the Inner Basin is in part due to macroalgal accumulations and a cause of habitat stress. The gradient in habitat quality is inversely related to the levels of watercolumn nitrogen and phytoplankton biomass levels and magnitude and frequency of oxygen depletion.

The Inner Basin supports areas of significant macroalgal accumulations (up to 80% cover), with high phytoplankton biomass, particularly in the inner (above Little Island) and western coves, averaging 19.9 and 10.9 ug L-1 respectively, consistent with the elevated TN levels in the upper portions of this basin (AB1a and AB4, 1.48-0.53 mg L-1). While periodic oxygen depletion occurs in these "cove regions" it appears that only moderate stress results as

147

MASSACHUSETTS ESTUARIES PROJECT

oxygen generally remains >4 mg L-1 in >97% of the samples from both the BayWatcher grab samples and the mooring record. The benthic animal community metrics reflect these stresses with a reduced number of species (13), individual (214), Diversity (2.14) and Evenness (0.63). Habitat impairment underscored by the general presence of significant numbers of stress indicator species (Capitellids and amphipods). Impairment in the open water Inner Basin declines approaching the bridge opening where tidal exchange occurs. The main basin and region near the bridge show moderate to high quality benthic habitat with moderate to high numbers of species (23-25), diversity (2.70-2.92), with high species numbers (604-1024) and the low numbers of stress indicator species.

In contrast, the Outer Basin is currently supporting high quality benthic habitat based upon the key community indices, the Weiner Diversity Index (H') and Evenness consistent with its lower TN, chlorophyll a and generally high oxygen levels (>6 mg L-1, 96% of BayWatcher samples, AB-7). Only the in the uppermost reach nearest the bridge are there periodic oxygen declines with 4% of samples <4 mg L-1, most likely due to ebb tidal waters from the Inner Basin. The overall Outer Basin is currently supporting high quality benthic habitat with high numbers of species (27-32), high numbers of individuals (636-867), moderate to high diversity (2.26-2.74). Equally important the species dominating the communities were generally representative of non-stressed environments (diverse communities of polychaetes, mollusks and crustaceans).

The salt marsh creek of Dike Marsh is also supporting high quality benthic animal habitat associated with salt marshes as seen in the numbers of species (17), and numbers of individuals (260), high Diversity (2.87), and Evenness (0.72), and the low numbers of stress indicator species.

Overall, the Outer Basin supported diverse communities of polychaetes, mollusks and crustaceans typically associated with high quality coastal environments. The levels of diversity and the species numbers present in Outer Basin of Apponagansett Bay are among the highest encountered throughout the MEP region. While the Inner Basin is supporting a productive benthic animal habitat, it is clearly showing moderately impaired habitat, particularly in the upper most and western cove regions. This spatial pattern of habitat quality parallels nitrogen and associated organic enrichment, with lower habitat quality in the inner versus outer regions, is common to estuaries in general.

The high quality benthic animal habitat regions of the Outer Basin of Apponagansett Bay are similar to other low nutrient estuarine basins. For example, the outer stations within Lewis Bay in Barnstable currently support high quality benthic habitat as seen in the numbers of individuals (502 per sample), number of species (32). Also, high quality habitat was observed in Quissett Harbor with similarly high numbers of individuals (412) and species (28). Equally important, in all cases of high quality habitat the community is not consistent with nutrient enrichment and is composed of a variety of polychaete, crustacean and mollusk species, as opposed to stress tolerant small opportunistic oligochaete worms. All of the benthic animal habitat metrics are significantly lower in the upper regions of the Inner Basin, particularly above Little Island, consistent with the lower water quality and higher nitrogen and organic enrichment. All of the habitat and environmental metrics are consistent with impaired benthic animal habitat.

The observed water column metrics (oxygen, chlorophyll-a, total nitrogen) are consistent with the infaunal communities within the Apponagansett Bay Estuary. The infauna surveys indicated that the Apponagansett Bay Estuarine System is currently supporting highly productive and diverse benthic communities throughout its Outer Basin and salt marsh areas,

148

MASSACHUSETTS ESTUARIES PROJECT

but has clear impairment of communities within the upper reaches of the Inner Basin. The moderate gradient in benthic community metrics, with highest quality habitat in the Outer Basin closest to the high quality waters of Buzzards Bay, to the moderate quality habitat in the main basin of the Inner Basin and impaired habitat in the uppermost reach of the Inner Basin, furthest from Buzzards Bay, reflects the nitrogen gradient and resulting gradients in oxygen depletion, organic enrichment and macroalgal accumulation.

Integration of all of the metrics clearly indicates that the Outer Basin is generally supporting high quality benthic animal habitat while the Inner Basin is just beyond its capacity to assimilate nitrogen loads without impairment (i.e. it is above its nitrogen threshold). Since the Outer Basin of Apponagansett Bay is also just beyond its nitrogen threshold to support healthy eelgrass habitat (it does have high quality benthic habitat) as noted by the recent gradual losses, a slight reduction in nitrogen enrichment to enhance the benthic habitat in the Inner Basin will also lower nitrogen in the Outer Basin on the ebbing tides and should also restore the moderate impairment of the eelgrass habitat. The overall level of impairment of each basin (Table VIII-1) and the level of nitrogen needed for restoration of impaired areas is discussed in Section VIII.3, below.

The relative pattern of habitat quality based upon the eelgrass and benthic community data is consistent with the results of the oxygen and chlorophyll time-series data (Section VII.2) and nitrogen levels within the inner and outer basins (Section VI). The absence of eelgrass beds from the inner basins and the coverage in the outer basin is supported by the low phytoplankton levels (low turbidity) and low nitrogen levels in the outer versus inner basin waters. Both the pattern of coverage and the associated levels of the key nutrient related water quality parameters are typical of nutrient enriched shallow embayments (see below). Overall, it appears that the Apponagansett Bay Estuary has slightly exceeded its assimilative capacity for nitrogen with the resulting recent gradual decline in eelgrass coverage and impaired benthic animal habitat in the Inner Basin. Determining the nitrogen target to restore the impaired eelgrass habitat and protecting infauna habitat in the inner basins is the focus of the nitrogen management threshold analysis, below.

VIII.2 THRESHOLD NITROGEN CONCENTRATIONS Overall, the Outer Basin supported diverse communities of polychaetes, mollusks and crustaceans typically associated with high quality coastal environments. The levels of diversity and the species numbers present in Outer Basin of Apponagansett Bay are among the highest encountered throughout the MEP region. While the Inner Basin is supporting a productive benthic animal habitat, it is clearly showing moderately impaired habitat, particularly in the upper most and western cove regions. This spatial pattern of habitat quality parallels nitrogen and associated organic enrichment, with lower habitat quality in the inner versus outer regions, is common to estuaries in general.

The approach for determining nitrogen loading rates that will support acceptable habitat quality throughout an embayment system is to first identify a sentinel location within the embayment and secondly, to determine the nitrogen concentration within the water column that will restore the location to the desired habitat quality. The sentinel location is selected such that the restoration of that one site will necessarily bring the other regions of the system to acceptable habitat quality levels. Once the sentinel site and its target nitrogen level are determined (Section VIII.2), the Linked Watershed-Embayment Model is used to sequentially adjust nitrogen loads until the targeted nitrogen concentration is achieved (Section VIII.3).

149

MASSACHUSETTS ESTUARIES PROJECT

Table VIII-1. Summary of nutrient related habitat quality within the Apponagansett Bay Estuary within the Town of Dartmouth, MA. and based upon assessments described in Section VII. WQMP indicates BayWatcher Water Quality Monitoring Program.

Apponagansett Bay Estuarine System Health Indicator Inner Basin Lower Basin Dike Marsh Dissolved Oxygen H/MI1 H2 H3 Chlorophyll MI4 H5 H6 Macroalgae MI/SI7 H8 H9 Eelgrass --10 MI11 --10 Infaunal Animals MI12 H13 H14 Overall: MI15 H/MI16 H17 1- oxygen depletion to 4 mg l-1 common but short duration (upper , mid mooring and WQMP) short, upper reach oxygen <4 mg L-1, 6% of time-series and similarly ~4% of WQMP long-term record. Lower time-series influenced by ebbing waters from salt marsh. 2 - relatively high dissolved oxygen with levels >5 mg L-1 in 96% of samples throughout the Outer Basin (N>100/site, 3 sites WQMP), few declines to 4 mg L-1, infrequent declines may relate to outflowing Upper Basin waters. 3 - salt marsh creeks are naturally organic and typically have oxygen depletion (even anoxia) without habitat impairment as communities are adapted to these conditions, unlike in open water basins. 4 - levels high in the Inner Basin averaging 11-20 ug L-1 for the mid and upper moorings and ~9 ug L-1 in WQMP. Blooms to 40 ug L-1 in upper reach. 5- low to moderate for a coastal basin, averaging 6.5 - 7.0 ug L-1 WQMP. (WQMP 2000-2014, N=54) with time-series (Squeteague) <20 ug L-1 90% of record, average 11.6 ug. 6- tidal water with moderate chlorophyll levels averaging 8.4 ug L-1 and 7.7 L-1 at mooring and WQMP 7- drift algae accumulations significant in upper most and western areas, coverages reaching 80%. Gracillaria, Ulva (generally associated with nitrogen enrichment). Likely impairment benthic habitat. 8- drift algae generally absent, hard substrates with attached algae 9- sparse to no drift macroalgae in the tidal creeks or deposited on creek banks. 10- no documented (verified) evidence of eelgrass "presence" in this basin historically. 11- most of basin margin supports eelgrass habitat, loss of some coverage nearest the upper tidal reach improving toward the inlet. Spatial pattern of loss from the upper and deeper margin of the beds is typical of nitrogen enrichment and indicates moderate impairment. 12 - Inner: low numbers of individuals (214), species (13), diversity (2.06) and Evenness (0.63), stress and transition indicator species prevalent (Capitellids and Amphipods),crustaceans and mollusks and polychaetes. Main basin: moderate/high numbers of individuals (604), species (23), diversity (2.93) and Evenness (0.67), mostly dominated by polychaetes, crustaceans and mollusks but with patches of stress indicators (Capitellids, Tubificids). 13 - polychaetes, crustaceans and molluscs with some deep burrowers; sparse organic enrichment species, moderate-high diversity (2.26-2.74) and high# of species (27-32) and individuals (636-867). 14 - community metrics consistent with an unimpaired salt marsh sandy creek with moderate to high numbers of species (17), individuals (260) and diversity (H') (2.87), with high Evenness (>0.7). Stress indicator species sparse to absent, community dominated by polychaetes with filter feeders common. 15 -no significant eelgrass habitat documented, but with areas of macroalgal accumulation in the northern and western cove areas. Benthic habitat impaired by oxygen depletion (to 4 mg L-1, (mooring and WQMP), high phytoplankton biomass with organic enrichment in coves (mooring mean chlorophyll mid-upper reaches = 11-20 ug L-1), with moderately impaired benthic habitat , moderate numbers of species (13), individuals (214), diversity (2.06) with some organic enrichment indicator species. 16 - relatively stable eelgrass habitat in shallow margin areas, recent gradual loss of eelgrass indicates moderate impairment, high quality benthic habitat with high individual and species numbers, without stress indicator species. Sparse macroalgae (attached, not drift) present with low-moderate phytoplankton biomass (chlorophyll = 6.5-7.0 ug L-1; WQMP) and high dissolved oxygen, > 5 mg L-1, 96% of records, 5.0 - 4,0 mg L-1 4% of records. 17 - no eelgrass habitat consistent with a salt marsh, with typical unimpaired benthic animal habitat with moderate to high numbers of species (17), diversity (H'=2.90) and high numbers of individuals (260) and Evenness (E>0.70), without stress indicator species. Consistent with the oxidized sediments and absence of macroalgal accumulations. H = High quality habitat conditions; MI = Moderate Impairment; SI = Significant Impairment; SD = Severely Degraded; -- = not applicable to this estuarine reach

150

MASSACHUSETTS ESTUARIES PROJECT

Determination of the critical nitrogen threshold for maintaining high quality habitat within the Apponagansett Bay Embayment System is based primarily upon the nutrient and oxygen levels, temporal trends in eelgrass distribution and current benthic community indicators. Given the information on a variety of key habitat characteristics, it is possible to develop a site-specific threshold, which is a refinement upon more generalized threshold analyses frequently employed.

The Apponagansett Bay Estuarine System presently shows a moderate impairment to eelgrass habitat within its Outer Basin (below bridge). The impairment is based upon the recent temporal trend in loss of eelgrass from the coverage areas at the margins of the basin. Both the location and the temporal trend is consistent with nitrogen enrichment. However, as the rate of loss has been gradual and is relatively recent, that indicates that this estuary is only just beyond its nitrogen threshold (i.e. the level of nitrogen a system can tolerate without impairment). Similarly, the benthic animal habitat in the upper reaches of the Inner Basin is clearly impaired based upon the low number of species and diversity and presence of organic enrichment indicator species. As with eelgrass in the Outer Basin, benthic habitat within the Inner Basin is moderately impaired, just above its ability to assimilate more nitrogen without impairment (e.g. above its nitrogen threshold). The observed gradual decline in eelgrass coverage is consistent with the moderate level of chlorophyll-a and generally high oxygen levels and moderate nitrogen enrichment. Eelgrass habitat is much more sensitive to the associated effects of nitrogen enrichment due to its need for light, hence sensitivity to increased turbidity or elevated phytoplankton levels in the overlying water. Benthic animals do not require light, so nitrogen enrichment needs to begin to increase organic matter deposition and oxygen conditions to impact the habitat quality. Based on the field measurements both the Inner (benthic habitat) and Outer (eelgrass) basins are impaired by nitrogen enrichment. However, both basins are only moderately impaired for their respective habitats and so that nitrogen management should provide for a rapid improvement of the nutrient impaired habitats throughout the estuary.

Evaluation of the nitrogen levels associated with the impaired benthic habitat and eelgrass habitat are currently 0.53 mg N L-1 in the cove region north of Little Island which is representative of the uppermost tidal reaches of the Inner Basin and 0.38 mg N L-1 in the areas of eelgrass decline in the Outer Basin (tidally averaged). The observed levels of impairment in these Apponagansett Bay basins are comparable to other estuaries with similar TN levels.

For example, tidally averaged total nitrogen in areas of eelgrass decline in Megansett Harbor were 0.36 mg L-1 with high quality benthic communities at 0.45-0.47 mg N L-1 within Megansett Harbor and Squeteague Harbor, respectively. In addition, within the Nantucket Harbor Estuary tidally averaged levels in the lower reach of Head of the Harbor (0.340-0.353) were associated with recent loss of eelgrass coverage, while eelgrass was lost from West Falmouth Harbor when tidally averaged TN exceeded 0.35 mg L-1. The recent relatively small loss (as a percentage of total coverage) of eelgrass from Quissett Harbor was seen at tidally averaged nitrogen (total nitrogen, TN) levels of 0.354 mg N L-1. A threshold of 0.35 mg N L-1 was selected to restore eelgrass in these basins and also Phinneys Harbor. Benthic animal habitat quality versus watercolumn nitrogen has consistently indicated impairment above TN levels of 0.50 mg N L-1. For example, this level was found to be associated with high quality benthic habitat in Popponesset Bay based upon the infaunal analysis coupled with the nitrogen data (measured and modeled), nitrogen levels on the order of 0.4 to 0.5 mg TN L-1 were found supportive of high infaunal habitat quality in this system. Similarly, in the Three Bays System, healthy infaunal areas are found at nitrogen levels of TN <0.42 mg TN L-1 (Cotuit Bay and West Bay), with impairment in areas where nitrogen levels of TN >0.5 mg TN L-1 (North Bay), and

151

MASSACHUSETTS ESTUARIES PROJECT

severe degradation at nitrogen levels of TN >0.6 mg TN L-1. Similarly, impaired benthic habitat was quantified within the Fiddlers Cove and within the upper terminal basins of Rands Harbor at TN levels of 0.56 mg TN L-1 and 0.57 mg TN L-1, respectively, supporting the contention that levels <0.50 mg N L-1 are needed for restoration of impaired benthic animal habitat in s.e. Massachusetts estuaries.

It should also be noted that In numerous estuaries it has been previously determined that 0.500 mg TN L-1 is the upper limit to sustain unimpaired benthic animal habitat (e.g. Eel Pond (Waquoit), Parkers River, upper Bass River, upper Great Pond, Rands Harbor and Fiddlers Cove). Present TN levels within the upper reaches of the Inner Basin of Apponagansett Bay are >0.50 mg N L-1, consistent with the observed lack of eelgrass beds and impaired benthic animal habitat. Based upon comparisons to other systems, the TN levels in the Inner Basin, the periodic oxygen depletions and phytoplankton blooms, it appears that a watercolumn nitrogen threshold for this basin of 0.50 mg TN L-1 is required for restoration. All habitat metrics indicate a moderate to moderate level of habitat impairment (Table VIII-1).

Based upon these results, a threshold for tidally averaged TN at long-term monitoring station AB-4 in the Inner Basin and 0.35 mg L-1 (average AB-2 and AB-6) in the Outer Basin was selected to restore eelgrass habitat. In this system both conditions need to be met, but meeting the 0.50 mg N L-1 (tidally averaged) at station AB-4 for benthic habitat restoration should ensure meeting the eelgrass target for the Outer Basin. The nitrogen loads associated with the threshold concentration at the sentinel location and secondary infaunal check stations are discussed in Section VIII.3, below.

VIII.3. DEVELOPMENT OF TARGET NITROGEN LOADS The nitrogen thresholds developed in the previous section were used to determine the amount of total nitrogen mass loading reduction required for restoration of eelgrass and infaunal habitats in the Apponagansett Bay System under Build-Out Conditions. Tidally averaged total nitrogen thresholds derived in Section VIII.1 were used to adjust the calibrated constituent transport model developed in Section VI. The Build-Out watershed nitrogen loads were sequentially lowered, using reductions in septic effluent discharges only, until the nitrogen levels reached the threshold level at the sentinel stations chosen for the Apponagansett Bay System (AB-4 is located approximately at the midpoint of the upper Apponagansett Bay north of the causeway). It is important to note that load reductions can be produced by reduction of any or all sources or by increasing the natural attenuation of nitrogen within the freshwater systems to the embayment. The load reductions presented below represent only one of a suite of potential reduction approaches that need to be evaluated by the community. The presentation is to establish the general degree and spatial pattern of reduction that will be required for restoration of this nitrogen impaired embayment.

As shown in Table VIII-2, the nitrogen load reductions for Build-out conditions within the system necessary to achieve the threshold nitrogen concentrations required using: 1) removal of 60% of the septic nitrogen load from Buttonwood Brook watershed (WS1) and, 2) the removal of 15% of septic nitrogen loading from Apponagansett Brook Gaged Watershed (WS2). It should be noted the removals of septic loading from watersheds 1 and 2 occurred prior to the natural attenuation that occurs. Therefore the septic loads were further reduced on the approach to Apponagansett Bay. The distribution of tidally-averaged nitrogen concentrations associated with the above thresholds analysis is shown in Figure VIII-1.

152

MASSACHUSETTS ESTUARIES PROJECT

Figure VIII-1. Contour plot of modeled average total nitrogen concentrations (mg/L) in Apponagansett Bay System under Build-out, for threshold conditions (0.50 mg/L at water quality monitoring station AB-4). The approximate location of the sentinel threshold station for Apponagansett Bay System (AB-4) is shown.

153

MASSACHUSETTS ESTUARIES PROJECT

Table VIII-2. Comparison of sub-embayment watershed septic loads (attenuated) used for modeling of build-out and threshold loading scenarios of the Apponagansett Bay System. These loads do not include direct atmospheric deposition (onto the sub-embayment surface), benthic flux, runoff, or fertilizer loading terms. Build-out threshold threshold sub-embayment septic load septic load septic load % (kg/day) (kg/day) change Apponagansett Harbor 12.05 12.05 +0.00% Apponagansett Bay 0.67 0.67 +0.00% Dike Marsh 2.81 2.81 +0.00% Surface Water Sources Buttonwood Brook 9.44 2.68 -71.60% Apponagansett Brook 1.30 1.00 -23.11%

Tables VIII-3 and VIII-4 provide additional loading information associated with the thresholds analysis. Table VIII-3 shows the change to the total watershed loads, based upon the removal of septic loads depicted in Table VIII-2. Removal of septic loads from Buttonwood Brook and Apponagansett Brook Gaged watersheds results in the total nitrogen loads presented in Table VIII-4. Table VIII-4 shows the breakdown of threshold sub-embayment and surface water loads used for total nitrogen modeling. In Table VIII-4, loading rates are shown in kilograms per day, since benthic loading varies throughout the year and the values shown represent ‘worst-case’ summertime conditions. The benthic flux for this modeling effort is reduced from build-out conditions based on the load reduction and the observed particulate organic nitrogen (PON) concentrations within each sub-embayment relative to background concentrations in Buzzards Bay.

Table VIII-3. Comparison of sub-embayment total attenuated watershed loads (including septic, runoff, and fertilizer) used for modeling of build-out and threshold loading scenarios of the Apponagansett Bay system. These loads do not include direct atmospheric deposition (onto the sub-embayment surface) or benthic flux loading terms. build-out threshold threshold % sub-embayment load load (kg/day) change (kg/day) Apponagansett Harbor 23.42 23.42 +0.00% Apponagansett Bay 1.48 1.48 +0.00% Dike Marsh 7.96 7.96 +0.00% Surface Water Sources Buttonwood Brook 26.02 19.27 -25.96% Apponagansett Brook 3.24 2.94 -9.31%

154

MASSACHUSETTS ESTUARIES PROJECT

Table VIII-4. Threshold sub-embayment loads and attenuated surface water loads used for total nitrogen modeling of the Apponagansett Bay system, with total watershed N loads, atmospheric N loads, and benthic flux direct benthic flux threshold load atmospheric sub-embayment net (kg/day) deposition (kg/day) (kg/day) Apponagansett Harbor 23.42 4.48 -3.96 Apponagansett Bay 1.48 3.64 -13.83 Dike Marsh 7.96 0.58 0.04 Surface Water Sources Buttonwood Brook 19.27 -- -- Apponagansett Brook 2.94 -- --

Comparison of model results between build-out loading conditions and the selected loading scenario to achieve the target TN concentrations at the sentinel stations is shown in Table VIII-5. To achieve the threshold nitrogen concentrations at the sentinel station, a reduction in TN concentration of approximately 7% was required at station BR-7.

Table VIII-5. Comparison of model average total N concentrations from present loading and the modeled threshold scenario, with percent change, for the Apponagansett Bay system. Sentinel threshold station is in bold print. monitoring build-out threshold Sub-Embayment % change station (mg/L) (mg/L) Head of Bay AB-1A 1.38 1.13 -18.7% North Little Island AB-4 0.54 0.50 -7.0% Upper Basin-Lower AB-3 0.45 0.43 -4.7% Lower Basin-Upper AB-2 0.38 0.37 -3.1% Lower Basin AB-6 0.34 0.34 -1.9%

The basis for the watershed nitrogen removal strategy utilized to achieve the embayment thresholds may have merit, since this example nitrogen remediation effort is focused on watersheds where groundwater is flowing directly into the estuary. For nutrient loads entering the systems through surface flow, natural attenuation in freshwater bodies (i.e., streams and ponds) can significantly reduce the load that finally reaches the estuary. Presently, this attenuation is occurring due to natural ecosystem processes and the extent of attenuation being determined by the mass of nitrogen which discharges to these systems. The nitrogen reaching these systems is currently “unplanned”, resulting primarily from the widely distributed non-point nitrogen sources (e.g. septic systems, lawns, etc.). Future nitrogen management should take advantage of natural nitrogen attenuation, where possible, to ensure the most cost-effective nitrogen reduction strategies. However, “planned” use of natural systems has to be done carefully and with the full analysis to ensure that degradation of these systems will not occur. One clear finding of the MEP has been the need for analysis of the potential associated with restored wetlands or ecologically engineered ponds/wetlands to enhance nitrogen attenuation. Attenuation by ponds in agricultural systems has also been found to work in some cranberry bog systems, as well. Cranberry bogs, other freshwater wetland resources, and freshwater ponds provide opportunities for enhancing natural attenuation of their nitrogen loads.

155

MASSACHUSETTS ESTUARIES PROJECT

Restoration or enhancement of wetlands and ponds associated with the lower ends of rivers and/or streams discharging to estuaries are seen as providing a dual service of lowering infrastructure costs associated with wastewater management and increasing aquatic resources associated within the watershed and upper estuarine reaches.

Although the above modeling results provide one manner of achieving the selected threshold level for the sentinel site within the estuarine system, the specific example does not represent the only method for achieving this goal. However, the thresholds analysis provides general guidelines needed for the nitrogen management of this embayment.

156

MASSACHUSETTS ESTUARIES PROJECT

IX. MANAGEMENT SCENARIO

IX.1 BACKGROUND During the course of the development of the land use information for Apponagansett Bay, MEP and Town of Dartmouth staff worked extensively to develop an accurate date of first connection for each parcel currently connected to the municipal sewer system. As mentioned in Section 4.1, this effort was initiated to create a more reasonable match between measured nitrogen loads in Buttonwood Brook and Apponagansett Brook and nitrogen loads based on land use within those subwatersheds. Since the brook loads were measured in 2003-2004, staff needed to determine which currently sewered parcels were utilizing septic systems for wastewater treatment prior to that time period. In addition, it was thought that a date prior to the stream measurements was more appropriate, in order to account for a reasonable estimate of groundwater travel time to the brooks and their tributaries. MEP staff selected 2001 as the appropriate year to match the measure brook nitrogen loads.

With the help of town staff, a sewer connection database was developed from the Town of Dartmouth, Department of Public Works betterment database. However, initial MEP staff review of the dates in the betterment database found that more than half of the listed sewered parcels had a placeholder date assigned rather than the actual date of sewer connection. With further town help, staff reviewed the DPW betterment paper records and determined actual connection dates for 44% of the placeholder parcels within the Apponagansett Bay watershed. Using the updated digital database, MEP staff used the average percentage of pre-2002 connections to estimate the sewer connection year for the remaining placeholder parcels. Using the actual and estimated dates of sewer connections for 2001 and before, MEP staff determined the subwatershed nitrogen loads for existing conditions.

During the course of developing a more refined betterment database, Town and MEP staff discussed developing a nitrogen loading scenario that would utilize current (2014) sewer connections with existing land use development patterns (Table IX-1). These 2014 conditions are also the basis for the sewer connections used in the buildout conditions.

IX.2 SCENARIO RESULTS

The Town of Dartmouth request a nitrogen loading scenario based on 2014 Conditions based on the collaborative work between the Town and MEP staff. The scenario produces reductions in Nitrogen across all the watersheds as shown in Table IX-1 and is based on land use development and sewer connections in Dartmouth as of 2014. This scenario only includes development that has occurred since the existing conditions scenario and does not include all development included in the buildout scenario (see Chapter 4). New Bedford land use and sewer connections are the same as those listed for the existing conditions scenario. Table IX-2 and Table IX-3 illustrate the overall change to septic and watershed loads resulting from this alternative scenario. Sewer connections/septic system removal results in significant reductions in the watershed loads in specific sub-embayments, particularly the Dike Marsh sub-watershed which has less loads than under existing conditions. Based on the assumptions developed for this alternative, Table IX-4 presents the various components of nitrogen loading for the Apponagansett Bay system. The proposed reductions in load related to the 2014 Scenario meet the threshold target (0.5 mg/L TN at AB-4) at the sentinel station.

157

MASSACHUSETTS ESTUARIES PROJECT

Table IX-1. Apponagansett Bay Watershed Nitrogen Loads incorporating current (2014) Dartmouth Sewer Connections. Sewer connections are based on parcels currently receiving sewer bills. Slight land use changes are incorporated based on reconciling the sewer billing database and parcel classifications. Attenuation of system nitrogen loads within Buttonwood Brook and Apponagansett Brook are based on measurements discussed in Chapter 4. All values are kg N yr-1. 2014 scenario Apponagansett Harbor N Loads by Input (kg/y): Scenario N Loads % of Water Body Watershed Farm Impervious River "Natural" Pond UnAtten N Atten Atten N Wastewater Fertilizers Surface ID# Animals Surfaces Wetlands Surfaces Outflow Load % Load Area Name Apponagansett Harbor System 8862 4112 566 4498 3217 3219 684 25159 23514 Buttonwood Brook 1 3543 1999 136 2698 1029 8 190 9603 11% 8547 Apponagansett Brook Gaged 2 883 200 1 292 152 0 62 1590 37% 1002 Apponagansett Bay Inner E 3 2301 575 125 764 227 0 76 4069 4069 Apponagansett Bay Inner W 4 1531 889 163 448 738 37 136 3941 3941 Dike Marsh 5 425 336 140 152 1068 0 186 2306 2306 Apponagansett Bay Outer W 6 106 92 0 88 3 0 22 311 311 Apponagansett Bay Outer E 7 72 22 0 56 0 0 14 164 164 Dike Marsh Estuary Surface 213 213 213 Apponagansett Bay Outer W Estuary Surface 26 26 26 Apponagansett Bay Inner Estuary Surface Area 1633 1633 1633 Apponagansett Bay Outer Estuary Surface Area 1301 1301 1301

Table IX-2. Comparison of sub-embayment watershed septic loads (attenuated) used for modeling loading conditions for 2014 Scenario. These loads do not include direct atmospheric deposition (onto the sub-embayment surface), benthic flux, runoff, or fertilizer loading terms. Build-out scenario septic load % sub-embayment septic load septic load change (kg/day) (kg/day) Apponagansett Harbor 12.05 10.58 -12.3% Apponagansett Bay 0.67 0.49 -26.8% Dike Marsh 2.81 1.16 -58.5% Surface Water Sources Buttonwood Brook 9.44 6.83 -27.6% Apponagansett Brook 1.30 0.81 -37.8%

158

MASSACHUSETTS ESTUARIES PROJECT

Table IX-3. Comparison of sub-embayment total attenuated watershed loads (including septic, runoff, and fertilizer) used for modeling of conditions for 2014 Scenario. These loads do not include direct atmospheric deposition (onto the sub- embayment surface) or benthic flux loading terms. build-out scenario sub-embayment load load (kg/day) % change (kg/day) Apponagansett Harbor 23.42 21.95 -6.3% Apponagansett Bay 1.48 1.30 -12.2% Dike Marsh 7.96 6.32 -20.6% Surface Water Sources Buttonwood Brook 26.02 23.42 -10.0% Apponagansett Brook 3.24 2.75 -15.2%

Table IX-4. Sub-embayment loads used for total nitrogen modeling of the Apponagansett Bay system for present loading scenario with loading conditions for 2014 Scenario, with total watershed N loads, atmospheric N loads, and benthic flux. direct benthic flux scenario load atmospheric sub-embayment net (kg/day) deposition (kg/day) (kg/day) Apponagansett Harbor 21.94 4.48 -5.03 Apponagansett Bay 1.30 3.64 -13.88 Dike Marsh 6.32 0.58 0.04 Surface Water Sources Buttonwood Brook 23.42 -- -- Apponagansett Brook 2.75 -- --

Table IX-5. Comparison of model average total N concentrations from 2014 scenario, with percent change, for the Apponagansett Bay System. The threshold station is shown in bold print. monitoring Build- scenario Sub-Embayment % change station out(mg/L) (mg/L) Head of Bay AB-1A 1.38 1.22 -11.8% North Little Island AB-4 0.54 0.50 -7.1% Upper Basin-Lower AB-3 0.45 0.43 -5.6% Lower Basin-Upper AB-2 0.38 0.37 -4.0% Lower Basin AB-6 0.34 0.33 -2.5%

159

MASSACHUSETTS ESTUARIES PROJECT

X. LIST OF REFERENCES

AFCEE (with Howes, B.L. & Jacobs Engineering). 2000. Ashumet Pond Trophic Health Technical Memorandum. AFCEE/MMR Installation Restoration Program, AFC-J23- 35S18402-M17-0005, 210pp. Anders, F.J., and M.R. Byrnes. 1991 Accuracy of Shoreline Change Rates as Determined from Maps and Aerial Photographs. Shore and Beach 16:17-26. Aravena, R. and W.D. Robertson. 1998. Use of Multiple Isotope Tracers to Evaluate Denitrification in Ground Water: Study of Nitrate from a Large-Flux Septic System Plume. Ground Water, 36(6):975-982. Bazely, D.R. and R.L. Jefferies, 1985. Goose Faeces: A source of nitrogen for plant growth in a grazed salt marsh. Journal of Applied Ecology. 22:693-703. Bent, G.C., 1995, Streamflow, groundwater recharge and discharge, and characteristics of surficial deposits in Buzzards Bay Basin, Southeastern Massachusetts: U.S. Geological Survey Water-Resources Investigations Report 95-4234. Brawley, J.W., G. Collins, J.N. Kremer, C.H. Sham, and I. Valiela, 2000. A time-dependent model of nitrogen loading to estuaries form coastal watersheds. Journal of Environmental Quality 29:1448-1461. Brigham Young University, 1998. “User’s Manual, Surfacewater Modeling System.” Buzzards Bay National Estuary Program, 1991. Buzzards Bay Comprehensive Conservation and Management Plan, 8/91 Final. 246p. Cambareri, T.C. and E.M. Eichner, 1998. Watershed Delineation and Ground Water Discharge to a Coastal Embayment. Ground Water; 36(4): 626-634. Cape Cod Commission Water Resources Office, 1991. Technical Bulletin 91-001, Nitrogen Loading. Cape Cod Commission Water Resources Office, 1998. Cape Cod Coastal Embayment Project Interim Final Report. Cape Cod Commission, 1998. Cape Cod Coastal Embayment Project: A Nitrogen Loading Analysis of Popponesset Bay. Cape Cod Commission Technical Report. Cape Cod Commission, 1998. “Cape Cod Coastal Embayment Project.” Barnstable, MA. Chow, V. T. (1959). Open Channel Hydraulics, McGraw-Hill, NY. Coalition for Buzzards Bay 2003. State of the Bay 2003. Coalition for Buzzards Bay, 2006. Unpublished Baywatchers data summaries. Conover, J.T., 1958. Seasonal Growth of Benthic Marine Plants as Related to Environmental Factors in an Estuary. Institute of Marine Science, 5:97-147. Costa, J.E., B.L. Howes, D. Janik, D. Aubrey, E. Gunn, A.E. Giblin, 1999. Managing anthropogenic inputs to coastal embayments: Technical basis and evaluation of a management strategy adopted for Buzzards Bay. Buzzards bay Project Technical Report. 62 p. Draft Final 9/24/99.

160 MASSACHUSETTS ESTUARIES PROJECT

Costa, J.E., B.L. Howes, I. Valiela and A.E. Giblin. 1992. Monitoring nitrogen and indicators of nitrogen loading to support management action in Buzzards Bay. In: McKenzie et al. (eds.) Ecological Indicators Chapter. 6, pp. 497-529. Costa, J.E., G. Heufelder, S. Foss, N.P. Millham, B.L. Howes, 2002. Nitrogen Removal Efficiencies of Three Alternative Septic System Technologies and a Conventional Septic System. Environment Cape Cod 5(1): 15-24. Costello, Charles. Section Chief, Wetlands Conservancy Program. Director, Eelgrass Mapping Program 617-292-5907. Crowell, M., S.P. Leatherman, M.K. Buckley. 1991. Historical Shoreline Change: Error Analysis and Mapping Accuracy. Journal of Coastal Research 7(3):839-852. D’Elia, C.F, P.A. Steudler and N. Corwin. 1977. Determination of total nitrogen in aqueous samples using persulfate digestion. Limnology and Oceanography 22:760-764. DeSimone, L.A. and B.L. Howes. 1996. Denitrification and nitrogen transport in a coastal aquifer receiving wastewater discharge. Environmental Science and Technology 30:1152-1162. DeSimone, L.A., D.A. Walter, J.R. Eggleston, and M.T. Nimiroski. 2002. Simulation of Ground- Water Flow and Evaluation of Water-Management Alternatives in the Upper Charles River Basin, Eastern Massachusetts. USGS Water-Resources Investigations Report 02-4234. 103 p. US Geological Survey. Northborough, MA. Dyer, K.R., 1997. Estuaries, A Physical Introduction, 2nd Edition, John Wiley & Sons, NY, 195 pp. E&A Environmental Consultants, Inc. 2000. Evaluation and Prioritization of Compost Facility Runoff Management Methods. Bothell, WA. Available at: http://cwc.org/organics/organic_htms/cm002rpt.htm (accessed 9/11/07). Eichner, E.M. and T.C. Cambareri, 1992. Technical Bulletin 91-001: Nitrogen Loading. Cape Cod Commission, Water Resources Office, Barnstable, MA. Available at: Eichner, E.M., T.C. Cambareri, G. Belfit, D. McCaffery, S. Michaud, and B. Smith, 2003. Cape Cod Pond and Lake Atlas. Cape Cod Commission. Barnstable, MA. Eichner, E.M., T.C. Cambareri, K. Livingston, C. Lawrence, B. Smith, and G. Prahm, 1998. Cape Coastal Embayment Project: Interim Final Report. Cape Cod Commission, Barnstable, MA. Ellis, M.Y., 1978. Coastal Mapping Handbook, Department of the Interior, U.S. Geological Survey and U.S. Department of Commerce, National Ocean Service and Office of Coastal Zone Management, U.S. GPO, Washington, D.C. Fischer, H. B., List, J. E., Koh, R. C. Y., Imberger, J., and Brooks, N. H. (1979). Mixing in inland and coastal waters. Academic. San Diego. Fiske, J.D., J.R. Curley and Robert P. Lawton, 1968. A study of the marine resources of the Westport River. Monograph Series Number 7. Div. of Marine Fisheries, Department of Natural Resources, Commonwealth of Massachusetts. 52p. FitzGerald, D.M., 1993. “Origin and Stability of Tidal Inlets in Massachusetts.” In: Coastal and Estuarine Studies: Formation and Evolution of Multiple Tidal Inlets, Volume 29, Symposium on Hydrodynamics and Sediment Dynamics of Tidal Inlets (D. G. Aubrey and G.S. Geise, eds.). American Geophysical Union, Washington, D.C. pp. 1-61.

161 MASSACHUSETTS ESTUARIES PROJECT

FitzGerald, D.M., C.T. Baldwin,., N.A. Ibrahim, and D.R. Sands, 1987, Development of the Northwestern Buzzards Bay Shoreline, Massachusetts, in: FitzGerald D.M., and Rosen, P.S., (eds.), Glaciated Coasts, Academic Press, San Diego, CA, p. 327-357. Fleming, R., P. Eng and H. Fraser, 2001. The Impact of Waterfowl on Water Quality – Literature Review. University of Guelph - Ridgetown, Ontario, Canada, September 2001. Geise, G.S., 1988. “Cyclical Behavior of the Tidal Inlet at Nauset Beach, Massachusetts: Application to Coastal Resource Management.” In: Lecture Notes on Coastal and Estuarine Studies, Volume 29, Symposium on Hydrodynamics and Sediment Dynamics of Tidal Inlets (D. Aubrey and L. Weishar, eds.), Springer-Verlag, NY, pp. 269-283. Glennon, B. 2001. Dartmouth: Theearly history of a Massachusetts Coastal Town. Garrison Wall Publishers. Hampson, G.R. 1989. A REMOTS Survey of Buzzards Bay with Ground Truth Verification, EPA Report Region I Water Management Division, CR-8142976-01 Hampson, G.R., E.T. Moul. 1978. No. 2 Fule Oil Spill in Bourne, Massachusetts: Immediate Assessment of the Effects on Marine Invertebrates and a 3-Year Study of Growth and Recovery of a Salt Marsh. Journal of Fisheries Research Bd. Canada 35(5):731-744 Harbaugh, A.W. and McDonald, M.G., 1996. User’s Documentation for MODFLOW-96, an update to the U.S. Geological Survey Modular Finite-Difference Ground-Water Flow Model: U.S. Geological Survey Open-File Report 96-485, 56p. Henderson, F. M., 1966. Open Channel Flow. Macmillan Publishing Company, New York. pp. 96-101. Hoff, J.G. and R.M. Ibara, 1977. Factors affecting the seasonal abundance, composition and diversity of fishes in a southeastern New England estuary. Estuarine and Coastal Marine Science (1977)5: 665-678 Hoff, J.G. P. Barrow and D. A. MCGill, 1969. Some Aspects of the hydrography of a relatively unpolluted estuary in Southeastern Massachusetts. Proceedings of 24th Purdue University Industrial Waste Conference. Part I, p87-98. Howes B., S. W. Kelley, J. S. Ramsey, R. Samimy, D. Schlezinger, E. Eichner (2003). Linked Watershed-Embayment Model to Determine Critical Nitrogen Loading Thresholds for Stage Harbor, Sulphur Springs, Taylors Pond, Bassing Harbor, and Muddy Creek, Chatham, Massachusetts. Massachusetts Estuaries Project, Massachusetts Department of Environmental Protection. Boston, MA. Howes B.L., N.P. Millham, S.W. Kelley, J. S. Ramsey, R.I. Samimy, D.R. Schlezinger, E.M. Eichner. 2012. Linked Watershed-Embayment Model to Determine Critical Nitrogen Loading Thresholds for the Slocum’s and Little River Estuaries, Dartmouth, Massachusetts. SMAST/DEP Massachusetts Estuaries Project, Massachusetts. Department of Environmental Protection. Boston, MA. Howes, B. and L. White. 2005. Massachusetts Estuaries Project, Watershed Nitrogen Loading from Lawn Fertilizer Applications with the Town of Orleans, Massachusetts. Coastal Systems Program, School for Marine Science and Technology, University of Massachusetts Dartmouth. Massachusetts Estuaries Project. Massachusetts Department of Environmental Protection. Howes, B., Kelley, S., Ramsey, J., Samimy, R., Eichner, E., Schlezinger, D., and Wood, J., 2004. Linked Watershed-Embayment Model to Determine Critical Nitrogen Loading Thresholds for Popponesset Bay, Mashpee and Barnstable, Massachusetts.

162 MASSACHUSETTS ESTUARIES PROJECT

Commonwealth of Massachusetts, Department of Environmental Protection, Massachusetts Estuaries Project, 138 pp. + Executive Summary, 10 pp. Howes, B.L., D.D. Goehringer, N.P. Millham, D.R. Schlezinger, G.R. Hampson, C.D. Taylor and D.G. Aubrey. 1997. Nantucket Harbor Study: A quantitative assessment of the environmental health of Nantucket Harbor for the development of a nutrient management plan. Technical Report to the Town of Nantucket, pp. 110. Howes, B.L., H.E. Ruthven, J.S. Ramsey, R.I. Samimy, D.R. Schlezinger, E.Eichner. 2013. Linked Watershed-Embayment Model to Determine Critical Nitrogen Loading Thresholds for the New Bedford Inner Harbor Embayment System, New Bedford, MA (Updated Final Report). SMAST/DEP Massachusetts Estuaries Project. Massachusetts Department of Environmental Protection. Boston, MA. Howes, B.L., J.S. Ramsey and S.W. Kelley, 2000. Nitrogen modeling to support watershed management: comparison of approaches and sensitivity analysis. Final Report to MA Department of Environmental Protection and USEPA, 94 pp. Published by MADEP. Howes, B.L., R.I. Samimy and B. Dudley, 2003. Massachusetts Estuaries Project, Site-Specific Nitrogen Thresholds for Southeastern Massachusetts Embayments: Critical Indicators Interim Report Howes, B.L., R.I. Samimy, D.R. Schlezinger, S. Kelley, J. Ramsey, T. Ruthven, and E. Eichner, 2004. Linked Watershed-Embayment Model to Determine Critical Nitrogen Loading Thresholds for the Quashnet River, Hamblin Pond, and Jehu Pond, in the Waquoit Bay System of the Towns of Mashpee and Falmouth, MA. Massachusetts Estuaries Project Final Report, pp. 147. Howes, D.L, D.D. Goehringer, 1996. Water Quality Monitoring of Falmouth’s Coastal Ponds: Results from the 1994 and 1995 Seasons http://ma.water.usgs.gov/ground_water/ground-water_data.htm http://www.capecodcommission.org/regulatory/NitrogenLoadTechbulletin.pdf http://www.state.ma.us/dep/brp/dws/techtool.htm Huefelder, G. R. 1988. Bacteriological Monitoring in Buttermilk Bay. U.S. Environmental Protection Agency. Region 1 Boston, MA. BBP-88-03 Johnson, R. L., K. T. Perez, E.W. Davey, J.A. Cardin, K.J. Rocha, E.H Dettmann, J.F. Heltsche. 2000. Discriminating the benthic effects of anthropogenic point sources from salinity and non-point nitrogen loading. Unpublished draft journal submission. US EPA Office of Research and Development, National Health and Environmental Effects Research Laboratory, Atlantic Ecology Division, Narragansett, RI 41p. Jorgensen, B.B. 1977. The sulfur cycle of a coastal marine sediment (Limfjorden, Denmark). Limnology Oceanography, 22:814-832. Kelley, S.W., Ramsey, J.R., Côté, J.M., Wood, J.D. (2001). “Tidal Flushing Analysis of Coastal Embayments in Chatham, MA” Applied Coastal Research and Engineering, Inc. report prepared for the Town of Chatham. 115 pp. King, Ian P. (1996). "Users Guide to RMA2 Version 4.2." US Army Corps of Engineers – Waterways Experiment Station Hydraulics Laboratory. King, Ian P., 1990. "Program Documentation - RMA2 - A Two Dimensional Finite Element Model for Flow in Estuaries and Streams." Resource Management Associates, Lafayette, CA.

163 MASSACHUSETTS ESTUARIES PROJECT

Klump, J. and C. Martens. 1983. Benthic nitrogen regeneration. In: Nitrogen in the Marine Environment, (Carpenter & Capone, eds.). Academic Press. Larson, G.J., 1982, Nonsynchronous retreat of ice lobes from southeastern Massachusetts, in Larson, G.J. and Stone, B.D., eds. Late Wisconsinan Glaciation of New England: Dubuque, Iowa, Kendal/Hunt Publishing Co.p.101-114. Lindeburg, Michael R., 1992. Civil Engineering Reference Manual, Sixth Edition. Professional Publications, Inc., Belmont, CA. Manny, B.A., R.G. Wetzel and W.C. Johnson, 1975. Annual contribution of carbon, nitrogen, and phosphorous by migrant Canada Geese to a hardwater lake. Ver. Internat. Verein. Limnol.19:949-951. Maryland Department of Natural Resources. Maryland Mute Swan Task Force Recommendations. Chesapeake Bay Foundation., January 2001. Massachusetts Department of Environmental Protection, 1999. DEP Nitrogen Loading Computer Model Guidance. Bureau of Resource Protection. Boston, MA. Available at: Massachusetts Department of Revenue. November, 2002. Property Type Classification Codes. Massachusetts Water Resources Authority, 1983. Water supply study and environmental impact statement for the year 2020, Task I: Water demand projections. MWRA Report, Boston. Masterson, J.P., Walter, D.A., Savoie, J., 1996, Use of particle tracking to improve numerical model calibration and to analyze ground-water flow and contaminant migration, Massachusetts Military Reservation, western Cape Cod, Massachusetts: U.S. Geological Survey Open-File Report 96-214, 50 p. Mello, M.J. and B. Remington, 2003. Macroinvertebrate inventory within the Buzzards Bay basin focusing on the Westport, Paskamansett and Weweantic River watersheds. Lloyd Center Report #2003-2. submitted to MA Natural Heritage & Endangered Species Program. 13p., 22 fig., 11 tables Melvin, R.L., V. de Lima, and B.D. Stone. 1992. The stratigraphy and hydraulic properties of tills in southern New England. USGS Open-File Report 91-481. 53 p. US Geological Survey. Hartford, CT. Millham, N.P. 1993. Groundwater flow to a shallow coastal embayment: Little Pond, Cape Cod, Massachusetts. Ph.D. Thesis, Boston University, Boston, pp. 237. Millham, N.P. and B.L. Howes, 1994a. Freshwater flow into a coastal embayment: groundwater and surface water inputs. Limnology and Oceanography 39: 1928-1944. Millham, N.P. and B.L. Howes, 1994b. Patterns of groundwater discharge to a shallow coastal embayment. Marine Ecology Progress Series 112:155-167. Millham, N.P. and B.L. Howes. 1994. A comparison of methods to determine K in a shallow coastal aquifer. Groundwater. 33:49-57. Millham, N.P. and B.L. Howes. 1994. Nutrient balance of a shallow coastal embayment: I. Patterns of groundwater discharge. Marine Ecology Progress Series 112:115-167. Murray, D.P. O.D. Hermes and T.S. Durham, 1990. The New Bedford area; A preliminary assessment. Geol. Soc. of Am. Special Paper 245. Normandeau Associates, Inc., 1995. Biological assessment of minimum flow requirements for the Paskamanset River, MA. Report #R-15859.000 prepared for Woodward & Curran. 27 October 1995. 30p.

164 MASSACHUSETTS ESTUARIES PROJECT

Norton, W.R., I.P. King and G.T. Orlob, 1973. "A Finite Element Model for Lower Granite Reservoir", prepared for the Walla Walla District, U.S. Army Corps of Engineers, Walla Walla, WA. O’Hara, C.J., Oldale, R.N., 1987. Geology, Shallow Structure, and Bedform Morphology, Nantucket Sound, Massachusetts. 1:125,000. United States Geological Survey Miscellaneous Field Studies Map MF 1911. Oldale, R. N., 1992, Cape Cod and the Islands: The geologic history: East Orleans, MA, Parnassus Imprints, 208 p. Pacheco, D.M.M., 1993. Temporal changes in fish communities in a southeastern New England estuary: species diversity and multivariate analyses. University of Massachusetts Dartmouth. A Thesis in Marine Biology. Submitted in partial fulfillment of the Requirements for the degree of Master of Science. June 1993. 114p. Pleasant Bay Technical Advisory Committee, and Ridley & Associates, Inc., 1998. Pleasant Bay resource management plan. Report to the Pleasant Bay Steering Committee, 158 pp + app. Pollock, D.W., 1994. User’s Guide to MODPATH/MODPATH_PLOT, version 3 – A particle tracking post-processing package for MODFLOW, the U.S. Geological Survey modular three dimensional finite-difference ground-water-flow-model: U.S. Geological Survey Open-File Report 94-464, [variously paged]. Ramsey, J.S., B.L. Howes, S.W. Kelley, and F. Li (2000). “Water Quality Analysis and Implications of Future Nitrogen Loading Management for Great, Green, and Bournes Ponds, Falmouth, Massachusetts.” Environment Cape Cod, Volume 3, Number 1. Barnstable County, Barnstable, MA. pp. 1-20. Reinert, S.E. and J.O. Hill, Jr., 2001. The Birds of Allens Pond:Ecology of a Coastal Avifauna. Lloyd Center for Environmental Studies, April 2001. Reinert, S.E. and M.J. Mello, 1995. Avian community structure and habitat use in a southern New England estuary. Society of Wetland Scientists.15(1):9-19. Robertson, W.D., S.L. Schiff, and C.J. Ptacek. 1998. Review of Phosphate Mobility and Persistence in 10 Septic System Plumes. Ground Water, 36(6):1000-1010. Ryther, J.H., and W.M. Dunstan. 1971. Nitrogen, phosphorous and eutrophication in the coastal marine environment. Science, 171:1008-1012. Scheiner, D. 1976. Determination of ammonia and Kjeldahl nitrogen by indophenol method. Water Resources 10: 31-36. Schlosser, I.J., 1987. A conceptual framework for fish communities in small warmwater streams, p. 17-24. IN: Community and Evolutionary Ecology of North American Stream Fishes. W.J. Matthews and D.C. Heins, ed. University of Oklahoma Press. Norman, OK Shalowitz, A.L., 1964. Shore and Sea Boundaries--with special reference to the interpretation and use of Coast and Geodetic Survey Data. U.S. Department of Commerce Publication 10-1, Two Volumes, U.S. GPO, Washington, D.C. Smith, K. 1999. Salt Marsh Uptake of Watershed Nitrate, Mashapaquit Creek Marsh, West Falmouth Harbor, Falmouth, Cape Cod, Massachusetts. Masters Thesis, Boston University Department of Earth Sciences, Boston, pp. 1-76.

165 MASSACHUSETTS ESTUARIES PROJECT

Smith, K.N. and B.L. Howes. Manuscript. Attenuation of watershed nitrogen by a New England salt marsh: a buffer for cultural eutrophication of coastal waters. Smith, R.L., B.L. Howes and J.H. Duff. 1991. Denitrification in nitrate-contaminated groundwater: occurrence in steep vertical geochemical gradients. Geochimica Cosmochimica Acta 55:1815-1825. Standbridge, J.H., J. Delfine, J. Kleppe and R. Butler, 1979. Effects of waterfowl (Anas platyrhynchos) on indicator bacteria populations in a recreational lake in Madison, Wisconsin. Applied and Environmental Microbiology. 38(3):547-550. Station, Coastal and Hydraulics Laboratory, Users Guide To RMA4 WES Version Stearns and Wheler. 2001. Wastewater Facilities Plan and Final Environmental Impact Report for the Town of Falmouth, Massachusetts. Hyannis, MA. Stone, B.D. and J.D. Peper. 1982. Topographic control of the deglaciation of eastern Massachusetts: ice lobation and the marine incursion, pp. 145-166 in Larson, G.J. and Stone, B.D., (eds.), Late Wisconsinan Glaciation of New England. Kendall/Hunt Publishing Co. Dubuque, IA. Taylor, C.D. and B.L. Howes, 1994. Effect of sampling frequency on measurements of seasonal primary production and oxygen status in near-shore coastal ecosystems. Marine Ecology Progress Series 108: 193-203. Thieler, E.R., J.F. O’Connell, C.A. Schupp, 2001. The Massachusetts Shoreline Change Project: 1800s to 1994. Technical Report, 60 p. U.S. Army Corps of Engineers (1964). “Beach Erosion Control Report on Cooperative Study of Falmouth, Massachusetts.” Headquarters, Department of the Army, Office of the Chief of Engineers, Washington, D.C. U.S. Army Corps of Engineers, New England Division, Tidal Flood Profiles, New England Coastline, September 1988. U.S. Army, Engineer Research and Development Center, Waterways Experiment Station, Coastal and Hydraulics Laboratory, Users Guide To RMA4 WES Version 4.5, June 05, 2001. Ullman, W.J., K.C. Wong, J.A. Madsen, J.R. Scudlark, D.E. Krantz, A.S. Andres and T.E. McKenna. 2001. Nutrient transport and cycling in an agriculturally impacted coastal watershed: Multidisciplinary approaches to interdisciplinary environmental problems. Proceedings of the 5th annual Environmental Engineering Research Event (20-23 November 2001, Noosa, Queensland). 10 pages. USGS web site for groundwater data for Massachusetts and Rhode Island: Valiela, I and M.L. Cole, 2002. Comparative evidence that salt marshes and mangroves may protect seagrass meadows from land-derived nitrogen loads. Ecosystems (2002) 5:92- 102. Valiela, I, M. Alber and M. LaMontagne, 1991. Fecal coliform loadings and stocks in Buttermilk Bay, Massachusets, USA, and Management Implications. Env. Mngmt. 15(5): 659-674. Van de Kreeke, J., 1988. “Chapter 3: Dispersion in Shallow Estuaries.” In: Hydrodynamics of Estuaries, Volume I, Estuarine Physics, (B.J. Kjerfve, ed.). CRC Press, Inc. pp. 27-39.

166 MASSACHUSETTS ESTUARIES PROJECT

Walter, D.A. and Whealan, A.T. 2005. Simulated Water Sources and Effects of Pumping on Surface and Ground Water, Sagamore and Monomoy Flow Lenses, Cape Cod, Massachusetts. US Geological Survey Scientific Investigations Report 2004-5181, 85 p. Weiskel, P.K. and B.L. Howes, 1991. Quantifying Dissolved Nitrogen Flux Through a Coastal Watershed. Water Resources Research, Volume 27, Number 11, Pages 2929-2939. Weiskel, P.K. and B.L. Howes, 1992, Differential Transport of Sewage Derived Nitrogen and Phosphorous through a Coastal Watershed. Environmental Science and Technology, Volume 26, No. 2, pp. 352 - 360 White, L.M. 2003. The Contribution of Lawn Fertilizer to the Nitrogen Loading of Cape Cod Embayments. A Thesis submitted in the partial fulfillment of the requirements for the degree of Master of Arts in Marine Affairs, University of Rhode Island. Wilhelm, S.R., S.L. Schiff, and W.D. Robertson. 1996. Biogeochemical Evolution of Domestic Waste Water in Septic Systems: 2. Application of Conceptual Model in Sandy Aquifers. Ground Water, 34(5):853-864. Williams, J. R. and Tasker, G.D., 1978, Water resources of the coastal drainage basins of Southeastern Massachusetts, Northwest shore of Buzzards Bay. U.S. Geological Survey Hydrologic Investigations Atlas HA-560. Wood, J.D., J.S. Ramsey, and S. W. Kelley, 1999. “Two-Dimensional Hydrodynamic Modeling of Barnstable Harbor and Great Marsh, Barnstable, MA.” Applied Coastal Research and Engineering, Inc. report prepared for the Town of Barnstable. 28 pp. Woodward & Curran and Normandeau Associates, Inc., 1999. An assessment of the fish assemblage and habitat quality of the Paskamansett River near the Town of Dartmouth, Massachusetts. Report # R-17710.000 to the Town of Dartmouth, MA. January, 1999. 9p., 3 tables, 11 fig., 3 appendices. Zen, E-an, editor, 1983, Bedrock geologic map of Massachusetts: U.S. Geological Survey, scale 1:250,000. Zimmerman, J.T.F., 1988. “Chapter 6: Estuarine Residence Times.” In: Hydrodynamics of Estuaries, Volume I, Estuarine Physics, (B.J. Kjerfve, ed.). CRC Press, Inc. pp. 75-84.

167

APPENDIX 1

Excerpts from:

Turn the Tide Project - Natural Resource Assessment of the Slocums River, Little River and Apponagansett Bay Watersheds

Chapter VII - Shellfish Chapter VIII - Birds

A-1

VII.6 SHELLFISH SURVEYS IN DARTMOUTH ESTUARIES (APPONAGANSETT BAY)

As part of the Turn the Tide Natural Resources Assessment for the Town of Dartmouth, a series of detailed shellfish surveys were completed in 2004 and 2005 and used to augment the habitat assessment of the Apponagansett Bay estuary under the MEP. A detailed examination of the shellfish resources of the Town of Dartmouth was undertaken because shellfish beds are an integral part of the coastal economy, a recreational pastime, and, like other members of the benthic invertebrate community, an indicator of estuarine habitat quality.

Shellfishing has historically provided substantial commercial benefits to the Town of Dartmouth (Figure VII-15). However, as can be seen from Figure VII-15, the value of the landings has fallen in most years since the peak landing year of 1985. The amounts of individual species harvested have varied widely from year to year and are influenced by many factors including: availability, market price, fishing effort; cycles of reproduction, loss of habitat, predation, disease, closures due to toxic algal blooms, oil spills, hurricanes and rainfall intensity; and poaching. Of these factors, loss of habitat can be directly related to excessive nutrient inputs, while toxic algal blooms may also be related to increased nutrients in coastal waters. Chronic bacterial contamination of Dartmouth waters is primarily due to stormwater runoff from watersheds.

Dartmouth Shellfish Landings Annual Total value

$3,000,000

$2,500,000

$2,000,000

$1,500,000

$1,000,000

$500,000

$0 1938 1944 1949 1954 1959 1964 1969 1974 1979 1984 1989 1994 1999

Figure VII-15 Annual value in dollars of the commercial and recreational shellfish harvest in the Town of Dartmouth (Source: Dartmouth Town Reports).

The annual harvests statistics have been dominated by the hard shell clam (=quahog) with oysters, bay scallops, soft shelled clams and surf clams being landed in lesser amounts or for some years not at all (Figure VII-16).

The decline in the scallop harvests since the last significant harvest in the mid-1980’s may be in part due to decline in the extent of eelgrass beds within the Town waters (Figure VII-17). (For discussion of eelgrass distribution in The Town of Dartmouth Apponagansett Bay, please

A-2

refer to Section VII.3 above. Eelgrass beds provide the primary habitat for bay scallops and reductions in aggregate eelgrass bed coverage imply smaller scallop populations available for harvest. Today, remaining eelgrass beds in Dartmouth occur predominantly in outer Apponagansett Bay and along the shores of Ricketsons Point, Smith Neck, Mishaum Point and Barneys Joy Point.

DARTMOUTH TOTAL LANDINGS IN POUNDS

3,500,000

3,000,000 Bay Scallops Oysters Quahogs Surf Clams 2,500,000

2,000,000

POUNDS 1,500,000

1,000,000

500,000

0

3 7 7 57 59 61 6 65 67 69 71 73 75 7 79 81 83 85 8 89 91 93 95 97 99 1955 YEAR

Figure VII-16. Shellfish harvest by type for the period 1955-1999. The quahog dominates the shellfish landings. (Source: Dartmouth Town Reports; 2002 Shellfish Management Plan, SMAST).

Oyster harvests in Dartmouth waters also have been highly variable over the past 60 years, reaching a peak in the mid 1980’s and again in the late 1990’s (Figure VII-18). Inner Apponagansett Bay was a primary source for the 1980’s harvest, which was enhanced by transplanting oysters from the closed east side of the inner bay to purify in open areas on the west side of the bay. Other areas that produced oysters in Dartmouth during this period include the lower Slocums River, Little River near the Little River Bridge, Allens Pond and a small area south of the Padanaram causeway on Smith Neck Road. The variability of the oyster harvest is likely to have complex causes. Oysters are sensitive to environmental conditions such as salinity and bottom sediment type and they are prey to multiple diseases and predators, so it is difficult to say which environmental factors are the most important cause of the oyster harvest variability and decline in Apponagansett Bay specifically or in a more general sense across all Dartmouth waters over the past 20 years.

A-3

DARTMOUTH SCALLOP LANDINGS IN POUNDS

450,000

400,000

350,000

Recreational 300,000 Commerciall

250,000

POUNDS 200,000

150,000

100,000

50,000

0

9 1 57 5 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 9 93 95 97 99 1955 YEAR

Figure VII-17. Scallop landings by weight. The largest landings were during the period 1964-72 when annual catch reached 54,000 lbs in 1965 and in 1971. The last large catch of bay scallops was in 1984-1985 when about 11,000 and 18,000 lbs, respectively were landed. Small catches of 150 to1400 lbs were landed between 1994 and 1997.

The availability of shellfish for harvest is also affected by the closure of shellfishing areas due to high bacterial counts. In Buzzards Bay as a whole, shellfish closures increased during the thirty-year period from 1960-1990 from less than 5,000 acres to more than 15,000 acres (Buzzards Bay National Estuary Program). Significant areas within Apponagansett Bay and all of the Slocums River are permanently closed (prohibited) due to chronic bacterial levels. Closures in inner Apponagansett Bay began in the 1920’s and since the 1970’s have been more or less continuous.

The bacterial-related closures of these Dartmouth water bodies are caused by moderate to high bacterial counts in some of the streams flowing into them. Closures along the north and east side of Apponagansett Bay are due to primarily to relatively high bacterial counts in Buttonwood Brook, Apponagansett Bay Brook and several storm drains that drain Padanaram village areas. A detailed discussion of the bacterial conditions of the streams and embayments of Dartmouth is available in the Total Maximum Daily Load reports prepared by SMAST for these water bodies.

Together, quahogs and soft-shell clams are two staple species in the region. Soft-shell clams are a primary species for recreational harvests, while quahogs have been the dominant commercial shellfish species landed in Dartmouth estuaries for decades. Areas BB12.3 and

A-4

BB12.2 (Figure VII-19) in Apponagansett Bay have been especially important, sometimes producing up to 98% of the commercial quahog catch in Dartmouth (SMAST 2000).

DARTMOUTH OYSTER LANDINGS IN POUNDS

900,000

800,000

700,000 Total Recreational 600,000 Commerciall

500,000

POUNDS 400,000

300,000

200,000

100,000

0

57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 1955 2001 2003 YEAR

Figure VII-18. Recreational and commercial oyster harvests in pounds since 1955. Data source: SMAST 2002; Dartmouth Town Reports.

The decline in hard shell clam (quahog) harvest and other species has occurred despite the efforts of the Shellfish Department to manage the resources by both transplanting mature stock for purification and to seeding other areas with immature stock. In Apponagansett Bay north of the Padanaram Bridge, large-scale propagation of quahogs was carried out between 1958 and 1973 (Dartmouth Town Reports). After an overall decline in quahog landings from 1964-1980 (Figures VII-15 & 16) the mid to late 1980’s marked a peak in Dartmouth quahog landings, helped by planting activities in closed beds. But this was followed by a large decrease and associated reduction in quahog value by 1999 (SMAST, 2000).

From Figures VII-15 and VII-16 it is clear that the status and productivity of shellfish beds has changed constantly with time over the past 60 years. During the early 1960’s annual harvests varied in value from about 1.5 million pounds to 3.5 million pounds. As water quality data for Dartmouth embayments for the period prior to 1991 is scant, it is not clear to what degree declines in water quality during the same period affected shellfish harvests, though the loss of eelgrass beds in inner Apponagansett Bay was likely due to excess nitrogen entering the embayment from the watershed. However, it is clear that improvements in water quality from nitrogen management can lead to restoration of eelgrass beds and thus bay scallop habitat.

A-5

Likewise, bacterial source reduction in the embayment watersheds can decrease both permanent and conditional shellfish bed closure rates, leading to increased availability of shellfish stocks for harvest.

Records for Dartmouth’s shellfish harvests are not separated by estuary. Therefore, in order to provide data for assessing the existing shellfish distribution and abundance and to provide a baseline measurement for documenting future shellfish distribution and abundance in Apponagansett Bay, as well as Slocums River and Little River, CSP and Lloyd Center scientists conducted shellfish surveys of clams and oysters over the period 2003-2005. The results of the surveys (specifically those for Apponagansett Bay) are presented as a supplement to the typical MEP habitat assessment.

A-6

BB-12.4 BB-12.1

BB-12.20

BB-12.3 BB-12.2

BB-12.7

BB-12.5

Figure VII-19. Shellfishing closures in Apponagansett Bay. Areas BB-12.4, 12.1 and 12.2 are prohibited for shellfishing. The locations of the closed areas are related to surface water sources of stormwater runoff from Padanaram village areas and the upper Apponagansett Bay watershed. (Map source MA Division of Marine Fisheries)

A-7

VII.6.1 Survey Methodology

Apponagansett Bay

Shoreline transects. Seven transects were established in Apponagansett Bay. (Figure VII- 20; Table VII.7.5). Stations at transects in Apponagansett Bay were established in a similar manner as for the transect stations in Slocums River. Sample locations and methodology were the same as for the survey undertaken in the Slocums River. An additional station was added to the seaward end of Transects 1 & 6 due to their gradual slope. Sub-tidal sampling stations were also established for the Apponagansett Bay survey and are depicted in Figure VII-21.

Table VII.7.5. GPS coordinates for Apponagansett Bay shellfish transects.

Transect North West survey dates (not incl. oyster survey) AP-1 41° 35' 00.72" 70° 57' 19.02" 28-Oct-04 AP-2 41° 35' 5.94" 70° 57' 16.44" 10-Nov-04 AP-3 41° 35' 12.18" 70° 56' 45.3" 11-Nov-04 AP-4 41° 35' 31.14" 70° 57' 7.62" 15-Nov-04 AP-5 41° 35' 46.5" 70° 57' 25.8" 9-Nov-04 AP-6 41° 35' 34.8" 70° 57' 33.9" 26, 27-Oct-04 AP-7 41° 35' 8.64" 70° 57' 44.52" 26-Oct-04

AP5

AP6

AP4

AP3 AP7 AP2

AP1

Figure VII-20. Location of transects established for shellfish survey in Apponagansett Bay during 2004-05.

A-8

A characterization of the sediments at the stations at each transect follows in the table below.

Transect Station 1 Station 2 Station 3 Station 4 AP-1 sand & gravel sand & gravel Sand sand

AP-2 sand & gravel muddy sand & gravel muddy sand Sand, gravel, cobble, AP-3 Sand, gravel, cobble, shells shells AP-4 muddy sand, gravel muddy sand, gravel muddy sand, gravel muddy sand, gravel

AP-5 muddy sand muddy sand muddy sand muddy sand

AP-6 fine sand over peat fine sand over peat muddy sand, peat Muddy sand, peat sand and gravel over sand and gravel over AP-7 Fine sand over peat peat peat

1

2

4 3

5

6 7 16

8 10 15 9 14 11 13

12

18

19

20

Figure VII-21. Location of subtidal stations surveyed for shellfish in Apponagansett Bay during 2004.

A-9

Oysters in Apponagansett Bay were surveyed in summer, 2005 at the four transect sites in the same manner as in the Slocums River (Figure VII-20) making survey results in Dartmouth's estuaries cross comparable.

Subtidal stations. Twenty subtidal stations were randomly chosen throughout Apponagansett Bay north of Padanarum Bridge, which were located in the field and sampled in the same manner as in the Slocums River (Figure VII-21; Table VII.7.6).

Table VII.7.6. GPS coordinates, dates and sediment characteristics for Apponagansett Bay subtidal stations. Station North West Date Sediment characteristics AP1 41° 35' 48.0" 70° 57' 34.6" 21-Sep-04 organic mud AP2 41° 35' 42.4" 70° 57' 32.1" 21-Sep-04 organic mud AP3 41° 35' 35.9" 70° 57' 31.4" 21-Sep-04 organic mud AP4 41° 35' 33.9" 70° 57' 20.4" 5-Oct-04 mud AP5 41° 35' 27.3" 70° 57' 24.7" 5-Oct-04 silty mud AP6 41° 35' 22.5" 70° 57' 29.9" 5-Oct-04 silty mud, some gravel AP7 41° 35' 20.5" 70° 57' 46.7" 22-Sep-04 organic mud AP8 41° 35' 16.0" 70° 57' 48.3" 22-Sep-04 organic mud AP9 41° 35' 10.7" 70° 57' 36.2" 22-Sep-04 organic mud AP10 41° 35' 15.6" 70° 57' 22.1" 22-Sep-04 organic mud AP11 41° 35' 8.6" 70° 57' 26.1" 22-Sep-04 organic mud AP12 41° 35' 00.4" 70° 57' 25.4" 5-Oct-04 muddy sand AP13 41° 35' 4.8" 70° 57' 9.8" 21-Oct-04 muddy gravel AP14 41° 35' 9.6" 70° 57' 6.9" 21-Oct-04 mud AP15 41° 35' 14.5" 70° 57' 7.7" 21-Oct-04 mud AP16 41° 35' 18.6" 70° 57' 8.1" 21-Oct-04 mud AP17 41° 35' 24.3" 70° 57' 2.3" 21-Oct-04 mud AP18 41° 34' 56.8" 70° 57' 25.5" 13-Oct-04 gravelly sand AP19 41° 34' 48.7" 70° 57' 25.3" 13-Oct-04 fine sand AP20 41° 34' 38.3" 70° 57' 18.0" 8-Oct-04 gravelly sand

Because larger invertebrates may have been under-sampled by the size of the grab sampler used during the benthic invertebrate survey, invertebrates retained within the mesh basket were identified and counted during the shellfish survey for all three Dartmouth estuaries. These organisms, including marine worms, were preserved in alcohol for later counting and identification.

VII.6.2 Results and Discussion

Quahog/Soft-shelled Clam Surveys (Apponagansett Bay)

A total of 117 quahogs and 12 soft-shelled clams were encountered along the transects, and 17 quahogs were encountered in the subtidal samples. Quahogs were most abundant in the transects nearest the bridge (Figure VII-22), but were scant in the subtidal stations (Figure VII-23) except for Station 20 south of Gulf Hill Road bridge. Soft-shelled clams were in low numbers in the transects (Figure VII-22), and they were absent from the subtidal stations.

A-10

Key

30 20 10 0 mercenaria mya AP5 30 20 10 30 0 20 AP6 10 0 AP4

60 30 50 40 20 30 20 10 30 10 20 0 0 AP3 AP7 10 0 2 AP2

30 20 10 0 AP1

Figure VII-22. Distribution of quahogs (Mercenaria) and soft-shelled clams (Mya) in Apponagansett Bay transects during 2004.

A-11

3

2

1

2

1

2

6

Figure VII-23. Distribution of quahogs in the subtidal Apponagansett Bay stations during 2004.

Quahogs reached the highest densities at Transect 3 (29/m2) and Transect 2 (12/m2) respectively (Figures VII-22 & VII-24). The sediments at these stations were comprised of a mix of sand and gravel, with a lesser portion of silt/organic mud present. Sediments at subtidal Stations 18-20 south of the Gulf Hill Road bridge were a mix of fine sand and gravel, but the remaining stations were primarily silt/organic mud. This at least partially explains the paucity of quahogs in the subtidal samples.

Within transects, quahogs favored the deeper stations over mean low water Station 1 only at Transect 3, the opposite trend seen in Slocums River (Figure VII-25).

A-12

MASSACHUSETTS ESTUARIES PROJECT

30 quahog

soft-shelled clam 25

20

15 density (mm2 density

10

5

0 1234567 Transect

Figure VII-24. Distribution of quahog and soft-shelled clam densities on transects sampled in Apponagansett Bay north of Padanarum Bridge.

25 Station 1

Station 2

Station 3

20 Station 4

subtidal

15 number

10

5

0 1-20 21-40 41-60 61-80 81-100 >100 size class (mm)

Figure VII-25. Size class distribution for quahogs sampled from Apponagansett Bay north of Padanarum Bridge.

However, size class distribution reflects a trend also seen in the Slocums River towards larger sizes from the subtidal stations and lower portions of the transects (Figure VII-25). Model II ANOVA (Table VII-7) results in an estimate that 94% of the variance in size distribution in the

A-13

MASSACHUSETTS ESTUARIES PROJECT transect and subtidal stations in Apponagansett Bay is due to within station variance, which is consistent with the results from Slocums River. Thus, as in Slocums River, relatively more variation occurs within than among stations.

Table VII-7. Model II ANOVA calculations for quahog size distribution in the Slocums River among transects 1 – 3 and subtidal stations.

Source df SS MS F F(.05,[3,12 F(.01…

among 3 6849.002 2283.001 8.940362 2.68 3.95

within 127 32430.58 255.3589

total 130

n0 1/(4-1)([35+28+51+17]-[1225+784+2601+289]/[35+51+17+28]) = 118.5344

(MS groups - MS within)/No = 17.10594

s2 + s2A = 255.36 + 17.11 = 272.47

s 2= 94% s2A = 6%

An in both Slocums and Little Rivers, young quahogs are nearly absent in the samples (Table VII-8). Fewer than 19% (25 out of 134) are less than 40mm in diameter. The benthic invertebrate survey samples analyzed by Cove Corporation produced only a single quahog. In cannot be concluded from our survey, however, whether the causes of low recruitment in Dartmouth’s three estuaries (Figure VII-26) are related.

Table VII-8. Size distribution of quahogs in Apponagansett Bay documented during the shellfish survey. size Station 1 Station 2 Station 3 Station 4 subtidal total 1-20 400004 21-40 8660121 41-60 17 13 18 1 4 53 61-80 672311047 81-100 024028 >100 000000 crushed 000101

total 35 28 51 3 17 134

During September, 1998 as part of the Regional Shellfish Restoration Plan an inventory of shellfish using a suction dredge in Apponagansett Bay was contracted by the Town of Dartmouth. Ten of the sampling stations fell within Shellfish Area BB12.3 (refer to Figure VII-19 for location of BB12.3), which allowed for temporal comparison with data collected at eleven subtidal stations in the same area during 2004. The comparative data indicates a decline from 2.2 quahogs/m2 in 1998 to 1.3/m2 during our survey. As in our survey, juveniles were also lacking in the 1998 survey, which produced a ranges of quahog sizes from 60 – 125 mm in

A-14

MASSACHUSETTS ESTUARIES PROJECT diameter. Because of the large sizes, it is unlikely that year to year variability in recruitment accounts for this decline, however, it should be noted that sampling methodologies incorporated in the two surveys were not necessarily equivalent. Possible causes of larval and/or juvenile mortality include predation on larvae by fish, newly settled juveniles by crabs, or an inhospitable substrate settlement environment.

45 Slocums

Little 40 Apponagansett

35

30

25

number 20

15

10

5

0 1-20 21-40 41-60 61-80 81-100 >100 size class (mm)

Figure VII-26. Size class distribution for quahogs in three Dartmouth estuaries sampled during 2003-2004.

Absence of quahogs and soft-shelled clams from the mid to upper portions of each estuary may, however, be correlated at least indirectly with water quality. High nutrient loads produce yearly macroalgae blooms that eventually decompose, adding to the organic mud and creating a substrate that is inhospitable for shellfish. The macroalgae surveys (Section VII.4 above) confirm the high density of macroalgae in the areas virtually devoid of shellfish.

Oyster Survey (Apponagansett Bay)

In Apponagansett Bay (Figure VII-27) 13 live and 137 dead oysters were counted during this survey.

Results of the survey in Apponagansett Bay (inclusive of Slocums and Little River) are summarized in Table VII-9. Live oysters were noted at only five of the nineteen sites, with two sites holding only one oyster. The three other live oyster sites contained between 12 and 17 oysters. Highest density in a single quadrat was 10 oysters at Station 8 in the Slocums River. Mean shell length at this site was 129 mm (s.d., 22 mm). The other sites with live oysters had generally lower mean lengths: 85, 53, 52 mm. One site with one live oyster had a length of 123 mm.

A-15

MASSACHUSETTS ESTUARIES PROJECT

Oyster shells (dead) were noted at 18 of the 21 sites sampled. At sites with live oysters, the number of dead shells was only somewhat correlated with the number of live oysters, with the highest live and dead being at the same site (Slocums #1). Slocums Transect #8 also had the highest number of dead oysters with a maximum of 43 shells found in a single quadrat.

Key

40 AP5 30 Live Dead 20 10 0 AP 5

60 40 30 40 20 20 10 0 0 AP 6 AP 4

40 30 20 10 0 AP 7

40 30 40 20 30 40 10 30 20 0 20 10 10 0 AP 3 0 AP 2 AP 1

Figure VII-27. Locations of Apponagansett Bay oyster survey sites with counts of live and dead oysters for each site. Blank graphs indicate no presence of either live or dead oysters.

There are few sites with live oysters in all three water bodies examined and only 3 of 21 sites examined held more than 1 live oyster. The sites with live oysters also had distinct differences in average length, with the site (Slocums #8) having the largest concentration and largest mean length, while the other two sites with more than 1 live oyster had about 66% and 40% mean lengths of Slocums #8.

The presence of dead oysters at 13 of the nineteen sites suggests that oysters were more widely distributed in Apponagansett Bay and the Slocums River, while the small number of dead shells in Little River seems to indicate that there has not been a substantial recent population

A-16

MASSACHUSETTS ESTUARIES PROJECT there. Anecdotal information from the present and past Town Shellfish Constables indicates that Apponagansett Bay was the main source of oysters in the Town-wide total oyster harvest with the Slocums River, Little River and Allens Pond providing additional amounts. The peak oyster harvests of the mid-1980’s indicate substantial resources at that time, but also a high variability in harvest totals from year to year (Figure VII-28). The size and number of individual oysters at Slocums River #8 indicates that oysters can reach relatively large size today. The lack of smaller individuals in the population at Slocums #8 may indicate that there is no replacement cadre of younger oysters to replace those larger ones lost to predation, parasites and disease. On the other hand the lack of larger individuals in the two other multiple-individual sites may indicate a younger population, high stress levels or a selective loss of larger individuals to predation.

Table VII-9. Live and dead oysters found in Apponagansett Bay (AP), Slocums River (SR) and Little River (LR). Live Live Live Mean Mean Dead Live Density Length Std DevDead Density Station Count (indiv/m^2) (mm) (mm) Count (indiv/m^2) SR 1 Not sampled SR 2 17 1.7 84.8 20.7 33 3.3 SR 3 0 - - - 0 - SR 4 0 - - - 16 1.6 SR 5 0 - - - 2 0.2 SR 6 0 - - - 14 1.4 SR 7 0 - - - 7 0.7 SR 8 40 4 129 22 236 23.6 SR 9 0 - - - 0 - LR 1 0 - - - 0 - LR 2 0 - - - 0 - LR 3 0 - - - 0 - LR 4 1 0.1 53 4 0.4 AP 1 0 - - - 6 0.6 AP 2 0 - - - 29 2.9 AP 3 0 - - - 2 0.2 AP 4 1 0.1 0 59 5.9 AP 5 12 1.2 52 10 27 2.7 AP 6 0 - - - 0 - AP 7 0 - - - 14 1.4 Notes: 40 meter survey length 10 - 1 m^2 quadrats

The causative agent for the absence of few live oyster sites with correspondingly high occurrences of dead oysters is not easy to determine. Oysters are sensitive to environmental conditions such as salinity and bottom sediment type and they are prey to disease and predators. Anecdotal evidence suggests that there has been an increase in the amount and extent of highly organic bottom sediments in the central area of the Slocums River, likely due to nitrogen enrichment. Changes in bottom sediment character in Apponagansett Bay may also be a contributing factor to reductions in favorable habitat. Such an increase in unfavorable sediment conditions would reduce available oyster habitat in the embayments. Oysters require salinities ranging from 10 to 32 ppt to survive, while an optimal range for oyster growth is about

A-17

MASSACHUSETTS ESTUARIES PROJECT

16 to 26 ppt. Average salinity in the upper portion of the Slocums River in 2004-2005 was 18.2 ppt (S.D. 10.1), mean salinity in the middle portion of the embayment was 25.7 ppt (S.D 5.9) and 30.1 ppt (S.D. 1.9) in the lower embayment, suggesting that oyster reproduction and growth would not be limited by salinity in large areas of the Slocums embayment. Salinities in inner Apponagansett Bay over the period 2004-2005 averaged between 30 and 31 ppt which is above the optimal range for oysters, however lower salinities near the shorelines in Apponagansett Bay due to local groundwater discharge may create small areas of somewhat lower, more favorable salinity. In any case, we have no evidence that salinities in Apponagansett Bay have changed over the past 15 years and since a large proportion of the peak harvests were obtained from Apponagansett Bay, it seems likely that salinity was not a controlling factor in the oyster population dynamics then or that it would be so now. Thus, it is difficult to say from the limited scope of this survey which environmental factors are the most important influences upon the oyster harvest variability in Dartmouth waters over the past 40 years.

DARTMOUTH OYSTER LANDINGS IN POUNDS

900,000

800,000

700,000 Total Recreational 600,000 Commerciall

500,000

POUNDS 400,000

300,000

200,000

100,000

0

3 5 9 1 3 5 9 3 7 9 3 7 9 1 3 55 57 59 61 6 6 67 6 7 7 7 77 7 81 8 85 8 8 91 9 95 9 9 0 0 9 0 1 20 2 YEAR

Figure VII-28. Recreational and commercial oyster harvests in pounds since 1955. Data source: SMAST 2002; Dartmouth Town Reports.

A-18

MASSACHUSETTS ESTUARIES PROJECT

VII.7 WATERBIRD SURVEYS OF APPONGANSETT BAY AND WATERSHED

Estuaries, such as Apponagansett Bay (Figure VII-29) serve as important habitats for many bird species on a year-round basis and therefore are a critical component of the region’s avian diversity. For example, the American Black Duck (Anas rubripes) exemplifies the most estuary-dependent wintering waterfowl, one that relies nearly exclusively on intertidal habitats as a winter food source (Reinert and Hill, 2001. In addition to waterfowl, gulls, terns, shorebirds, and waders also rely on estuarine wetlands as migratory stopover sites, wintering grounds, or nesting areas (Reinert and Mello, 1995).

AB2

AB1 AB3 AB4

AB5 AB6

Figure VII-29 Avian survey zones for Apponagansett Bay.

Birds are also a potential source of water quality contamination from bacterial and nutrient loads. Canada geese, mallards and swans, and ring-billed gulls in particular, as well as waterfowl and gulls in general have been implicated as major contributors to nutrient and/or bacterial pollution both locally (Buttermilk Bay: Heufelder,1988; Valiela et al.,1991) and regionally (Chesapeake Bay, Maryland: Maryland Division of Natural Resources, 1989; Fleming et al., 2001; Manny et al.,1975, 1994 in a 15 hectare lake in Michigan; Standridge et al, 1979 in a lake in Wisconsin; Basely and Jefferies,1985 in Hudson Bay). Gulls have been implicated as major bacterial contaminators in lakes and ponds in Westchester County, NY (NY DEP). Thus, water birds are both an important resource and under certain circumstances, they can be a potential source of elevated nutrient loading. This assessment was undertaken for this survey was designed to address both issues:

1) Biodiversity – Determine overall species richness and abundance and identify records of state and federally listed species and regionally important species within these systems.

2) Water quality – Determine the areas of highest abundance and density of waterbirds to identify potential “hotspots” for bacterial and nutrient loading.

A-19

MASSACHUSETTS ESTUARIES PROJECT

(The full water bird report is included in the Turn the Tide Project: Natural Resources Assessment of the Slocums River, Little River and Apponagansett Bay Watersheds (April 2009) which was completed for the Lloyd Center for the Environment.)

Two hundred-one hectares of the Apponagansett Bay estuary (Figure VII-29) and the 2.4 hectare Buttonwood Pond (Figure VII-30) were inventoried. Apponagansett Bay was subdivided into six zones: one south of Padanarum Bridge, one south of Gulf Road, and four north of the road and bridge. Inventories were conducted on approximately a weekly basis.

Vantage points for survey Vantage Points for survey

Figure VII-30 Vantage points for Buttonwood Pond avian survey.

Bird results are expressed in terms of total number, relative abundance (mean number per survey) and density (number per hectare per survey). Birds were lumped seasonally by winter (December-February), spring (March-May), summer (June-August), and fall (September- November). Data from an independent study, the Lloyd Center’s winter waterfowl count, are based on two winter counts per season: I –early December; II – last weekend in January or first weekend in February from 1988 through 2007.

A-20

MASSACHUSETTS ESTUARIES PROJECT

VII.7.1 Abundance and seasonal distribution

Apponagansett Bay: One hundred-eight surveys were conducted in Apponagansett Bay between 2002 and 2006 during which time 32,939 birds comprising 58 species were counted: 16,428 during winter; 7,169 during fall; 6,229 during spring; and 3,113 during summer (Table VII-10, Figure VII-31). Ducks (14,125 – 43%) gulls (10,500 – 32%) and geese (3,590 – 11%) were the dominant bird groups in Apponagansett Bay. Waterfowl (ducks and geese) were most abundant in winter. Gull numbers were higher in winter than summer, but less dramatically than for waterfowl. Less abundant groups included passerines (2,791), which were relatively evenly distributed year-round; and cormorants (1,077), which peaked during fall.

The top five species (Figure VII-31) in order of abundance were: ring-billed gull (5,764 – 17%), bufflehead (5,759 – 17%), herring gull (4,223 – 13%), mallard (3,516 – 11%), and Canada goose (3,056 – 9%). Buffleheads and ring-billed gulls were co-dominant and showed nearly identical seasonal trends except for summer, when buffleheads were absent. Herring gulls were most abundant during winter and spring. Mallards and Canada geese showed similar seasonal trends, and both peaked during winter.

A-21

MASSACHUSETTS ESTUARIES PROJECT

Table VII.8.1 Seasonal totals for birds censused in Apponagansett Bay over a 4 year period (2002-2006).

winter spring summer fall Family/Group common name 23 surveys 27 surveys 28 surveys 30 surveys total Anatidae bufflehead 3320 1013 1426 5759 mallard 2754 207 279 276 3516 American black duck 822 250 147 1219 scaup species 662 130 30 822 red-breasted merganser 200 432 107 739 dabbling duck 389 4 32 425 duck spp. 340 16 8 24 388 diving duck spp. 201 91 28 320 gadwall 139 6 147 292 domestic duck 45 53 25 123 common goldeneye 97 16 1 114 long-tailed duck 101 6 1 108 surf scoter 59 37 96 greater scaup 79 0 79 american wigeon 2 59 61 scoter species 40 40 common eider 7 2 9 hooded merganser 3 4 7 wood duck 5 1 1 7 black duck-mallard hybrid 1 1 ducks 9221 2217 342 2345 14125

Canada goose 1612 310 768 366 3056 domestic goose 124 129 160 413 brant 113 7 120 snow goose 1 1 geese 1850 439 768 533 3590

swans mute swan 102 48 8 53 211

Laridae gull spp. 3249 1392 285 944 5870 ring-billed gull 960 369 239 773 2341 herring gull 449 630 500 285 1864 great black-backed gull 78 119 93 70 360 laughing gull 38 26 64 bonapartes gull 1 1 gulls 4736 2510 1155 2099 10500

common tern 29 56 85 least tern 1 4 5 terns 0 30 60 0 90

Columbidae rock dove 465 747 298 1108 2618 mourning dove 5 7 4 16 Corvidae American crow 9 47 26 49 131 Passeridae house sparrow 11 11 Alcedinidae belted kingfisher 1 8 9 Icteridae common grackle 5 5 Sturnidae European starling 1 1 passerines, etc. 474 799 338 1180 2791

Phalacrocoracidae double-creasted cormorant 1 77 155 540 773 cormorant spp. 29 18 257 304 cormorants 1 106 173 797 1077

Ardeidae great blue heron 5 2 118 70 195 great egret 22 55 30 107 snowy egret 9 25 12 46 green heron 4 4 heron spp. 2 2 little blue heron 1 1 waders 5 34 204 112 355

Haematopdidae American oystercatcher 1 1 Charadriidae semipalmated plover 10 9 19 black-bellied plover 1 1 Scolopacidae yellowlegs spp. 12 2 11 25 semipalmated sandpiper 13 6 19 shorebird spp. 17 2 19 dunlin 7 7 14 willet 2 2 shorebirds 7124536100

Podicipedidae horned grebe 1 1 Gavidae common loon 26 22 8 56 loon spp. 4 5 9 red-throated loon 1 1 grebes/loons 31 28 0 8 67

Accipitridae osprey 5 15 4 24 turkey vulture 3 1 4 northern harrier 1 1 red-tailed hawk 1 1 sharp-shinned hawk 1 1 hawk spp. 1 1 Strigidae barred owl 1 1 raptors/owls 1 6 20 6 33

total individuals 16428 6229 3113 7169 32939 mean/survey 714 231 111 240 324 total number species 58

Table VII-10 Seasonal Totals for birds censused in Apponagansett Bay over 4 year period (2002-2006).

A-22

MASSACHUSETTS ESTUARIES PROJECT

150 bufflehead 130 ring-billed gull herring gull 110 mallard

90 canada goose

70

50 mean # of birds per survey per # of birds mean 30

10

winter spring summer fall -10 season

Figure VII-31 Seasonal abundance for the dominant bird species associated with Apponagansett Bay over a 4-year period, 2002-2006.

Less abundant species included scaup (primarily, if not exclusively greater scaup), which were common during winter and red-breasted merganser in spring, winter and fall. Swans were present in modest numbers, while shorebirds were relatively scarce. Double-crested cormorants were common during fall and moderately abundant during summer. Great blue heron were common during summer. Common loons were seen more frequently than horned grebes. American crow were moderately abundant.

Buttonwood Pond: Fifty-five surveys were conducted at Buttonwood Pond from 2004-2006, during which time 11,947 birds consisting of 25 species were counted: 3,839 during fall; 3,623 during summer; 2,252 during spring; and 2,234 during winter (Table VII-11, Figure VII-32). Geese (3,905 – 33%), gulls (3,701 – 31%) and to a lesser extent, ducks (3,395– 28%) were co- dominant. Unlike the estuarine portion of the Apponagansett watershed, geese and ducks were most abundant during summer and fall, reflecting a pattern of semi-domestication. Gulls, however, were most abundant during winter and spring. Passerines (830) made up a smaller, but significant group that peaked during the fall and winter months.

The top five species (Figure VII-32) in order of abundance were: Canada goose (3,648 – 31%), (3,226 – 27%), ring-billed gull (3,174 – 27%), mallard, rock dove (804 - 7%), and herring gull (432 – 4%). Canada goose and ring-billed gull numbers were similar, but geese peaked in summer and fall, gulls in winter and spring. Mallards were significantly more abundant in summer and fall than in the winter and spring months.

A-23

MASSACHUSETTS ESTUARIES PROJECT

Table VII.8.2 Seasonal totals and mean numbers for bird species and groups at Buttonwood Pond over a 2-year period, 2004-06. winter spring summer fall 10 surveys 15 surveys 17 surveys 13 surveys total Anatidae Canada goose 410 345 1352 1542 3648 domesticated goose 36 80 71 70 257 geese 446 425 1423 1612 3905

mallard 178 282 1632 1134 3226 hybrid or domestic duck 33 44 39 33 148 domesticated muscovy 1 8 9 bufflehead 6 6 American black duck 2 2 1 5 hooded merganser 1 1 ducks 214 333 1671 1176 3395

swans mute swan 217392381

Laridae ring-billed gull 962 1287 279 646 3174 herring gull 266 65 39 62 432 great black-backed gull 69 11 3 7 90 laughing gull 1 3 4 lesser black-backed gull 2 2 gulls 1299 1362 322 718 3701

Columbidae rock dove 270 92 139 303 804

Icteridae common grackle 10 4 14 red-winged blackbird 3 2 5 Sturnidae European starling 7 7 passerines 13 13 26

Ardeidae great blue heron 5 3 3 11 great egret 2 6 7 waders 69318

Phalacrocoracidae double-crested cormorant 42 6

Rallidae American coot 325

Charadriidae killdeer 2 2 Scolopacidae semipalmated sandpiper 2 2 shorebird spp. 1 1 shorebirds 4 1 5

Podicipedidae pied-billed grebe 11

total individuals 2234 2252 3623 3839 11947 mean/survey 223 150 213 272 215 toal number species 25

Table VII-11 Seasonal totals and mean numbers for bird species and groups at Buttonwood Pond over a 2 year period (2004-2006)

A-24

MASSACHUSETTS ESTUARIES PROJECT

150

130 canada goose ring-billed gull mallard 110 rock dove herring gull 90

70

50 mean # of birds per survey per # of birds mean

30

10

winter spring summer fall -10 season Figure VII-32 Seasonal abundance for the dominant bird species associated with Buttonwood Pond over a 2-year period, 2004-2006.

Domesticated waterfowl are present in both Apponagansett Bay and Buttonwood Pond. White domestic geese and ducks live along the Apponagansett Bay shoreline in close proximity to residences. Non-native species that are periodically to regularly being fed by the public at Buttonwood Pond include mute swans, domestic ducks and geese of indeterminate lineage, starlings, and rock doves. Native species, particularly Canada geese, mallards and gulls also flock to areas where humans are feeding the birds. The southeast and particularly the west shoreline of Buttonwood Pond that closely borders an impervious road, have become devoid of grasses due to a combination of bird and human trampling as a result of feeding. Although signs are posted to discourage feeding the birds, feeding is particularly intensive along this west shore of the pond, so the system’s water quality is expected to be impacted.

VII.7.2 Comparison with 20-years of winter waterfowl surveys

Thirty-one species of waterfowl (not including domesticated or hybrid birds) have been documented during the Lloyd Center’s 20-year winter waterfowl count at 25 estuarine sites from Dartmouth, MA to Tiverton, RI. During this period, Canada geese and black ducks comprised over half (54%) of the waterfowl in the December counts and nearly two-thirds (64%) during the January/February counts. Canada Geese increased dramatically from the late 1900’s into the early 2000’s, with over 10,000 counted in January 2000; however numbers have dropped off recently (Figure VII-33). Conversely black ducks have declined steadily during the December counts, albeit less so during the January/February counts. The total waterfowl number, however, closely follows the trends for the geese, suggesting that “good” years for geese are

A-25

MASSACHUSETTS ESTUARIES PROJECT good years for waterfowl overall, and that the number of geese is the major contributor in the overall numbers of waterfowl counted.

12000 CAGO - I CAGO - II 10000 ABDU - I ABDU - II TOTAL - I 8000 TOTAL - II

6000 bird count

4000

2000

0 87/88- 89/90- 91/92- 93/94- 95/96- 97/98- 99/00- 01/02- 91/92 93/94 95/96 97/98 99/00 01/02 03/04 05/06 5-year running blocks

Figure VII-33 Twenty years of Canada goose (CAGO), American black duck (ABDU) and total waterfowl abundance in 25 SE Massachusetts and Rhode Island estuaries. (I = December count II = January count)

Twenty-two waterfowl species were found in Apponagansett Bay (71% of the regional total). The dominant species in order of abundance were: bufflehead, mallard, Canada Goose, American black duck, and red-breasted merganser. Buffleheads and mallards were the most abundant species (Figure VII-34), reflecting a system with different habitat composition that the Slocums-Little River estuary. The relative abundance of buffleheads, plus red-breasted mergansers and scaup are indicative of a more open estuarine system, which Apponagansett Bay in fact is. Buffleheads appear to be declining in abundance, except for the period during the late 1990’s and early 2000’s when most waterfowl temporarily increased in abundance. Mallards, declined sharply during the first 10 years of the survey then have rebounded since the early 2000’s.

A-26

MASSACHUSETTS ESTUARIES PROJECT

200 bufflehead mallard 180 Canada goose American black duck 160 red-breasted merganser greater scaup /scaup spp.

140

120

100

80 mean # birds/survey

60

40

20

0 87/88- 88/89- 88/90- 90/91- 91/92- 92/93- 93/94- 94/95- 95/96- 96/97- 97/98- 98/99- 99/00- 00/01- 01/02- 02/03- 91/92 92/93 93/94 94/95 95/96 96/97 97/98 98/99 99/00 00/01 01/02 02/03 03/04 04/05 05/06 06/07 running 5-year blocks

Figure VII-34 Twenty-year record of dominant waterfowl species in Apponagansett Bay.

Canada goose numbers reflect the regional trends but less dramatically than in the Slocums- Little River estuary. However, the region-wide decline in black ducks is not reflected in the data from Apponagansett Bay (Figure VII-34). The next tier (in abundance) of waterfowl, red- breasted merganser and scaup species (primarily greater scaup), have experienced a slight but steady decline in number over the past two decades.

The dominant waterfowl (bufflehead, mallard, Canada goose, American, black duck, scaup, and red-breasted merganser) documented during the Turn the Tide survey were the most abundant waterfowl documented over the past two decades. The relative abundance of these species was also the same in both surveys, despite significant changes in the abundance of the most dominant species (bufflehead and mallards) over the past two decades.

VII.7.3 Federal and State-listed Species

No species listed in the Federal Endangered Species Act was encountered in either Apponagansett Bay or Buttonwood Pond. Four species listed in the Massachusetts Endangered Species Act were encountered. Common (85 observations) and least (5 observations) terns forage in Apponagansett Bay, however there currently is no appropriate nesting habitat within this system for either species. Common loons (56 observations) are regular winter visitors, particularly south of Padanarum bridge. Northern harrier (1 observation) is rarely encountered within this system. One pied-billed grebe was recorded at Buttonwood Pond, likely a fall migrant.

A-27

MASSACHUSETTS ESTUARIES PROJECT

VII.7.4 Bird Densities within the Apponagansett Bay System

Apponagansett Bay: Apponagansett Bay had an overall density of1.6 birds/ha (Table VII-12). However, 4 birds/ha for a 20.9 ha area of AB2 (Figure VII-29), the upper end of the Bay is significant because mixed congregations of gulls and dabbling waterfowl at the shallow north tip of this zone comprise a large portion of these birds. The flocks are mostly ring-billed gull and mallards. Rock doves were present year-round, nesting under Padanarum Bridge, frequently roosting on rooftops near the shoreline, and taking food from the public at the landing in zone AB3. The small resident flock of mallards and white domestic ducks and geese were often seen feeding with the doves. Also of concern in AB3 are gulls, mostly ring-billed, constantly present on the impervious surface of Apponagansett Park along with occasional flocks of Canada Geese and white domestic geese on the lawn of the park. The white domestic waterfowl were also seen periodically feeding on lawns along the east shore of the bay.

Table VII.8.3 Density and relative distribution of waterbird groups within the Apponagansett Bay system over a 4-year period (2002-2006)*. Groups zone AB1 AB2 AB3 AB4 AB5 AB6 total area(ha) 11.5 20.9 63.3 23.5 11.6 70.3 201.1 bird group ducks 0.89 1.79 0.77 0.41 0.21 0.50 0.71 gulls 0.37 1.24 0.48 0.41 0.17 0.43 0.51 geese 0.24 0.94 0.10 0.04 0.07 0.07 0.18 passerines 0.03 0.01 0.36 0.01 0.01 0.02 0.13 cormorants 0.05 0.01 0.01 0.01 0.00 0.11 0.05 waders 0.05 0.02 0.01 0.01 0.05 0.01 0.02 swans 0.02 0.03 0.01 0.02 0.01 0.00 0.01 shorebirds 0.01 0.01 0.01 0.00 terns 0.01 0.01 0.00 loons, grebes 0.01 0.00 raptors 0.01 0.01 0.00 birds/hectare/survey 1.7 4.1 1.7 0.9 0.5 1.2 1.6

Species AB1 AB2 AB3 AB4 AB5 AB6 total 11.5 20.9 63.3 23.5 11.6 70.3 201.1 dominant species bufflehead 0.31 0.10 0.49 0.29 0.11 0.18 0.29 ring-billed gull 0.90 0.30 0.14 0.02 0.22 0.28 herring gull 0.34 0.33 0.16 0.24 0.14 0.19 0.21 mallard 0.01 0.98 0.11 0.01 0.01 0.12 0.18 Canada goose 0.23 0.88 0.04 0.04 0.07 0.07 0.15 * density values expressed in mean # of birds /hectare/survey (108 total surveys) averaged by season

Table VII-12 Density and relative distribution of waterbird groups within the Apponagansett Bay system over a 4 year period (2002-2006)

Buttonwood Pond: At Buttonwood Pond, overall density was 92 birds/ha, a density that likely creates a nutrient loading and bacteria problem (Table VII-13). Gulls and waterfowl all occurred at relatively high densities, the most concentrated species being Canada goose (27 birds/ha), ring-billed gull (25 birds/ha) and mallard (21 birds/ha). Rock doves existed at a modest density

A-28

MASSACHUSETTS ESTUARIES PROJECT of 7 birds/ha but were constantly present. Although not a concern relative to density, mute swans and muscovy ducks could be problematic in this small pond due to their unique capacity to destroy vegetation.

Table VII.8.4 Density of waterbird groups and dominant species in ButtonwoodPond over a 2 year period (2004-2006).

bird group density* dominant species density* gulls 30.72 Canada goose 27.30 geese 29.23 ring-billed gull 25.84 ducks 24.21 mallard 21.04 passerines 6.90 rock dove 6.73 swans 0.56 herring gull 2.77 waders 0.12 cormorants 0.04 shorebirds 0.03 rails 0.03 grebes 0.01 birds/hectare/survey 92 * density values expressed in mean # of birds/hectare/survey for a 2.4 ha waterbody (55 surveys) averaged by season

Table VII-13 Density of water bird groups and dominant species in Buttonwood Pond over a 2 year period (2004-2006)

VII.7.5 Summary and Conclusions

Canada geese, mallards, and ring-billed gulls, all documented contributors of nutrients and bacteria to surface waters, were among the most abundant birds in Apponagansett Bay and Buttonwood Pond. Where tidal flushing dilutes nutrients and bacteria, we can’t conclude that high enough concentrations of water birds (1.6 birds/ha) exist over time to be a primary contributor to nutrient or bacteria loads in Apponagansett Bay. However, local concentrations primarily associated within areas where feeding the birds is an ongoing pastime (Apponagansett Park and town landings at each end of Padanarum bridge) may create locally high bacterial and nutrient levels.

The most critical area of concern was Buttonwood Pond, where an average of 92 birds/ha were documented; the majority of which (mallards, Canada geese, ring-billed and herring gulls, and rock doves) are actively fed by visitors to the pond on a regular and year-round basis despite signs discouraging this practice. Documentation of the amount of nutrients and/or bacteria ultimately reaching Apponagansett Bay via Buttonwood Brook was beyond the scope of this project, but Buttonwood Pond cannot be discounted as a source of contamination from bird feces.

Common and least terns forage during the summer in Apponagansett Bay, and common loons winter in the Bay. However, neither the Bay nor Buttonwood Pond is currently providing significant habitat for any other state-listed avian species.

A-29