Tidal Salt Marsh Mitigation in the New Jersey ...

Tidal Salt Marsh Mitigation in the New Jersey Meadowlands: Using Long-term Monitoring Data from Three Mitigated, Urban Salt Marshes to Develop Guidelines for Future Mitigation Efforts

Prepared for The New Jersey Meadowlands Commission and United States Environmental Protection Agency

Prepared by JeanMarie Hartman, PhD Niki H. Learn, MS Mark C. Wong, MS and Jennifer L. Momsen

93 Lipman Drive Blake Hall, Cook Campus Rutgers, the State University of New Jersey New Brunswick, NJ 08901-8524

August 2006

Executive Summary WILL BE UPDATED AFTER GUIDELINES ARE ADDED We sought to determine the utility of several assessment methods for use in the Hackensack Meadowlands. To do so, we surveyed the literature for both rapid assessment methods and other methods, including HGM. Using criteria set forth by the U.S. EPA, we chose a total of five assessment methods, including three rapid assessments, as appropriate for use in the Meadowlands. These five were incorporated into the detailed review that follows. Using our extensive data set from five years of monitoring, we applied each of the five assessment methods to Mill Creek, a mitigated salt marsh in a highly urbanized setting. Our summary of trends in vegetation, avian and fish communities was then used for validation purposes. Our results indicate that rapid assessment and other assessment methods may not be suitable for use in the Meadowlands. Specifically, we found that these methods: (1) did not fully capture the condition or functions of Mill Creek, of three mitigated, urban wetlands; (2) were unable to capture variation at the marsh over time; and (3) did not capture the landscape or urban setting of the salt marsh. We therefore recommend limited use of these assessment methods, as they are currently designed. An analysis of the monitoring data shows that not all of it is necessary for accurate assessment, however, a scaled-back assessment protocol including detailed monitoring should continue to be used to accurately assess the condition and function of wetlands in the Hackensack Meadowlands. There is a need for the development of a specific assessment protocol for wetlands in the Hackensack Meadowlands. Such an assessment must clearly take into account the complexities brought on by the high level of urbanization in the area.

DRAFT

2

TABLE OF CONTENTS 1.0

INTRODUCTION............................................................................................... 11 1.1 Background and Scope of Report ............................................................. 12 1.2 The Meadowlands District, Typical Salt Marshes, and Site Descriptions 14 1.3 Literature Review...................................................................................... 22 1.3.1 Function and Condition................................................................. 22 1.3.2 Issues to Be Considered During Planning, Site Selection, and Design ........................................................................................... 24 1.3.2.1 Planning: Goals and Objectives ..................................... 24 1.3.2.2 Site Selection ................................................................. 24 1.3.2.3 Design ............................................................................ 25 1.3.2.4 Design Issue: Hydrology................................................ 25 1.3.2.5 Design Issue: Landscape Factors ................................... 26 1.3.2.6 Design Issue: Urban Environment ................................. 26 1.3.2.7 Design Issue: Geomorphology....................................... 27 1.3.2.8 Design Issue: Vegetation ............................................... 27 1.3.3 Issues to Be Considered During Construction .............................. 27 1.3.4 Issues and Condition Indicators to Be Considered During Monitoring .................................................................................... 28 1.3.4.1 Adaptive Management ................................................... 28 1.3.4.2 Timeframe...................................................................... 29 1.3.4.3 Condition Indicators....................................................... 29 1.3.4.4 Condition Indicator: Hydrology..................................... 31 1.3.4.5 Condition Indicator: Vegetation .................................... 31 1.3.4.6 Condition Indicator: Avian Habitat Use ........................ 32 1.3.4.7 Condition Indicator: Fish Habitat Use ........................... 33 1.3.4.8 Condition Indicator: Benthic Invertebrate Habitat Use . 34 1.3.4.9 Baseline Data ................................................................. 34

2.0

MILL CREEK – A CASE STUDY IN DESIGN AND IMPLEMENTATION 34 2.1 Rationale for Selection, Design, and Construction................................... 35 2.1.1 Design Objectives ......................................................................... 36 2.1.1.1 Water Quality Improvement .......................................... 36 2.1.1.2 Fisheries and Wildlife Habitat Improvement................. 37 2.1.1.3 Recreation/Education Opportunities Improvement ....... 38 2.1.2 Baseline Studies ............................................................................ 38 2.1.2.1 Topographic Survey....................................................... 39 2.1.2.2 Vegetation Mapping....................................................... 39 2.1.2.3 Soil, Sediment, and Surface Water Sampling................ 40 2.1.2.4 Avian Studies ................................................................. 40 2.1.2.5 Hydrological Analysis ................................................... 40 2.1.3 Role of IVA and Other Functional Assessment Methods............. 41 2.1.4 Review of Permits for Compensatory Mitigation......................... 42 2.1.5 Design Objectives and As-built Conditions.................................. 44 2.2 Monitoring Objectives, Protocols, and Implementation ........................... 46

DRAFT

3

2.2.1

2.3 3.0

Requirements of the Memorandum of Understanding.................. 46 2.2.1.1 Flora ............................................................................... 47 2.2.1.2 Avian.............................................................................. 48 2.2.1.3 Geomorphology/Hydrology/Salinity ............................. 48 2.2.1.4 Sediment and Soils......................................................... 48 2.2.1.5 Contaminants ................................................................. 48 2.2.1.6 Aquatic Animals ............................................................ 49 2.2.1.7 Insect Communities ....................................................... 49 2.2.2 Development and Implementation of Monitoring Protocols ........ 49 2.2.2.1 Flora ............................................................................... 49 2.2.2.2 Avian.............................................................................. 50 2.2.2.3 Geomorphology/Hydrology/Salinity ............................. 51 2.2.2.4 Sediment and Soils......................................................... 51 2.2.2.5 Contaminants ................................................................. 51 2.2.2.6 Aquatic Animals ............................................................ 52 2.2.2.7 Insect Communities ....................................................... 53 Identification of Potential Causes for Mitigation Failure ......................... 53

LONG-TERM MONITORING DATA ANALYSIS ....................................... 54 3.1 Vegetation Data Analysis ......................................................................... 55 3.1.1 Mill Creek ..................................................................................... 55 3.1.1.1 Methods.......................................................................... 55 3.1.1.2 Results and Discussion .................................................. 56 3.1.2 Harrier ........................................................................................... 68 3.1.2.1 Methods.......................................................................... 68 3.1.2.2 Results and Discussion .................................................. 68 3.1.3 Skeetkill ........................................................................................ 79 3.1.3.1 Methods.......................................................................... 79 3.1.3.2 Results and Discussion .................................................. 82 3.1.4 Conclusions................................................................................... 91 3.2 Sampling Size Assessment Using Rarefaction Curves............................. 92 3.2.1 Methods......................................................................................... 93 3.2.2 Results and Discussion ................................................................. 93 3.2.3 Conclusions................................................................................... 95 3.3 Spatial and Temporal Trend Analysis of Vegetation Using Coefficients of Areal Association (CAA).......................................................................... 95 3.3.1 Methods......................................................................................... 96 3.3.2 Results and Discussion ................................................................. 97 3.3.3 Conclusions................................................................................... 99 3.4 Avian Data Analysis ............................................................................... 106 3.4.1 Mill Creek ................................................................................... 106 3.4.1.1 Methods........................................................................ 106 3.4.1.2 Results and Discussion ................................................ 107 3.4.2 Harrier ......................................................................................... 122 3.4.2.1 Methods........................................................................ 122 3.4.2.2 Results and Discussion ................................................ 125 3.4.3 Skeetkill ...................................................................................... 141

DRAFT

4

3.5

3.6

3.7

3.4.3.1 Methods........................................................................ 141 3.4.3.2 Results and Discussion ................................................ 143 3.4.4 Conclusions................................................................................. 152 Fish Data Analysis .................................................................................. 161 3.5.1 Mill Creek ................................................................................... 161 3.5.1.1 Methods........................................................................ 161 3.5.1.2 Results and Discussion ................................................ 162 3.5.2 Harrier ......................................................................................... 170 3.5.2.1 Methods........................................................................ 170 3.5.2.2 Results and Discussion ................................................ 172 3.5.3 Skeetkill ...................................................................................... 178 3.5.3.1 Methods........................................................................ 178 3.5.3.2 Results and Discussion ................................................ 178 3.5.4 Conclusions................................................................................. 181 Benthic Invertebrate Data Analysis ........................................................ 183 3.6.1 Mill Creek ................................................................................... 183 3.6.1.1 Methods........................................................................ 183 3.6.1.2 Results and Discussion ................................................ 184 3.6.2 Harrier ......................................................................................... 192 3.6.2.1 Methods........................................................................ 192 3.6.2.2 Results and Discussion ................................................ 194 3.6.3 Skeetkill ...................................................................................... 201 3.6.3.1 Methods........................................................................ 201 3.6.3.2 Results and Discussion ................................................ 202 3.6.4 Conclusions................................................................................. 210 Evaluation of Marsh Condition............................................................... 216 3.7.1 Mill Creek ................................................................................... 216 3.7.2 Harrier ......................................................................................... 218 3.7.3 Skeetkill ...................................................................................... 220

4.0 TESTING ASSESSMENT METHODS FOR DEVELOPMENT OF SYSTEM FUNCTION GUIDELINES........................................................................ 221 4.1 Functional and Conditional Assessments ............................................... 222 4.2 Assessment Method Selection, Application, and Evaluation ................. 223 4.2.1 Delaware Rapid Assessment Procedure V.2.0............................ 225 4.2.1.1 Background .................................................................. 225 4.2.1.2 Results and Discussion ................................................ 229 4.2.2 Florida Wetland Rapid Assessment Procedure (WRAP)............ 230 4.2.2.1 Background .................................................................. 230 4.2.2.2 Results and Discussion ................................................ 230 4.2.3 Massachusetts Assessment Approach......................................... 231 4.2.3.1 Background .................................................................. 231 4.2.3.2 Results and Discussion ................................................ 232 4.2.4 Method for the Evaluation and Inventory of Vegetated Tidal Marshes in New Hampshire (The Coastal Method) ................... 235 4.2.4.1 Background .................................................................. 235 4.2.4.2 Results and Discussion ................................................ 235 DRAFT

5

4.2.5

4.2.6

Hydrogeomorphic Assessment of Tidal Fringe Wetlands (HGM) ..................................................................................................... 236 4.2.5.1 Background .................................................................. 236 4.2.5.2 Results and Discussion ................................................ 237 Conclusions................................................................................. 241

5.0

SYNTHESIS ...................................................................................................... 243 5.1 Planning, Site Selection, Design, and Construction ............................... 244 5.2 Monitoring .............................................................................................. 249 5.3 Hydrology ............................................................................................... 254 5.4 Vegetation ............................................................................................... 256 5.5 Fauna....................................................................................................... 259

6.0

GUIDELINES ................................................................................................... 263 6.1 Planning .....................................................Error! Bookmark not defined. 6.2 Site Selection .............................................Error! Bookmark not defined. 6.3 Design ........................................................Error! Bookmark not defined. 6.4 Construction...............................................Error! Bookmark not defined. 6.5 Reference Sites...........................................Error! Bookmark not defined. 6.6 Monitoring .................................................Error! Bookmark not defined. 6.7 Assessment.................................................Error! Bookmark not defined.

7.0

LITERATURE CITED .................................................................................... 264

DRAFT

6

LIST OF TABLES Table 1.3.4-1. Ranking of indicators from literature review. ...................................................... 30 Table 2.1.4-1. Mill Creek Permit List of Compensatory Mitigation. .......................................... 44 Table 2.1.5-1. Mill Creek Mitigation As-built Acreages............................................................. 45 Table 3.1.1-1. Invasive and weedy species found at Mill Creek. ................................................ 58 Table 3.1.1-2. Species with mean cover greater than 10% found at Mill Creek, Summer 2004. 59 Table 3.1.1-3. Species richness for Mill Creek over the five-year monitoring period. ............... 62 Table 3.1.1-4. Species richness of each vegetative form over the five-year monitoring period at Mill Creek. .............................................................................................................. 65 Table 3.1.1-5. Mill Creek vegetation cover in acres.................................................................... 67 Table 3.1.2-1. Invasive and weedy species found at Harrier. ...................................................... 70 Table 3.1.2-2. Species with mean cover great than 10% found at Harrier, Summer 2004.......... 72 Table 3.1.2-3. Species richness for Harrier over the five-year monitoring period. ..................... 75 Table 3.1.2-4. Species richness of each vegetative form over the five-year monitoring period at Harrier. .................................................................................................................... 78 Table 3.1.2-5. Harrier vegetation cover in acres.......................................................................... 79 Table 3.1.3-1. Invasive and weedy species found at Skeetkill. ................................................... 82 Table 3.1.3-2. Species with mean cover greater than 10% found at Skeetkill, Summer 2001.... 83 Table 3.1.3-3. Species richness for Skeetkill over the three-year monitoring period.................. 87 Table 3.1.3-4. Species richness of each vegetative form over the three-year monitoring period at Skeetkill................................................................................................................... 90 Table 3.1.3-5. Skeetkill vegetation cover in acres. ...................................................................... 91 Table 3.3-1. Number of Species Losing Mean Percent Cover Between 2001 and 2002............. 97 Table 3.3-2. Spring Low Marsh mean % cover and CAA values (+/- standard error) for typical forbs and graminoids at Mill Creek....................................................................... 100 Table 3.3-3. Spring High Marsh mean % cover and CAA values (+/- standard error) for typical forbs and graminoids at Mill Creek....................................................................... 101 Table 3.3-4. Spring Upland mean % cover and CAA values (+/- standard error) for typical forbs and graminoids at Mill Creek................................................................................ 102 Table 3.3-5. Summer Low Marsh mean % cover and CAA values (+/- standard error) for typical forbs and graminoids at Mill Creek....................................................................... 103 Table 3.3-6. Summer High Marsh mean % cover and CAA values (+/- standard error) for typical forbs and graminoids at Mill Creek....................................................................... 104 Table 3.3-7. Summer Upland mean % cover and CAA values (+/- standard error) for typical forbs and graminoids at Mill Creek....................................................................... 105 Table 3.4.1-1. Number of avian surveys performed at Mill Creek by season and year. ........... 106 Table 3.4.1-2. Avian species observed at Mill Creek in 1997 and 2000-2004.......................... 110 Table 3.4.1-3. Mean species richness + standard error per survey at Mill Creek by season and year. ....................................................................................................................... 114 Table 3.4.1-4. Mean species richness + standard error per survey at Mill Creek by foraging guild....................................................................................................................... 115 Table 3.4.2-1. Number of avian surveys performed at Harrier by season and year. ................. 122 Table 3.4.2-2. Avian species observed at Harrier from 1996-1997 and winter 1999-2004. ..... 127 Table 3.4.2-3. Mean species richness + standard error per survey at Harrier by season and year. ............................................................................................................................... 132

DRAFT

7

Table 3.4.2-4. Mean species richness + standard error per survey at Harrier by foraging guild. Only the spring season was sampled in 2004........................................................ 133 Table 3.4.3-1. Number of avian surveys performed at Skeetkill by season and year................ 141 Table 3.4.3-2. Avian species observed at Skeetkill from 1999-2001. ....................................... 145 Table 3.4.3-3. Mean species richness + standard error per survey at Skeetkill by season and year. ............................................................................................................................... 148 Table 3.4.3-4. Mean species richness + standard error per survey at Skeetkill by foraging guild. ............................................................................................................................... 148 Table 3.4.4-1. Number of species in each foraging guild at the three marshes. ........................ 154 Table 3.4.4-2. State endangered and threatened species in New Jersey that were present at one or more of the three marshes. .................................................................................... 156 Table 3.5.1-1. Species of fish collected at Mill Creek between 2000 and 2004........................ 164 Table 3.5.1-2. Percent of total catch represented by each species caught at Mill Creek between June and September/early October for the years 2000 to 2004............................. 166 Table 3.5.1-3. Mean catch per unit effort for each species caught at Mill Creek between June and September/early October for the years 2000 to 2004. .................................... 167 Table 3.5.2-1. Species of fish collected at Harrier between 1999 and 2004.............................. 173 Table 3.5.2-2. Percent of total catch represented by each species caught at Harrier between June and September for the years 1999 to 2004............................................................ 174 Table 3.5.2-3. Mean catch per unit effort for each species caught at Harrier between June and September for the years 1999 to 2004................................................................... 175 Table 3.5.3-1. Fish and snapping turtles captured at Skeetkill. ................................................. 180 Table 3.6.1-1. Benthic invertebrate taxa found in sediment samples at Mill Creek.................. 186 Table 3.6.1-2. Number of taxa, mean total density, and three most abundant major taxonomic groups by month at Mill Creek. ............................................................................ 187 Table 3.6.1-3. Relative abundance of major taxonomic groups at Mill Creek.......................... 187 Table 3.6.2-1. Benthic invertebrate taxa found in sediment samples at Harrier........................ 195 Table 3.6.2-2. Number of taxa, mean total density, and three most abundant major taxonomic groups by month at Harrier. .................................................................................. 197 Table 3.6.2-3. Relative abundance of major taxonomic groups at Harrier................................ 198 Table 3.6.3-1. Benthic invertebrate taxa found in sediment samples at Skeetkil. ..................... 204 Table 3.6.3-2. Number of taxa, mean total density, and three most abundant major taxonomic groups by month at Skeetkill................................................................................. 205 Table 3.6.3-3. Relative abundance of major taxonomic groups at Harrier................................ 206 Table 4.2-1. Resulting scores from application of assessment methods.................................... 226 Table 4.2-2. Comparable scores for WEA sites from HA and NPSI methods. ......................... 233 Table 4.2-3. HGM comparison of Mill Creek scores with Meadowlands reference sites......... 240

LIST OF FIGURES Figure 1.2-1. Figure 3.1.1-1. Figure 3.1.1-2. Figure 3.1.1-3. Figure 3.1.1-4.

DRAFT

The Meadowlands District showing six mitigation sites .................................... 21 Mill Creek permanent vegetation sampling plots ............................................... 57 Mean relative cover of each vegetative form at Mill Creek................................ 59 Mean relative frequency of each vegetative form at Mill Creek. ....................... 59 Mean cover of each vegetative form at Mill Creek............................................. 61

8

Figure 3.1.1-5. Mean frequency of each vegetative form at Mill Creek. .................................... 61 Figure 3.1.1-6. Average species richness per plot at Mill Creek over the five-year monitoring period....................................................................................................................... 63 Figure 3.1.1-7. Overall diversity of each zone within the marsh at Mill Creek. ......................... 64 Figure 3.1.1-8. Average diversity per plot of each zone at Mill Creek. ...................................... 64 Figure 3.1.2-1. Harrier permanent vegetation sampling plots ..................................................... 69 Figure 3.1.2-2. Mean relative cover of each vegetative form at Harrier. .................................... 72 Figure 3.1.2-3. Mean relative frequency of each vegetative form at Harrier. ............................. 73 Figure 3.1.2-4. Mean cover of each vegetative form at Harrier. ................................................. 74 Figure 3.1.2-5. Mean frequency of each vegetative form at Harrier. .......................................... 74 Figure 3.1.2-6. Average species richness per plot at Harrier over the six-year monitoring period. ................................................................................................................................. 76 Figure 3.1.2-7. Overall diversity of each zone within the marsh at Harrier. ............................... 77 Figure 3.1.2-8. Average diversity per plot of each zone at Harrier. ............................................ 77 Figure 3.1.3-1. Skeetkill permanent vegetation sampling plots................................................... 81 Figure 3.1.3-2. Mean relative cover of each vegetative form at Skeetkill................................... 84 Figure 3.1.3-3. Mean relative frequency of each vegetative form at Skeetkill............................ 85 Figure 3.1.3-4. Mean cover of each vegetative form at Skeetkill................................................ 86 Figure 3.1.3-5. Mean frequency of each vegetative form at Skeetkill......................................... 86 Figure 3.1.3-6. Average species richness per plot at Skeetkill over the three-year monitoring period....................................................................................................................... 88 Figure 3.1.3-7. Overall diversity of each zone within Skeetkill. ................................................. 89 Figure 3.1.3-8. Average diversity per plot of each zone at Skeetkill........................................... 89 Figure 3.2-1. Species-area or rarefaction curves for permanent monitoring plots at Mill Creek. ................................................................................................................................. 94 Figure 3.2-2. Aggregated species-area or rarefaction curves for permanent monitoring plots at Mill Creek. .............................................................................................................. 95 Figure 3.4.1-1. Mill Creek pre-mitigation avian sampling locations......................................... 108 Figure 3.4.1-2. Mill Creek post-mitigation avian sampling locations. ...................................... 109 Figure 3.4.1-3. Mean abundance per survey at Mill Creek by guild and season....................... 116 Figure 3.4.1-4. Mean abundance per survey at Mill Creek by season and year. ....................... 117 Figure 3.4.2-1. Harrier pre-mitigation avian sampling locations............................................... 123 Figure 3.4.2-2. Harrier post-mitigation avian sampling locations. ............................................ 124 Figure 3.4.2-3. Mean abundance per survey at Harrier by guild and season............................. 134 Figure 3.4.2-4. Mean abundance per survey at Harrier by season and year. ............................. 134 Figure 3.4.3-1. Skeetkill avian sampling locations.................................................................... 142 Figure 3.4.3-2. Mean abundance per survey at Skeetkill by guild and season. ......................... 149 Figure 3.4.3-3. Mean abundance per survey at Skeetkill by season and year. .......................... 151 Figure 3.5.1-1. Mill Creek fish sampling locations. .................................................................. 163 Figure 3.5.1-2. Species richness of fish at Mill Creek by month and for the full year.............. 165 Figure 3.5.1-3. Estimated Shannon-Wiener diversity at Mill Creek by month and for the full year. ....................................................................................................................... 168 Figure 3.5.1-4. Mean salinity by month at Mill Creek as measured at eight sampling locations on the dates of seining................................................................................................ 169 Figure 3.5.2-1. Harrier fish sampling locations. ........................................................................ 171 Figure 3.5.2-2. Species richness of fish at Harrier by month and for the full year.................... 174

DRAFT

9

Figure 3.5.2-3. Estimated Shannon-Wiener diversity at Harrier by month and for the full year. ............................................................................................................................... 176 Figure 3.5.2-4. Mean salinity by month at Harrier as measured at three sampling locations on the dates of seining...................................................................................................... 177 Figure 3.5.3-1. Skeetkill fish sampling locations. ..................................................................... 179 Figure 3.6.1-1. Mill Creek benthic invertebrates sampling locations........................................ 185 Figure 3.6.1-2. Number of benthic invertebrate taxa collected each month at Mill Creek by habitat. ................................................................................................................... 190 Figure 3.6.1-3. Mean density of benthic invertebrates collected each month at Mill Creek by habitat. ................................................................................................................... 191 Figure 3.6.1-4. Salinity at Mill Creek benthic invertebrate sampling stations in 2002. ............ 192 Figure 3.6.2-1. Harrier benthic invertebrates sampling locations.............................................. 193 Figure 3.6.2-2. Number of benthic invertebrate taxa collected each month at Harrier by habitat. ............................................................................................................................... 200 Figure 3.6.2-3. Mean density of benthic invertebrates collected each month at Harrier by habitat. ............................................................................................................................... 201 Figure 3.6.3-1. Skeetkill benthic invertebrates sampling locations. .......................................... 203 Figure 3.6.3-2. Number of benthic invertebrate taxa collected each month at Skeetkill by habitat. ............................................................................................................................... 209 Figure 3.6.3-3. Mean density of benthic invertebrates collected each month at Skeetkill by habitat. ................................................................................................................... 210

LIST OF APPENDICES Appendix A. Annotated Bibliography of Literature Cited Appendix B. Memorandum of Understanding between Rutgers University and the Hackensack Meadowlands Development Commission Appendix C. Soil Data Collected in 2005 Appendix D. Vascular Vegetation Identified in the Meadowlands Appendix E. Detailed Avian Data Appendix F. Detailed Fish Data Appendix G. Detailed Benthic Data Appendix H. Detailed Rapid Assessment Results

DRAFT

10

1.0 INTRODUCTION Since European settlement, roughly half of all US wetlands have been lost, through agriculture, urbanization and other land-use processes (Dahl and Allord 1996). In the latter half of the 20th century, several shifts in wetland policy, including Section 404 of the 1972 Clean Water Act and the ‘no net loss’ policy resulted in the need for an accurate inventory and assessment of the nation’s wetlands, for assessments measuring the impacts to individual sites caused by filling wetlands, and for a means of measuring wetland functions such that lost functions and values could be replaced through mitigation. In the years following passage of the Clean Water Act, several methodologies have been developed in response to these needs. The hydrogeomorphic (HGM) method was developed by Brinson and the U.S. Army Corps of Engineers and focuses on wetland function, assessing this through physical features of the wetland (Brinson 1993). The linchpin of the HGM method was hydrology (Galatowitsch et al. 1998a). Over the years, however, problems with HGM have emerged, especially with use for restored sites. In these cases, not all functions are clearly related to hydrogeomorphic properties and issues of altered soils and invasive plants can affect wetland function and limit restoration success (Galatowitsch et al. 1998a). Additionally, Brinson (1993) acknowledges the limitation that maximizing one function will cause a concomitant decrease in one or more others, such that not all functions can be maximized at once. Rapid assessment techniques were developed with the intention of offering a quick method to consider wetland condition and function in policy-making decisions. Rapid assessment methods (RAMs) rely on indicators to determine wetland function or condition. For example, soil organic matter can be an indicator of denitrification, success of plant establishment, or invertebrate production. However, there is a paucity of research linking easily obtained metrics to ecosystem function (Galatowitsch et al. 1998a) and it is difficult to determine rapid, accurate broad-scale measures that can be used consistently as proxies for more detailed data. Effective wetland monitoring methods are traditionally described by a “three tier framework” (Fennessy et al. 2004). Broad, large-scale, landscape-level assessments reside at one end of the spectrum (level 1 assessments), while intensive, site-specific biological, chemical and physical measures (level 3 assessments) occupy the opposite end. RAMs represent an intermediate assessment method (level 2), which, while quick and cost-effective, still provides regulators and decision makers with a scientific wetland evaluation. Detailed wetland monitoring, in contrast to RAMs, directly monitors biological conditions. While this method is seen as the most complete and effective way to assess wetlands (U.S. EPA 2002a), it is labor and time intensive and often generates staggering amounts of data that are difficult to synthesize. In any study, it is important to match the scope of a project with appropriate assessment methods (U.S. EPA 2002a). There are pros and cons to both long-term and rapid assessment methods and there should be a middle ground when developing assessment methods and guidelines. Obviously, rapid assessment methods are unsuitable for long-term site monitoring and yet, indepth monitoring is too bulky for making policy decisions since it produces extensive data sets that take too long to analyze and summarize. In addition, rapid assessments suffer from a lack of

DRAFT

11

research into indicators, while long-term, intensive data sampling usually is not possible at mitigation sites (Galatowitsch et al. 1998a). This indicates the need for easily measured, reliable indicators of wetland function. Clearly, there is a need for assessment methods to utilize aspects of both types of methods and be based on sound science. Careful testing of assessment methods and direct comparisons of those methods must be completed in a variety of settings to better understand the variability and robustness of these methods. There are many types of wetlands in a variety of different environments (e.g., forested, coastal, freshwater, saltwater, urban). Applicability of the method must be tested against these varying types to determine if the method is useful in a broad or specific sense. 1.1 Background and Scope of Report On January 30, 2003, the New Jersey Meadowlands Commission (NJMC) submitted a proposal to the U.S. EPA, to conduct an extensive evaluation of existing data collected under a Memorandum of Understanding (MOU). This was approved by the U.S. EPA on September 8, 2003. The approved scope of work called for the development of a systems function assessment guideline for wetlands mitigation design and monitoring. These guidelines were to include information on selection of potential mitigation sites, design and construction of the mitigation, and development of a monitoring protocol to assess if the post-mitigation marsh is selfsustaining. To help develop these guidelines, an extensive literature review was conducted, detailed analysis of long-term monitoring data was performed, and an evaluation of RAMs and other assessment methods was made. The literature review is necessary to learn from past restoration projects so as not to repeat mistakes of the past. There are many well-researched and well-documented projects from which to cull information. Learning from past experience will contribute new knowledge to the field of restoration. The assessment methods used here are presently in use or in development and pertain to the current state-of-the-art when it comes to wetland assessment. This is relevant to any discussion on the development of guidelines. It deals, for example, with issues of cost (in money and time) and level of detailed information that comes out of the assessment. In developing guidelines, a discussion of RAMs and other assessment methods, relative to long-term monitoring is relevant because any viable, scientifically accurate, and relatively quick assessment method will require aspects of both rapid and long-term assessments. The following lists the report deliverables as outlined in the original January 30, 2003 proposal: 1. Review current literature and identify the pertinent issues regarding the effectiveness of current wetland monitoring and assessment methods. (see Section 1.3 and Appendix A) 2. Assess the evidence of these issues in the evaluation of mitigation projects in the Meadowlands. (see Sections 4.0 and 5.0) 3. Identify the system functions that are essential for the development of evaluation criteria for site selection, design, as-built project review and project performance. (see section 1.3) 4. Utilize the Mill Creek Enhancement Site as the primary study site focusing on biological assessments used in evaluating site selection, design and ecological performance of the restoration. (see Section 2.0) a. Review design rationale for design objectives b. Review baseline studies and assess how information influenced design DRAFT

12

5. 6. 7.

8.

9.

c. Review role of IVA and other function assessment in determining design objectives d. Review permits for compensatory requirements e. Review as-built conditions for achievement of design objectives f. Review monitoring protocols Review monitoring program objectives and implementation. (see Section 2.2) Identify potential causes of mitigation failures and assess causes. (see Sections 2.3 and 5.0, respectively) Identify and rank system functions for assessment of restoration success and determine if these are the same functions and rankings for assessing site selection. (see Sections 1.3.2 through 1.3.4 and 5.1 through 5.5) Apply evaluation process to other wetlands mitigation sites. (see Sections 3.0 through 5.0) a. Skeetkill Creek Marsh b. Harrier Meadow c. Western Brackish Marsh d. Eastern Brackish Marsh e. Lyndhurst Marsh Preserve Based on study, recommend guidelines for comprehensive approach to wetlands restoration. (see Section 6.0)

La Peyre et al. (2001) state that "Despite the emphasis of many environmental programs on the need for evaluation, there is a paucity of studies identifying potential indicators for evaluation of specific resources or for specific management programs". What is detailed in the first part of this report are the findings of a literature review conducted to answer the following broad question: How does one define “success” of a wetlands mitigation? To answer this, other questions need to be asked: What problems have been identified with past and current mitigation projects?; Are there wetland system functions that can be used to evaluate success?; What parameters have been and should be used to gauge success and compliance with regulations? While answers to some of these questions have been identified in the literature, there is still vague generalization made regarding specific assessment indicators. The findings of this literature review will be used to evaluate the mitigation and monitoring of three marshes and will be used in the development of assessment guidelines presented in Section 6.0. It is hoped that recommendations made in this document will be used to standardize methods and metrics used to assess mitigation sites. As part of this review, 102 references were reviewed with publication dates ranging from 1987 to 2005. Appendix A contains an annotated bibliography of the literature cited in this report. Sources came from journals, reports, internet websites, and books. These documents were examined to identify the pertinent issues regarding the appropriateness of current mitigation site selection, design, implementation, and monitoring practices in the Meadowlands, as well as the effectiveness of current wetland monitoring and assessment methods. Included in this review were state and national documents regarding wetland assessment including rapid assessment methods (RAMs) and other methods such as hydrogeomorphic (HGM) methods. In addition to the literature review, this report details the findings of an extensive evaluation of existing data from three salt marsh mitigation sites. The most extensive evaluation was applied to Mill Creek, the largest of the three sites, which had a complete and consistent data set for all

DRAFT

13

five years of monitoring. Details regarding site selection, design, implementation, and monitoring at Mill Creek were then examined in light of recommendations found in the literature. Long-term monitoring data for Mill Creek, Harrier, and Skeetkill is presented. This is followed by an evalutation of rapid and semi-rapid assessment methods. Such methods, due to the lower cost and time investment required for their use, have been increasingly used as regulatory tools and in evaluating the condition and/or function of marshes. Due to their rapid nature, such methods tend to be quite general and their appropriateness for use in the Meadowlands or for assessment of mitigation success is questionable; thus, testing these methods against long-term monitoring data allows examination of their strengths and weaknesses. Those assessment methods that were deemed most likely to be appropriate for use in the tidal salt marshes of the Meadowlands were applied to Mill Creek, Harrier, and Skeetkill. Assessment of the various methods for sensitivity and appropriateness for use in the Meadowlands is accomplished through evaluation using the long-term data set. The results of the literature review; the review of the site selection, design, and construction process; the long-term monitoring; and the evaluation of assessment methods are then combined to produce detailed guidelines for the mitigation or enhancement of salt marshes in the New Jersey Meadowlands. These guidelines include issues relevant to every stage of the process from planning and site selection through monitoring and adaptive management. They deal with issues specific to the Meadowlands, including the unique nature of tidal salt marshes and the highly urban setting within which these wetlands occur. 1.2 The Meadowlands District, Typical Salt Marshes, and Site Descriptions The Meadowlands District is part of the Hudson-Raritan Estuary and encompasses 19,730 acres of urban landscape interspersed with waterways and wetlands that today cover over 40% of the district’s area. The district straddles the Hackensack River and includes portions of both Bergen and Hudson Counties. The environmental history of the district is a tumultuous one. Prior to European settlement, large portions of the Meadowlands were covered in freshwater Chamaecyparis thyoides (Atlantic white cedar) bogs interspersed with other fresh and salt marsh vegetation, including vast expanses of wetland meadow grasses (Bontje et al. 1991, Hartman et al. 2003, Marshall 2004, Hackensack Riverkeeper Inc. 2006). As the area became increasingly populated with the burgeoning metropolises of New York and Newark, impacts on the river and its wetlands increased. The cedars and other forest trees were increasingly harvested for use as construction materials (Hartman et al. 2003), the growing population drew more and more drinking water from the Hackensack River while depositng more and more industrial and sewage waste into or adjacent to the river (Hackensack Riverkeeper Inc. 2006), harvested salt hay extensively (through the early 1900s), and increasingly ditched and drained wetlands for construction of buildings, roads, and railroads, as well as mosquito control (Maguire Group Inc. 1989, Kiviat and MacDonald 2002, Marshall 2004). The invasive Phragmites australis (common reed), which thrives upon drier wetland areas, had become dominant in many areas by the 1920s (Maguire Group Inc. 1989); this tall, aggressive plant grows at great density, crowding out other plant species and excluding many animals.

DRAFT

14

Beginning in the mid-1800s freshwater inputs into the Meadowlands began to be diverted and in the early 1900s the Hackensack River was dredged to increase its depth; this led to increasing intrusions of salt water into the estuary (Marshall 2004). Most of the cedar swamps were already gone by 1922, when the regional demand for water resulted in the building of a reservoir for drinking water. Thus, the Oradell Dam was built, diverting much of the flow from the upper reaches of the Hackensack, and increasing the salinity in its lower reaches, quickly irradicating most of the remaining freshwater wetlands (Bontje et al. 1991, Hackensack Riverkeeper Inc. 2006). Meanwhile pollution and overharvesting had decimated fish populations (Kiviat and MacDonald 2002). By the time the Hackensack Meadowlands Commission (HMDC) was created in 1969 wetland acreage had been dramatically reduced from over 42 square miles to only about 13 square miles (Marshall 2004). The cedar swamps were gone, development was rampant, and landfills dotted the landscape. At its creation, the HMDC was initially given three tasks by the New Jersey State Legislature: 1. Oversee orderly development of the Meadowlands 2. Manage solid waste disposal and eventually close landfills 3. Implement environmental regulations for ecosystem protection These tasks were somewhat contradictory and initial efforts of the HMDC focused on the first two portions of its job – regulating development and getting the landfills under control, with the ecosystem protection portion gradually becoming more important over time as national and local appreciation for wetlands increased. The commission eventually acquired numerous wetlands and began to be involved in the process of improving degraded wetlands through mitigation projects. To reflect its changing focus, in 2001 the commission was renamed and the new name, the New Jersey Meadowlands Commission (NJMC), will be used throughout the rest of this report. Though urban sprawl is an ongoing problem in the Meadowlands, many aspects of the environment have improved since the founding of the NJMC. Following passage of the 1972 Clean Water Act, municipalities were forced to clean up their sewage output (Hackensack Riverkeeper Inc. 2006). A shift toward service and technological businesses reduced industrial pollution and most of the landfills have been closed or are currently inactive. Yet, by the late 1980s Phragmites dominated most of the wetland area remaining in the wetlands and game fish were still largely absent from the Hackensack River and its tributaries. In 1995 PSE&G’s Bergen power generating station stopped discharging heated water into the Hackensack River; this appears to have led to improvements in the fish populations in portions of the Meadowlands, as increasing numbers of game fish and greater fish biomass were found in 2001-2003 relative to 1987-1988 (Bragin et al. 2005). The lower part of the river is still subject to heated discharge and many sources of pollution remain (Kiviat and MacDonald 2002). But mitigation projects carried out in the late 1980s and others in the late 1990s began to remove the Phragmites and restore tidal flow to some wetland areas, restoring fish access to these areas and creating better habitat for a variety of bird species, including waterfowl and migrating shorebirds. These tidal saline marsh mitigations are the subject of this report. Tidal salt marshes are dependent on tidal flooding frequency and duration which affects sedimentation rates and the establishment of salt-tolerant vegetation. Over time, salt marshes

DRAFT

15

develop a complex zonation and structure of plants, animals, and microbes, responding to changes in tidal inundation, salinity, and temperature (Bertness and Ellison 1987, Mitsch and Gosselink 2000). The vegetation of these marshes is characterized by low diversity and distinct vegetation zones that occur within specific tidal ranges (Mitsch and Gosselink 2000). The marshes in the Meadowlands are intertidal, meaning they have areas that are flooded twice daily at high tide and areas that are completely drained at low tide (where draining is not impeded by tide gates, as it is in the impoundments at Harrier). Typically, intertidal marshes are characterized by gentle slopes that allow for distinct zones characterized by different flooding regimes and are dominated by clonal graminoids with small populations of other species occurring in low densities or patches (Niering and Warren 1980, Mitsch and Gosselink 2000). In the mid-Atlantic region (as in New England), low marsh areas (which receive substantial flooding twice daily) are typically dominated by Spartina alterniflora (saltmarsh cordgrass) while high marsh areas (which occur at higher elevations and receive less flooding) are dominated by a mixture of short-form S. alterniflora, Spartina patens (salt hay), and Juncus gerardii (saltmeadow rush), often interspersed with Distichlis spicata (seashore spikegrass). Short-form S. alterniflora may form near the border of the low and high marsh. S. patens dominates the lower portion of the high marsh and J. gerardii dominates the upper portion of the high marsh, which is least often flooded. D. spicata is typically the first grass species to fill in barren areas left by wrack (dead stems of other vegetation that float in on the tide and are deposited on the marsh) or other debris. Annuals from the genus Salicornia (pickleweed) are also well-known to establish in these salty, barren patches. Other species commonly found interspersed among the dominant species of these marsh zones include Eleocharis parvula (small spikerush), Pluchea odorata (saltmarsh fleabane), Amaranthus cannabinus (saltmarsh water hemp), Solidago sempervirens (seaside goldenrod), other Solidago species, Aster tenuifolius (perennial saltmarsh aster), and Aster subulatus (annual saltmarsh aster). The wetlands in the Meadowlands are of much lower salinity than sea water. Such brackish marshes are also typically dominated by the above graminoid species and may harbor additional species with a low level of salt tolerance, such as Typha angustifolia (narrow leaf cattail), may also be found. Between the high marsh and upland vegetation, a scrub/shrub zone can be found, with typical common species including Baccharis halimifolia (groundsel tree) and Iva frutescens (big-leaf sumpweed). Because so few species can tolerate anything beyond the lowest levels of salt, these distinct vegetation zones can be found in salt marshes around the world and even in brackish marshes of relatively low salinity. Thus, salt marshes are known for their distinct bands of vegetation where each band is composed of a specific set of species (though the exact species vary somewhat from one region to another). These zones include low marsh, high marsh, scrub/shrub, upland transition, and upland. All marshes discussed in this report are currently owned by the NJMC and were acquired between 1996 and 1999 (U.S. ACOE 2004) (Figure 1.2-1). Some mitigation projects were carried out in the Meadowlands prior to the acquisition of the properties by the NJMC. Among these were the last three sites listed in section 1.1: Western Brackish Marsh, Eastern Brackish Marsh, and the Lyndhurst Marsh Preserve. These last three sites are not actively managed by the NJMC and the NJMC has no data concerning their current states. Rutgers has no field experience with these three sites and limited data concerning these sites is available; most available data also dates back many years to near the time of the mitigations. Therefore, these

DRAFT

16

three sites will not be fully evaluated in this report, but relevant data will be considered where appropriate. Five of the six marshes were covered with a monoculture of the invasive grass Phragmites australis (common reed), a grass that grows to several meters in height and in very dense, uniform stands while gradually raising the elevation through higher deposition rates (Rooth and Stevenson 2000). Consequently this species interferes with marsh hydrology, reduces fish nursery habitat (Weinstein and Balletto 1999, Raichel et al. 2003), and excludes nesting by many birds. Harrier Meadow, however, was a mixture of Phragmites monoculture and other wetland species, including some of the grasses typical of salt marshes in New Jersey. In all cases Phragmites was sprayed with the water-soluble herbicide Rodeo (initially applied aerially on a large scale, with smaller ground-based applications as needed later in the process) and the marshes were reshaped to change the hydrology. In some cases tidal flow was fully restored and in others tidal flow is restricted in at least some portion of the marsh. Additionally, due to differences in design, each marsh differs in the proportion of area belonging to each vegetation zone. In order of the onset of mitigation activities, Western Brackish Marsh was the first site to be enhanced. This site is located immediately to the north of Mill Creek Marsh. Sixty-three acres were enhanced by Hartz Mountain Industries, Inc. as part of the permit requirements for building of the nearby Mill Creek Mall and adjacent Harmon Meadow office complex, which involved filling of 127 acres of similarly degraded wetland. The design focused on creating low marsh areas approiate for sustaining S. alterniflora. As such, much attention was paid to marsh elevation and slope to ensure proper tidal inundation. The herbicide Rodeo was applied in October of 1985 (with additional applications each October through 1988) to kill the existing stand of Phragmites and construction acitivites took place between March 1985 and July 1987 (Berger 1992). Channels and other areas were excavated to create open water. Special excavators were imported from England to produce the fine gradations in elevation needed by Spartina alterniflora in the intertidal marsh. Excavated soil was used to create raised island areas intended for habitation by woody species. This left the marsh with little high marsh acreage. More than 80% of the site is subject to tidal inundation. The marsh was seeded with S. alterniflora in May 1986, April 1987, and spring 1988 (Berger 1992, McCormick and Cantelmo 1995). Seeding was successful and this species became established in greater than 75% of the low marsh zone (French et al. 1996). Some other native marsh species recruited to the site naturally (Berger 1992). Upland areas were planted with trees, shrub root stocks, and herbaceous species in 1986 but suffered from heavy mortality due to the high salt content of the soils (Berger 1992). Survival improved by 1988 following leaching of the salt from the soil. Additionally, fencing was installed to decrease muskrat herbivory. By 1987 far more birds and bird species were using the newly enhanced Western Brackish Marsh than were using the adjacent, Phragmites-dominated Mill Creek Marsh (Bontje 1988). Fish populations showed less response, still being largely dominated by the pollution-tolerant Fundulus heteroclitus (mummichog), but water quality in the neighboring Hackensack River, though improved relative to the very poor quality seen in the 1950s and 1960s, was still somewhat poor in the late 1980s. No more recent data is available concerning the fauna at the Western Brackish Marsh; however, a 2004 report contains a few more recent observations concerning the vegetation; while some areas are dominated by S. alterniflora, others are dominated by the short annual Eleocharis parvula (small spikerush) and that high marsh areas are dominated by Phragmites (U.S. ACOE

DRAFT

17

2004). Upland areas support a variety of species but are dominated by Betula populifolia (grey birch) and Phragmites. Eastern Brackish Marsh is separated from Western Brackish Marsh and the northern portion of Mill Creek Marsh by the eastern spur of the New Jersey Turnpike, which comprises the western border of Eastern Brackish Marsh. Sixty acres of this marsh was also mitigated by Hartz Mountain Industries, Inc. in relation to the building of the Mill Creek Mall and Harmon Meadow office complex (McCormick and Cantelmo 1995). Again the focus was on creating habitat for S. alterniflora so emphasis was placed on creating the needed slopes to create the correct regime of tidal inundation. The Phragmites monoculture of this marsh was sprayed with Rodeo during the same months that Western Brackish Marsh was sprayed. The design for this site was similar to that for Western Brackish Marsh, with removal of soil to create tidal low marsh areas and deposition of this soil to create upland island areas within the marsh. Construction began in April 1987 near the time of completion of construction acitivities at Western Brackish Marsh. Construction of Eastern Brackish Marsh was completed in October 1988. Seeding efforts during spring in 1988, 1989, and 1990 were unsuccessful. Germinated seeds failed to overwinter and no S. alterniflora became established at Eastern Brackish Marsh. Transplant experiments conducted in the area in 1994 suggested that transplanted S. alterniflora might survive in Eastern Brackish Marsh (McCormick and Cantelmo 1995) but no record of any actual planting efforts has been found. The last known observations of the vegetation at this site indicate that the low marsh areas still do not support S. alterniflora and function as mudflats (U.S. ACOE 2004). The upland islands support a mixture of woody and nonwoody vegetation. No data concerning faunal use of this site has been located. The Lyndhurst Marsh Preserve is a 21-acre property of the NJMC, which apparently has been subject to two stages of enhancement, the first of which involved creation of 9 acres of wetland bordered by two acres of channels, which connect the marsh to adjacent creeks (Bontje et al. 1991). This mitigation was carried out by the Bellemead Development Corporation in exchange for filling 3.73 acres needed to build an office building (U.S. FWS 1994) and, like the two mitigations before it, was intended to be a low marsh dominated by S. alterniflora. Rodeo was applied to the site aerially in spring 1989, killing about 75% of the Phragmites; a second application in the fall killed the remaining stands (Bontje et al. 1991). Construction began in January 1990 and was undertaken using conventional earthmoving equipment, which required building of various bridges, finger roads, and dikes to keep out water. It involved removal of densely packed fine sediments that had been deposited there as hydrolic dredge spoils from turnpike construction activities. Numerous obstacles were overcome or worked around by adjusting the marsh design. A 2% grade was created (U.S. ACOE 2004) and planted with S. alterniflora in peat pots in June and July 1990 (Bontje et al. 1991). Pots contained three to four stems and were placed on three-foot centers with fertilizer added to the planting hole. Survival was good and a 1992 monitoring report covering monitoring of ten plots indicated that about 88 percent of the site was covered by S. alterniflora (U.S. FWS 1994). Within a few months of completing the site egrets, sanderlings, gulls, ducks, and the state-endangered Circus cyaneus (northern harrier) could be seen foraging and resting at the site (Bontje et al. 1991). Some Phragmites returned and was sprayed by hand, a ritual that was expected to be required each year. An August 1993 site visit by the Fish and Wildlife Service (FWS), however, indicated that many areas of the site were not receiving enough tidal inundation to prevent encroachment by

DRAFT

18

Phragmites and that only about 4 acres were actually covered in target vegetation (U.S. FWS 1994). The FWS also noted that the site did not fully comply with the mitigation plan and that a planned tern nesting island was absent, but that other marsh plant species had colonized the area; several species of birds were also observed, along with muskrat holes and raccoon tracks. At some later point (ca. 1996) the adjacent high marsh area (along the eastern edge of the preserve) was mowed and replanted with S. patens (U.S. ACOE 2004). Later observations of the entire site indicated that the low marsh Bellemead site continues to be dominated by S. alterniflora but with incursions of Phragmites. The high marsh portion, however, was dominated by Phragmites and D. spicata with some of the planted S. patens remaining, as well as some cover by Spartina cynosuroides (tall cordgrass), S. alterniflora, and Salicornia spp. (U.S. ACOE 2004). No additional data on wildlife was found. An adjacent site, the Lyndhurst Riverside Marsh, is currently under consideration for future enhancement by the NJMC. In the late 1990s the NJMC, began carrying out mitigation projects on recently acquired properties using funds provided by entities that were developing wetland areas elsewhere in the Meadowlands. Under the November 1998 Memorandum of Understanding (MOU) between Rutgers University and the Hackensack Meadowlands Development Commission (Appendix B), monitoring of three wetland mitigation sites (Mill Creek Marsh, Harrier Meadow, and Skeetkill Creek Marsh) was conducted in the Meadowlands District. Wetland preservation and enhancement are some of the main goals of the NJMC. As part of the MOU, monitoring of biotic and abiotic factors was proposed and implemented. At the time, each of the mitigation sites was dominated by a monoculture of Phragmites australis, which altered tidal flow and contributed to decreased wetland function. The intent of these mitigation projects was to promote selfsustaining tidal salt marsh communities with improved wetland structure and function and increased species diversity. The goals of each required removal and later control of Phragmites, an increase in both open water areas and tidal inundation, creation of bird habitat, an increase in fish access, and an increase in mosquito control Mitigation design for each of the sites required some grading of the marsh surface to create additional meanders in existing tidal channels and excavation of shallow pools to accomplish these goals. Harrier Meadow (hereafter, Harrier), located in North Arlington, NJ, was the first site Rutgers worked at and construction was completed in September 1998. The site encompasses a total of approximately 77.5 acres, only about 50 acres of which were included in the mitigation. Prior to mitigation, the site served as a depository for shot-rock from the construction of nearby US Interstate Route 280, which was performed in the mid-1960s. The Erie Landfill is located just north of the site; it had been inactive for some years and started serving as an active landfill again in November 2002 when the 1-E Landfill (located south of Harrier, across an open water/mudflat area) was closed. Prior to mitigation, portions of the site supported a mixture of native salt marsh species typical of a high marsh, including Spartina patens and Distichlis spicata. Other areas, however, were covered by a monoculture of the invasive grass, Phragmites australis with some areas of the invasive Lythrum salicaria (purple loosestrife) intermixed (Feltes and Hartman 2002). While there were some areas of tidally-influenced open water in the interior of the site, the vast majority was densely vegetated with little water exchange. Hydrology in the restored portion of the site is restricted by two managed tide gates, though water often flows over the inner gate, which lies between the northern and western impoundments. The gates allow several tidal management scenarios to be employed. These range from completely opened

DRAFT

19

gates to allow for relatively full tidal exchange to almost completely closed, in order to contain water within the marsh for wildlife management purposes. An upland trail steeply grades to lower elevation areas leading through narrow mudflats to open water channels. These steep slopes consist generally of rocks embedded in a sandy substrate, leaving little room for the development of low marsh areas. Few areas at elevations higher than the mudflats and low marsh areas are regularly flooded even at high tides. There are a few upland areas dispersed throughout the site allowing for some topographic variability ranging from upland to low marsh elevations. Monitoring under the MOU was conducted for Harrier from 1999 through 2003. Additional funding allowed for some monitoring activities to continue in 2004 and additional plantings took place in 2004 and 2005. Monitoring results are detailed in Section 3.0. Mitigation at the approximately 16-acre Skeetkill Creek Marsh (hereafter, Skeetkill) located in Ridgefield, NJ, followed with construction being completed in December 1998. The site is located in a heavy commercial/industrial area. Similar to Harrier, the site generally consisted of a dense monoculture of Phragmites australis, with few open water areas and limited tidal exchange. Since restoration, this site has experienced complete tidal exchange throughout much of the marsh, and some degree of exchange in the impounded area. There is a central drainage channel running through the site that allows for filling and drainage. The majority of the site is directly connected to this channel hydrologically and the portion farthest from the road is connected via a pipe and drains significantly during low tide. The marsh surface has a generally flat topography with many areas of ponded water evident on the marsh surface at low tide. Topographic variation comes from sloping created by channels. There are a few upland areas dispersed throughout the site allowing for some topographic variability ranging from upland to low marsh elevations. Soils are characterized as a mix of sand and organic matter. Construction activities were completed in 1998. Monitoring under the MOU was conducted for Skeetkill from 1999 through 2001 and results can be found in Section 3.0. Mill Creek Marsh (hereafter, Mill Creek) in Secaucus, NJ, was completed in late 1999. The approximately 137-acre site is bound to the west by Mill Creek (Figure 1.2-1). Prior to mitigation, historic mosquito ditches were present, allowing for some degree of tidal exchange. The site was dominated by a monoculture of Phragmites australis. While the tidal creeks and ditches attracted wildlife, these conditions provided limited wildlife habitat for the majority of species. Tidal flushing via the network of existing mosquito ditches was generally good in the areas it affected; however, the majority of the site experienced tidal action only with the highest tides each month or during storm tides. The restored site hydrology is characterized by fully tidal, open exchange in channels throughout the marsh. Full tidal exchange is restricted by weirs in the two large open water impoundments. These drain much slower relative to the rest of the site. Soils are characterized by a significant amount of organic matter. Slopes and elevation gradually increase from open water through the mudflats and into the higher elevation low marsh area. Since this site was modeled largely after Western Brackish Marsh, it was designed to include only a very narrow band of high marsh with a much wider low marsh zone. Thin upland areas snake around the impoundments adding some topographic variability. There area several islands located within the site; these also contain some upland areas. Over time, erosion processes have produced a very sharp change in elevation that is observable leading from the mudflat/low marsh areas to the uplands. Monitoring under the MOU was conducted for Mill Creek from 2000 through 2004. Monitoring results are detailed in Section 3.0.

DRAFT

20

Figure 1.2-1.

DRAFT

The Meadowlands District showing six mitigation sites

21

1.3

Literature Review

1.3.1 Function and Condition A primary objective of wetland assessment is reporting on ambient ecological condition (U.S. EPA 2003). Methods that evaluate condition should serve this purpose (Fennessy et al. 2004). As highlighted by (Fennessy et al. 2004)), current assessment efforts, especially rapid assessment methods, should focus on conditional over functional assessment. They define condition as “the relative ability of a wetland to support and maintain its complexity and capacity for self-organization with respect to species composition, physicochemical characteristics and functional processes as compared to wetlands of a similar class without human alterations” (Fennessy et al. 2004, p. 3). Ideally, the U.S. EPA holds that monitoring programs should assess wetland condition. They suggest that at the broadest level, monitoring should include detecting and characterizing ambient condition, describing whether condition is improving, defining seasonal patterns in condition, and identifying thresholds for stressors (U.S. EPA 2002a). The increasing focus on comprehensive watershed management and protection underscores the need to look, not only at functional value of individual sites, but at the condition of all wetlands in a watershed (Theising 2001). Wetlands work together to support biodiversity, and impacts upon individual wetlands cumulatively affect the overall condition of a watershed. Though increasing attention has been paid in recent years to wetland condition, Section 404 of the Clean Water Act and the “no net loss” policy, with their focus on impact assessment of individual sites, initially led to a focus on wetland function. Thus, much of the literature available focuses on functional assessments. Niedowski (2000, p. 7) defines function as “a physical, chemical, or biological process which takes place in wetland areas.” She provides the following examples of commonly recognized functions: food chain production, provision of fish and wildlife habitat, barrier to waves and erosion, storm and flood water storage, and nutrient and chemical uptake. Other functional assessments can include the potential for use of habitat by birds and mammals, the production of invertebrates, and the ability to improve water quality (Borde et al. 2004). Many functions are value-laden, which limits their objectivity. These human-related values (e.g., aesthetics, recreation use, pollutant removal) are appropriate in some cases, especially in urban settings, where the societal value of a restored site can garner public interest and support (Niedowski 2000), but such functions should not be the sole consideration. The link between condition and function lies in the assumption that ultimately, a wetland in excellent condition will optimally perform wetland functions (Fennessy et al. 2004). Performing optimally is not the same as performing at the highest possible level, however. Performance at the highest wetland function is an ideal case and possible only in a hypothetical setting. Optimal wetland performance requires tradeoffs among multiple functions where none of them are performing at the highest possible level. Increasing function A can bring about a concurrent

DRAFT

22

decrease in function B. For example, Phragmites marshes score very highly on sedimentation functions, but the high sedimentation rates result in decreases in other functions, such as wildlife habitat. Optimal performance takes constraints like landscape into consideration. When determining how to measure the success of mitigation, consideration must be given to the feasibility of the system returning to reference conditions and to the tradeoffs between and among condition and function and how these relate to overall marsh condition. Additionally, there is a tendency to look at function and condition separately, despite the links between the two and the different value provided by different functions and conditions. Assessment methods with greater emphasis on hydrology and water quality, with their many important functions, often ignore the condition of vegetation and wildlife and instead try to measure these as functions, either in terms of potential habitat or in terms of their asthetic value, which is determined in a very qualitative manner. Ideally, indicators of both function and condition would be considered when assessing wetlands (Hatfield et al. 2004). This would allow for the measurement of various important functions while also looking at the actual condition of the marsh and not just its potential to be in good condition. Particularly in an urban landscape a high value may be placed on certain functions above that placed on good wetland condition and different functions are performed better by some wetland types than others. For example, because of the social value placed on waterfowl, there has been emphasis placed on open water wetlands in the Meadowlands, to the detriment of other wetland types with greater tidal influence and different effectiveness at fulfilling various hydrological functions. A similar trend has been seen in Massachusetts where many created wetlands were very dissimilar to the riverine forested wetlands they were built to replace (Brown and Veneman 2001). While it may be easier, cheaper, and more expedient to build an impoundment or a wet meadow than a riverine forested wetland or a fully tidal salt marsh with normal topography, there is an inherent shift in the functional capacity of the watershed that accompanies changes in wetland type. Due to the complicated urban environment, there are many factors that influence mitigation in the Meadowlands. Cost, space, pool of potential mitigation sites, fragmented habitat, public input, and many other factors play a role in the determination of goals for the mitigation. Mitigation in the Meadowlands is dependent early in the planning stages on the functional processes that are meant to be replaced to satisfy mitigation goals and objectives. Functionalbased considerations are necessary in the planning, site selection, and design phases of the project. Upon construction of the mitigation site, however, condition and structure of the mitigated wetland become the driving force that leads to sustainability. The National Research Council (1992) suggests that structural (condition) and functional components should make up the basis for measuring restoration success. The discussion that follows draws functional and conditional aspects from the literature to provide a list of issues that are relevant to the process of wetland mitigation.

DRAFT

23

1.3.2

Issues to Be Considered During Planning, Site Selection, and Design

1.3.2.1 Planning: Goals and Objectives Central to any mitigation plan is an ecological or biological target (Diefenderfer et al. 2003). What are the goals and objectives of the mitigation? To protect rare, threatened or endangered species? To restore the condition and function of a degraded or lost habitat? Ecological, economic, public benefit and aesthetic goals should be identified if applicable (Niedowski 2000). Clear goals and objectives are needed early in the planning stages to allow room for trade-offs. For example, although it is much more costly, should plantings be conducted instead of reliance on natural colonization? If a goal is to increase the abundance of species A, what are the detrimental effects upon species B and C, if any? A determination should be made as to what types of biotic (plants, animals) and abiotic (hydrology, ecosystem functions) goals are targeted for the mitigation. Because the Meadowlands District is located in a heavily urbanized landscape, anthropogenic influences bear heavily on the development of mitigation goals and success criteria (Ehrenfeld 2000a). Successful urban wetland mitigations require a determination of the kinds and intensities of urban influences affecting the site and examination of ways to assess wetland function within that context (Ehrenfeld 2000b). In order to set achievable goals for the mitigation project, ecological constraints of the urban environment must be identified, accepted, and, if possible, capitalized upon (Baldwin 2004). In a sense, the degraded urban wetland is considered incapable of achieving ecosystem function like that of pre-disturbance conditions so the best must be made with what is available. Zedler and Callaway (1999) recognize the difficulties and limitations of restoring urban wetlands, stating that in an extreme case, mitigations in urban environments are the most difficult to develop into natural systems. Acquiescence to development and pollution is not suggested here, but rather an awareness of the relentless forces inherent in urbanization. A theme that should be persistent in these discussions is sustainability of the mitigated site (Niedowski 2000). Based on the literature, knowledge of several key ideas (e.g., hydrology, landscape, flora and fauna) is necessary to develop ecologically-based, scientifically-sound goals, plans and designs. 1.3.2.2 Site Selection Choosing the location of a mitigation site, likewise, depends on many variables based on ecological principles. However, the primary factors that influence site selection are biological importance, potential for restoration sustainability (Diefenderfer et al. 2003), and satisfaction of mitigation goals. The selected site should be found in the same watershed where the degraded/filled site is located. This will allow for replacement of like habitat to restore the balance of hydrologic, soil, and biotic aspects of that watershed. NOAA (2003) recommends consulting local, regional/watershed, and state lists of priority wetland restoration sites with the best possible sites found near to wetlands similar to the target type. They also provide a list of factors to consider when selecting a site: hydrology, topography, geology, soils, biotic components, land ownership, and local agency requirements. Borde et al. (2004) recommend that a site assessment be conducted during the site selection process. The site assessment is used for collection of baseline data regarding pre-mitigation conditions.

DRAFT

24

Once a range of potential mitigation sites has been identified, prioritization is required to allow for the best use of resources. In 2002, Restore America’s Estuaries developed a four-step process to prioritize projects for restoration site selection (Restore America's Estuaries 2002). Step 1 identifies the pool of potential sites targeted for restoration. Step 2 involves ranking the project sites based on: a) the benefits the project site would provide to the region, b) the degree to which the project site encourages coordination among federal, state, and private groups, c) the level of innovation shown in the technical aspects of the project, d) the expected success of the project, and e) the capacity to cover project costs. Step 3 requires grouping of potential project sites at two levels: according to six national regions, and then by estuary within each region. Using additional parameters to evaluate the highest ranking sites, Step 4, uses the ranked list developed in Step 3 to determine the final site selected for restoration. 1.3.2.3 Design Mitigation project design applies not only to site-specific construction plans but to the entire project at large. It requires the development of a conceptual model prior to development of detailed plans. The conceptual plans help to identify the factors that dictate the development, maintenance, condition, and function of the restored habitat (Diefenderfer et al. 2003). It also requires development of a numerical model for selection of performance criteria, for sensitivity analysis and prediction (e.g., hydroperiod), and for prediction of successional trajectories (Diefenderfer et al. 2003). Both models are then used to develop preliminary and alternative designs. The different designs should be able to meet the specific objectives of the mitigation project while posing different management methods and corresponding costs. The selected site should be designed to replace another similar site so that there are no changes in the distribution of habitat types (e.g., an upland mitigation meant to replace lost mudflat habitat) within the landscape (Zedler 2000). 1.3.2.4 Design Issue: Hydrology The importance of hydrology in determining the success of a mitigation project cannot be overstated and is widely recognized by the scientific community, yet there is a paucity of guidance on how to restore sustainable hydrology (Montalto and Steenhuis 2004). Appropriate hydrologic function is imperative to a healthy, growing wetland ecosystem and is a major determinant of wetland processes. Hydrology affects the abiotic characteristics of the wetland such as salinity, water temperature, dissolved oxygen, and nutrient availability. These in turn, affect the biotic aspects of the marsh and determine what taxa are present in the system. It greatly influences vegetation structure via seed dispersal, species establishment and diversity, sedimentation, and nutrient availability (Environmental Law Institute 2004). The biota affects structural components of the marsh and cycle back nutrients into the system (Mitsch and Gosselink 2000). Hydrology also controls movement of material between and within wetlands. Many mitigation “failures” have been attributed to improper hydrology where tidal inundation is excessive and/or there are periods of extreme dryness (Mitsch and Wilson 1996). In mitigation design, it is necessary to be able to predict the tidal range expected after construction. Both insufficient and excessive tidal flooding can lead to wetland loss. Hydrologic modeling should be used in planning and design (Montalto and Steenhuis 2004). Hydroperiod (the frequency and duration of flooding) must be designed into the mitigation. Although generalized tidal data is

DRAFT

25

available, site-specific variation is too great. Field data must be collected prior to construction to gauge local variability. 1.3.2.5 Design Issue: Landscape Factors Although site-specific factors are quite important in selecting, planning, and designing a mitigation site, there has been a re-focus from site-specific conditions to the larger landscape. Sustainability of the mitigation site is dependent on the larger system processes (e.g., hydrology) that control function at the site itself. This is especially pertinent when mitigations target motile species (e.g., birds) - concentrating only on local factors ignores habitat requirements at a larger scale (Borde et al. 2004). Project planning and design should require research into the potential effects of landscape factors such as size, shape, adjacent land use, connectivity, and distance to other similar natural areas (for seed bank, dispersal, and population sustainability reasons) (Diefenderfer et al. 2003). Breaux et al. (2005) found that larger sites situated within a matrix of other wetland sites were more sustainable and offered more habitat benefits to wildlife than smaller, isolated sites. Site selection should take the landscape position of the potential mitigation site into consideration. Placing a mitigation site adjacent to other wetland sites (restored or natural) expands the natural area, connecting disjunctive wetlands (Shisler 1990). This will increase corridors for seed bank dispersal and migration of organisms between sites. A limited seed bank can slow recovery of plant species (Zedler 2000), requiring costly planting activities. In general, locating a mitigation site adjacent to a proper seed bank may increase natural recruitment and allow for “self-design” to take hold at the mitigated site. Additionally, it will provide buffers to potential disturbance (natural and anthropogenic). 1.3.2.6 Design Issue: Urban Environment For a wetland mitigation project to be successful, especially in an urban environment, integration of the concerns and needs of the general public must be taken into account. Public support for a mitigation project can be more easily obtained when there is a connection made between natural wetland function and the people; that is, if the people can see one or more functions as being beneficial to society. Drawing the support of the local community will benefit any mitigation effort. Emphasis should be placed on restoring function not only to better the wetland system itself and the surrounding landscape, but also to contribute toward meeting the goals that society and the public have deemed worthwhile (Cairns 2000). Some of the most successful mitigation projects have been supported by and have integrated the concerns of people in the vicinity of the mitigation. In many of these cases, it was the feeling of loss suffered from degradation of the ecosystem by the people living there that spurred the mitigation effort (Cairns 2000). Hatfield et al. (2004), in their evaluation of eight wetland assessment methods, noted that assessment scores of a degraded wetland in an urban environment can actually be greater than scores for a pristine wetland in non-populated rural areas when using methods that place emphasis on the value of the marsh to society because of the higher functional contribution the degraded site has to the urban society. Montalto and Steenhuis (2004) comment that a better understanding of the linkage between hydrology and function of natural aquatic and terrestrial systems could contribute to an enhanced appreciation by the public as well as policy makers, for the functional services which natural systems provide in urban environments.

DRAFT

26

1.3.2.7 Design Issue: Geomorphology Site geomorphology is an important variable to consider in the design process. Sustainable mitigations depend on sufficient sedimentation rates that allow for the establishment and maintenance of appropriate geomorphology and wetland plants (Boumans et al. 2002). In many of the wetland mitigations that have occurred in the Meadowlands, increased tidal exchange is encouraged by removal of monotypic stands of Phragmites australis. The removal of this tide restriction increases the inundation period of the mitigated areas, and opens up areas to sediment availability for the establishment of target wetland plants. However, with the increased tides comes the chance for unexpected sedimentation processes to occur. Potential changes to the geomorphology (e.g., elevation changes such as erosion or sedimentation, spurred by increased tidal flux) should be accounted for. Sedimentation depends on the overall hydrologic regime, which should be appropriate to provide sufficient sedimentation while maintaining the necessary tidal flux to sustain wetland vegetation (Boumans et al. 2002). Elevation is a considerable factor in determining the types and amounts of planted vegetation that could survive (Delaney et al. 2000). Additionally, substrate elevation determines vegetation zonation (Environmental Law Institute 2004). Differences in constructed shoreline geometry result in different vegetation growth patterns suggesting that wetland design significantly affects function (Environmental Law Institute 2004). In a study of a constructed coastal marsh in Virginia, Havens et al. (1995) found that a more successful restoration would have been achieved if substrate from existing wetlands was used in construction, if microtopography such as rivulets had been considered, and if direct routing of drainage ditches to the head of constructed wetlands had been considered. These recommendations would probably enhance the attractiveness of the constructed site to a variety of marsh species. 1.3.2.8 Design Issue: Vegetation Havens et al. (1995) also found vegetation to be important. In the same constructed marsh study in Virginia, they found that greater success would have been achieved if vegetation had initially been established with high stem densities and wide (at least 12 m) fringes, if there had been an establishment of mature shrub growth, and if there had been a mix of different habitat types incorporated into the design. A study of 13 restored wetlands in New York showed that areas that were already vegetated were more readily replaced with desirable species when site hydrology was restored, while areas where vegetation was removed during the process of restoration and then left barren were more likely to become dominated by an undesirable species (Brown 1999). Planting as an alternative to “self-design” dispersal processes from adjacent wetlands is an effective, albeit costly tool to increase vegetation cover at the mitigation site. 1.3.3 Issues to Be Considered During Construction Construction plans allow visualization of physical structure and are used to determine costs and schedule of implementation. They should include detail on slope, elevation, erosion protection, seeding and planting requirements, hydrology and substrata composition (Diefenderfer et al. 2003). The plans should allow for adjustment in the case of unexpected field conditions requiring plan modification. Any alterations would require field documentation and notification on postconstruction as-built drawings. In all but the simplest mitigation projects, an engineer should be

DRAFT

27

involved from the planning phase through construction (Diefenderfer et al. 2003) and the project manager or engineer should be on-site to monitor progress. Since large earth-moving activities take place during this phase, it is necessary to minimize disturbance created by movement of heavy vehicles and machinery (NOAA Interagency Workgroup on Wetland Restoration 2003). Construction monitoring should be performed to document that construction is proceeding as planned (Shisler 1990, NOAA Interagency Workgroup on Wetland Restoration 2003) and that no adjacent natural habitat or private property is excessively damaged in the construction process. Seasonal considerations should be made for plantings (if any) and construction activities. A comparison of pre-/post-construction condition can be performed using postconstruction as-built drawings and the pre-construction plans. 1.3.4 Issues and Condition Indicators to Be Considered During Monitoring Following construction of the mitigation site, performance of plantings and development of asbuilt drawings, it is necessary to track changes that are occurring at the site and to see if the site is maintaining itself and meeting the goals set forth in the planning stages of development. Conditional rather than functional issues become paramount in the monitoring stage to ensure the sustainability of the mitigation. 1.3.4.1 Adaptive Management The monitoring program should be developed in the planning phase of the project. It is there that project goals are identified and corresponding monitoring protocols can be developed that consider the requirements of the goals (Diefenderfer et al. 2003). A good monitoring program provides a comprehensive look at post-construction site conditions and allows for tracking of biotic and abiotic changes to the mitigated wetland. The primary purpose of the monitoring program is to evaluate progress towards specified goals and to determine any adjustments necessary to bring the project back on track if there is deviation from the prescribed goals (Diefenderfer et al. 2003). The monitoring program should be able to identify any breaks or obstacles that hinder the progress of the mitigation. Once deviations are identified, it should be able to compensate by undergoing adjustment, i.e., it should be adaptive. Many mitigation projects do not reach desired targets during prescribed monitoring times. With an adaptive outlook, success criteria can be developed with an eye towards trends and target goals that are both incremental and adjustable (Hackney 2000, Kentula 2000, Hobbs and Harris 2001, Diefenderfer et al. 2003). An adaptive management plan acknowledges that there is no way to fully predict all aspects of the interdependencies between natural systems and that setting static target goals does not allow for the stochastic nature of succession in natural systems. It is an interactive process where past management choices are redesigned using knowledge gained from prior monitoring activities. Although monitoring is typically the least supported part of a restoration project, it is the most important aspect of an adaptive management program (Steyer and Llewellyn 2000). Existing ecological successional theory can provide a framework for desired outcomes and often, these are predictable. But unforeseen disturbances from the environmental matrix (e.g., climatic, anthropogenic) can divert the successional trajectory of the system in many and varied directions (Borde et al. 2004, Choi 2004). Devising modifications and alternatives and considering more

DRAFT

28

than one possible outcome should be required as part of a realistic, long-term management plan (Lindig-Cisneros et al. 2003, Choi 2004) that is robust to unforeseen changes in the successional process. Clearly identified goals developed in the planning and design process, acknowledgment of uncertainties and development of alternatives, and a strong monitoring program are all elements of an adaptive management plan (Borde et al. 2004). 1.3.4.2 Timeframe The temporal scale of required monitoring is difficult to approximate because there has been no definitive research done on timeframes involved with achievement of full functionality or optimum condition. Many studies show that the industry-accepted minimum of 5 years is too short to demonstrate a base level of performance (Mitsch and Wilson 1996, Zedler and Callaway 1999, Petranka et al. 2003, Environmental Law Institute 2004) but Thayer et al. (2003) suggest that monitoring continue for a minimum period of five years post-construction. Ideally, the monitoring program should be in place until the system is self-sustaining. Longer timeframes, possibly on the order of decades, would be preferred (Mitsch and Wilson 1996, Henry et al. 2002, Roman et al. 2002, McKenna 2003) although it is not always feasible, cost-wise. Diefenderfer and Thom (2003) and Pacific Estuarine Research Laboratory (1990) suggest that 10 and 20 years, respectively, is not an unreasonable amount of time to perform monitoring. They propose decreasing annual monitoring frequency by 2 to 4 years after initial construction. They state that there will be evidence that mitigation will “work” within the first three years after construction. Decreasing monitoring frequency after year three appears appropriate. Since target timeframes are difficult to prescribe, performance criteria should be stated in terms of targeted trends and ranges. Short timeframes are inadequate to judge success as it pertains to wetland mitigations. 1.3.4.3 Condition Indicators Monitoring parameters should provide a quantitative measure of progress towards specified goals. They are dependent on the type of restoration being performed, the system in which the restoration will occur, and the project goals. It is imperative that a standardized method of evaluation be determined so that results from different mitigation sites can be compared (Niedowski 2000). Variability in methods, while at times necessary due to logistical constraints, should be kept at a minimum. Monitoring data should be compared to reference sites (Ruiz-Jaen and Aide 2005). Thom (2000) states that the problem with traditional restoration goals is the performance criterion of “matching conditions in a natural reference system”. This is especially true in an urban landscape. As is the case with the Meadowlands, finding a natural reference system may not be realistic or appropriate (Windham et al. 2004). Even if a comparable wetland reference system were available, it is rarely, if ever, possible to return a degraded wetland system to its undisturbed state (Thom 2000, Ehrenfeld 2000a, 2000b). Niedowski (2000) states that salt marsh restoration to pre-disturbance conditions in the northeastern US is not even possible since human activity has acted in the region since before colonial settlement. Still, there are sites within the Meadowlands that are less impacted than others and these should be used as reference sites. Without comparison to reference sites, it is impossible to determine practical and realistic mitigation goals or to track progress (Niedowski 2000, Ruiz-Jaen and Aide 2005).

DRAFT

29

Of the 102 references examined for this literature review (Table 1.3.4-1), only 28 of them directly mention appropriate indicators to measure success. Appendix A contains an annotated bibliography of references cited in this report, as well as a list of additional references. Ninteen indices were identified with the following five metrics occurring most frequently: hydrology (13 instances), vegetation community dynamics (9 instances), benthic invertebrate community dynamics (8 instances), avian community dynamics (7 instances), fish community dynamics (7 instances). Collection of baseline data occurred 4 times and gets special treatment as it enables a comprehensive view of pre- and post-mitigation function and condition. Table 1.3.4-1. Ranking of indicators from literature review.

Indicator Number of Occurrences hydrology 13 vegetation community dynamics 9 benthic invertebrate community dynamics 8 avian community dynamics 7 fish community dynamics 7 baseline data 4 soil 4 ecosystem scale 3 topography 3 "biological criteria" = chemical and biotic factors, 2 energy source, flow regime, habitat structure construction monitoring 2 sedimentation rates 2 water chemistry 2 vegetation structure (cover, biomass, density of 1 woody species, vegetation profiles) biological indicators vs. chemical indicators for 1 watershed health diversity – richness and abundance of organisms 1 from >1 group, preferably from >1 trophic level nutrient cycling (decomposition, mineralization, etc.) 1 biological interactions (e.g., mycorrhizae, herbivory) habitat for motile species

1 1

microbial communities N-retention sediment nutritional content

1 1 1

soil chemistry

1

DRAFT

30

The following discussion focuses on the top five indicators, each of which occurred at least seven times in this literature review. Addtional consideration is given to the collection of baseline data as it is an important standard for which to assess the mitigation after construction. Prioritizing these parameters can provide a streamlined methodology for assessment. It is imperative that a standardized method of evaluation be determined so that results from different restoration sites can be compared. Variability in methods, while at times necessary due to logistical constraints, should be kept at a minimum. 1.3.4.4 Condition Indicator: Hydrology For the same reasons that hydrology is so important in the planning and design process, it is equally critical in post-construction monitoring. Monitoring is necessary since wetlands are highly dependent on proper hydrology. It also influences vegetation zonation (e.g., invasives) (Garde et al. 2004). Monitoring should include tidal measurements in deepwater channels and within the wetland area (Montalto and Steenhuis 2004). Montalto and Steenhuis (2004) documented linkages between hydrology and tidal marsh function and restoration in the New York/New Jersey estuary. They cite the critical need for an in-depth understanding of local hydrological conditions in mitigation projects. The following site-specific indicators have been suggested by Neckles et al. (2002): • • • •

cross-sectional profiles of major tidal creeks measured at permanent locations groundwater level from permanent wells installed within the marsh and along upland marsh edge basic water quality parameters sampled in main tidal channels (e.g., dissolved oxygen, salinity, temperature and pH) tidal current in main channels

The Pacific Estuarine Research Laboratory (PERL) recommends pre- and post-construction measurements (depth and duration) of tidal inundation (Pacific Estuarine Research Laboratory 1990). Elevation is also thought to provide a good metric of hydrologic function as it is a general indicator of inundation. At a minimum, 1-foot contours are suggested. Additional indicators include water column stratification (via surface and bottom measurements of temperature and salinity) which indicates impaired tidal flushing. 1.3.4.5 Condition Indicator: Vegetation Wetland vegetation has been widely accepted as an indicator of wetland success but there are no set criteria from which to perform comparisons. Development of a standard metric has not been identified in the literature; however some have suggested the use of species diversity and richness (Environmental Law Institute 2004). At the same time, vegetation cover and composition are two of the most commonly used metrics in monitoring of mitigation sites and in some cases, they are the only indicators used (Windham et al. 2004). Boumans et al. (2002) stated that long-term mitigation success can be expected if emergent vegetation is present. Wissinger et al. (2001) compared plant communities in restored wetlands to those in hydrogeomorphically similar reference sites and found that vegetation cover, plant species and

DRAFT

31

plant diversity were similar to that in reference wetlands within six years after construction. They attributed their results to, among other things, remnant populations of wetland plants at the restoration sites. They concluded that the presence of a diverse remnant community at restoration sites should avoid the necessity of inoculation with plant propagules. This method of “self-design” was also touted by Mitsch and Wilson (1996). Other authors have also recognized the importance of vegetation in gauging success (Zedler and Callaway 2000, Boumans et al. 2002). Specific vegetation parameters have been proposed by Niedowski (Niedowski 2000): • • • • • •

species presence stem density plant height disease, predation, or other disturbance mapping of vegetation zones fixed-point photo stations

Zedler and Callaway (2000) list other metrics used in previous investigations: • • • • • • •

cover density biomass species richness plant height stem length areal change in habitat and/or vegetation classes

Breaux et al. (2005) recommend using absolute rather than relative cover to explicitly represent vegetation structure. Studies have shown that native species richness and diversity increases with wetland age, size, and distance to nearest established wetland. Density of tree and herbaceous cover may also be useful (Environmental Law Institute 2004), especially as it relates to provision of cover for birds. Management for invasive species such as Phragmites australis is essential for maintenance of habitat and species diversity (Garde et al. 2004). 1.3.4.6 Condition Indicator: Avian Habitat Use Research into the connection between avian assemblages and wetland age (VanRees-Siewart and Dinsmore 1996, Buffington et al. 1997) suggests use as an indicator with most of these studies proposing that the correlation stems from vegetative community dynamics; increased community complexity implies increased wetland maturity (Environmental Law Institute 2004). Others recognize avian use as an indication of function (Niedowski 2000, Wissinger et al. 2001, Neckles et al. 2002, Frederick and Ogden 2003, Borde et al. 2004, Windham et al. 2004). Specific avian indicators have been proposed by Niedowski (2000). These include: • •

DRAFT

saltwater-fish-feeding birds benthic-invertebrate-feeding birds

32

• •

wintering waterfowl (if applicable) other species

PERL (1990) specify that sampling locations should be placed in areas of varying habitat, with relatively homogeneous vegetation and topography. In a study conducted by Seigel et al. (2005), monitoring of avian usage at two mitigation sites in the Meadowlands (Mill Creek and Harrier) was conducted to assess the effects of mitigation of largely Phragmites-dominated, non-tidal wetlands. Pre-construction monitoring was conducted at Harrier in 1997 with post-construction monitoring conducted from 2001 through 2003. At the other site, Mill Creek post-construction monitoring was conducted from 2001 through 2003. Seigel et al. found that there was considerable change to the avian community after mitigation. With the mitigation of varied habitat, there was a concomitant increase in avian abundance and an increase in species that utilized differing habitat (e.g., mudflat and open water) other than just Phragmites habitat. 1.3.4.7 Condition Indicator: Fish Habitat Use Studies have shown that fish use marsh surfaces for foraging, spawning and as areas for predator refuge (McIvor et al. 1989). Fish also serve as an important trophic link in salt marsh communities with adults preying upon invertebrates and larval conspecifics (Weinstein and Balletto 1999, Zhou and Weis 1999). Fish are in-turn preyed upon by other nekton such as crabs and other fish, as well as waterfowl. Low species richness indicates stressed environmental conditions. Much of the literature focuses on taxonomic metrics (e.g., density, size, species composition, which is an early warning indicator of community change in response to disturbance) and does not address wetland function. However, it has been proposed to use taxonomic metrics in conjunction with more function-based metrics (e.g., change in biomass, trophic interactions, feeding patterns, residence time, recruitment, chemical retention and percentage of pollution tolerant species) to describe how fish use wetlands (Environmental Law Institute 2004, Windham et al. 2004). Fish are widely studied because their high motility enables them to access restored wetlands more quickly than other taxa. Others have suggested looking at fish habitat use to indicate function (Niedowski 2000, Borde et al. 2004). PERL (1990) specify methods for fish sampling and recommend: • • • • • • • • •

DRAFT

varying habitat should be sampled quarterly sampling at moderate tide levels using blocking nets and seines (minimum 3 mm mesh size) seine hauls of 10-15m parallel to channels and creeks until the total number of fish caught, declines two consecutive times a determination of species composition a determination of numbers of each species a determination of fish length a determination of catch-per-unit-effort a determination of total population size a determination of density-per-surface-area

33

1.3.4.8 Condition Indicator: Benthic Invertebrate Habitat Use Benthic invertebrates are good indicators of wetland function because they are the trophic connection between detritus and vertebrates, playing an important role in food webs. Breaking down detritus, they cycle organics and contribute to the mixing of sediments. Invertebrate density and taxonomic composition can serve as indicators of a functioning ecosystem. In addition, with larger invertebrates that have long life cycles, age structure can be used as an indicator of function and can show the effects of environmental disturbance (Environmental Law Institute 2004). Due to their highly motile early life stages, they are good landscape-scale indicators (Windham et al. 2004). Wissinger et al. (2001) compared the invertebrate community at a 3-year old mitigation site with reference wetlands and found them to be similar. Others have recognized invertebrate diversity to indicate function (Borde et al. 2004). Specific benthic invertebrate indicators have been proposed by Niedowski (2000): • • •

ribbed mussels fiddler crab burrows other benthic invertebrates

Windham et al. (2004) have suggested looking at invertebrate nursery habitat rather than population dynamics as an indicator of progress. Sampling for rates of colonization in restored habitat in comparison with reference sites can be indicative of target succession trajectories. PERL (1990) recommend using sediment cores to sample burrowing benthos and litterbag traps for mobile invertebrates on the marsh surface. RAMs, with their less intensive sampling procedures, underestimate impacts due to insufficient accounting of rare species (Environmental Law Institute 2004). 1.3.4.9 Baseline Data While not a condition indicator, per se, collection of baseline data during the site assessment enables characterization of the past (historical/predisturbance condition), present (degree of altered condition), and future (sustainability) of the mitigation site (Diefenderfer et al. 2003). Most studies, however, do not address this when assessing mitigation success. Surprisingly, only four authors in this literature review cited the need for adequate pre-construction, baseline reference data (Kondolf 1995, Niedowski 2000, Thayer et al. 2003, Environmental Law Institute 2004). The Environmental Law Institute found that one of the biggest hurdles to overcome in achieving a successful mitigation was the lack of baseline data. With no reference point to look to, there is no way to determine if site conditions and wetland function have improved. Baseline data enables a comparison with post-construction condition to determine the extent of biotic and abiotic change that has occurred as a result of the mitigation.

2.0 MILL CREEK – A CASE STUDY IN DESIGN AND IMPLEMENTATION The Mill Creek site was used as the primary study site focusing on biological assessments used in evaluating site selection, design, and ecological performance of the restoration. Deliverable #4 of the NJMC’s proposal to the U.S. EPA, called for reviews of the design rationale, baseline studies, and role of the Indicator Value Assessment method (IVA) (and other functional

DRAFT

34

assessments) to determine how these influenced the design for the Mill Creek site. Deliverable #4 also included a review of the permits for compensatory mitigation requirements, a review of the as-built conditions, and a review of the monitoring protocols for the Mill Creek site. Each of these reviews will be used to consider the process that is currently in place for site selection, design, implementation, and monitoring of wetland restoration sites in the Meadowlands. The approximately 137-acre Mill Creek site is located adjacent to the eastern side of Mill Creek, west of the eastern spur of the New Jersey Turnpike, in the town of Secaucus (Figure 1.2-1). Prior to restoration, the wetlands at the Mill Creek site exhibited reduced wetland functions. Because of elevations above the mean high water line, much of the site did not experience diurnal tidal inundation. A dense monoculture of common reed (Phragmites australis) dominated the site. This combination of site conditions resulted in very minimal fisheries habitat and production, aquatic diversity and abundance, general waterfowl and wildlife habitat, nutrient retention and transformation, sediment stabilization, flood storage functions, and opportunities for recreation and education. Directly to the north of the Mill Creek site lies the Western Brackish Marsh, a 65-acre parcel restored by Hartz Mountain Industries, Inc. (Hartz Mountain) as mitigation for wetland fill which facilitated the construction of the nearby Mill Creek Mall in the late 1980s. Observations of increased wildlife activity since mitigation, the success of historic Spartina alterniflora seeding and planting efforts, the adjacency of the Western Brackish Marsh to the proposed Mill Creek site, and the general impression of a sustainable restoration led the Mill Creek design team to use the Western Brackish Marsh site as a reference marsh for the Mill Creek site restoration design. 2.1 Rationale for Selection, Design, and Construction As the Mill Creek site was located adjacent to a “successful” wetland mitigation, i.e., Western Brackish Marsh, various site selection issues (see Section 1.3.2.2) were indirectly assessed, including: • • • • • •

Does the proposed site have the potential to satisfy mitigation goals? Is it located in a similar watershed as other sites so that they share the same hydrologic, soil, and biotic requirements? Have local, regional, and state lists for priority restoration sites been consulted? Who owns the land? Are there local agency requirements? Has there been or will there be a site assessment for baseline data collection? Related to this, is there or will there be an assessment of hydrology, topography, soils, geology, and biotic components?

The rationale was that pre-mitigation site conditions at the Mill Creek site were similar to those at a pre-mitigation Western Brackish Marsh. Since mitigation at Western Brackish Marsh was considered “successful”, it was reasoned that mitigation at Mill Creek would proceed likewise. To address Deliverable #7 of the NJMC’s proposal to the U.S. EPA, it should be noted that of these site selection issues, only hydrology ranks definitively as a “success” measure (Table 1.3.41). DRAFT

35

2.1.1 Design Objectives The purpose of the Mill Creek restoration was to improve the ability of the site to provide water quality improvement, fisheries and wildlife habitat, and recreation/education opportunities. The general objectives for the restoration design are described in the following paragraphs. More specific details on each of the design components (i.e., channels, edge, mudflat and emergent zones, uplands and islands, ponds, brackish impoundments, stormwater treatment system, wetland vegetation, and upland planting) are contained in the Mill Creek Design Rationale report (Hackensack Meadowlands Development Commission 1997). 2.1.1.1 Water Quality Improvement Since tidal flushing in Mill Creek and the network of existing mosquito ditches and creeks was already deemed to be good, the overall objective of the design was to grade the rest of the site so that it, too, experienced daily tidal fluctuations. Water quality improvement was to be realized with an increase of marsh/water interaction that was absent due to the dense Phragmites monoculture and resulting higher elevations that predated mitigation activity. In the mitigation process, low marsh emergent, mudflat, and open water areas were to be created. System flushing was to be maintained through the construction of a network of primary and secondary channels throughout the site. Primary channels were designed to meander through the site, linking the site to Mill Creek and to existing channels extending out of the Western Brackish Marsh. The channels were to incorporate portions of existing mosquito ditches and topographical depressions in order to minimize excavation. Secondary channels were designed to aid in draining the floodplain during low tide and to provide natural barriers against the intrusion of Phragmites rhizomes. Emergent areas and mudflats were to be constructed along the edges of all primary and secondary channels and along the perimeter of all islands and impoundments. Two large, sheltered impoundments were included in the design to provide waterfowl habitat and fish breeding areas. These impoundments were designed to be relatively shallow so that they also provided dabbling duck, shorebird, and wading bird feeding habitat. Water control structures were designed to be used to hold water in the impoundments at low tide, thereby minimizing excavation and offering the ability to control impoundment water levels. These managed open water impoundments were envisioned to have timed drawdowns (in the spring and fall) to accommodate the shorebird migration season. Some stormwater detention and treatment was designed into the system. Stormwater enters the site from the west side of the eastern spur of the NJ Turnpike as well as from the Mill Creek Mall found to the south of the mitigation site. The incoming stormwater was predicted to be discharged into a long channel running from the southwest corner of the site to the southeast, and then north along the eastern edge of the site, emptying near the northeastern corner, near the Western Brackish Marsh. It was envisioned that marsh vegetation along the channel would serve as a form of biological treatment and filtration for some of the runoff water. These features would increase the quality of water draining from the adjacent built environment (Hackensack Meadowlands Development Commission 1997).

DRAFT

36

2.1.1.2 Fisheries and Wildlife Habitat Improvement Built into the design were specific habitat features intended to make the site more desirable for wildlife. As previously mentioned, two brackish impoundments were constructed on the southern portion of the site. Along with the improved water quality benefits of these areas, the impoundments were expected to provide valuable roosting and foraging habitat for waterfowl and shorebirds, habitat conditions currently limited in the Meadowlands. Uplands were designed throughout the site to block wind and noise from the open water and emergent habitat areas. The uplands also served as on-site disposal sites for sediment removed during the excavation of the open water and emergent zones. Two islands within the impoundments were included in the design to provide nesting habitat for waterfowl and resting and foraging habitat for shorebirds. In one of the brackish impoundments, a least tern (Sterna antillarum) habitat island was planned. These state endangered birds are decreasing in numbers as a result of the disappearance of their nesting habitat. Least terns lay their eggs on isolated beaches with pebbles or shells on sand and short, sparse vegetation. They are easily preyed upon in this area by common crows, seagulls, and small mammals like Norway rats and raccoons. To create least tern habitat, construction was planned using spoils from excavation activities covered with sand and broken shells. Surrounding this area by water was expected to protect it from intrusion by small mammals. Previous attempts to deter crows and seagulls from entering least tern habitat have been largely without effect and have, in some cases, even deterred the least tern from laying eggs in the area. However, observations made in the Meadowlands area report least tern nests on the gravel roofs of many buildings, indicating that least terns will attempt to use all habitats made available to them. This island was planned to also serve as a high tide resting area for shorebirds. Other than provision of habitat for avian species, there was no specific purpose to the planned second island as there was in the case of the “least tern island”. Although the main method for vegetation of the restoration site was through self-design (Mitsch and Wilson 1996) with recruits coming from surrounding areas, some plantings in the upper elevations of the site were planned. Along with these upland plantings, the following are other special habitat features that were included in the original design (according to IVA, sites greater than 100 acres with 3 or more of these get higher scores): • • • • •

plants bearing fleshy fruit (e.g., cherry, persimmon) mast-bearing hardwoods (e.g., oak, beech, hickory) cone-bearing trees or shrubs native prairie (e.g., warm season grasses on uplands for nesting cover for waterfowl) exposed bars (e.g., mudflats)

The intent of these special habitat features was that wildlife would benefit from the dense planting of native woody plants, shrubs, groundcovers, and warm season grasses on the uplands created from the wetland excavation spoils. The proposed upland plants were selected for their adaptability to harsh sites with potential high salinity, their wildlife food value, and their wildlife cover value. For instance, numerous berry producing plants including, Sambucas canadensis (elderberry), Viburnum trilobum (cranberry bush viburnum), and Myrica pensylvanica (bayberry), were selected for the song bird population that they can attract. The song birds will in

DRAFT

37

turn introduce countless other plant species into the system. It is unclear what other species were planted, but the planting arrangement, groupings, and orientation were intended to mimic those commonly found in natural systems. The intent was to provide a familiar structure to wildlife. Limited supplemental maintenance was planned so that additional successional plants would self-establish. The installation of dead snags on the upland was designed to provide roosting and hunting perches for raptors which were expected to hunt the islands and low marsh areas of the restored site. 2.1.1.3 Recreation/Education Opportunities Improvement Public access to the Mill Creek site was planned to include a 1.5-mile long pedestrian trail network beginning at the Mill Creek Mall. This trail was to be an extension of another planned trail system which was supposed to link the site to neighborhood residential and commercial centers. In addition to providing a recreational amenity for area residents and workers, the trail system would also provide access to the site for ongoing monitoring and maintenance. As was the case with the islands, the trails were designed to make on-site use of the spoils expected to result from dredging of the channels and other areas of the site. The layout of the trail system was limited to the proposed contiguous upland portions of the site. The planned trail began at the Mill Creek Mall, winding through the site eventually forming two loops, providing access to the water’s edges and upland lookout areas. The first loop, circling a smaller impoundment, will be 0.5 miles long. The second, larger loop was planned to be 1.1 miles long, extending around the larger impoundment. Feature areas, including bird blinds, interpretive signage, and a canoe launch were planned to enhance the site’s potential recreational use. The trail surface was to be topped with fine red gravel, making the site unrestricted to those with disabilities. The width, slopes, and added features of the trail were also to be maintained to allow access to all individuals. 2.1.2 Baseline Studies Various baseline studies and site preparation activities were performed at the Mill Creek site to assist with the development and implementation of the mitigation design and monitoring plan. During the mid-1980s, the site was known as the “IR-2” site. This parcel was owned by Hartz Mountain and was the proposed location for approximately 2,750 units of low-rise, multifamily housing referred to as The Villages at Mill Creek. This development would have resulted in the destruction (i.e., filling) of approximately 110 acres of estuarine wetlands, and the alteration of an additional 49 to 54 acres of wetlands for the creation of canals, lagoons and enhanced marsh area (Hartz Mountain Industries 1985). Due to the permitting issues involved in a development proposal of such magnitude, many environmental and engineering studies were conducted on this wetland parcel (Jack McCormick & Associates 1978, Environmental Resources Management 1984, U.S. EPA and Gannett Fleming 1992). When the Hackensack Meadowlands Development Commission (currently the New Jersey Meadowlands Commission) purchased this parcel from Hartz Mountain, all of the background studies were provided to the Hackensack Meadowlands Development Commission (HMDC). The HMDC biologists and wetland restoration design staff reviewed all of this existing information prior to developing the wetland restoration design for the Mill Creek site. The review of existing information was completed by

DRAFT

38

the spring of 1997. The results of these studies showed that the sediments in the area of the proposed mitigation were at the same level as other areas typical in the Meadowlands. Because there were no factors limiting use of the site for the proposed mitigation, plans were made to proceed. Because of the age of the data provided by Hartz Mountain at the time of purchase, it was decided to collect a new set of data to better reflect up-to-date design objectives specific to the site. As part of the data collections, a topographic survey was performed; vegetation mapping was conducted; soil, sediment, and surface water samples were collected; avian and hydrologic studies were performed; and a baseline IVA was completed at the site. The IVA was specifically developed for the evaluation of wetland functions and values in the Hackensack Meadowlands for purposes of determining impacts and mitigation ratios (U.S. EPA and U.S. Army Corps of Engineers 1995). The assessment process involves answering numerous questions pertaining to the characteristics of the assessment area and its watershed. At the time, IVA was the accepted method for determining the relative impacts to wetlands within the Meadowlands District. Additional detail regarding the role of IVA in developing site selection criteria and design ideas is presented in Section 2.1.3. Information on the other baseline studies is presented in the following paragraphs. 2.1.2.1 Topographic Survey Along with current data on the tidal characteristics that exist on-site, accurate elevation data is critical for the proper design of tidal wetland restorations. Therefore, a topographic survey was completed in April 1997. Topographic data was obtained for the entire site on a 50-foot grid. This resulted in a topographic map for the site that displayed land surface elevations at 1-foot contours, with interpolated 0.5-foot contour lines. The adjacent Western Brackish Marsh, to the north of the Mill Creek site was also surveyed to aid in hydrologic analysis and for vegetation/elevation mapping, as Western Brackish Marsh served as the reference site for the Mill Creek mitigation design. Channel and emergent wetland cross sections developed during the design of the Mill Creek site restoration were based on the survey of land surface and vegetative community type elevations conducted in the Western Brackish Marsh. Field observations were correlated with elevations surveyed on several indicative cross sections. Fieldwork was performed to determine the relative elevation at which different wetland subclasses occurred in the Western Brackish Marsh, where mudflats were observed between +0.7 and +1.7 feet NAVD 88, and low marsh between +1.7 and +2.4 feet NAVD 88. Phragmites invasion at the Western Brackish Marsh was observed where elevations exceeded +1.5 feet NAVD 88. Unfortunately, this was the observed upper end of the mudflat elevation and precisely in the range of the observed low marsh. 2.1.2.2 Vegetation Mapping Mapping of baseline vegetation conditions was completed during June 1997. This information was used to document pre-mitigation conditions and was used in the IVA analysis. During the topographical survey and the vegetation mapping, the presence of a large quantity of Atlantic white cedar (Chamaecyparis thyoides) stumps and logs were found in the southeastern corner of

DRAFT

39

the site, covering an area of approximately 35-40 acres. Since site design was not yet completed at the time that the cedars were discovered, and because excavation in this area was expected to be difficult and costly, the preliminary design situated the two brackish water impoundments in the area dominated by the cedar stumps and logs. 2.1.2.3 Soil, Sediment, and Surface Water Sampling Considering the history of significant manufacturing activity in the Meadowlands, the analysis of soil and sediment samples for potential pollutants was conducted to provide baseline information for determining the possible extent and concentration of contaminant movement to the site from off-site sources (Hackensack Meadowlands Development Commission 1997). If red flags were raised due to highly elevated pollutant concentrations, mitigation at Mill Creek would not have proceeded. Soil and sediment samples were collected and analyzed for pH, priority pollutants + 40, and total petroleum hydrocarbons in April 1997. Soil and sediment samples were also collected and analyzed as baseline data on the on-site benthic macroinvertebrate community. Concurrent with the benthic invertebrate sampling, baseline surface water sampling for contamination was also conducted. Results of the analyses did not raise any concerns regarding pollution at the site; conditions were not atypical of those in the Meadowlands overall. As such, construction activities at the site were allowed to progress unimpeded once a formal design plan was approved. 2.1.2.4 Avian Studies Baseline avian studies were conducted at the Mill Creek site between August 1997 and February 1998. A total of five observation towers were constructed at the site to provide unimpeded views for monitoring. Each observation site was visited at irregular intervals prior to restoration. While data was collected for this work, none of it impacted the mitigation site design. 2.1.2.5 Hydrological Analysis During the initial phase of the restoration design, existing hydrological analyses performed during the earlier studies for the IR-2 site investigations, as well as existing tidal data available from NOAA were reviewed. Given the importance of understanding the tidal regime at the site to the success of the restoration design, however, the HMDC decided that it would be advantageous to collect new tidal data for the site, rather than rely on the analysis of tidal information that was 15 years old. Initial tide gauge installation attempts met with some difficulty as equipment malfunctioned. Therefore, tide gauges were not placed within Mill Creek until May 1998. According to its permit, construction had to begin by June 1, 1998. The HMDC collected continuous tidal data (at 30-minute intervals) in Mill Creek from May to September 1998. This data set was meant to be analyzed by the Woods Hole Group in order to provide a full column of tidal datums, based on the observed data. Although this site-specific, recently collected tidal data represented the best information for input into the restoration design, the permits for the restoration contained strict time deadlines regarding when the restoration construction had to begin. In order not to be in violation of the permit, construction had to begin by June 1, 1998, allowing only one month of tidal data to be collected. This tight deadline did not allow enough time between tidal data collection and the analysis of that data by the Woods Hole Group for the

DRAFT

40

data to be used in the restoration design. After the restoration construction was completed, it was discovered that the tidal elevation used for mean high water in the restoration design was approximately one foot lower than what was measured during the 1998 observed data. There has not been any study to assess the effect that the miscalculation of the high tide level has had on the post-mitigation condition of Mill Creek. 2.1.3 Role of IVA and Other Functional Assessment Methods The IVA was specifically developed for the evaluation of wetland functions and values in the Hackensack Meadowlands (U.S. EPA and U.S. Army Corps of Engineers 1995). It is a semiquantitative wetland assessment methodology developed as a tool to assess functional values of wetlands within the Hackensack Meadowlands. The method assigns functional index scores, relative to the best performing wetlands in the area. The IVA method is derived from the Wetlands Evaluation Technique (WET) that was completed for the Meadowlands’ wetlands in the mid-1980s (Maguire Group 1989). The method was designed to assess a large number of wetlands in the Meadowlands for use in making management decisions. For the IVA method, questions that were deemed to be unnecessary were culled from the long list of WET questions, leaving a list of 196 indicators (Hruby et al. 1995). As such, many of the questions used in the Meadowlands IVA directly reference their corresponding WET number. Once the WET questions pertaining to the characteristics of the assessment area and its watershed have been answered, scores for each function are determined. Each indicator was assigned to the function(s) that it was deemed to be relevant to. In IVA methods, indicators can be additive, multiplicative, or fractional and greater weight can be given to indicators that are deemed to have a stronger impact on a given function (Hruby et al. 1995). Function scores are then normalized by dividing each wetland’s score by the maximum score attained by any wetland in the region and multiplying by 100 so that all scores ranged between 0 and 100. Value can also be assigned to each wetland by multiplying its IVA score for each function by the wetland’s area to give regulators an idea of the amount of functioning that will be lost by filling or gained by mitigation. Thus the Meadowlands IVA produces a relative score of the value of ten wetland functions for each assessment area and these functions are grouped into three functional groups as follows: Social significance: Recreation Floodflow alteration Conservation potential Wildlife habitat:

Aquatic diversity and abundance General fish habitat General waterfowl habitat General wildlife habitat Export of primary production

Water quality improvement:

DRAFT

Nutrient retention/transformation Sediment toxicant retention

41

At the time of the NJMC mitigations, this version of the Meadowlands IVA was the accepted method for determining the relative impacts to wetlands within the Meadowlands. The method was later revised and used to some degree but has fallen out of favor. Being specifically tailored for use in the Meadowlands, the IVA was the only functional assessment method used in the development of the design of the Mill Creek site. It helped to steer the design in several areas, in order to maximize improvements to wetland functions. For example, if the IVA method attributed higher scores to open water than emergent wetlands, the design reflected this by increasing the area of open water compared to emergent marsh. Specifically, IVA had an impact on the design in the following areas; the ratio of open water to emergent marsh areas, interspersion of vegetation, wetland/upland edge, water depth, and special habitat features, as detailed below. For example, because wetlands with a ratio of 61:39 (open water: emergent area) or greater, are awarded higher point values in IVA assessments, the ratio between open water and emergent subclasses was designed to be approximately 70:30 for the Mill Creek restoration. Several questions pertaining to vegetation appear in the IVA evaluation and were addressed in the HMDC design. The horizontal pattern of vegetation classes is addressed by looking at vegetation class interspersion. Because of the dense Phragmites monoculture dominating the site prior to mitigation, the baseline condition was considered a relatively homogeneous area of a single dominant species. A mosaic pattern was planned for the final design, in which relatively small areas supporting different vegetation classes were to be interspersed throughout the site. In the interspersion, vegetation form richness was emphasized by introduction of four or more vegetation classes or at least eight vegetation subclasses (e.g., forest, scrub/shrub). HMDC strove to achieve a mosaic of relatively small patches of vegetation interspersed with open water impoundments, channels, and mudflats. An effort was made to design an irregular margin between the upland and wetland zones of the site to open up more habitat edge; it was hoped that this would directly benefit foraging waterfowl. Vegetation was used to create wind sheltered open water and emergent areas throughout the site. The incorporation of special habitat features, such as plants bearing fleshy fruit, mast-bearing hardwoods, cone-bearing trees or shrubs, native prairie (e.g., warm season grasses on uplands for nesting cover for waterfowl), and exposed bars (e.g., mudflats) were incorporated into the design. Mitigation sites greater than 100 acres in area that display three or more of these features were awarded high point values in IVA assessments. Dominant water depth of a wetland was addressed in IVA in that at least 50% of tidal wetlands should be flooded to at least 21 inches at high tide. The HMDC attempted to maintain this inundation depth in the Mill Creek design. 2.1.4 Review of Permits for Compensatory Mitigation Between 1994 and 1997, the U.S. Army Corps of Engineers (ACOE) and the NJ Department of Environmental Protection (NJDEP) issued various permits and approvals to four agencies and/or companies which authorized the filling of a total of 31.8 acres of wetlands. As a condition of the permits which allowed the wetlands to be filled, all four of the permittees were required to perform compensatory wetlands mitigation (totaling 96.8 acres) as a result of unavoidable

DRAFT

42

wetlands impacts. The four permittees include the New Jersey Department of Transportation (NJDOT), the New Jersey Transit Corporation (NJTC), Jet Aviation (JA), and The Port Authority of NY & NJ (PA). In order to resolve outstanding permit compliance issues, starting in July 1996 with NJTC, and followed later by the remaining three permitees, each provided HMDC with sufficient funding to carry out the required mitigations. These mitigations culminated in the restoration plan for the Mill Creek site. In addition to the required mitigation of the four permittees, in August 1998, the ACOE issued another permit which authorized additional acreage for an “advance mitigation” area to be constructed within the Mill Creek site. Portions of this advance mitigation parcel were used to provide acreage for mitigations required due to the filling of wetlands that were authorized by the ACOE and/or NJDEP for an additional seven permittees. Although the Mill Creek site ultimately provided wetland compensatory mitigation for wetland fills resulting from 12 different projects involving 11 different permittees, the design of the site was done in a holistic manner. From the outset, the wetland ecologists and engineers used the information gathered during the baseline studies to design a wetland restoration for the entire approximately 137-acre site. It was only after the design for the entire site was finalized, that the apportionment of the site into parcels of appropriate size to meet the conditions of the various permittees was done. As such, permit requirements for compensatory mitigation played no role in the design of the mitigation project. Additionally, as the permits did not contain detailed language regarding design or monitoring requirements, or any other success criteria, the permits did not play an important part in the initial site design. (The only requirement found in the permits was a requirement for 85% vegetation cover.) Carving the entire site into smaller areas for each of the permittees was performed only after the site design was already finalized. This was done only to show which portions were “set aside” to satisfy the mitigation requirements for each permittee. These boundaries were overlaid onto the design drawing, and later onto the asbuilt map of the site. These boundaries were never surveyed in the field. A table summarizing the compensatory mitigation that was satisfied by the Mill Creek site is shown in Table 2.1.4-1.

DRAFT

43

Table 2.1.4-1. Mill Creek Permit List of Compensatory Mitigation. PERMITTEE New Jersey Transit New Jersey Transit

Advance Mitigation

New Jersey DOT

Route 120 Relocation

New Jersey DOT

Advance Mitigation Hanger Facility at Teterboro Airport

Jet Aviation Port Authority I Port Authority II Port Authority HNSE Rutherford Lodging Conrail/CSX

PROJECT Secaucus Railroad Transfer Station

Runway Improvements at Newark Airport Parking Structure Improvements Advance Mitigation Repairs to Claremont Terminal Parking Lot Development

HMDC

Rail Line Connection Allied Project Site Development Remediation Project along the Passaic River HMEC Entrance & Visitor Center Egress

JCMUA

CSO Project

AlliedSignal CLH

PERMIT NUMBERS 93-03412 (Corps) 0909-92-0003 (NJDEP) 93-03412 (Corps) 0909-92-0003 (NJDEP) 0205-960001.4(WFD/WQC/SE) 96-01160(Corps) 0205-960001.4(WFD/WQC/SE) 96-01160(Corps) 0237-90-001.4(WQC) 94-05821-RS(Corps) 97-09200(Corps) 0000-900022.14(IP/WQC) 0000-900022.20(IP/WQC) 97-09150(Corps) 0909-910001.5(WFD/WQC) 0906-96-0001.5(WFD) 0256-98-0004.1(WQC) 98-00620(Corps) 0249-93-0003-4(WQC) 98-21290-J2(Corps) 0212-95-0001.8(WFD) 0212-95-0001.9(GP4) 0714-99-003.2(WFDE) 99-14310(Corps) 0232-90-0001.7(WQC) 0906-99-0008.2(SE) 0906-99-0008.5(WQC) 0906-990008.6(WFD/WQC)

WETLAND IMPACTS 13.14

ACRES REQUIRED

ACRES MITIGATED

45

45 6.91

8.73

30.55

30.55 2.72

0.24

0.48

0.61

6.82

14

14

1.5

4.5

4.5 4.772

0.6

0.6

0.6

0.004

0.008

0.008

0.06

0.12

0.12

0.38

0.76

0.76

0.08

0.16

0.16

0.105

0.277

0.277

0.15

0.3

0.3

2.1.5 Design Objectives and As-built Conditions In 2001, a review of the wetland mitigation plan (i.e., the design drawing, dated September 16, 1997) and the as-built map (dated June 1, 2000) was performed to determine if the design objectives had been met (TAMS Consultants 2001). This was performed to determine if the design drawing matched the as-built map of the site. The two drawings were overlaid and a comparison of the features was made. Generally, the as-built map closely matched the wetland mitigation plan. The two impoundment areas were built in the designed locations as well as to the designed areal extent and elevations. The emergent fringe/mudflats were designed and built to surround the interior of the impoundments as well as along the tidal channels. The as-built

DRAFT

44

shows that the channels were built in the proposed locations and elevations; moreover, the configuration with respect to the interconnections with other channels, closely matches that shown in the mitigation plan. Two islands were originally designed in the mitigation plan – one in each impoundment. However, only one (the island designed to attract nesting least terns) was actually constructed. This island was built in the northern impoundment, as intended, and in a generally comparable configuration and elevation. The enhanced uplands, which include the walking trails, were constructed in the proposed locations as shown on the mitigation plan. Although the Mill Creek site was designed as a holistic wetland mitigation, the monitoring required by each permit issued by the ACOE and NJDEP did not allow the HMDC to examine the site as a whole. Each permitted area had to be examined and reported on as a separate entity. In order to meet the permit requirements, the area apportioned to each permittee was examined by itself, without any consideration of the adjacent portions of the site or how the entire site functions within the larger context of the Mill Creek watershed or the greater Meadowlands system. For that reason, the review of the as-built drawing to the restoration design drawing was done for each of the smaller areas covered by each of the various wetland permits that were issued for this site. A comparison of the acreages of uplands and wetlands as depicted on the mitigation plan (i.e, the design drawing) and the as-built drawing, is provided in Table 2.1.5-1. Table 2.1.5-1. Mill Creek Mitigation As-built Acreages. Mitigation Permittee

Proposed Acreage (ac.) Total Upland Wetland

As-built Acreage (ac.) Total Upland Wetland

New Jersey Transit New Jersey DOT Jet Aviation Port Authority of NY&NJ Mitigation Bank* Totals

59.5 43.8 1.8 19.6 14.3 139

63.3 42.4 1.8 18.7 14.3 140.5

Mitigation Bank Allocations HNSE Port Authority Rutherford Lodging Conrail/CSX AlliedSignal CLH JCMUA

14.5 13.3 0.8 5.6 3.5 37.7

45 30.6 1 14 10.8 101.4

7.6 6.1 0.8 4.9 2.9 22.3

55.7 36.3 1 13.8 11.4 118.2

Wetland Surplus/ Deficit (+/ac.) +10.7 +5.7 0 -0.2 +0.6 +16.8

0.6 9.272 0.008 0.12 0.76 0.16 0.3

* Mitigation Bank is the name given to what were the unallocated acreages of the Mill Creek site. Several entities have subsequently purchased mitigation acre-credits from this area.

This comparison reveals that except in the case of Jet Aviation and the Port Authority of NY & NJ, more wetlands were restored than what was anticipated by the design drawing. The reason

DRAFT

45

that additional wetlands were created was due to the fact that the actual amount of material dredged from the channels and/or marsh plain was less than what was calculated during the design process. This meant that there was less dredge spoil to be disposed of. The design called for disposing of dredge spoils on-site, to be used in the creation of upland trails and islands. The shortage of dredged materials resulted in only one of two planned upland islands being constructed. The shortage of materials also caused a change in the design relative to the upland walking trails. The lack of material with which to construct the pedestrian trail resulted in two “low areas”, one adjacent to each impoundment. Two boardwalks were later built to span these low areas, which were expected to act as “emergency” spillways should higher than normal tides overfill the impoundments. The lack of spoils for the trail system, coupled with the actual level of mean high water being approximately one foot above the design elevation caused the two spillways to erode. This erosion made the water control structures that were designed to allow for water level management within the impoundments obsolete, as both impoundments are currently subject to the tide. The mitigation requirements varied somewhat from one permit to another. In general the permits contained some deed requirement to ensure that the mitigation site was conserved as wetland. These requirements were met through the fact that the mitigations took place on property owned by the NJMC. The permits assumed that planting would occur and required 85% survival of these plantings. However the NJMC’s design plan for Mill Creek did not require initial plantings and this design was approved by the permitting agencies. The permits also required 85% areal cover of hydrophytic plants by the end of the second growing season. As the most specific and quantifiable requirement, this became the focus criterion for measurement of project success. If the 85% cover requirement or other requirements such as wetland acreage, as-built compliance with design plans, provision of wildlife habitat, floodwater retention, and water quality enhancement were not met rectifying action was required until the project could be deemed successful. The permits required three to five years of monitoring, though monitoring could be extended if the site was not in compliance at the end of the period. Some of the permits also specified vegetation monitoring methods (10 x 10 foot plots employed in each habitat type with relative frequency and cover measured for each species, vegetation mapping once per year, and photographs) – these specifications were incorporated into the vegetation monitoring program.

2.2

Monitoring Objectives, Protocols, and Implementation

2.2.1 Requirements of the Memorandum of Understanding Monitoring under the MOU was conducted for Mill Creek from 2000 through 2004. According to the MOU (Hackensack Meadowlands Development Commission 1998), “The criterion for success [at each of the sites was 85% relative vegetative] cover of target (non-Phragmites) vegetation within two years post-construction.” General goals for the marsh (from the mitigation plan) also included control of Phragmites; restablishment of tidal flow; creation of open water areas; creation of dabbling duck, shorebird, and wading bird foraging, breeding, and overwintering habitat; greater fish access; and mosquito control. As such, monitoring of various wildlife and other aspects of wetland function were DRAFT

46

stipulated in the MOU (see below). However, aside from wetland area, the 85% cover of wetland species criterion was the only specific, easily measurable criterion listed in either the permits or the design plan; it thus became the focus of monitoring and determination of whether the mitigation was a success. Since only two years were given to reach this criterion, much of Mill Creek had not achieved this compliance criterion by that time and planting with Spartina alterniflora commenced. In subsequent years, however, it was evident that these deficient areas were generally increasing in relative cover of target species. By the last year of monitoring, the vast majority of the site had reached the 85% criterion; however, there were still areas that had not met the compliance criterion. Specifically, four permitted areas were deficient in relative coverage. Within these permitted areas, a total of five monitoring plots were deficient in coverage, including two in the low marsh and three in the high marsh. It must be noted that relative cover in these areas was generally increasing. No habitat zones had a relative cover of less than 62.5% in 2004. Mill Creek was designed with impounded water and tidal channel habitat, created by excavation of the Phragmites monoculture. As Phragmites was removed from the sites, site elevation was lowered due to removal of some of the Phragmites root zone. Tidal flushing was restored throughout the Phragmites-dominated marshes with the goal that natural colonization by wind and water-dispersed native salt marsh species, such as Spartina alterniflora, would inhibit reestablishment of Phragmites. To document vegetation change and evaluation of “success” of the mitigation in terms of compliance with jurisdictional requirements, annual monitoring reports were submitted to the NJMC from 2000 through 2004. These reports included documentation, such as vegetation cover maps, and data tables detailing analysis of changes in vegetative species and cover within permanent monitoring plots. 2.2.1.1 Flora “The criterion for success is cover of target (non-Phragmites) vegetation two years postconstruction,” as stated in the MOU. Additional stipulations for monitoring of flora include: • • • • •

species, relative frequency, and percent cover determined within each habitat using 50+ 10’x10’ plots and a systematic stratified sampling design; monitor in mid-June & August vegetation cover maps (1” is > 50’) prepared for each season photographs of all representative areas taken at least once each year between June 1 and August 15 at least 85% vegetation survival or cover by the end of the second growing season (if not met the NJMC was required to plant and/or re-grade as necessary to reach the required percentage by year 3) vegetation changes monitored with GIS and composite maps would be built showing trends in vegetation change (e.g., Phragmites to Spartina); base map would be created from aerial infrared photographs; post-construction maps would be made from aerial photos

Additionally, the MOU states that “If significant differences in vigor between plots or species become apparent, biomass measurements will be added to the sampling protocol. Plant tissue analysis and insect activity will also be surveyed in this situation.”

DRAFT

47

2.2.1.2 Avian The MOU requires documentation of the use of the restored marsh by nesting and migrating birds. This was to be accomplished by field-based methods, noting the following: • •

breeding and migratory bird populations using line transect sampling (established at 250meter intervals) Weekly monitoring during spring migration and summer breeding seasons (May 1 to July 15) and during fall migration (Sep 1 to Oct 15); otherwise at 3-week intervals

The observer was to walk along the transect and record the number of individuals seen or heard by species. Also, the distance and orientation to each individual, and the vegetation type was to be recorded. 2.2.1.3 Geomorphology/Hydrology/Salinity The following measurements were to be conducted four times during each growing season: • • •

elevation, tide height, and pore water salinity were to be measured in patches representative of the major vegetation types with at least three patches per vegetation type and five random points per patch tide height measurements (and elevation) were meant to be obtained with dyed wooden lathes (as the tide washes away the dye, it leaves a mark of the maximum tide height with elevation determined from tide height measurements) pore water salinity was to be measured using PVC sippers with intake ~7cm below surface

2.2.1.4 Sediment and Soils Soil structure was to be monitored three times after construction: immediately after construction, two years post-construction, and five years post-construction. All other tests were to be performed on an annual basis. • •

sediment pH and redox potential were to be measured with a portable meter the Rutgers Soils Laboratory were to perform the following tests: 1. soil structure (for sand, silt, and clay content) 2. organic matter determination by loss on ignition 3. soluble salts (determined by electrical conductivity) 4. standard fertility (including P, K, Ca, Mg, Cu, Mn, and Zn) 5. Kjeldahl nitrogen

2.2.1.5 Contaminants Concentrations of contaminants in the sediment, uptake by vegetation, and mobilization of contaminants were to be measured during the monitoring period. Intensive monitoring was to DRAFT

48

take place during the first year of sampling in order to identify patterns of contamination. During the second through fifth years, monitoring was to be conducted at permanent sites representing the range of conditions found in year one. Mobilization of contaminants was to be monitored through repeated sampling over time. The following measurements were to be taken: • • •

sediment cores were to be taken from all vegetation types and from creek beds concentrations of heavy metals and organics (including pesticides and PCBs) were to be measured, with a determination of relative concentrations of mobile and immobilized forms of heavy metals uptake of contaminants by vegetation required collection of plant tissue samples from the same sites as the sediment cores with roots and shoots analyzed separately

2.2.1.6 Aquatic Animals The following measurements were to be conducted periodically throughout the monitoring period: • fiddler crab populations were to be estimated according to Bertness and Miller (1984), using 0.25 m2 quadrats randomly placed at 10-meter intervals in four veg zones (tall & short S. alterniflora, Phragmites, and S. patens) (this was to be performed annually) • fiddler crab burrows were to be counted to estimate crab density with select burrows excavated to determine species composition (this was to be performed annually) • benthic invertebrates were to be sampled using sediment cores with specimens preserved (this was to be conducted annually between the months of July through November) • fish species were to be sampled from two discrete zones: subtidal creeks (sampled with seines at low tide) and the intertidal marsh (sampled with pit traps to determine if juveniles were using the marsh for feeding); sampling was to be conducted annually during the growing season. 2.2.1.7 Insect Communities The insect community was to be sampled using black-light traps, yellow-pan traps, yellow sticki strips, and pit fall traps. Sampling was to be conducted in the upland and emergent marsh areas 3 to 4 times per growing season. 2.2.2 Development and Implementation of Monitoring Protocols While monitoring of various biotic and abiotic conditions were required as stated in Section 2.2.1, reporting to regulatory agencies was only required for flora. The success criterion of 85% relative cover of non-Phragmites wetland vegetation was the regulatory goal of the mitigation. Monitoring for other biotic and abiotic conditions was performed but not annually reported as was the case with vegetation. 2.2.2.1 Flora Vegetation change was documented in annual monitoring reports prepared from 2000 through 2004. To show progress towards the 85% criterion, vegetation monitoring plots were established DRAFT

49

in the various vegetation zones present at the site with multiple replicate plots established in each zone. Annual mapping of vegetation zones was also conducted to quantify changes in acreage of the different vegetation zones. 62 permanent plots were established in 2000 after mitigation construction was completed. Based on flooding frequency and elevation, 15 plots were established in the channels/mudflats, 17 in the low marsh, 15 in the high marsh and 15 in the upland portion of the Mill Creek site. An additional eight channels/mudflats plots were established in 2002, however, one of the original 62 plots was lost permanently, giving a total of 69 plots (61 original + 8 additional) used for monitoring purposes at the Mill Creek site. From the monitoring plot data, species, abundance, and percent cover were collected. The plots were 10 ft2 in size with subplots of 1 ft2 giving a total of 100 subplots. The data was later used to derive numbers for species richness, frequency, relative frequency, and relative cover. For all years of monitoring, data was collected twice per year: in the late spring and again in the summer. At each sampling event, photographs were taken from permanent photostations designated during plot establishment. Vegetation cover maps were produced using a mixture of color infrared aerial photographs, digital aerial photographs, and detailed field observations. Field surveys were conducted in the mid to late growing season. Baselines and transects were used for field measurement and identification of vegetation zones. These baselines and transects were also mapped by acquiring GPS points. From field data and interpretation of aerial photographs, vegetation cover polygons were created in a geographic information system (GIS). Acreage and final layouts were calculated and produced in the GIS. These protocols match well with the requirements stipulated in the MOU and described in Section 2.2.1.1 of this report. Because the 85% target vegetation coverage was not reached by the end of the second growing season, the NJMC initiated a Spartina alterniflora planting program for the low elevation areas of the site, as required by the MOU. 2.2.2.2 Avian Point-count survey stations like those used to obtain the baseline avian data were used rather than transects. Instead of walking along a transect, observers would situate at these stations and observe birds quietly from a fixed position. An observational radius was estimated from each station and no direction was noted from observer to individual bird. Avian surveys consisted of the sum of counts from all survey stations. Surveys were performed between sunrise and 10 am and were not conducted in rain or heavy wind. Birds flying over the marsh were generally not recorded, except for foraging raptors and aerial insectivores, which hunt while flying. Postmitigation monitoring was conducted year-round, with surveys typically conducted once every week or two depending on season and weather conditions, with fewer surveys being conducted during the winter months. Individual birds were identified to species whenever possible. Rather than noting vegetation type in general, the habitat type being used by each bird was noted. The birds were then classified into foraging guilds in order to characterize habitat use by the community.

DRAFT

50

Although there are some differences in the protocols used in the field and those stated in the MOU (see Section 2.2.1.2 of this report), they are insignificant as the field methods capture the same data. 2.2.2.3 Geomorphology/Hydrology/Salinity Although there were a few periodic monitoring terms described in the MOU (elevation, tide height, and pore water salinity), only one of them, elevation, was successfully performed after construction of the Mill Creek site. An as-built site survey was performed after construction where elevation measurements were collected. Tide height and pore water salinity measurements were not collected at any time after construction was completed as stated in the MOU (see Section 2.2.1.3 of this report). Hydrologic monitoring was attempted soon after construction but failed due to equipment problems. No other attempt was made to monitor hydrology. 2.2.2.4 Sediment and Soils While soil structure measurements were to be monitored three times after construction (immediately after, two years after, and five years after) and other sediment and soil metrics were scheduled to be measured annually (see Section 2.2.1.4 of this report), none were monitored on a periodic basis. Soil samples were, however, collected in summer 2005, along with soil pH measurements. Percent moisture, percent organic matter, soil color, and soil texture were determined. Results can be found in Appendix C. 2.2.2.5 Contaminants Although contaminant concentrations were to be measured throughout the monitoring period (see Section 2.2.1.5 of this report), there was only one instance of sampling. As described by Feltes and Hartman (2002), in August 2001, soil and sediment samples were collected from 20 sites at Mill Creek. Samples included 8 low-marsh/mudflat samples, 8 open water samples from impoundments (usually covered by water and at the edge of water cover at low tide), and 4 upper-transitional samples (vegetative transition zones on the slopes of uplands, around the highest tide water marks). The occurrence of contaminants in the sample sites was organized by chemical group and arranged into ecological ranges (i.e., the ecological range or ER) in order to estimate the probability that effects from each specific contaminant would occur. If a sample concentration exceeds the ER median (ERM), effects will frequently occur >50% of the time. If a contaminant level occurs below the ER low (ERL), the effects would rarely be seen,