Restoring New York City
Proposals for Improving Ecological and Human Health
Edited by Dr. James A. Danoff-Burg
Department of Ecology, Evolution, and Environmental Biology, Columbia University

Projects Available
Harlem River Yard, Bronx, NYC

Project Home

Listed By Borough
Listed By Ecosystem

Bayside Acacia Cemetery
Central Park's North End
East River Oyster Beds
Fresh Creek Marshland
Governors Island
Harlem River Yards
Inwood Forests
Inwood Marsh & Nature Center
Manhattan Forests
Manhattan Marsh Re-creation
Manhattan Maritime Holly Forest
Northern Manhattan Forests
Operation Renovo Gardens
Rockaway Beach
Time Capsule NYC
West Harlem Marshlands


Harlem River Yard Restoration Plan

Matthew Foster




The restoration of degraded habitat, especially in urban areas, is as an important component in the stewardship of urban ecosystems.  Harlem River Yard is a 96 acre site in the Bronx area of New York City that has been degraded through many years of industrial use.  Although much of the site is utilized for a solid waste transfer station, the remainder of the site is undeveloped and environmentally degraded.  It is envisaged that the site be restored by removing non-native vegetation and planting native species, with the aim of improving the ecological function of the area, enhancing aesthetics and creating a stepping stone that will ecologically connect surrounding parkland sites.  The restoration will be completed for three different areas including a stand of eastern cottonwood trees, an area of native grass and a salt marsh area.


Harlem River Yard                                                                              (

Table of Contents


Title Page……………….............………………………..…………….     

Abstract ……………….........................……………………….………     

Table of Contents ……………..........………………………….………    


Site Description and History…………….….......................….………     

Rationale for Restoration……………….............................….………     

Restoration Plan ……….........................….......................….………      

Restoration Areas - Species Selection and Site Layout.....…...…....     

Evolutionary Ecology and Population Genetics ………....……..…....     

Metapopulation Dynamics…………………..………...…………..…....    

Community Ecology………………..…............…….......………..…....    

Abiotic Filters .........………………..…............…....……………..…....    

Measuring Success……………..…..…...............………………..…....   

Time Line ................………………..….............…………………..…....  

Funding and Budget.…….......……………....…………………..…….     





Text Box: Figure 1: Location of Harlem River YardThe Harlem River Yard (HRY) is a 96 acre site located in the Mott Haven area of the Bronx, New York City.  The site is owned by the New York City Department of Transportation and leased to Harlem River Yard Ventures, a private developer.  Use of the site includes a 13.9 acre solid waste transfer station known as the Harlem River Yard Facility.  In the late 1990’s, a 28 acre intermodal transfer facility was developed at the site as part of the New York City Solid Waste Management Plan (SWMP).  However, a large segment of the property is undeveloped and under environmental stress (SWMP FEIS, 2005).  In keeping with the philosophy of the New York State Governor’s Hudson River Estuary Action Agenda, it is envisaged that the part of the property not utilized by the SWTS and intermodal facility be restored and maintained as a green space.  The area would be restored in the sense that native plantings that resemble the known original vegetation of the area would be utilized with the goal of creating a buffer between the Harlem River and the facility.  This proposal discusses the advantages of such an undertaking and also examines the problems that the restoration may face in the context of the environmental degradation of the site.  Scenarios for the selection of native and non-native species important in the ecological function of the site are also discussed.


Site Description and History


HRY is sited on the north bank of the Harlem River, with a river frontage of approximately 0.6 miles (Figure 1).  The northern edge of the facility is on East 132nd Street, where rail tracks, an access roadway and the rail/road intermodal transfer station are sited.  The eastern end of the facility runs under the Willis Avenue Bridge.  Historical data for nearby sites suggest salt marsh was a common ecosystem for the Harlem river and Spuyten Duyvil Creek (Loeb, 1986), so it may be inferred that the HRY was also originally salt marsh.  The Harlem River Yard was constructed by the New York, New Haven and Hartford Railroad Company during a period of expansion beginning in 1887 (Condit, 1980).    Historic photographs (Figure 2) show the rail yard in 1900, flanked by gently sloping river banks and low forest.  However, extensive filling of the banks has taken place since that time, significantly altering the area (SWMP, FEIS, 2005).  A photograph from 1959 (Figure 3) shows the area heavily built, with hardening of the banks throughout the site.

Text Box: Figure 2: Harlem River Yard c.1900 (Lehman College CUNY)During the construction of the rail/road intermodal facility, areas of the site were covered with fill and paved to contain contaminants discovered during hazardous waste assessments (Fiteni 2001).  Contaminants discovered included abandoned underground storage tanks (UST), lead and other metals, and polyaromatic hydrocarbons (PAHs), apparently from coal and ash (Fiteni 2001).  Non-native grasses maintained by mowing have been utilized in unpaved areas.  There is also a Text Box: Figure 3: Harlem River Yard c.1959 (Flagg, 2000)stand of trees including tree-of-heaven (Ailanthus altissima) and eastern cottonwood (Populus deltoids) present (SWMP FEIS, 2005).


Rationale for Restoration


The Harlem River Yard Waste Transfer Station forms an integral part of the New York City Waste Management Plan and the company managing the site has been given a ninety-nine year lease.  It is therefore unlikely that a large component of the site will ever be used for anything but industrial purposes.  However, the New York State Governor’s Hudson River Estuary Action Agenda states that river and shoreline habitats should be enhanced, biodiversity & habitat diversity be conserved, and Hudson River tributaries be restored.  It therefore seems reasonable that the unutilized foreshore portion of the Harlem River Yard be restored as nearly as possible to native habitat.  Additionally, the site lessee has been in breech of the lease by failing to fulfill its obligation to develop the site, possibly giving a legal basis for pursuing alternative development projects such as the proposed restoration.

Restoring even the foreshore portion of HRY would result in numerous aesthetic, habitat health and environmental services improvements.  The aesthetic improvement would consist of the visual screening the HRY Waste Management Facility from the nearby Randall’s Island Park by maintaining and enhancing the existing stand of trees on the south side of the waste management facility building.  If possible, plants used in this area and along the foreshore would also be selected for their potential in creating a natural line of phytoremediation defense for any possible waste leakage from existing contaminants.

The river edge would be restored, in so far as is possible, by removing non-native grasses and planting native semi-aquatic herbaceous plants with the aim of creating natural stabilization and enhancing habitat for aquatic fauna.  Another important area for action is treatment of the site for compaction.  The US Army Corps of Engineers (USACE) Critical Restoration Implementation Plan for the bordering lower Hudson River also calls for a softening of banks and the restoration of wetland where possible.

            In addition to aesthetic enhancement and improved environmental services, the restored habitat at HRY would form part of the urban forest described by Dahlin in the Crotana Park Urban Forest Management Plan (Dahlin, iii).  HRY would form an important link between nearby greenspace areas including the previously mentioned Randall’s Island Park, Highbridge Park, Pulaski Park and St Mary’s Park and many others.

Restoration Plan


Setting and meeting goals is one of the key factors in the development of a successful restoration project.  As already stated, one of the main goals for this project is the replacement of non-native species with native species.  However, there are several possible scenarios for the final appearance of HRY, each presenting different advantages and problems with regard to final appearance, utility and cost.  Table 1 summarizes possible scenarios along with advantages and disadvantages of each.



Plant types

Examples of Plants Used




Non native herbaceous plants

Pennycress (Thlaspi caerulescens)

Low cost, potential for phytoremediation including heavy metals

Potential to be invasive in surrounding areas.  Does not meet stated goal of screening site structures and using native species


Native herbaceous plants

Salt marsh Cord grass (Spartina alterniflora)

Relatively low cost, in keeping with stated goal of using native plants

Does not meet stated goal of screening site structures


Non-native trees

London Plane (P. × hispanica)

Extant in surrounding areas, lower potential to introduce diseases etc. to metapopulation

Does not meet stated goal of screening site structures and using native species


Native trees

Eastern Cottonwood (Populus deltoids)

Meets goal of using native species.  Cottonwoods have phytoremediation potential for TCE

Trees not appropriate for all areas of site, e.g. river edge


Combination of 1&2

As above

Low cost, closer to extant communities in surrounding area

Does not meet stated goal of using native plants


Succession of 1 to 2 & 4

As above

High potential for phytoremediation, will eventually meet stated goal of using native plants

Succession may be difficult to implement


Combination of 2&4

As above

Meets goal of using native plants

Lower potential for phytoremediation

Table 1: Summary of restoration scenarios for HRY


The most appropriate restoration scenario described in Table 1 is scenario 6.  However, due to the ongoing presence of contaminants at HRY, the large scale planting of large trees would conflict with the remediation strategy currently in place.  Therefore, the best option will be the maintenance and enhancement of the existing cottonwood stand and the replacement of non-native grasses with native grasses in the landscaped areas of the site.  The next section describes details of species selection and propagules for the envisaged plan and is followed by a summary of several theoretical issues in context of this plan.


Restoration Areas - Species Selection and Site Layout


The HRY restoration area is divided in to three main areas as shown in Figure 4.  The areas were conceived based on the main habitat type; however the climax point for each area will depend to a certain degree upon how each area is colonized during the restoration process.


Figure 4:  Restorations areas.  Site 1, eastern cottonwood stand; Site 2, native grassland; Site 3, salt marsh


Area 1:  Consists of the existing cottonwood (Populus deltoids) stand, which will be maintained and expanded in the future if possible.  Due to the size of this stand, development will depend on a number of issues such as interaction with surrounding populations.  These issues are discussed in greater detail in the following sections summarizing ecological theory.


Area 2:  Envisioned as a native grassland area.  Grasses in this area would initially be selected from those appropriate for phytoremediation such as switch grass (Panicum virgatum), followed by the introduction of native grasses such as Pennsylvania sedge (Carex pensylvanica) and Sideoats grama (Bouteloua curtipendula).  Propagules for this area may be sourced from commercial suppliers


Area 3:  The former barge docking area, which is currently silted to a shallow depth, would be further filled to create a salt marsh habitat. The suite of grasses and wet land plants most appropriate for this area include salt marsh cord grass (Spartina alterniflora), saltwater cord grass (Spartina cynosuroides) and two species of bulrush, water hemp (Acnida cannabinus) and salt marsh bulrush (Scirpus robustus), marsh elder (Iva frutescens), swamp rose-mallow (Hibiscus palustris), and groundsel bush (Baccharis halimifolia).  This list of species is extant in the salt marsh areas of Shorakapok Preserve area of Inwood Park located approximately 4.5 miles from the HRY site.  Some of these species including cattail and bulrush also have rhizoremediation potential (Dietz and Schnoor 2001).   Inwood Park may be a potential source of propagules for this area.


Evolutionary Ecology and Population Genetics


Current research suggests evolutionary pressures on a population should be considered in the formulation of a restoration project due to the importance of such pressures in the long term viability of the population (Stockwell et al., 113).  From the classic Darwinian evolutionary perspective, the existence of heritable traits and the action of selective pressures on those traits will cause the loss of individuals less well adapted for the subject habitat and a resulting selection for the best adapted population.  Traditionally, this process has been understood to take place over long time periods, often in the geologic time scale.  In terms of restoration projects, however, there is a need to consider what has been termed “contemporary evolution” (Stockwell et al., 113).  Contemporary evolution is the action of evolutionary forces on a much shorter time scale, and for the restoration ecologist, this means that what has been determined as the existing set of pressures facing the species of interest may alter as the environment changes during the restoration process.

In the HRY project, restoration would alter many evolutionary agents currently found at the site, thereby changing the fitness of the existing eastern cottonwood population.  Examples of evolutionary agents that might be changed include soil pH, soil composition and water content, type and populations symbiotic soil bacteria and changes in the shade regime due to the introduction of other plant types.  For example, using non-native herbaceous plants for bioremediation may alter the abiotic nature of the site.  Even if members of populations originally found at the site are maintained, such changes may present selective pressures for them.  For example, the eastern cottonwood has a pH tolerance in the 5.5-7.5 range and prefers moist soil.  Introduction of competing species may alter the soil acidity or moisture content, placing an evolutionary pressure on this population.  Decisions as to the importance of the survival of the extant population would need to be made.  If maintaining the specific population is deemed important, it may be necessary to undertake common garden experiments to ascertain the tolerance of the existing plants and plan restoration efforts accordingly.

Another of key factor in the survival of a population is the existing genetic diversity (Falk et al., 15).  Interaction with the environment places selective pressures on populations and in order for the population to succeed it must contain sufficient heritable traits to allow adaptation to environmental changes (Falk et al., 14).  Such selective pressures should also mean an existing population will have a high frequency of alleles (genotype) associated with the physical traits (phenotype) adapted for the existing environment (Falk et al., 15).  However, if sufficient genetic diversity does not exist in a population it will be not be able to adapt to disturbance events such as extreme weather, disease infestation, or the introduction of predators (Falk et al., 15).  This problem is of special concern in small areas such as HRY, where limited area restricts population size.

When small populations inbreed, there is an increase in the probability of homozygousity.  While heterozygous individuals carry different copies of a specific gene, homozygous individuals have two identical copies.  Genetic diversity is therefore reduced in populations with a higher level of homozygousity, and the availability of potentially adaptive traits is reduced (Falk et al., 16).  The negative effects of homozygousity may be solved by introducing genetic material from outside populations; a process termed genetic rescue (Falk et al., 16).  Due to the probabilistic nature of genetic variation, only a small number of outside genes are required to propagate genetic rescue in a population.  For this reason, it is important for any restored species at HRY to have a connection with other sources of genetic material, i.e. nearby populations.  The existing species in surrounding areas such as Inwood Park will therefore have an impact on the selection of species for the HRY project.

In addition to increased homozygousity as a result of inbreeding, restoration projects that introduce outside genetic material must consider the possibility of introducing alleles that produce phenotypes poorly adapted to the restoration site.  By reducing the number of potentially positive existing alleles, such genes will reduce the adaptive ability in the same manner as inbreeding (Falk et al., 16).

Since the eastern cottonwood population at HRY is very small, it may be feasible to carry out genetic testing of this and members of nearby populations to determine relatedness.  Male clones of the species are often planted for soil stabilization purposes, so the breeding status of the existing population should also be determined.

The area available and the higher relative population density for both terrestrial and aquatic native grasses as well as the proximity of nearby populations may allow these species to be more self-sustaining with regard to gene flow.


Metapopulation Dynamics


            The spatial arrangement of urban environments means that where populations of plants and animals occur, they are generally small and isolated to a lesser or greater degree from other populations of the same species.  Together these smaller populations form metapopulations.  Metapopulation theory offers a tool to help understand the interaction of several populations of the same species and the survival of the species at the local and regional level (Maschinski, 64).  A population’s chance of survival in a specific area increases as the size of the area, or patch, increases, or when the number of patches increases (Maschinski, 66).  While direct analysis of metapopulations is difficult due to the need for long term data and the complex biology of most species (Maschinski, 62), metapopulation analysis, with the aid of modeling, makes it is possible to define the minimum number of interacting populations necessary for the metapopulation to persist.

In the context of the HRY restoration project and the eastern cottonwood, metapopulation analysis would be a useful tool for determining whether site is of sufficient size to support an isolated population, or if interaction with other populations will be necessary.  The estimated minimum viable population (MVP) ranges 50-1000 individuals (Maschinski, 60), so a large increase in the extant population would be necessary.  From a metapopulation standpoint, 15 to 20 well connected local populations are necessary to maintain a viable metapopulation.  Census of other populations of eastern cottonwood in nearby areas such a Randall’s Island Park would be necessary to determine the degree to which a population at HRY could be expected to persist without intervention.  As with the discussion of population genetics above, the higher relative populations of terrestrial and aquatic grasses may simplify the management of these species.


Community Ecology


            By definition, a large fraction of the available land in an urban area is utilized for buildings, roads and other structures, severely reducing the amount of space available for flora and fauna.  Therefore, one of the greatest challenges for an urban restoration project is managing the community ecology of the restored area.  Community ecology refers to the species richness of an area and is generally presumed to be a function of the competition and prey/predator relationships (Menninger and Palmer, 90).  Alternatively, Wilson and Simberloff, argue that species richness is a function of patch size combined with a species’ ability to disperse, while the most recent theories of community ecology are based on disturbance theory and stochasticity (Menninger and Palmer, 90).  Disturbance theory states that diversity is impacted by natural processes such as forest fires or hurricanes that some species require to reproduce or disperse.  Stochasticity is the effect of community demographics, where, for example, a population size is reduced due to natural occurrences, and if it falls below a certain size it goes extinct.

            For the eastern cottonwood population and other species that would form the eventual community as part of this restoration project it is presumed stochastic and disturbance events would be minimized by management.  However, community ecology theory would be useful for modeling the number of species that could be successfully introduced to the site.


Abiotic Filters


The survival of a species at specific site is affected by factors such as the amount of light and shade and soil chemistry and texture (Menninger and Palmer, 95).  Due to the relatively high degree of development at HRY, with large paved areas and buildings, consideration and monitoring of abiotic factors will be important.  As mentioned previously, changing soil composition due to soil remediation efforts may affect the extant eastern cottonwood population.  The flow of water during periods of high rainfall will be of particular importance to the health of salt marsh areas created by softening the riverbanks and the main salt marsh area.  Water flowing from access roads may potentially cause rapid changes in water salinity which would be detrimental to semi-aquatic plants such as spartina.  While one of the main advantages of marsh grasses is their ability to stabilize river banks and decrease point source pollution, high water flow from paved areas may also cause excessive erosion and changes in water chemistry outside of the plants’ level of tolerance.  It will therefore be necessary to control erosion by engineering methods such as the provision of erosion control matting at least in the short term.

Selection of plant species may also impact decisions relating to the management abiotic factors, such the control of soil pH.  For example, the eastern cottonwood and the native grass, sideoats grama, both have ph ranges around 5.5-7.5 (USDA, 2006), simplifying the management of the contiguous area these species will share.


Measuring Success


A comprehensive assessment of biodiversity and population estimates will be required at the outset and at intervals to be determined in order quantify the success of the project and allow for modification of restoration plans and/or goals if necessary.  Such experiments should be carried out using a control with before and after measurements as described by designs such as Before-After Control-Impact (BACI) design.  BACI studies consider both the area of interest and a control site both before and after a perturbation (i.e. the restoration activity) takes place (Osenberg et al., 2006).

For HRY specific measurement regimes would be required for the different restoration areas.  For example the eastern cottonwood may be sampled for population density, while the native grass areas would additionally be sampled for soil contaminant concentrations.


Time Line


The time line for this project will initially be reliant to a great extent on the phytoremediation aspects of the project.  For example, a significant reduction of residual PAHs using grasses requires treatment times greater than one year (Parrish, Banks et al., 2004).  If it is possible to retain extant cottonwood populations, the time necessary to reach a sustainable community will be greatly reduced for Area 1.  Experience from other restoration projects has shown that grasses and wetland planting generally takes in the region of six months.  Projected times for the completion of various aspects of the project are detailed in Table 2.



Time for Completion



6 months

Commence BACI testing for all areas.  Removal of non-native turf grass & trees, Areas 1 & 2.  Softening of banks by filling along river edge.  Filling of Area 3


6 months

Planting of phytoremediation grasses, Area 2. Planting of aquatic and semi-aquatic plants, Area 3.  Monitoring of Area 1


1-4 years

Removal of phytoremediation grasses and planting of native grasses, Area 2.  Monitoring of Areas 1 & 3



Monitoring of all areas for ecosystem functioning, with additional monitoring for contaminant levels

Table 2: Project phases and expected completion times



Funding and Budget


Restoration cost can vary greatly depending of the complexity and scale of the restoration project.  Berger and associates (1995) found the restoration cost of 11 salt marsh restoration projects range from 1000 to 1 million dollars adjusted to 2005 dollars (RIHR, 2001).  Projects at the lower end of the range involved simple removal of phragmites using volunteer labor, while the most expensive project involved the removal of ten feet of fill over and ten acre site with the subsequent planting and monitoring of native grasses (RIHR, 2001).  A project involving the construction of dykes cost $20,000 ($22,000 in 2005) per acre (RIHR, 2001).

The proposed restoration at HRY involves engineering of the riverbank to restore wetland areas, suggesting the cost of restoration would be in the range of $22,000 per acre.  The total area of the waste management and transfer facilities at HRY is approximately 42 acres.  Therefore, the total area undergoing restoration is approximately 54 acres with a restoration cost, assuming $20,000 per acre, of approximately 1 million dollars.  Unfortunately, because the HRY site is an operating industrial facility under lease, it is unlikely that volunteer labor would be appropriate.

A potential source of funding will be Federal Brownfields funds allocated for the remediation of the site.  If the site is to be treated to as parkland, and will be actively managed there will be a shift in focus from a single restoration budget to an ongoing maintenance budget.  Additionally, economies of scale could be achieved by sharing resources, for example suppliers of propagules with similar restoration projects that occur in nearby areas such as Inwood Park.




            The restoration of HRY, including the conservation of native species such as the eastern cottonwood, in accordance with initiatives to improve the general environmental health of New York City’s rivers and associated habitats, offers an opportunity to achieve multiple benefits from a single project.  Such benefits include increased public enjoyment by general beautification, improved ecosystem services provided by restored vegetation and expansion of the local urban forest.





Condit, C. W. (1980). The Port of New York - A History of the Rail and Terminal System from the Beginnings to Pennsylvania Station. Chicago, The University of Chicago Press.


Dahlin, Kyla.  Crotana Park Urban Forest Management Plan.  City of New York. 2004.


Dietz, A. C. and J. L. Schnoor (2001). "Advances in Phytoremediation." Environmental Health Perspectives 109: 163-168.


Flagg, Thomas R.  New York Harbor Railroads. Scotch Plains: Morning Sun Books.  2000


Fiteni.  (2001)


Falk, Donald A.,  Richards, Christopher M.,   Montalvo, Arlee M. and Knapp, Eric E.  “Population and Ecological Genetics in Restoration Ecology.”  Foundations of Restoration Ecology. Ed. Donald A. Falk, Margaret A. Palmer, and Joy B. Zedler.  Washington: Island Press, 2006.


Lehman College Library (CUNY)


Loeb, Robert E.  Plant Communities of Inwood Hill Park, New York County, New York.”  Bulletin of the Torrey Botanical Club, Vol. 113, No. 1. (Jan. - Mar., 1986), pp. 46-52.


Maschinski, Joyce.  “Implications of Population Dynamic and Metapopulation Theory for Restoration.” Foundations of Restoration Ecology. Ed. Donald A. Falk, Margaret A. Palmer, and Joy B. Zedler.  Washington: Island Press, 2006.


Menninger, Holly L.,  and Palmer, Margaret A.  “Restoring Ecological Communities: From Theory to Practice.”   Foundations of Restoration Ecology. Ed. Donald A. Falk, Margaret A. Palmer, and Joy B. Zedler.  Washington: Island Press, 2006.


New York City Solid Waste Management Plan, Chapter 9.  Final Environmental Impact Study Harlem River Yard.  2005


Osenberg Craig W., Bolker, Benjamin M., White, Jada-Simone S., St. Mary, Colette M., Shima, Jeffrey S. “Statistical Issues and Study Design in Ecological Restoration: Lessons Learned from Marine Reserves.” Foundations of Restoration Ecology. Ed. Donald A. Falk, Margaret A. Palmer, and Joy B. Zedler.  Washington: Island Press, 2006.


Parrish, Z. D., M. K. Banks, et al. (2004). "Effectiveness of phytoremediation as a secondary treatment for polycyclic aromatic hydrocarbons (PAHs) in composted soil." International Journal of Phytoremediation 6(2): 119-137.


Rhode Island Habitat Restoration. (2001). "Restoring Coastal Habitats for Rhode Island's Future - Cost Analysis." 2006.


Stockwell, Craig A.,  Kinnison, Michael T.,  and Hendry, Andrew P.  “Evolutionary Restoration Ecology.”  Foundations of Restoration Ecology. Ed. Donald A. Falk, Margaret A. Palmer, and Joy B. Zedler.  Washington: Island Press, 2006.


USDA (2006)

Last Updated by James Danoff-Burg, 20 Dec 06