2013 Water Chestnut OARS report to CISMA

OARS’ Report to the SUASCO CISMA

Project title: “Mapping Water Chestnut in the Sudbury, Assabet, and Concord River System”

Date: October 31, 2013

Contact information:

Address: OARS Inc., 23 Bradford Street, Concord, MA 01742 (www.oars3rivers.org) Executive Director: Alison Field-Juma ([email protected]) Staff Scientist: Suzanne Flint ([email protected]) Tel: 978-369-3956

Project objectives: To document the location and extent of water chestnut populations in the Sudbury, Assabet, and Concord River system to inform, support and, in the long-term, track the progress of remediation efforts. This work will be coordinated with the initial mapping of the distribution of purple loosestrife along the Sudbury River from Ashland to Concord proposed as part of the purple loosestrife control project. This project was proposed in response to the RFP titled: Mapping Water Chestnut According to the Nyanza Restoration Plan, issued by the SUASCO Cooperative Invasive Weed Management Area Steering Committee.

Project Summary: This project was designed to support the preferred restoration alternative identified in the Nyanza Restoration Plan (May 2012) under Aquatic Biological Resources: control of aquatic weeds in the Sudbury River watershed to benefit freshwater wetlands and riverine habitat. Specifically, this project mapped distribution of invasive water chestnut in the Sudbury, Assabet, and Concord River system (including Heard Pond and four ponds of the Hop Brook system) to inform, support and, in the long-term, start to track the progress of remediation efforts. Data collection for water chestnut was completed between June 3rd and August 16th, 2013, using GIS-enabled GPS units to ensure that the information could be readily shared between invasive plant control projects and could be presented as a “snapshot” of the whole watershed for that season. A total of 2,133 acres of river area was surveyed from the headwaters of the Assabet and Sudbury Rivers in Westborough to the Centennial Island dam on the Concord River in Lowell. In the upper sections of both the Assabet River and the Sudbury River where boating was not possible, spot surveys were conducted from road crossings where the rivers were visible; 164 acres of river area was not surveyed.

Water chestnut was found in sections of all three rivers throughout the system. The furthest upstream that water chestnut were found on the Assabet was in the impoundment upstream of the Rte 85 dam in Hudson and on the Sudbury was below the dam between Framingham Reservoirs #2 and #1. On the downstream end, water chestnut were found along the edges of the Concord River to the downstream-most end of the survey section at the Centennial Island dam in Lowell.

Methodology: Project area: The project area (overview map in Appendix) for water chestnut mapping included all boatable sections of the Sudbury, Assabet, and Concord Rivers from Westborough to Lowell, the four ponds of the Hop Brook system (Hager, Grist Mill, Carding Mill, and Stearns ponds), and Heard Pond in Wayland. Upper sections of both the Assabet River (upstream of Rte 20 in Northborough) and the Sudbury River (upstream of Cordaville Road, Ashland, including Cedar Swamp Pond) were, as anticipated, very difficult to access because of low flows in early August. Therefore, spot surveys were

conducted from road crossings where the rivers were visible for those sections. Hager Pond, where there is no public access, was not surveyed. Equipment: OARS borrowed two Magellan MobileMapper units from CISMA for data input in the field. OARS supplied one kayak, paddle, and the PFDs. Purchased for this project were: one kayak and paddle, roof rack and tie-downs for boat transport, two waterproof cameras to document conditions, additional memory cards for the MobileMapper units to accommodate loading background maps, and water-resistant binoculars. Grid system and data input: OARS staff worked with Mass Audubon’s Jeff Collins to adapt and extend the GPS/GIS data input system developed in 2012 comprising: (1) GIS survey “grids” of the project area hand -drawn based on ortho-photos and elevation maps obtained from Mass GIS (datum: GCS North American 1983) (Figure 1); (2) data input screens for the MobileMappers to accommodate both water chestnut and purple loosestrife survey data. Grids for the water chestnut data collection were drawn taking into account the likelihood of having vegetation: the near-shore grids are smaller, while grids in deeper water are larger for more efficient mapping. In the course of developing the grid system, it was determined that having separate water chestnut and purple loosestrife grids would be the most efficient way to collect information. The grid system is expandable for future projects (for example, if mapping were extended to the tributaries). Background maps (ortho-photos for example) were added to the mappers to assist orientation in the field. Field data collection: Potential access points were identified and added to a Google Map for use by the surveyors; permissions for access were obtained as needed. OARS staff coordinated with US Fish and Wildlife and groups in Lincoln, Concord, and Wayland to ensure that the sections where water chestnut harvesting was anticipated were prioritized for mapping first, so that mapping occurred before harvesting. OARS hired, trained, and supervised two summer employees to conduct the survey work. Working mainly from kayaks, the surveyors paddled assigned sections visually assessing the presence, location and density of water chestnut. With the MobileMapper GPS on, surveyors could paddle the length of aquatic grids to estimate the size of the grid. For each grid cell where water chestnut were present, plants were visually identified, the diameter of a typical rosette measured, and the number of rosettes and percent cover in the grid square estimated. Ten percent of the grids in sections that have water chestnut were re-surveyed by OARS staff for quality control. Data handling, analysis, and reporting: At the end of each week of field work, data was downloaded from the MobileMappers to the central GIS database at the OARS office to ensure that data was backed up and data in the two MobileMapper units was synced. Field data were mapped in ArcGIS and exported for use by towns and groups conducting water chestnut remediation. Data input fields included (a complete list of the GIS data fields is included with the GIS files metadata):

OAR_ID (grid identification number)

Impoundment name (common name of the impounded area, if impounded)

Figure 1: MobileMapper map screen

Area (grid area in square meters)

Date (date of field survey)

Surveyor (initials of surveyor)

Number of water chestnut rosettes (0 = none, 1=1 to 5; 2 = 5 to 10; 3 = 10 to 25; 4= 25+)

Estimated percent cover (0 = 0%; 1 = 1 to 10%; 2 = 10 to 20%; 3 = 20 to 30%... etc. )

Measured rosette diameter (0 = not recorded; 1 = < 2”; 2 = 2 to 4”; 4 = 4 to 6”; 6 = 6 to 8”; 8 = 8 to 10”; 10 = 10 to 12”; 12 = 12 to 14”; 14 = 14”+)

Notes (field notes)

Water body (Sudbury River, Assabet River, Concord River, or pond name)

Complete (0 = not assessed; 1 = assessment complete) Findings: Survey: A total of 2,133 acres of river area was surveyed from the headwaters of the Assabet and Sudbury Rivers in Westborough to the Centennial Island dam on the Concord River in Lowell. Spot surveys were conducted from road crossings in the upper sections of the Assabet River upstream of Rte 20 in Northborough, and the Sudbury River upstream of Cordaville Road, Ashland, including Cedar Swamp Pond, where boating was not possible. Hager Pond, where there is no public access, was not surveyed. A total of 164 acres of river area was not surveyed. Raw GIS data will be provided in a separate file. Maps of all river sections are included in a zipped file. Plant density and growth: Of the 2,133 acres surveyed, 76% (1,629 acres) had no water chestnut while 3.7% (78 acres) had moderate to heavy density of plant coverage. The remaining 20.3% (426 acres) had emerging or light-moderate density populations (Table 2). Timing: Water chestnut is most readily observed and mapped once the plants have reached the water surface. Within the watershed, rosettes reached the water surface between mid-May and mid-June, measuring 2-4 inches. The average rosette diameter increased 5 inches between the end of June and the middle of August (Table 1). Unusually high streamflows at the end of June re-submerged plants and halted field work for more than a week. By September 30th, observations in the Saxonville impoundment showed that some plants still had seeds, ready to drop, on the rosettes.

Figure 2: Billerica Impoundment, May 28th.

Size: A minimum of ten rosette diameters were measured each field day, with the goals tracking plant growth over the season and of standardizing percent-cover estimates over the 10 weeks of the surveys. (Theoretically, it should take twice as many 6-inch rosettes as 12-inch rosettes to completely cover the same grid area.) Because there was only a weak correlation between plant diameters measured and percent coverage estimated, water chestnut density is reported only in field-observations of percent coverage (Table 2). Table 1: Rosette Diameters

Date Surveyed Average Maximum Diameter (inches) Max Maximum Diameter (inches)

June 1 – 15 6 10

June 15 - 30 6 10

July 1 - 15 7 12

July 15 - 31 9 14

August 1 - 16 11 14

Table 2: Assessed plant coverage

Assessed % Coverage Acres

80 - 100% “heavy density” 33

60 - 80 % 45

40 - 60% “moderate density” 22

20 - 40% 79

1 - 20 % “light density” 325

0% 1,629

TOTAL surface area surveyed 2,133

Locations: Of the designated Wild & Scenic sections of the rivers, only the Assabet was free of water chestnut; major infestation continues on the Sudbury and Concord Rivers. The furthest upstream that water chestnut was found by the survey team on the Assabet was in the impoundment upstream of the Rte 85 dam in Hudson and on the Sudbury was below the dam between Framingham Reservoirs #2 and #1. Many of the areas with significant infestations of water chestnut were already known (although not systematically mapped before this) including: the Billerica impoundment (upstream of the Faulkner Mill dam, North Billerica); the Concord Impoundments off the Concord River in Concord; the Sudbury River from Rte 27, Wayland to Sudbury Road, Concord; the Hop Brook ponds (Stearns Mill, Carding Mill and Grist Mill ponds); the Assabet River between Gleasondale Road and Sudbury Road in Stow and the impounded sections upstream of the Rte 85 dam in Hudson, near Crow Island in Stow, and the Ben Smith Impoundment (upstream of the dam at Rte 117, Maynard).

The Assabet River sections in Stow and Maynard that are being managed (hand pulling) were mapped in 2012 and again 2013, using the same methodology. The number of water chestnut rosettes recorded in the Crow Island section was lower compared with 2012 (500-1300 in 2012 vs. 300-700 in 2013). However, in the section between Gleasondale and Sudbury Roads the number of rosettes recorded in 2013 was higher compared with 2012 (60 -200 in 2012 vs. 200-400 in 2013). Water chestnut in both of these sections were pulled by hand in 2012 and 2013, with a similar level of effort.

Newly documented locations: There was a heavy infestation of water chestnut found in the Sudbury River upstream of the Saxonville dam (Framingham) that had not been previously documented. The total surface area of the impoundment between Wickford Road and Central Street, Framingham, is 39 acres. The surface area surveyed having 80 – 100% water chestnut coverage (as of July 17, 2013) was 13 acres.

Framingham Reservoir #1 on the Sudbury River has moderate to heavy densities of water chestnut along the edges and the northwest arm of the impoundment. Upstream of the Sudbury River areas being mapped in this project, MWRA’s contractor reported “light to moderate” densities of water chestnut on the Sudbury Reservoir in Southborough. That infestation was being managed by the contractor with hand-pulling in 2012 and 2013. This is worth noting as a potential source of seeds to the Sudbury River, particularly Framingham Reservoir #1.

Downstream of the Billerica Impoundment, the edges of the Concord River from Faulkner Street to the downstream-most extent of the survey at the Centennial Island dam in Lowell have light to moderate densities of water chestnut.

The Powdermill Impoundment of the Assabet River (upstream of the dam at Old High Street, Acton) also has a light density of water chestnut. This impoundment was drawn down from 2004 to about 2011, so the reappearance of water chestnut in this location is noteworthy.

Recommendations:

This mapping effort showed that water chestnut is spreading in the river system, with many acres of small infestations (325 acres of “light density”) that are likely to rapidly increase in size and density unless efforts are made next year to bring them under control. These areas are best suited to hand-pulling and efforts should be made by river/reservoir managers, non-profits, and others to organize and, where necessary seek funding, to put control efforts in place. Priority should be on areas (1) where emerging populations are in shallow water and thus harvester operation is not an option, and (2) upstream of river sections that are clear or currently under management.

1. Mapping: The boatable sections of the river system should be re-mapped in 2014 to confirm the findings of 2013 and to assess the variability of growth timing and locations. Comparison with results of the 2012 survey of the Assabet River using the same methodology showed variations in populations that could not be explained by harvesting and may be due to weather or other environmental variables. It is clear that a data series of at least two years would be valuable to inform management decisions and

Figure 3: Sudbury River at Saxonville, Framingham

assess management effectiveness. In addition, we recommend that the Foss Reservoir (Reservoir #3), between the Sudbury Reservoir and Framingham Reservoir #1 also be mapped, since this section may contribute to downstream infestations.

2. Research: There are gaps in the available knowledge of plant life cycle and biology that need to be filled to design effective management activities. In addition there is no consolidated source of guidance for water chestnut management or permitting of control activities.

3. Management areas: The areas of heavy density water chestnut in the Saxonville Impoundment of the Sudbury River and the Billerica Impoundment of the Concord River will require mechanical harvesting to control the problem. The Saxonville section is particularly significant because of the density of the infestation and because it can serve as a seed-bank to re-infest the downsteam areas that are currently being managed. Because this area is downstream of the Nyanza Superfund Site, water chestnut should be tested for mercury and other heavy metals to assess potential disposal options for harvested plants.

The areas of the Concord River downstream of the Billerica Impoundment might be controlled by hand-pulling once the problems at the Billerica Impoundment have started to be addressed. Areas currently being managed through mechanical harvesting appear to continue to need mechanical harvesting: Stearns Mill Pond; Carding Mill Pond; the Sudbury River from Rte 27, Wayland, to Sudbury Road, Concord. Sections of the Assabet River being managed by hand-pulling, continue to need hand-pulling and the effort needs to be extended to the Hudson/Rte 85 Impoundment and further on the Gleasondale Impoundment (upstream of the dam at Rte 62, Stow), and the Powdermill impoundment in Acton/Maynard.

Project Spending (Final):

PROJECT BUDGET for Water Chestnut Mapping Budget ($) Spending($)

Personnel

Field surveyors (10 weeks) 5,760 4,044

OARS Staff 4,360 6,810

Employee benefits 872 1,362

Personnel Subtotal 10,992 12,216

Contracts

GIS/GPS technical assistance (Mass Audubon bill projected) 900 900

Travel 990 652

Equipment 1,660 1,344

Supplies 650 78

Overhead 3,038 3,038

Total 18,230 18,228

Appendix: Overview Map showing map sections (zip file separate)