Historical Changes of Eelgrass in Buzzards Bay
Composite map of current eelgrass distribution (2001-2010).
2001-10 Composite Coverage as Google Earth file
2001-10 Composite Coverage as Zipped Shapefile
Historical Changes in Eelgrass Abundance and State of the Bay Scores
For the 2003 State of Buzzards Bay Conference, the Buzzards Bay NEP estimated historical changes of eelgrass cover in Buzzards Bay from the 1930s to the present day based on aerial photographs, and anecdotal information contained in earlier reports. In the 2003 report, we also included a more speculative assessment of what eelgrass habitat area might have been in pre-colonial time when Buzzards Bay was pristine. This pre-colonial estimate of eelgrass bed habitat area was assigned a score of 100, and in the 2003 report we assigned a score for eelgrass abundance in Buzzards Bay by dividing then current eelgrass bed habitat area by the hypothetical eelgrass habitat area (x100). In the 2007 and 2011 Buzzards Bay Coalition State of the Bay reports, the BBNEP updated the eelgrass score based on the latest available information.
In the 2003 State of the Bay report, we also postulated the general decadal trends of changing eelgrass habitat area in Buzzards Bay during the 20th century based on a handful of embayments that were documented in the Costa 1988 reports using sediment cores and aerial photographs. Our analysis took into account the 1931-32 the wasting disease that destroyed at least 90% of eelgrass beds in the North Atlantic, but in Buzzards Bay likely destroyed closer to 99% of its eelgrass cover.
For the 2003 State of the Bay report, the score was based on DEP's 1995 eelgrass survey, which was made available publically about 1999 (the DEP 2001 survey data was still under review). For the 2007 report, we used the DEP 2001 survey coverage, which was available about 2004. For the 2011 State of the Bay report, the score was based on eelgrass beds mapped by DEP in selected embayments using a 2006 aerial survey, and using pre-released draft data from DEP for a 2010 aerial survey of selected Buzzards Bay embayments (courtesy Charles Costello). Because there was an insufficient number of embayments mapped by DEP for a new score for the 2011 State of the Bay report, the BBNEP undertook a review of other publically available recent aerial imagery, to expand the geographic extent of eelgrass maps on which to estimate a new eelgrass score, as described in more detail below.
At the same time, the BBNEP also analyzed for the first time, additional archival photographs from the 1950s, 1960s, and 1970s, and newer aerial photography from 2005-2010 to develop a more comprehensive trend of changing abundance of eelgrass in embayments around Buzzards Bay. This allowed us to define with more precision the historical recovery of eelgrass in Buzzards Bay after the wasting disease, detail trends for specific embayments and outer coast areas, and new declines evident during the past 20 years. This new analysis helped us refine average abundance of eelgrass for each decade between the 1950s and today. All these new maps and aerial surveys will be contained in a new report of historical changes in eelgrass in Buzzards Bay to be released later in 2012, with additional maps showing assumed eelgrass boundaries on the original source photographs to facilitate comment and review of our approaches.
In the sections below, we highlight aspects of some of the data used, our methods, and other considerations that are relating to mapping eelgrass in Buzzards Bay.
Costa 1988 EPA Report in Thesis DataWith funding from the Buzzards Bay Program in 1985, Costa mapped current and historical distribution of eelgrass in Buzzards Bay based on field surveys between 1983 and 1987, historical aerial photographs between 1938 and 1983, and sediment cores (Costa 1988a, 1988b). A description of the methodologies used and the limitations of using older aerial surveys are found on the eelgrass GIS data page. Important considerations when evaluating old aerial photographs include whether or not the photographs were taken too early in the growing season to enable easy visualization of annual eelgrass beds, whether or not a recent hurricane or winter icing affected eelgrass abundance in shallow waters, and whether or not water transparency is adequate to enable visualization of the lower edge of the beds.
The 1988 eelgrass reports included historical changes of eelgrass abundance for selected sites. More recently, the Buzzards Bay NEP has been scanning and registering original source photographs and acquiring others, and using current GIS software and methods, creating new GIS coverages of eelgrass change that will be posted on this website.
1931-1932 wasting diseaseA "wasting disease" destroyed virtually all eelgrass in Buzzards Bay (and elsewhere in North America) during 1931-32. The disease appears to have been caused by a virulent cellular slime mold that might have been introduced to the Atlantic or facilitated by exceptionally warm weather during those summers. All documentation suggests that eelgrass populations equaled or exceeded present-day abundance prior to this catastrophe. Photographs taken 6 to 10 years after the disease show that eelgrass covered less than 10% of the present-day habitat area in Buzzards Bay (although many of these areas may have been affected by the Hurricane of 1938), and many areas appear not to have been recolonized for decades. Most areas of Buzzards Bay appear to have fully recovered from the wasting disease by the mid 1960s.
The process of recolonization was similar in many areas: new beds initially appeared on bare substrates, beds expanded, additional new beds appeared, and some beds were removed by disturbance. In this way eelgrass population saturated small areas (1-25 acres) 5 to 15 years after initial colonization. Rates of eelgrass colonization over larger regions (100's of acres) depended on distance from refuge populations and heterogeneities of the environment. The greatest rates of eelgrass expansion occurred during the 1950's and 1960's. Most available substrate was saturated by the 1980s, but eelgrass continued to expand in some areas, or show new pollution-related declines in others.
First DEP Eelgrass Survey vs. the 1980s SurveyDEP first conducted an aerial survey to map eelgrass in Buzzards Bay in 1995 (a recommendation in the 1991 Buzzards Bay Comprehensive Conservation and Management Plan) using photographs from about June 1, 2005. One limitation of an early aerial survey is that functional annual eelgrass beds (that is, those growing from seed) may not be established in certain shallow areas. While DEP did conduct field surveys during the following summer, field surveys can only corroborate the presence or absence of eelgrass, although DEP did visit hundreds of sites in Buzzards Bay.
The first DEP survey showed considerably less eelgrass in Buzzards Bay than the 1980s reports. Conclusions about the locations of eelgrass in new DEP surveys as compared to the 1980s map were difficult because with maps only available for two dates, it was unclear at the time whether the absence of eelgrass at specific locations was due to eutrophication, date of aerial imagery, storms, disease, natural variability, or other factors, or from differences in the time of year or limitations or visual limitations of the aerial imagery. For example, the 1980s surveys were largely based on a October 1981 coastal aerial survey. In such early fall aerial photographs, annual and perennial eelgrass beds show the greatest contrast with background sediments because of high leaf and epiphyte biomass. This 1981 aerial survey was also undertaken mostly during a low tide, and with exceptional water clarity conditions in most areas, so that deep beds were also easily discerned. Because of these differences in the quality and conditions of the aerial surveys, it was difficult to determine whether the apparent absence of eelgrass at some locations in the 1996 survey was based on distribution of annual beds, pollution, disease, storm losses, or limitations of the aerial survey.
It was not until DEP repeated its eelgrass aerial and field survey in 2001 (released about 2004), and when we began reviewing other sets of aerial photographs, that it became apparent that much of the eelgrass losses documented in 1996 survey were genuine and likely due to eutrophication (like in the Wareham River and off Wareham's Great Neck) and worsening (read this Codium versus Eelgrass off Great Neck article).
The subsequent 2001 aerial survey was flown somewhat later (mid-June), and this survey also had only average water transparency, so the full extent of deeper beds in offshore areas and deeper parts of harbors were remained unclear until more aerial surveys became available and were evaluated.
Development of New CoveragesIn 2003, DEP also mapped eelgrass at a large number of sites in the Commonwealth using a 1951 state highway department aerial survey photograph set. These photographs were taken ata low altitude during the fall, and generally show good water transparency. Because many areas still had not recovered from the wasting disease (there is a dearth of eelgrass in the Marion and Wareham areas in the DEP coverage), this eelgrass cover should not be used to determine the possible maximum extent of eelgrass in pollution and nitrogen TMDL studies, especially if aerial surveys from the 1960s, 70s, or 80s show a greater extent of eelgrass cover. DEP has not yet made the 1951 eelgrass survey coverage available through the MassGIS website.
Shortfalls in agency funding have curtailed baywide eelgrass mapping efforts in Buzzards Bay by DEP. About 2008, DEP released a new survey of most embayment areas in upper Buzzards Bay based on 2006 aerial imagery. This coverage is available at the MassGIS website. A similar effort in a smaller number of embayments was undertaken using 2010 aerial imagery, but these data have not yet been released.
To expand the assessment of historic and current trends of eelgrass cover, the Buzzards Bay NEP has begun georeferencing and enhancing historic and recent aerial imagery of Buzzards Bay. Included in this effort, we have scanned in detail for the first time, photos from Costa thesis aerial archive, as well as downloaded new and archival photographs from various sources. Using ArcGIS 9.3, images were easily imported and georeferenced generally to within less than 10 ft of current MassGIS ortho base maps. Images were typically enhanced within ARCGIS by stretching raster histograms and changing brightness and contrast. This ability to enhance aerial images enabled visualization of deeper beds and deep edge boundary locations that were impossible to evaluate using tracing and transfer techniques common a decade ago. Bed boundaries were "heads up" digitized in the GIS software with the imagery magnified to capture eelgrass habitat boundaries and bare patches within beds, often with a point spacing less than 10 meters. This approach enabled us to create eelgrass habitat maps with a high degree of precision to the imported base maps. Because this fineness of detail was not consistent with some earlier DEP eelgrass coverages, in a few embayments the DEP eelgrass bed boundaries were reevaluated to better capture bed patchiness to ensure better consistency with new coverages developed. Similarly, eelgrass coverages digitized by Costa in the 1980s lacked the fineness of detail achievable through a reanalysis of that study's original source photographs, so the newly digitized coverages on specific dates were substituted for the Costa survey files when possible.
Hypothetical Eelgrass Distribution during the Colonial PeriodFor the 2003 Coalition State of the Bay report, we estimated eelgrass cover circa 1600. This is a speculative exercise, but it was worthwhile to imagine the potential distribution of eelgrass in Buzzards Bay without human impacts such as nitrogen loading, increased water turbidity associated with urban runoff, resuspension of sediments from boat traffic, and other human disturbances.
Eelgrass distribution is largely dependent upon water transparency. This is evident in the clear waters of offshore areas of Cape Cod and Islands, where eelgrass often grows between 20 and 30 feet MLW, and deeper depths of eelgrass beds have been recorded by divers. In less polluted and better flushed harbors and coves, eelgrass beds can still be found to depths between eight and twelve feet. In contrast, in most polluted embayments, eelgrass, if present at all, may only grow to depths of 6 feet MLW or often much less.
These patterns of growth became the basis of assumed eelgrass distribution in pristine conditions. Using the depth of growth of eelgrass in clean waters, we can speculate what eelgrass abundance in Buzzards Bay may have once been, and a hypothetical historical eelgrass habitat in Buzzards Bay can be postulated. Such hypotheses can even be tested using sediment cores because Costa (1988, 1989) and others have shown that the remains of eelgrass seed coats (tests) are a valuable biostratigraphical marker that remain in the sediments of many bays and harbors, possibly for many centuries. For example, no eelgrass has been found in inner Apponagansett Bay (Dartmouth) for decades, but sediment cores show eelgrass was abundant for centuries based on assumed rates of sedimentation (Costa, 1988a).
For the estimation of eelgrass cover in 1600, the following two assumptions were made. For upper Buzzards Bay (roughly a line drawn from North Falmouth to Mattapoisett), we assumed eelgrass grew to 12 ft. MLW. For the lower half Buzzards Bay, we assumed eelgrass grew down to 20 feet MLW.
2011 State of the Bay report assumptions and data[Note: The BBNEP is still in the process of analyzing recent and historic photographs to expand our assessment of eelgrass trends in Buzzards Bay. As this work continues, draft maps will be posted as links on this page, and a final report will be released later in 2012. This section explains how data available January 2012 was extrapolated to the 2011 State of the Bay report.]
One way to characterize eelgrass cover data is assign a relative abundance score of eelgrass in each embayment relative to the peak coverage observed in any photograph. Thus, if eelgrass in 1951 covered 100 acres in a particular embayment in 1951, but only covered 60 acres in 1995, its score in 1995 would be 60. In the graph below, we show the average score for all embayments based on the average embayment score using all data during any particular decade.
The graph above incorporates certain biases. For example, embayments with large areas of eelgrass are weighted the same as embayments with only small amounts of eelgrass. Some parts of Buzzards Bay (such as the upper portions of Buzzards Bay) were mapped with more frequency, so these locations bias the trends. Finally, most of the recent data is for embayments. Eelgrass in well-flushed outer coastal areas does not show the same declines as many embayments, so these areas must be accounted for. Because the eelgrass score for Buzzards Bay is meant to track the trend of total eelgrass cover in Buzzards Bay (upper and lower bay, embayment and offshore beds), the data was characterized in a way to minimize some biases in the available data.
Specifically, data was aggregated so that only data for embayments that were in both the DEP 2001 and DEP 2006, or the DEP 2001 and DEP 2010 coverage. Because these represent too few sites, and few offshore areas, we mapped eelgrass in additional areas, including some isolated reference areas like East Cove on West island to include in the analysis. In some embayments, there was more eelgrass in 2006 or 2010, than observed in 2001. In other embayments, there was less. Looking only at embayments where both data pairs were present and ignoring embayments where there was no eelgrass in 2001 and later years, the ratio of 2006 to 2001 DEP eelgrass areas (at the time of our analysis in January 2012) was 0.988 (1385 versus 1355 acres), or only a 2.2% decline. If the average relative eelgrass bed cover for all areas between 2006 and 2010 (including new BBNEP eelgrass maps) is compared to the corresponding areas in the 2001 survey the ratio of new to old acres is 0.922, (1,891 versus 1716 acres), or a 9.3% decline of eelgrass in these mostly inner harbor areas. When the same analysis is conducted comparing 2001 to only maps based on 2009 or 2010 photographs, considering only areas available in both the 2001 and 2009-10 maps, total eelgrass in these areas (at the time of our analysis in January 2012) declined from 1168 acres in 2001 to 969 acres in the later period, or a 17.0% decline. These changes are summarized in the table below.
|Matched pair analysis of embayments with coverage on both dates, status as of January 2012 for State of the Bay report|
|Comparison||acres as of 2001||acres as of 2005-06||acres as of 2009-10||average 2005-2010||percent change|
|2001 versus 2005-2006||1384.9||1354.9||-2.2%|
|2001 versus average 2005-2010||1891.2||1715.7||-9.3%|
|2001 versus 2009-2010>||1167.5||968.7||-17.0%|
To extrapolate these declines to the rest of Buzzards Bay where eelgrass was not mapped, including outer coastal areas where we assumed that eelgrass was not declining, we made these additional assumptions. Based on the 1980s and DEP 1990s 2001 surveys of eelgrass in offshore areas, only about 30% of the 2001-era remaining eelgrass of Buzzards Bay was found within the embayments at that time. The remaining 70% of the surviving eelgrass in Buzzards Bay is found in the better flushed open coasts of the bay, and there is no evidence for a general loss of eelgrass in these outer areas. We extrapolated the 17.0% decline for the mapped areas and unmapped areas as summarized in the table below. Overall we concluded there was an average 6.8% decline , or a loss of 441 acres since the last state of the bay report. This decline can be used to scale the eelgrass score of 25.3 in 2007 to 23.2 in 2011.
|Areas in Acres||2001||2009-2010||Comments|
|bed area in reference sites||2,353||1,952||Applied the 17.03% decline from subsample of references to the 2001 total bed area in those same sites|
|bed area in outer areas||2,343||2,343||Presumed constant.|
|Elizabeth Islands||776||776||Presumed constant, data from DEP 1995 survey.|
|10% adjustment||547||507||10% adjustment of above totals for presumed missing deep and offshore beds, as per the 2007 report.|
|Score||25.3%||23.2%||Report rounded index score to 23, a decline of 2.|
Summary of past State of the Bay Eelgrass scores.
|SOB Report Year (year of data)||Score|
|NA (1980s, various dates)||42.5|
|2001 (1995 data)||33.6|
|2007 (2001 data)||25.3|
|2011 (2009 and 2010 data)||23.2|
While this exercise is useful for characterizing environmental degradation in Buzzards Bay as part of the Coalition's State of the Bay Reports, when it comes to taking action to evaluate nitrogen pollution impacts to Buzzards Bay embayments, scientists always review the specific known trend of eelgrass coverage in each estuary. Click the links of the embayment names at the top of this page to see the actual trends for individual embayments.
Buzzards Bay Related Eelgrass Links and Reports on Historical ChangesCosta, J. E. 1988a. Eelgrass in Buzzards Bay: Distribution, Production, and Historical Changes in Abundance. EPA 503/4/88-002 204 pp (5.4 MB)
This report contains much the same information as the thesis work, except that it contains additional detailed embayment specific site histories.
Costa, J. E. 1988b. Distribution, production, and historical changes in abundance of eelgrass (Zostera marina L. ) in Southeastern MA. Ph. D. Thesis, Boston University, 395 pp.
(7.5 MB pdf)
DEP eelgrass interactive maps.
GIS shape files from MassGIS.
A May 12, 2004 letter to the Wareham Marine Resources Commission about the distribution of eelgrass in Onset Bay in response to a request for information.
PDF fact sheet on Massachusetts DEP eelgrass mapping program, and the Wetland Conservancy Program
MassGIS - DEP Eelgrass Data layer Description and download
NOAA Project: Loss of eelgrass at Cape Pogue Bay on Martha's Vineyard, Massachusetts, 1988-1994