Chesapeake Bay Program Indicator Framework
Reporting Level Indicators
Indicator and Data Survey
A. Category/Name/Source/Contact
(1) Category of Indicator
_x_ Factors Impacting Bay and Watershed Health
___ Restoration and Protection Efforts
___ Watershed Health
___ Bay Health
(2) Name of Indicator: Bay Watershed Forest Cover
(3) Data Set Description:



For what purpose(s) were the data collected? (e.g., tracking, research, or longterm monitoring.) - Long-term monitoring.
Which parameters were measured directly? N/A.
Which were obtained by calculation? Estimates of Chesapeake Bay watershed
forest cover were developed for 1650, 1750, 1800, 1830, 1850, 1880, 1900, 1930,
1953, 1963, 1970, 1977 ,1980, 1990 and 2003 (The State of Chesapeake Forests,
2006, pp 14-17, http://www.na.fs.fed.us/watershed/socf.shtm), and 2011 USFS
Forest Indicator Analysis (FIA) analysis using EVALIDator tool and performed
for Chesapeake Bay watershed (Tonya Lister, personal comm).
(4) Source(s) of Data:
The State of Chesapeake Forests, 2006, pp 14-17
(http://www.na.fs.fed.us/watershed/socf.shtm) and FIA
Miles, P.D.(2011). EVALIDatorPC (Version 5.01.02) [Software]. Newtown Square, PA:
U.S. Department of Agriculture, Forest Service, Northern Research Station. Available
from http://apps.fs.fed.us/fiadb-downloads/datamart.html
Is the complete data set accessible, including metadata, data-dictionaries and embedded
definitions? If yes, please indicate where complete dataset can be obtained. The State of
Chesapeake Forests, 2006, Appendix A
(http://www.na.fs.fed.us/watershed/pdf/socf/Appendix%20A_Detailed%20Watershed%2
0Data.pdf) and Appendix B
(http://www.na.fs.fed.us/watershed/pdf/socf/Appendix%20B_Detailed%20County%20Da
ta.pdf)
General information on FIA and forest inventory online tool (http://www.fia.fs.fed.us/)
(5) Custodian of Source Data (and Indicator, if different): Sally Claggett 800-YOURBAY
ext 706
1
(6) CBPO Contact: Sally Claggett 800-YOURBAY ext 706
B. Communication Questions (complete either part 1, 2, or 3)
3. Factors Impacting Bay and Watershed Health indicators only
(7c) What is the long-term trend? (since start of data collection) In the 1600s, about 95
percent of the Chesapeake Bay watershed was forested. Forests now cover only 55
percent, or 24 million acres.
(8c) What is the short-term trend? (10-year trend) Between the years of 1990-2006, the
Bay watershed lost forestland at the rate of approximately 100 acres each day. Since
then, the loss of forests may have decreased because the slump in the economy meant
land was being developed at a slower rate (current estimates for 2005 on, is that 70 acres
of forest per day are lost, but it will take several more years to determine a true rate of
loss) but is still unsustainable.
(9c) What is the current status? Forests continue to be lost at an exceeding fast pace.
They now cover only 55 percent, or 24 million acres. The rate of forest loss could
increase again in the future if developers are incentivized, or the rate could slow if more
localities enact protective zoning.
(10c) What does this indicator tell us?

Nearly all forest land in the basin has been cut at one time or another. Isolated
remnants of 'virgin' forest (never cut) exist in very small quantities. From 1750 to
1890 most land in the watershed was cleared for farming, timber and fuel. Many
of these forests were allowed to grow back. Forest loss, in the past 30 years is
due primarily to urban sprawl. This loss is considered permanent deforestation.

More than 750,000 acres—equivalent to 20 Washington, DCs— have been
developed since the early 1980s, and the Bay watershed was losing forestland at
the rate of 100 acres each day. Current rate and trends for forest loss are not yet
available, but estimates are that loss has slowed about 30% in the last several
years.

Some areas facing rapid growth or supporting intense agricultural uses have lost
80 percent of their forest cover. Areas closest to the Bay have seen a rapid rate of
forest loss as these areas are prime areas for development.

Forest loss and fragmentation, created by urban sprawl, roads, and agriculture,
affects stream health, habitat quality and economic viability of the remaining
patches of forest.
(11c) Why is it important to report this information?
• As forests and wetlands are destroyed to make room for roads and buildings, their
ability to hold back and process pollutants and provide habitat are lost as well.
2
•
•
Forests protect and filter drinking water for 75 percent of the Bay watershed’s
residents and provide valuable ecological services and economic benefits
including carbon sequestration, flood control, wildlife habitat and forest products.
Retaining and expanding forests in the Chesapeake Bay watershed is critical to
our success in restoring the Chesapeake Bay. Forests are the most beneficial land
use for protecting water quality, due to their ability to capture, filter and retain
water, as well as absorb pollution from the air. In fact, our watershed forests can
process and/or retain up to 85 percent of the nitrogen they receive from air
emission sources such as motor vehicles and electric utilities. Conversely, a
reduction in forest area leads to a disproportionate increase in nitrogen loads to
our waterways.
(12c) What detail and/or diagnostic indicators are related to this reporting level indicator?
Bay Watershed Riparian Forest Buffer Cover; Water Quality indicator
4. All indicators (Answers to be written like you’re talking to either your grandmother
or a 10 year old. Analogies welcome!)
(7d) What did the most recent data show compared to the previous year? Since the
previous recording of 58% forest cover in 2003, the amount of forests in the watershed
has decreased. The two sources of information are both from FIA, however, the data was
collected in different ways. Because of this, we can not make a direct comparison
between the numbers (can’t say that 3% of the watershed’s forests were lost between
2003 and 20090.
(8d) If this was a significant increase/decrease:

To what do you attribute it? Decrease in forest cover is primarily due to
development.

Is this educated speculation or actual cause? There are data that show this is
actually happening.
(9d) What is the goal, target, threshold or expected outcome for this indicator? In 2007,
the Chesapeake Executive Council set an intention to “maximize the area of forest by
discouraging conversion of the most valuable forests and giving priority to forests in land
conservation programs. Further, we recognize the importance of working forests and will
ensure that public policies and market-based incentives help families retain and manage
these forests sustainably” (Forest Conservation Directive). Therefore, the unofficial goal
is to lose as few forests as possible. Maryland, for instance, has a policy for No Net Loss
of Forests. As stated earlier, 70% forest cover is a good target for healthy watersheds.
(10d) Was a new goal, target, threshold or expected outcome established since last
reporting? No
(11d) Did the methodology of data collection or analysis change from previous year(s)?
Yes Why and how? In 2003 and earlier years, the analysis from FIA determined that
the forest cover was 58%. Since that time, FIA data started to be collected on an annual
3
basis, versus every 5 years. In both 2003 and 2011, researchers and statisticians were at
least 68% confident in the data. However, when comparing these numbers in a change
analysis, the error bars go up significantly and there is less confidence. There is such an
uncertainty in determining an amount of change, that statisticians can not say with any
confidence what the trend in forest loss is using these data.

If so, how will this improve your/our future work? Going forward, the data points
in FIA will be more easily compared.
C. Temporal Considerations
(13) Data Collection Date(s): Estimates of watershed forest cover were developed for
1650, 1750, 1800, 1830, 1850, 1880, 1900, 1930, 1953, 1963, 1970, 1977 ,1980, 1990,
2003 (The State of Chesapeake Forests, 2006, pp 14-17), and 2011 (FIA data collection
and analysis).
(14) Planned Update Frequency (e.g. - annual, bi-annual): every 5 years
(a) Source Data: FIA
(b) Indicator: forest cover
(15) For annual reporting, month spatial data is available for reporting: N/A
D. Spatial Considerations
(16) Type of Geography of Source Data (point, line polygon, other): On-the-ground data
collection from fixed plots
(17) Acceptable Level of Spatial Aggregation (e.g. - county, state, major basin, tributary
basin, HUC):
All of the above, but in some cases, county-level aggregation may not be possible (there
may be too little data to obtain accurate information for the county level)
(18) Are there geographic areas with missing data? If so, where?
N/A
(19) The spatial extent of this indicator best described as:
(a) Chesapeake Bay (estuary)
(b) Chesapeake Bay Watershed
(c) Other (please describe): _______________________
Please submit any appropriate examples of how this information has been mapped or
otherwise portrayed geographically in the past. Several maps are featured in The State of
Chesapeake Forests, 2006 http://www.na.fs.fed.us/watershed/socf.shtm
4
(20) Can appropriate diagnostic indicators be represented geographically? yes
E. Data Analysis and Interpretation: (Please provide appropriate references and
location of documentation if hard to find.)
(21) Is the conceptual model used to transform these measurements into an indicator
widely accepted as a scientifically sound representation of the phenomenon it indicates?
(i.e., how well do the data represent the phenomenon?) Yes, good indicator for forest
cover. Partial indicator for forest health.
(22) What is the process by which the raw data is summarized for development and
presentation of the indicator? FIA process (http://www.fia.fs.fed.us/).
(23) Are any tools required to generate the indicator data (e.g. - Interpolator, watershed
model) A tool is available on the FIA website by state. To ”clip” the data to the
watershed boundary, a shapefile of the watershed must be used in GIS.
(24) Are the computations widely accepted as a scientifically sound? Yes.
(25) Have appropriate statistical methods been used to generalize or portray data beyond
the time or spatial locations where measurements were made (e.g., statistical survey
inference, no generalization is possible)?
1) Historic interpretation of forest cover was taken from Grace Brush’s palenology work
such as Reconstruction of Late Glacial and Holocene Forests in Middle Atlantic United
States.
2) Yes. The past 40 years of data on forest cover were derived directly from Forest
Inventory and Analysis, a research branch of the US Forest Service that uses fixed
sampling plots on a grid. Appropriate statistical sampling was involved for the region,
but somewhat specious at a resolution as fine as the county scale.
(26) Are there established reference points, thresholds or ranges of values for this
indicator that unambiguously reflect the desired state of the environment?
(health/stressors only) An estimate from the literature would have a healthy watershed
in this region be at least 70% forested.
F. Data Quality: (Please provide appropriate references and location of documentation
if hard to find.)
(27) Were the data collected according to an EPA-approved Quality Assurance Plan?
N/A—Data were collected via an equally-robust method for inventorying forests that was
developed by the US Forest Service.
If no, complete questions 28a – 28d:
5
(28a) Are the sampling design, monitoring plan and/or tracking system used to collect the
data over time and space based on sound scientific principles?
(28b) What documentation clearly and completely describes the underlying sampling and
analytical procedures used?
(28c) Are the sampling and analytical procedures widely accepted as scientifically and
technically valid?
(28d) To what extent are the procedures for quality assurance and quality control of the
data documented and accessible?
(29) Are the descriptions of the study or survey design clear, complete and sufficient to
enable the study or survey to be reproduced? Yes
(30) Were the sampling and analysis methods performed consistently throughout the data
record? Yes
(31) If datasets from two or more agencies are merged, are their sampling designs and
methods comparable? N/A
(32) Are uncertainty measurements or estimates available for the indicator and/or the
underlying data set? Yes
(33) Do the uncertainty and variability impact the conclusions that can be inferred from
the data and the utility of the indicator? No
(34) Are there noteworthy limitations or gaps in the data record? Please explain. Only
with the historical and pre-historical data as documented by Brush et al.
G. Additional Information (optional)
(35) Please provide any other information about this indicator you believe is necessary to
aid communication and any prevent potential miss-representation.
References:
1. The State of Chesapeake Forests, 2006 http://www.na.fs.fed.us/watershed/socf.shtm
2. Bibliography from The State of Chesapeake Forests:
http://www.na.fs.fed.us/watershed/pdf/socf/Bibliography.pdf
6