W. E. Frayer, Dean and Professor
                     School of Forestry and Wood Products
                      Michigan Technological University
                            Houghton, MI 49931 USA

ABSTRACT:  A network of 6195 sample units has been established throughout the
lower 48 states and Alaska for the purpose of establishing baseline data and
monitoring changes in wetlands.  Each sample unit is four square miles in area
and, for general usage, is classified with a minimum mapping unit of 0.1
acres.  Impacts of large-scale, catastrophic events can be examined and
quantified with the sample network.  Samples in the lower 48 states have been
routinely measured twice.  In addition to large changes that were somewhat
expected (conversion of wetlands to agriculture), unexpected changes of over a
million acres were experienced during the last period between measurements.


   The United State Fish and Wildlife Service has major responsibility for the
protection and proper management of migratory and endangered fish and wildlife
and their habitats.  Of particular concern are wetlands and associated
deepwater habitats.  Since 1974 the U.S. Fish and Wildlife Service, through
its National Wetlands Inventory Project, has been conducting an inventory of
the nation's wetlands.  The purpose is to develop and disseminate
comprehensive data concerning the characteristics and extent of wetlands.

   In addition to the all-encompassing, site-specific inventory of the
nation's wetlands, timely information at the national level -- or for other
large areas -- is often needed as a basis for policy decisions and
legislation.  This is potentially of an emergency nature, as might be the
case if large expanses of forested wetlands were suddenly and permanently
defoliated.  For these reasons, a sampling network was established which
allows for precise estimation of wetland changes.

   Two routine measurements have been made of the samples in the lower 48
states.  Wetland gains and losses between the 1950's and 1970's were studied
by Frayer et al. (1983).  The second study of wetlands in the lower 48 states
documented natural and human-induced wetland and deepwater habitat gains and
losses between the 1970's and the 1980's (Frayer et al., 1991).

   The statistical design used in the national trend study can be used with
intensified sampling to obtain reliable estimates for individual states or
other selected geographical areas.  This has been done for the Central Valley
of California (Frayer et al., 1989), Florida (Frayer et al., 1991), and Alaska
(Frayer et al., 1993).  The Alaska study is the first for that state, and it
provides the baseline data for future estimates of change.

                            WETLAND CLASSIFICATION

   The definitions, classifications and categories of wetlands and deepwater
habitats used are those described by Cowardin et al. (1989) and Anderson et
al. (1976).  In general terms, wetland is land where saturation with water is
the dominant factor determining the nature of soil development and the types
of plant and animal communities living in the soil and on its surface.  Common
terms used to describe various wetlands include marshes, swamps, bogs, small
ponds, sloughs, potholes, river overflows, mud flats, and wet meadows.
Deepwater habitats consist of certain permanently flooded lands.  Common names
used for deepwater habitats include bays, lakes, rivers, and reservoirs.

   The details of the classification system are beyond the scope of this
paper.  The kinds of wetlands and deepwater habitats studied and some common
examples include:


Marine intertidal (ocean beaches)
Estuarine intertidal unconsolidated shore (mud flats)
Estuarine intertidal aquatic beds
Estuarine intertidal emergent (salt marshes)
Estuarine intertidal forested and scrub/shrub (mangroves)
Palustrine unconsolidated shore
Palustrine open water (ponds)
Palustrine aquatic beds
Palustrine emergent (wet meadows)
Palustrine forested (swamps)
Palustrine scrub/shrub (bogs)

Deepwater habitats

Estuarine subtidal (bays)
Lacustrine (lakes and reservoirs)
Riverine (rivers)

Other surface area


More detailed discussions, exact definitions, and fuller descriptions are
presented by Cowardin et al. (1979) and Anderson et al. (1976).

                        SURVEY AND ANALYSIS PROCEDURE

   The sampling network was designed to provide the means of developing
baseline data for estimates of change.  We now have the network in place for
all states except Hawaii.  There are several ways in which the network can be
used.  Two examples are:

(1)  The samples can be measured at a current point in time and changes from
an earlier measurement can be estimated.

(2)  The samples can be measured before and after a specific event, in order
to estimate the changes caused by that event.  This was essentially the case
for the Mississippi Delta area during the first "routine" measurement of the
lower 48 states.  Aerial photo coverage was obtained before and after the
devastating land clearing of bottomland hardwoods for agriculture.

   These uses of the network have been made with high-altitutde aerial
photography which allows for minimum mapping units of 0.1 acres.  For some
events, the minimum mapping units might be larger and satellite data could be
used.  Other events might require the use of large-scale photography or ground
examination.  In any case, the same locations are in place and the user can
choose the measurement methods.

   A stratified random sampling design is used, with the basic strata being
formed by state boundaries.  In the lower 48 states, these are further divided
by the 35 physical subdivisions described by E. H. Hammond (1970).  In Alaska,
the 25 physical subdivisions described by Rieger et al. (1979) are used.  In
all coastal states, additional coastal strata are used to encompass most of
the marine and estuarine categories.  Special strata are also used in states
bordering the Great Lakes to separate those bodies of water.  This resulted in
over 230 strata.

   Sample units are allocated to strata in proportion to expected amounts of
wetlands and deepwater habitats as estimated by U.S. Fish and Wildlife
personnel.  The total number of sample units is 6195.

   Each sample unit is a four-square mile area, two miles on each side.  The
units are plotted on U.S. Geological Survey topographic maps.  In general,
aerial photographs are obtained for the samples for at least two points in
time.  The photography is then interpreted and annotated in accordance with
the classification system described earlier and procedures developed by the
U.S. Fish and Wildlife Service's National Wetlands Inventory Project.  The
results of the most recent classification are compared with earlier
classification(s), and any changes are annotated.


   It was expected that large losses in wetlands had taken place in the 1950's
and 1960's.  The first trends study of the lower 48 states was used to
estimate these changes.  Although the results are not indicative of a
catastrophe which happened overnight, the magnitude of the changes occurring
in a 20-year period are still rather "sudden" in terms of natural resource

   It was shown that 4.5 million acres of forested wetlands were lost in the
Mississippi Flyway.  Most of it was in the Mississippi Delta and was the
result of conversion to agriculture.  The national net loss in wetlands alone
during the 20-year period was 11 million acres, an area as large as the
combined states of Massachusetts, Connecticut and Rhode Island.  Nearly all of
the net loss was due to agriculture.  Substantial amounts of coastal wetlands,
especially in Florida, were converted to urban development.  In all, 20 states
showed significant wetland loss during the period.

   Documentation of these losses helped to pave the way for legislation and
policy directives aimed at curbing wetland losses.  A third measurement of the
sample network in the lower 48 states was conducted to estimate losses from
the 1970's to 1980's.  It was found that losses had continued.  In addition to
the 20 states having significant losses in the previous period, another 11
states now showed significant losses.  However, some of the losses were of a
different nature.

   Over one million acres were converted from forested swamps to "other" uses
during this recent period.  Most of this is along the southeastern coast.
Although the study did not define the specific cause, the author believes a
large amount is due to forestry operations.

   A loss of another nature - and what came as a surprise - was a tremendous
change in area of freshwater vegetated wetlands.  Net changes from large to
smaller vegetation were as follows:

   Palustrine forested to palustrine scrub/shrub (482.8 thousand acres) +
   palustrine forested to palustrine emergent (722.2 thousand acres) +
   palustrine scrub/shrub to palustrine emergent (68.6 thousand acres)
   = 1,273.6 thousand acres.

   This change of over 1.25 million acres in a nine-year period between 1974
and 1983 came as a surprise.  What is the cause?  It is not a sudden
occurrence, but it is of such magnitude that it should be of concern.  Could
it be a long-term effect of road construction?  That is, could subsurface
water flows have been altered by roads or other human-induced causes such that
effects are being noticed many years later?  Perhaps.  The data are available
to determine where and maybe why these changes have happened and may still be

                                IN CONCLUSION

   The sampling network described has been used for studies of the wetlands
and deepwater habitats of the lower 48 states and Alaska.  The design involves
careful measurement of a sample of the nation's surface area.  In general,
results are meaningful only at the national level or for broad areas.  With
intensification of samples, results can have adequate reliability to be useful
at flyway or state levels.

   Continual monitoring of surface area use and changes in use is needed to
provide the basis for wise decisions.  This report is the result of one such
method of monitoring initiated by the U.S. Fish and Wildlife Service.
Monitoring can be a routine, scheduled procedure.  When necessary, the network
provides the basis to assess changes caused by catastrophic events.

                               LITERATURE CITED

Anderson, James R., Ernest E. Hardy, John T. Roach and Richard E. Witmer.
   1976.  A land use and cover classification system for use with remote
   sensor data.  U.S. Geological Survey Prof. Paper 964.  22 p.

Cowardin, L. M., V. Carter, F. C. Golet and E. T. LaRoe.  1979.
   Classification of wetlands and deepwater habitats of the United States.
   U.S. Fish & Wildl. Serv.  103 p.

Frayer, W. E., T. J. Monahan, D. C. Bowden and F. A. Graybill.  1983.  Status
   and trends of wetlands and deepwater habitats in the conterminous United
   States, 1950's to 1970's.  Colo. State Univ.  32 p.

Frayer, W. E., Dennis D. Peters and H. Ross Pywell.  1989.  Wetlands of the
   California Central Valley:  status and trends, 1939 to mid-1980's.  U.S.
   Fish & Wildl. Serv.  28 p.

Frayer, W. E. and John Hefner.  1991.  Florida wetlands.  U.S. Fish & Wildl.
   Serv.  32 p.

Frayer, W. E.  1991.  Status and trends of wetlands and deepwater habitats
   in the conterminous United States, 1970's to 1980's.  Mich. Technological
   Univ.  32 p.

Frayer, W. E., Jonathan V. Hall and Bill O. Wilen.  1993.  Alaska wetlands.
   U.S. Fish & Wildl. Serv.  32 p.  In press.

Hammond, E. H.  1970.  Physical subdivisions of the United States.  In:
   National Atlas of the United States.  U.S. Geological Survey.  417 p.

Rieger, Samuel, Dale B. Schoephorster and Clarence E. Furbush.  1979.
   Exploratory soil survey of Alaska.  U.S. Dept. of Agr., Soil Cons. Serv.
   213 p.