WETLAND COMPENSATION COSTS
IN EPA REGION IX—THE SOUTHWEST
Dennis M. King, Ph.D.
and
Curtis C. Bohlen,Ph.D.
University of Maryland System,
Center for Environmental and Estuarine Studies,.
, Chesapeake Biological Laboratory
P.O. Box 38, Solomons, Maryland 20688
April 1, 1994
University of Maryland, CEES Technical Report UMCEES-CBL-94-051, April 1994,
Prepared under Cooperative Agreement Number CR818-227
• with • . ' ... ' •
• the U.S. EPA, Office of Policy'Analysis .
' - w,ith support from ••'..-
EPA Region IV (Atlanta) and Region IX (San Francisco).
EPA 230-R-96-004 '
-------
Southwest Region Compensation Costs
INTRODUCTION
Regional Climate, Ecology, and Wetlands
The southwestern United States, including the states of California,
Nevada, and Arizona, is characterized by great geographic, climatological, and
ecological diversity. Climate varies from the cool Mediterranean climates of
northern California and the montane climate of the Sierras to the cool desert
of Nevada's Great Basin, and the warm, desert climates of .the Mojave and
Sonoron deserts. This climatological diversity is Reflected in wide ecological
variation across the region, as well as in a diversity of wetland types.
Precipitation throughout the region is either rare and episodic or
strongly seasonal. Rainfall is most abundant (and most predictable) in the
northwest, with both quantity and predictability decreasing toward the south-
east. Mediterranean climates throughout much of California (with cool, wet
winters and warm, dry summers) ensure that the soil moisture and surface
water on which wetlands depend are also strongly seasonal,,except in north-
western- forests where coastal fogs and orographic precipitation maintain rela-
tively high soil moisture levels throughout much of the year.
In the Great Basin to the east, precipitation is generally lower (Reno,
Nevada averages 7.64 inches of rain a year, but precipitation on mountain
ridges may be several times that), and somewhat;unpredictable, although typ-
ically concentrated in the winter snowfall and to a lesser extent in torrential
summer rains. Surface waters in the Great Basin do not flow to the sea, but
instead drain locally to desert lakes, playas, and sinks. Year-to-year.variation
in precipitation is large, causing substantial interannual variation in water
depth in many of the lakes, rivers, and wetlands within ihe basin.
In the warm deserts to the south and east, rainfall is concentrated in
late summer thunderstorms, although a secondary rainfall peak sometimes
occurs during the winter. When rainfall comes,| it often is brief and intense,
producing flash floods, which in turn produce geomorphically distinct desert
streams with their diagnostic riparian vegetation.
Because of low and highly seasonal rainfall throughout much of the
region, riparian wetlands (sensu lato) are especially important. River and
stream systems collect water from large areas, import water from places with
higher precipitation, or transport it from areas of large snow accumulation.
They are islands of moisture in arid or seasonally arid landscapes and thus are
associated with most of the regions perennial water and wetlands. Small, iso-
lated wetlands are uncommon, except on the western slopes of mountain
ranges, where orographic precipitation is significant. Elsewhere, isolated wet-
lands tend to be ephemeral, appearing after rain! events only to quickly evapo-
rate, sometimes for years at a time (e.g., vernal pjools, playas, and,saline lakes).
The other major group of perennial wetlands in the region are the
coastal wetlands of California. Because of the regions' active geology and asso-
-------
Southwest Region Compensation Costs , . . " . 2 .
elated uplifting/ much of the .California coast is steep and rocky, with few shal-
low areas in which coastal wetlands could develop. Historically, therefore,
coastal wetlands were concentrated in estuaries and embayment, especially
San Francisco Bay. . ,
Endangered species are significant components of wetland ecosystems
throughout the region, for a variety of reasons.
• Because, of many thousands of years of evolution in isolated aquatic
habitats, a relatively large number of endemic species are found in wet-
lands and other aquatic habitats in desert regions of the Southwest. As
human activity has altered the hydrology of desert springs and major
river systems, the endemic species that depend on them have declined,
in many cases becoming endangered or threatened.
• In Nevada's Great Basin, extensive Pleistocene lakes .were replaced by
much smaller water bodies as evapotranspiration began-to outstrip pre-
cipitation. Once interconnected populations of aquatic organisms have
become fragmented as the lakes shrank. Aquatic refugia within the
basin are often saline and hot, presenting an especially harsh envi-
ronment for small aquatic organisms. Strong selection, isolation, and
thousands of years have produced several groups of relatively young
endemic species, sometimes adapted to very specific harsh environ-
mental conditions. ' .
• High degrees of endemism 'are also found throughout much of
California, apparently the result of biogeographic isolation provided by
surrounding mountains and- deserts. Here, too, as human activity has
destroyed most of the state's wetlands, and as the region's appetite .for
water has grown, species dependent on wetlands and other aquatic
ecosystems have been hit especially hard.
Wetland Losses
Estuaries and river basins have been favored areas for human settle-
ments since before the development of writing. The availability of water for
transportation, drinking, irrigation, and removal of wastes makes riparian ar-'
eas and coastal estuaries especially attractive areas for human settlement.
These areas are especially attractive in arid landscapes, in which.water is rare
or seasonally unavailable. Thus human 'settlement in arid, regions tends to
focus around water courses, wetlands, and coastal embayments. Indeed," the
impacts of humans on the wetlands of the Southwest have been substantial/
Significant wetland losses occurred very early in the European settlement of
the region.1
By the mid 1980s, wetland losses throughout the region were. substan-
tial. California had lost 91% of its original wetland inheritance (more than
-------
Southwest Region Compensation Costs
any other state) Arizona had. lost 36%, and Neva|da,-52% (Dahl 1990). While
the losses in the desert states are at or below the;national average, the losses
are notable for two reasons. First, population densities over most of these
states are very low (the majority of the population lives in a few urban areas).
Thus, relatively small rural populations have paused significant wetland
losses. Second, these states originally had relatively low percentages of their
surface area as wetland (1.3% for Arizona, 0|.7% for Nevada, 4.9% for
California), thus wetlands were never widespread, and continued losses have
made them even less so. . ;
Agriculture, irrigation, and massive wate|r projects have taken their
toll on wetlands throughout the region, not only jby'directly eliminating wet-
lands, but also by degrading those that are left. Large water projects have
flooded extensive riparian areas and removed large quantities of fresh water
from river flows throughout the region, (thusj drying out many riparian
wetlands, altering the timing and severity of peak stream flows, and
otherwise changing hydrologic and sediment transport characteristics of the
region's rivers and streams on which remaining wetlands depend).
Development of riparian areas for agriculture has been extensive throughout
the region; these areas were relatively flat, close to water, and easily irrigated.
Irrigation return flows have significantly altered the ecology of rivers and
other surface waters by increasing concentration? of dissolved solids in river
water. These arid-land impacts to wetlands and surface water are in addition
to the more widespread problems of nutrient enrichment, sediment loading,
and toxic discharges associated with human activities throughout North
America and the world. .
Major wetland losses in California have loccurred in the Central San
Joaquin and Sacramento River valleys (Herbold and Moyle 1989) and in the
San Francisco Bay region (Josselyn 1983). In the San Joaquin-Sacramento
River delta, many of the natural freshwater wetlands have been levied and
drained for agriculture. The first such levees were constructed as early as 1852,
and the 60 largest delta islands had all been converted to agricultural use
before the turn of the century. Small but widespread vernal pools once found
1 throughout California's central valley were often leveled or plowed, leading
to the endangerment of many typical vernal pool inhabitants.
Early losses of coastal wetlands in the San Francisco Bay region were to
agriculture, but losses to urban and industrial uses also,began early. Many
former wetlands and mud flats of northern San Francisco, San Pablo, and
Suisun Bays were diked before the turn of the century for agricultural pur-
poses, especially grazing. Many still provide fodder for local dairies. Land
speculation and inaccurate surveys helped drive the filling of wetlands in
port and urban areas. Whole communities along the coast of San Francisco
Bay are built on wetland fill: The tidal marshes of the southern San Francisco
•Bay area are naturally more saline than those in the north, and less suitable
for agriculture. However, many of the marshes and mud flats there were
-------
Southwest Region Compensation Costs . .4
diked for salt evaporation ponds.;The first local salt production occurred in
the 1850s. By the middle of this century, few natural marshes were left in the
southern Bay. • •' ' .;•-"''.- .'-•:•
Major wetlands losses in the desert states have Been more -widely dis-
tributed, in part because wetlands were more widely distributed. Wetland
losses have been the result of agricultural development, diversions of surface
water, and pumping of groundwater (Minshal et al. 1989). Direct wetland
losses to urban development are predominately a recent phenomenon.
Regional Economic Conditions ,
California's economy is large and diverse. Recent downturns in the de-
fense and aerospace industries, however, and a series of natural disasters
have hit California hard, and the state remains mired in recession despite the
national economic upswing. Southern California, long a center for aerospace
and military contracting, has been especially hard hit. The Los Angeles area
accounted for approximately half of the jobs lost in the .state during the height
of the recession. Property values, extremely high five years ago, have declined
by as much as 50% over the last few years, most significantly in southern
• California. California housing and construction industries are still in a
slump, reducing short-term pressure on the State's wetlands.
Paradoxically, the recent L,os Angeles earthquake may help pull
California out of its recent economic malaise'by forcing households and busi-
nesses to invest in rebuilding rather than making other investments put-of-
state. The housing and construction industries, in particular, are expected to
get a boost as earthquake victims rebuild. The influx of federal assistance to
the region, along with the multiplier effect of the increase in construction ac-
tivity may trigger an economic'resurgence and development throughout the
region. The effect of these investments on the region's wetland resources,
"however, remain unclear. Much will depend on the extent to which new
construction on previously undeveloped land replaces or is spurred along by
the redevelopment of areas damaged by the quake. .
The economic recovery was not as slow in coming to Arizona and
Nevada. Arizona and Nevada have grown rapidly over the last decade or so.
The two states have taken part in sunbelt growth made possible by the devel-
opment of inexpensive air conditioning and large water projects. While
retirement, tourism, and recreational development have been important to
growth in both states in the recent past, both states are making an effort to
attract manufacturing industries as well. Most recent growth has occurred in
and around major urban areas, a trend that is likely to continue, and has been
associated with large-scale land development rather than high density devel-
opment. This has put increasing pressure on wetlands and other features of
the -environmental landscape. ' „. .
-------
Simthuvst Region Compensation Costs
ANALYSIS OF REGION-SPECIFIC DATA
Met °ThP analysis of regional" wetland creation/ restoration, and
na"° dtvrdaetf°atoneP We herefore supplemented these records with cost
engineering cost-accouiumg ,,' u ..• ' TAr;tVi Qnhrontracted wetland
' 4.- oiects developed in collaboration witn suDcontiav.i.c\j.
-------
Soutltzuest Region Compensation Costs ' ]6
project costs change as project size -changes and to produce estimates of per
acre project, cost adjusted for project size. Reported results, except where
otherwise noted, are based on hypothesis tests with p<0.05.
There was also an extremely uneven distribution of cases within and
among project categories. Freshwater emergent wetland creation projects
were abundant in our .sample, for example, while projects to restore beds of
submerged aquatic plants were rare. This pattern, which reflects both the
frequency with which specific wetland types are restored or-"created
nationwide, and the vagaries of data collection, limits.the types of statistical.
• cost comparisons that are possible. The results presented here reflect the most
complete analyses possible with the existing databases.
'/
Nationwide Background
Wetland creation and restoration projects in the primary database were
separated, into eight project categories for analysis. These categories include:
(1) Aquatic Beds, consisting of tidal or nontidal communities of perma-
nently or nearly permanently submerged plants;
(2)' Complex Projects, incorporating three or more wetland types in a
single project; ''..-.'.
(3) -Freshwater Mixed Projects, consisting of nontidal projects in which
both- forested and emergent vegetation is produced;
(4) Freshwater Forested Projects, establishing woody vegetation (forest or
shrub) in nontidal wetlands;
(5) Freshwater Emergent Projects, establishing emergent wetlands in non-
tidal wetlands;
(6) Tidal Freshwater Wetlands Projects, often consisting of mixed emer-
gent and woody vegetation;
(7) Salt-marsh Projects and other marine or estuarine projects, establishing
wetlands dominated by emergent vegetation; and
(8) Mangrove Projects, establishing mangrove communities.
Differences in the costs of restoring different types of wetlands are not
large relative to the differences in costs within any one wetland category. This
reflects the enormous differences in the site and project design characteristics
within project categories-and frequent similarities among the tasks required
to restore wetlands in different categories. Median, mean, minimum, and
maximum per acre creation and restoration costs for the eight categories of
-------
Southwest Region Compensation Costs |
wetland projects just described and for agricultural conversion projects (from
the secondary database) are shown in Figure 1. .}
Cost Per Acre
(In 1993 $; excludes land costs)
"3T
1 (
! •
" high
> mean
i median
• low
1 ? «,?,
-1 ^ 1 — 1 i 1 ; " •
1 ' ^ >- /n
£ g "S "S 1 £ -e | .,
cg"g Q. J= ' § "5 5 P5 !-= 5 £
§-m£ 2£££|« S 5,
< 55 £EF||
S 0
^ , '^
"3 a)
0 >
•=: c
O) O
Wetland Type
i
I
Figure 1. Point estimates and ranges of project costs from the primary
database for specific project categories,;
Table 1 displays summary cost statistics by wetland category based on
the results of the nationwide study (primary data, except for agricultural
conversion data). Similar, detailed breakdowns of project costs from within
the region itself are potentially misleading because of small sample sizes
National-regional comparisons are given below: The table also includes cost
breakdowns by preconstruction, construction and postconstruction tasks, and
by input category (labor, materials, equipment and other). Region-specific
differences discussed elsewhere in this report will effect some of these values.
-------
Southwest Region Compensation Costs
Table 1. Cost Estimates and Cost Allocation (excludes land cost) From
the National Study.
- Aquatic. Bed
. Project Costs (Thousands)
Average ' $19.5
Minimum 18.3
Maximum 21.7
Median 18.6
Sample Size' ,, 3
Complex
$56.7
4.3 '
258.8
24.8
8
Project T\
FW ,
Mixed
$25.3
1.4
65.8
23.4
10
^pe ~
FW
Forest*
"$77,9 '
0.9
248.4
42.7
19
' FW
Emerg.
$48.7
1.7
170.6
35.2
28
Tidal
FW
$42.0
.0.6
92.6
32.9
3
Salt
Marsh
$18.1
1.0 .
43.6
10.2 '
9 •
Man-
grove
$18.0
2.1
42.8
13.6 ..
".4 -
Agric.
Conv**
$1.0 •
0.005
20.8
0.5
494
Breakdown by Tasks: .
Preconstruction ' 1 7%
Construction 63
Postconstruction 20
10%
74
16
5%
78
17
9%
74
18
13%
58 .
28
9%
87
4
16%
73
13% '
66
21
0%
100
•0
Breakdown by Input Category: , .
Labor 58%
Materials . 8
Equipment 34 . '
Other . 0 ' ' .
50%.
23
14
14
74%
10 :
16
0
51%
30
18
2
63% •
26
9
'1
31%
54
14
1
52%
27
•20* '•
2
51%
21
28
0
45%
. 0
.55
0
High end of rartge involves researching and restoring hydrology and planting; low end involves
restoring hydrology only. ' . '.
** - Agricultural Conversion data are derived from the secondary data. Cost breakdowns for
agricultural conversions are based on a project consisting of hydrologic modification without
. planting or formal plan development.
Regional Differences—Primary Data
Our Primary database contains 32 projects from the southwestern re-
gion, all from California. Approximately half of these projects are from
southern California and half from northern California. Projects from the
south were primarily intended to produce wooded wetland habitat for the
Least Bell's Vireo, but we also have data for a vernal pool creation project and
a few other emergent wetland projects. Northern California projects were
more diverse, including riparian restorations, and complex wetland en-
hancement projects in estuarine-palustrine wetland complexes along the San
Francisco Estuary. . .' ' -
An analysis of covariance was used to determine whether wetland cre-
ation and restoration projects in our Primary Data from California were dif-
ferent in cost from similar projects elsewhere in the country. They were.
Wetland projects in .California were approximately double the cost of similar
projects elsewhere in the country. A more complete analysis, adjusting for
wetland project type (creation, restoration, enhancement), shows a slightly
larger cost difference. California projects are about 2.3 times as .expensive as
projects elsewhere in the country. _ '. • • .
-------
Southwest Region Compensation Costs
Primary Data: Southwest vs. Other Regions
1,000,000 T
Southwest (•)
0.1
1000
Figure 1. Primary data—Southwest
Regional Differences—Secondary Data
Of our sample of 397 projects '(excluding; agricultural conversion pro-
jects), 46 were located in the Southwest region. All but two were from
California. Of the 46 projects, 20 were wetland creation projects, while 5 were
restoration, 5 enhancement, and 2 mixed projects. We could not determine
project type for the remaining 14 projects. The frequency of wetland creation
projects in this sample is lower than in the rest of the database (by chi-squared
test, X2 = 15.647, p=0.0013). Projects in the southwest were substantially larger
than projects elsewhere in the country (t-test on log-transformed project sizes,
t=4.034 p= 0.0001). The analysis of covariance showed no significant differ-
ences in per acre project costs between projects in the southwest, and those
elsewhere.
-------
Southiv'est Region Compensation Costs
10
Secondary Data: Southwest vs. Other Regions
10,000,000 T
1,000,000
100,000
10,000
If)
o
o
1
Southwest (•)
0 _\ _ »o#
r,000 -- Other (o)
100
10 -j-
-i
0.001 0.01 0.1 1 10 • 100 . 1000
Size (acre)
Figure 2. Secondary data—Southwest.
DISCUSSION
Characteristic wetland creation and restoration projects in the south-
-west United States are determined, in part by the meteorological and ecologi-
cal conditions in the region, and in part by the -limited opportunities that exist
for restoration and the high demand for mitigation sites. The need for wet-
land mitigation, by definition, is tied to the level of development activities,
both public and private, that harm wetlands. The politics of water projects
have changed dramatically, and it appears that the future will bring few
massive, federally subsidized water projects, which will limit wetland impacts
from water projects. The swampbuster provisions of the 1985 and. 1990 Farm
Bills, the wetlands reserve program, and changing priorities in agricultural
programs ^generally are likely to continue a nationwide trend away from
destruction of wetlands for agriculture. Urban development and associated
highway and utility construction, however, in general keeps pace with
development, and is likely to accelerate as the current economic recovery
expands throughout the Southwest.
All three states have very low percentages of their land area in wet-
lands (0.4% in California, 0.8% in Arizona, 0.3% in Nevada), which in princi-
ple should make avoidance of wetland impacts easier than in states with a
higher percentage of the land in wetland. However, many of the region's
metropolitan areas are near rivers and the sea,, so much wetland develop-
.ment will continue to be in areas with higher-than-average concentrations of
wetland. Many wetland impacts will come from linear projects like road and
-------
Southwest Region Compensation Costs
11
I
utility construction that must cross waterways in order to fulfill their basic
purpose. Demand for mitigation may also be increased by endangered species
impacts, some of which will be mitigated through creation or restoration of
wetland. ;
Restoration outside of the context of mitigation is likely to remain less
common in the Southwest than elsewhere in the country. Regional demand
for water withdrawn from rivers and stream systems for human uses has
grown explosively over the last several decades;. Reserving water from hu-
man distribution for environmental purposes will prove expensive, legally
difficult, and politically unpopular. However, without available water,
restoration of many former wetland areas will be difficult or impossible, and
those restoration projects that are possible will be relatively complex a'nd ex-
pensive. The Southwest has relatively few opportunities to pursue the kind
of inexpensive agricultural conversions that have proven so successful in the
Midwest and those few opportunities are relatively expensive, as agricultural
conversions go. In northern California, and along the coast, where restoration
is not hindered by a lack of water, property values are high, and many
degraded or converted wetlands are in profitable use for agriculture or salt
evaporation. Acquiring permission to restore wetlands on these lands, or
outright acquisition of restoration sites' in these areas are likely to be
expensive. Restoration and enhancement projects outside the context of
mitigation is likely to continue to focus on lands held by local, state, or, federal
governments, and on riparian restorations that can be carried out without
removing large areas of land from existing profitable uses.
Regional climate conditions and the history of human use of the land
combine to make typical wetland creation and restoration projects through-
out the Southwest unusually complex, and therefore expensive. The regional
context for wetland restoration in the Southwest can be summarized as fol-
lows: . ; , -
1
(1) Wetland creation and restoration projects in the region face unusual
technical challenges because of a lack of available surface water. Careful
planning, engineering, and construction are often necessary to ensure
simultaneous compliance with storm Water management require-
ments and provision of adequate water supplies -for the wetland itself.
"1 -
(2) Many wetland creation and restoration projects are undertaken, in part
to satisfy requirements for mitigation of losses of endangered species
habitat (e.g., the San Diego Mesa Mint, Light Footed Clapper Rail, and
Least Bell's Vireo). Endangered species compliance increases project
complexity, as well as monitoring and follow-up costs, and greatly in-
creases project costs. • . 1
(3) The concentration of wetland impacts in and around urban areas.also
increases project complexity. Creation or; restoration of natural or semi-
-------
Southtoest Region Compensation Costs 12
natural wetlands in or near urban centers is especially challenging.
While this is especially true of freshwater wetlands, for which provi-
sion of appropriate hydrologic conditions.may be impossible in an ur-
banized watershed, an urban,context also increases the difficulty of tidal
•restorations and limits the ability to restore populations of plant polli-
,, nators, predators, and .other key ecosystem links. -
(4) Project complexity may also be increased by the preponderance of ripar-
' • ian wetlands in the desert parts of the region. Riparian restorations can
,be expensive because of the complex engineering sometimes required
to ensure stable channels in flowing water systems. Restoration of low-
flow wetland systems seldom requires as much engineering expertise
as flood-prone riparian systems. Unfortunately, it is difficult to
compare the costs of riparian, restorations and typical wetland
restoration, since the size of riparian restorations is more often and
more appropriately measured by linear stream length .and stream
discharge, rather than area.
(5) Where restoration of coastal wetlands is undertaken, simple plantings
• are seldom sufficient. Restoration efforts often must take into consid-
eration extensive hydrologic modifications that have occurred in the
Southwest over nearly 150 years of development activity in the re-
gion's wetlands. A majority of the restoration and enhancement pro-
jects from the region in our Primary database included substantial hy-
drologic modification, such as installation, of culverts, re-establishment
of tidal guts, and so on! , •
(6) Project costs throughout much of the region are also inflated because
labor and equipment costs tend to be higher than those found in the
. rest of, the country. Construction costs in 12 California cities, as well as
in Reno and Las Vegas, Nevada, are slightly above the average for cities
nationwide (Smit and Waier 1991). The construction cost differential in
California varied from 7% above average in Vallejo to 25% above
average in San Francisco. In Las Vegas and Reno, construction typically
. costs 3% and 4% higher than the national average, respectively.
Construction costs in Arizona are 8% (Tucson) or 9% (Phoenix) below
national averages.
-------
Southwest Region Compensation Costs
13
BIBLIOGRAPHY
Dahl, T. E. 1990. Wetland Losses in the United States 1780's to 1980's. U.S.
Department of the Interior, Fish and Wildlife Service, Washington,
DC,21pp. I
Herbold, B., and P. B. Movie. 1989. The Ecology of: the Sacramento-San Joaquin
Delta: A Community Profile. U.S. Department of the Interior, Fish and
Wildlife Service Biological Report 85 (7.22). Washington, DC, xi + 106
pp.
Josselyn, M. 1983. The Ecology of San Francisco. Bay Tidal Marshes: A
^Community Profile. U.S. Department of the Interior, Fish and Wildlife
Service, Division of Biological Services, FWS/OBST83/23. Washington
DC, 102 pp. "I
Kiner D M and C. C. Bohlen. 1994a. Making Sense of Wetland Restoration
°' Costs. CEES Technical Report UMCEESrCBL-94-045, January 1994.
University of Maryland, Center for Environmental and Estuarine
Studies, Horn Point, MD.
King, D. M. and C. C. Bohlen. 1994b. A Technical Summary of Wetland
' Restoration Costs in the Continental United States. CEES Technical
Report UMCEES-CBL-94-048, April 1994. University of Maryland,
Center for Environmental and Estuarine Studies, Horn Point, MD.
Minshal, G. W., S. E: Jensen, and W. S. Platts. 1989. The Ecology of Stream and
Riparian Habitats of the Great Basin Region: A Community Profile.
U.S. Department of the Interior, Fish and Wildlife Service, Biological
Report 85(7.24). Washington, DC, 142 pp.
Smit, K., and P. Waier, eds. 1991. Means Landscape Cost Data. Kingston, MA:
R. S. Means Co. - .
-------
Southwest Region Compensation Costs
14*
APPENDIX A: ANALYSIS OF COVARIANCE TABLES
The following analysis of covariance tables provide statisticaldetails for
the conclusions presented in the main text. All analyses were performed on
logio-transfofmed data. The tables show partial sums of square and F ratios,
testing.the hypothesis that the particular source of variation is associated with
more of the variability in cost among projects than can be, accounted for by
chance. .' . ' . .
Table A.I. Analysis of Covariance for the Primary Data Comparing
Southwestern Wetland Projects and Projects from other Regions
of the Country.
ANCOVA Table ,
Source
Logio(Acres) .
Region
Logio(Acre)* Region
Model
Error .-
Total
DF
- 1
1
'1
'3
83
86
Sum of
Squares
7.2183
1.2839
0.1862
7.8771
19.8445
27.7216
Mean
Square
7.21831 '
1.28396
0.18625
2.62572
0.23909
F Ratio
30.1907
5.3702 '
0.7790
.10.8262
Prob»F
' 0.0000 >
0.0229
0.3800
0.0000 '
Parameter Estimates. - •
Logio(Acres)
Southwestern
Other
Slope
-0.4290
Std Error
Least Sq.
Mean
4.6149
4:3148
Std Error
0.093471
0.065105
N
88
32
56
Table A.2. Analysis of Covariance for the Secondary Data Comparing
Southwestern Wetland Projects and Projects from other Regions
of the Country
ANCOVA Table
Source
Logio(Acres)
Region
Log10(Acre)* Region
Model
Error
Total
DF
1
1
. 1
3
393
396
Sum of
Squares
12.280460
1.062403
0.155829
41.88826
213.79980
255.68806
Mean
Square
12.280460
1.062403
0.155829-
13.9628
0.5440
F Ratio
.22.5736
1.9529 .
0.2864
25.6659
Prob>F
0.0000
0.1631
0.5928
V
o.oodo ;
Parameter Estimates
Log10(Acres)
Southwestern • •
Other
Slope
-0.4112
Std Error
Least Sq.
, Mean
4.5686 .
4.3468
Std Error
0.13624
0.03946
N
397
46
- 351
-------
------- |