|
Items in cart [4] |
Questions? Call 888-624-8373 |
HARDBACK + PDF HARDBACK PDF BOOK PDF CHAPTERS Free Resources |
||||||||||||||||
|
||||||||||||||||
|
|
|||||||||||||||
| ||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||
| Copyright © 2009. National Academy of Sciences. All rights reserved. Terms of Use and Privacy Statement |
Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 116
5
Local and Regional Indicators
INTRODUCTION
Indicators are needed to inform us about ecological status and trends
at all spatial and temporal scales, and at a variety of levels of specific
ity, ranging from the status of local populations to the functioning of
large ecosystems. Because space and time are both continuous variables,
scales of applicability of indicators blend into one another. Indeed, many
indicators are useful at several scales. For example, the indicators we
recommend in this chapter for forest condition can be aggregated usefully
at regional, national, and continental scales.
In addition, most policy and management decisions are made at scales
defined by laws and regulations established by political entities, such as
local municipalities, counties, states, and the federal government. Although
the committee focused its attention on the national-level ecological indi-
cators recommended in Chapter 4, the methods used to select and formu-
late those indicators are equally applicable to indicators designed for use
at smaller spatial scales. Further, many national-level indicators can be
reported at various levels of disaggregation to serve as regional ecological
indicators. In this chapter, we examine a number of local and regional
indicators that we judge to be especially important, and show how they
can be computed and interpreted.
116
OCR for page 117
LOCAL AND REGIONAL INDICATORS
PRODUCTIVITY INDICATORS
117
In addition to a national-level indicator of ecosystem productivity, it
is also useful to have indicators specifically designed to capture the
performance of particular ecosystem types. In this discussion, we give
examples of indicators for forested ecosystems. Similar indicators can
and should be developed for other vegetation types, such as grasslands,
savannas, deserts, and wetlands.
FORESTS AS AN EXAMPLE
For regional forest indicators, we recommend indicators of produc-
tivity and species diversity, structural diversity, and sustainability. These
attributes support the continued provision of the following goods and
services from forests: wood and wood products, opportunities for recrea-
tion, tourism, and aesthetic enjoyment, maintenance of wildlife resources,
control of erosion and nutrient losses to surface waters, and mitigation of
. . .
greennouse-gas emissions.
The most valuable indicators for forests are those that can provide
early warning of adverse trends in productivity, species diversity, and
structural diversity. Productivity integrates the flow and storage of car-
bon with flows of nutrients, water, and light. It provides the sustained
yield of wood products. In addition, with the increased concentrations of
carbon dioxide in the atmosphere, management of forests for carbon stor-
age has assumed great importance (Cooper 1983, Harmon et al. 1990~.
Species diversity is also an important indicator of the condition of forests,
if for no other reason than that most species on the Endangered Species
List inhabit forests (Doyle 1998~. Structural diversity of forests includes
such features as crown condition, foliage-height profiles, and amounts
and status of coarse woody debris; these attributes are all important for
animal habitat (MacArthur et al. 1962; Franklin et al. 1981, 1989; Franklin
and Forman 1987; Spies et al. 1988~. The three features of forests that
indicators address also provide opportunities for recreation and tourism
and contribute towards maintaining the aesthetic quality of the nation's
forests.
The development of a program for monitoring the status and trends
of the nation's forested ecosystems is a continuing research effort that has
sound practical underpinnings. Continuous inventory programs of the
U.S. Forest Service form the basis of monitoring various aspects of forest
structure and productivity. These inventory programs are positioned to
take advantage of the existing theoretical base provided by individual-
tree simulation models (Shugart 1984~.
We first review the current forest inventory programs that provide
OCR for page 118
118
ECOLOGICAL INDICATORS FOR THE NATION
the empirical basis for forest-status monitoring. We then discuss those
current forest ecosystem models that may provide a theoretical basis for
evaluating forest functioning, as well as for forecasting future status and
trends in forest productivity and diversity. Based on this review, we then
recommend specific indicators of the status of forests, focusing on how
these indicators relate to current inventory programs and ecosystem
models.
Current Forest Inventory Programs
The U.S. Forest Service has periodically inventoried the status of the
nation's forests through its Forest Health Monitoring Program (FHMP)
and the older Forest Inventory and Analysis (FIA) Program. A nation-
wide network of plots for the FHMP has been partially implemented
(Anonymous 1996~. In this program, forests are inventoried in plots
spaced every 27 km on a nationwide grid network. The plot network
currently exists in 15 states, mainly in the Northeast and Southeast, and in
scattered areas in the rest of the country. At each grid point, various
subplots are sampled for a variety of attributes. The size of each subplot
is determined by the ecological scale of each attribute: the subplots range
in area from 2 m2 to 1 ha. These plots are sampled every four years for
traditional timber-yield data on tree-diameter distributions, tree species,
and site index. Canopy condition, leaf-area index, lichen communities,
scenic beauty, lichen chemistry, foliar chemistry, dendrochemistry, dendro-
chronology, branch evaluation, browse supply, and root condition are
also assessed (Anonymous 1996~. The FHMP currently covers 70 percent
of all forested lands in the coterminous United States. When fully imple-
mented in 2002, the FHMP system will provide detailed data with reason-
able spatial and temporal coverage to detect regional problems in the
nation's forests.
The FHMP system provides detailed data on relatively few plots. In
contrast, the FIA system provides extensive coverage on fewer attributes,
mainly those related to timber volume and forest productivity. In the FIA
program, permanent plots located on lands of all ownership types are
inventoried every decade, mainly to evaluate standing crop of timber, but
also in many cases for assessing understory vegetation, tree seedling
regeneration, disease indicators, and browse availability. For example, in
Minnesota alone, more than 10,000 plots have been inventoried in this
manner since before 1960 in a cooperative program between the U.S.
Forest Service and the Minnesota Department of Natural Resources.
These data form the basis for policy decisions in Minnesota regarding
timber supply from public lands and were the basis for the recently com-
OCR for page 119
LOCAL AND REGIONAL INDICATORS
119
pleted Generic Environmental Impact Statement for Expansion of the Pulp
and Paper Industry in Minnesota (Jaakko-Poyry Consulting 1992~.
The FIA plots provide information on diameter, height, and species of
all trees and numbers of seedlings by species. From these data, biomass,
tree species diversity, and mortality can be calculated. On some plots,
browse supply and condition are also measured to evaluate ungulate
habitat. These FIA data can corroborate trends measured in more detail
in the FHMP plots, and they provide more detailed coverage of the
productivity and diversity of forested lands.
To be of even more use, the FIA system requires more complete data
archiving and quality control. In particular, the locations of the FIA plots
need to be determined accurately using global-positioning systems, rather
than the current method of permanent stakes and survey markers, which
are sometimes lost. Finally, additional FIA plots need to be established to
reflect more accurately the distribution of land in various ownerships.
For example, although the FIA plots in Minnesota are distributed across
all ownership types, there are few FIA plots on national forest land in
other states, particularly in the Pacific Northwest.
Current Simulation Models of Forest Ecosystems
The raw field data collected by the FHMP and FIA programs can be
imported into individual tree-based ecosystem models (Shugart 1984) to
project potential trends, given hypotheses about how various stressors
(e.g., climate change, acid rain, and harvesting) affect tree physiology and
stand population dynamics. The development of these ecosystem models
in recent decades provides a theoretical basis for analysis and projection
of the data (see examples reviewed in Agren et al. 1991, Mladenoff and
Pastor 1993, Pastor and Mladenoff 1993~. These models project the
diameter and height growth of individual trees on a plot approximately
of 0.1 ha (approximately the same scale as the FIA and FHMP plots),
subject to constraints of growth limitations, including light limitations
through shading, temperature, water, and nutrients (Shugart 1984, Pastor
and Post 1986, Pacala et al. 1996~. The models have been tested against
independent data on successional trends, productivity, species diversity,
and nitrogen cycling throughout eastern North America (Shugart 1984,
Pastor and Post 1986, Pastor et al. 1987, Pacala et al. 1996~. Versions of
these models also exist for the Pacific Northwest (Keenan et al. 1995~. An
example of using one of these models to determine plot sampling regimes
for monitoring status and trends is detailed in Appendix A.
The combination of established, long-term monitoring programs and
a suite of extensively tested simulation models operating at the same scale
as the data provides a sound basis for an integrative program to assess the
OCR for page 120
20
ECOLOGICAL INDICATORS FOR THE NATION
status and trends of the condition of the nation's forests. We now turn to
specific recommendations for indicators that will enhance the usefulness
of these models and inventory programs.
Recommended Indicators for the Status of the Nation's Forests
We recommend that the following forest indicators be given high
priority: (1) productivity and tree species diversity, (2) soils, (3) light pen-
etration, (4) foliage-height profiles, (5) crown condition, and (6) physical
damage to trees. We recommend indicators that can be assessed with a
small amount of time spent collecting data on site, that would be ame-
nable to calculation of other synthetic indices (such as various diversity
indices) later in the laboratory or office, and that could be easily incorpo-
rated into existing inventory programs.
1. Productivity and Tree Species Diversity. Productivity and tree species
diversity form the basis of the forest food web; this web is sustained by
the ability of soils to provide water and nutrients and by the ability of the
crown to capture light. The FHMP and FIA programs already collect the
data required to assess the status and trends of productivity and tree
species diversity. These data include tree diameters at breast height, tree
heights, density by species, height classes at which species occur, and
canopy cover for each species within each height class. From these data,
carbon storage and net primary productivity of trees can be calculated, as
well as various diversity indices.
2. Soils. The soil profile should be characterized from a one-time
sampling to characterize structure, texture, and rooting depth (Soil Survey
Staff 1993~. These physical features determine the ability of the soil to
hold water at depths at which it is available to trees. Because these prop-
erties are relatively permanent, there is no need to reinventory them,
except perhaps after several decades. In contrast to water-holding capac-
ity, soil-nutrient availability can change fairly rapidly. Therefore, decadal
sampling of the soil is required to determine changes in soil organic mat-
ter, total nitrogen, exchangeable cations, and forest floor chemistry (C/N
ratio, lignin content, and P. K, Ca, and Mg contents). In addition, nitrogen
availability should be assessed by means of resin bags (Binkley 1984~.
Doing so would require that the plots be revisited the following year to
retrieve the bags, but this technique is able to assess rapid changes in
availability of limiting nutrients (Pastor et al. 1998~. Resin bag measure-
ments therefore can serve as an early warning system of trends in soil
productivity. The indicator of soil status (soil organic matter, recom-
mended in Chapter 4) could enhance the soils component of the FHMP
and FIA.
OCR for page 121
LOCAL AND REGIONAL INDICATORS
121
3. Light Penetration. Any disturbance to the canopy (or recovery of
canopies from prior disturbance) necessarily changes light penetration.
Light penetration can be measured easily by means of vertical photo-
graphs taken with a fish-eye lens, images that can then be analyzed later
in a laboratory after they have been scanned into a computer. Further-
more, because changes in light penetration are the major driving force
behind succession (Pacala et al. 1996), these data would be useful for
projecting status and trends by means of the individual-tree models.
4. Foliage-Height Profiles. Foliage-height profiles are relatively easy to
measure and provide a relative index of bird species diversity and possi-
bly insect diversity. They are therefore an important measure of structural
diversity. The vertical structure of the canopy, specifically foliage-height
diversity, is strongly correlated with bird species diversity (MacArthur
1959, MacArthur et al.1962, 1966~. Foliage-height diversity varies tempo-
rally with canopy development during succession (Aber 1979a) and spa-
tially along soil moisture gradients (Aber et al. 1982~. This diversity can
be measured rapidly by means of a camera that is mounted vertically and
used as a range-finder (Aber 1979b). The vertical distribution of leaf area
and therefore potential bird species diversity can then be calculated from
these data. Such data are also useful for assessing changes in growth as
well as its efficiency and allocation in forest stands (Ford 1982, Beadle et
al. 1982, Waring 1982~.
New technological developments may make it easier to collect such
data. In particular, NASA has developed an instrument, the Vegetation
Canopy Lidar (VCL), that can measure foliage-height profiles from above.
The functioning prototype has operated successfully from the Space
Shuttle and is now mounted on an aircraft. An updated version is being
constructed for launch on a small satellite with a projected launch date of
late 1999 or early 2000. This instrument will have a horizontal resolution
of 20 m and is intended to map canopy heights over the entire surface of
the planet once during its two-year lifetime. This goal may not be achieved,
but to compare aerial or satellite imagery with stand data, extremely
accurate spatial positions must be known for the stands, data that can be
obtained only with global positioning systems.
5. Crown Condition. Crown condition, which reflects the state of the
canopy that accounts for productivity, is correlated with tree energy
status. Trees with a history of poor crowns usually grow more slowly and
have diminished energy and carbon reserves. The latter characteristics
translate into less carbon being available for defense against insects and
pathogens and repair of damage caused by biotic or abiotic agents; en-
ergy reserves are critical in surviving periods of stress. Trends in damage
and crown condition are usually accurate indicators of trends in produc-
tivity and mortality. If severe enough, damage and crown condition may
OCR for page 122
22
ECOLOGICAL INDICATORS FOR THE NATION
be useful predictors of mortality (e.g., Silver et al. 1991), although more
general quantitative relationships have not been developed for those
parameters.
Useful measurements of crown condition are crown ratios (percent-
age of tree height that supports live foliage), crown diameters, density,
transparency, and dieback (progressive mortality of branches proceeding
from branch tips inward). These data may be combined to give composite
measures such as crown volume and crown surface area. Selected mea-
surements, such as crown transparency and dieback, may also prove to be
useful indicators. Continued evaluation may show how crown volume or
crown surface area values can be useful indicators of habitat quality,
especially for birds and insects.
6. Physical Damage to Trees. Trees are damaged by insects, pathogens,
poor management practices, weather-related stresses, and air pollution,
acting alone or in combination. Physical damage to trees after storms,
lightning strikes, fire, and logging provides entry points for insect pests
and diseases. The extent of damage and trends in damage, recorded by
species and age class, can be diagnostic of cause in certain cases, and can
provide a relative measure of likelihood that forest diversity, productivity,
sustainability, and aesthetic value will be compromised. Combining
quantitative measures of damage type and severity with mortality data
could eventually provide a quantitative basis for predicting trends in
valued forest attributes from damage indicators.
Damage categories used by foresters include wounds, evidence of
pathogen attack, brooms, broken branches, broken roots, and damaged or
discolored foliage, buds, and shoots. Weighting the components based on
how likely the damage will effect growth or survival provides an index of
the significance of the damage.
Implementation
The methods of collecting data on these indicators can be learned
easily by field foresters in one- or two-day workshops. Most of the data
processing would take place later in the laboratory through soil sample
analysis or computer analysis. A mere 1 to 2 hours per plot are required
to obtain these data (except for the initial, one-time soil profile character-
ization, which would require an additional 2 to 3 hours), beyond the time
currently spent on collecting the more traditional timber evaluation data.
Given the small increase in time spent in the field and the large advantage
that would accrue from obtaining these data, such a program should
receive high priority for development and implementation.
OCR for page 123
LOCAL AND REGIONAL INDICATORS
INDICATORS OF SPECIES DIVERSITY
123
In addition to the national indicators of the status of species diversity
recommended in Chapter 4, the nation needs indicators to evaluate the
diversity status of a local area, such as a national park or an area exploited
for human use. For evaluating the diversity status of such areas, we
recommend three indicators: Independence of the Area, Species Density, and
Deficiency of Natural Diversity.
Although we tried to reduce the number of these indicators of diver-
sity, and have grounded them in a single well-researched power law, all
three are needed because they each inform us about different aspects of
diversity. As Angermeier and Karr (1994) noted (about different levels of
taxonomic diversity), "no accepted calculus permits integration of counts
of elements across levels.... Arguably, no such calculus should be
sought." We believe this point applies to diversity measures as well. It
follows that the separate aspects of diversity need to be monitored and
reported separately.
Local assessment of species diversity presents a new problem, because
simple counts of species diversity have at least five weaknesses that make
them unreliable.
· Diversity counts are biased by sample size (Fisher et al. 1943~. The
larger the sample, the more species in the count. Simple counts are rarely
complete, and even when they are, one cannot be sure that they are.
Moreover, the species most likely to be missed are the rarest exactly
those that most need protection.
· Diversity counts vary with the extent of the area over which they
are measured. Larger areas have more species (Arrhenius 1921), not
because they are environmentally better, but simply because they contain
more habitats (Williamson 1943~.
· Diversity counts are biased by the length of the period over which
they are measured. More time leads to more species in the raw count
(Preston 1960~. Again, the greater number of species results not from
improved environmental quality, but simply because longer durations
yield greater heterogeneity. A longer period of observation is equivalent
to more habitats in space (Rosenzweig 1998), because different species
require various seasons and various kinds of years to succeed (Chesson
1994~.
· Diversity is a dynamic property of ecosystems (Rosenzweig 1995~.
Simple counts do not tell us whether diversity is sustainable.
· The diversity of any area within a continent depends partly on the
continental matrix in which that area is embedded (Rosenzweig 1995~.
Simple counts ignore that matrix. A species found in a place may persist
there only because favorable conditions are accessible elsewhere.
OCR for page 124
124
ECOLOGICAL INDICATORS FOR THE NATION
Therefore, simple species counts need to be processed and analyzed
before being incorporated into indicators. Fortunately, recent advances
in the study of community diversity provide us with a number of sophis-
ticated practical methods to minimize the sample-size bias (Burnham and
Overton 1979, Chao 1987, Chazdon et al. 1998, Colwell and Coddington
1994~. Our recommended local and regional indicators assume the use of
these methods, and they also correct for the other deficiencies of simple
counts.
In Chapter 4 we showed that samples of area contain a number of
species, S. that fits a power law, S = cAZ, where A is area, and c and z are
coefficients of the equation (Arrhenius 1921, Preston 1962~. A thousand
years ago, most of the sample areas monitored today were subsamples of
a contiguous whole. They exhibited characteristic z values between 0.10
and 0.20 (Rosenzweig 1995~. Now, however, they are likely to be isolated
remnants of the whole, which is a very consequential difference for main-
taining diversity.
Two types of species contribute to local diversity S (Shmida and Ellner
1984, Pulliam 1988, Pulliam and Danielson 1991~. The first are species
whose births exceed their deaths in the area. These are the source species of
the sample. Other species, known as sink species, maintain themselves in
a sample even though their average birth rate is less than their average
death rate, because they frequently immigrate into the area.
Isolating an area, as usually happens when a reserve is set aside, cuts
it off from the immigration that maintains its sink species (Rosenzweig
1995~. The sink species then eventually vanish from the isolate. This
reduces the c value and increases the z value characterizing the area's
species diversity. The c value is idiosyncratic to particular taxa and
regions, but z for isolates tends to be approximately 0.2 to 0.4, much
higher than the z of subsamples (Rosenzweig 1995~.
Knowing the relationship of S to area, the quasi-sustainable diversity
of an area can be estimated by cA03 (where c is the intercept coefficient of
the regional, logarithmic species-area pattern for the taxon being assessed,
and A is the size of the area being assessed.) Quasi-sustainable diversity
is diversity that should persist for many human generations (although
ultimately z rises and S declines, to a degree that is also predictable using
the species-area relationship).
An Indicator of Independence
The quasi-sustainable S suggests an indicator of independence based
on the z values most likely to characterize natural ecosystems. To com-
pute this indicator, first assess the diversity of an area, Si, and of its whole
province, Sw. Let the area of interest be Ai and that of the province Aw.
Ii, the measure of independence, is defined as:
OCR for page 125
LOCAL AND REGIONAL INDICATORS
Ii= [l°gSw - logSi] / 0.2[10gAW - logAi]
125
The value 0.2 in the denominator is the threshold z value: z > 0.2
means no sink species. The rest of Ii is the z value of the area. Thus, if
Ii > 1, the area probably contains few if any sink species and its diversity is
independent of other parts of the province. If, on the other hand, Ii < 1,
then some species living in the area rely on other areas for population
support. These species need to be identified with more traditional demo-
graphic techniques. Their source areas need to be located and preserved
as well.
An Indicator of Species Density
Managers typically wish to optimize the value of their reserves. It
might appear that the more species housed in a reserve, the better its
condition, but this is not necessarily true. As Chapter 4 showed, the form
of the species-area curve means that an adjusted species density reveals
more than raw species densities, Si/Ai. Recall from Chapter 4 that the
adjusted species density of an area is Di, where Di = Si/AiZ. The greater Di,
the more species the preserve maintains relative to the norm.
In calculating Di, use the prevailing or average z value for the biologi-
cal region. Because experience indicates that z is close to 0.3 in isolates, a
value of 0.3 can be used if data are unavailable to estimate z. As in
national assessments, high values of Di must be interpreted carefully
because they may reflect unsustainable overloading of the area. In par-
ticular, if high Di is accompanied by Ii < 1, the high species density is
unlikely to persist.
To see why this is true, consider an area that is not a reserve, but is
used for various residential and commercial purposes. Despite this situ-
ation, suppose the area supports many wild species as well. If changing
patterns of land use within the area squeeze those species into a more
restricted, smaller proportion of the whole, Di will rise and Ii will decline.
If it is known that changing uses will remove a certain amount of the area
from access by wildlife, the initial value of the higher Di and lower Ii can
be calculated in advance. But the increase in Di does not signal environ-
mental improvement because it is likely to decline to its former value.
Once the effective area diminishes, the only way for the system to return
to a sustainable diversity is through reduction in the actual number of
species it contains. The increased Di merely signals an impending loss
of sit
We do not recommend that local diversity managers calculate the M
value (see Chapter 4) of their areas instead of these areas' D value. At the
local level, the indicator needs dictate whether the area is under- or
overdiverse. M deliberately eliminates that distinction.
OCR for page 126
26
ECOLOGICAL INDICATORS FOR THE NATION
Indicators of Deficiency in Natural Diversity
When human uses dominate a landscape, natural assemblages of
species disappear, but they are in part replaced by exotic species. In
Chapter 4, we recommended a national indicator of native species diver-
sity, to indicate the degree to which exotics have replaced native species.
A local indicator that quantifies this tendency is also needed.
For example, consider the difference between the bird species of
Tucson, Arizona, and those of the surrounding natural landscape (Emlen
1974, Table 5.1~. The differences are typical of those seen in commensal
assemblages of most or all other taxa, so we describe Emlen's conclusions
and use them to design an index of deficiency in diversity, Ui.
Tucson sits in an Upper Sonoran Desert basin, surrounded by tracts
of natural vegetation and their resident bird species. The city itself is
mostly a vast suburb with expanses of vegetation supported by urban
irrigation. As a result of extensive watering, the total abundance and
biomass of all bird species has risen more than 26-fold, but most of the
individuals belong to a few commensal and exotic species, mostly house
sparrows, house finches, doves, starlings, and mockingbirds. Since
Emlen's study, more curve-billed thrashers, cactus wrens, Gila wood-
peckers, Gambel's quail, and pyrrhuloxias have moved into the city.
Phainopeplas are more abundant in both the city and the desert. Anna's
hummingbirds have virtually displaced the native black-chinned hum-
mingbirds, and great-tailed grackles, a new commensal, have become
quite common. Some raptors, such as Harris hawks, Cooper's hawks,
and great horned owls, are now regularly seen in the city. But the overall
difference Emlen observed has not changed in the two decades since he
wrote: birds are far more abundant in Tucson than in the surrounding
desert, but the city has fewer native species. Common, widespread oppor-
tunists and exotics account for most of the urban bird biomass.
There are several reasonable and complementary explanations for
why anthropogenic habitats often bring about the loss of many native
species and the burgeoning of commensals. First, anthropogenic habitats
have no evolutionary pedigree; species have not had a chance to adapt to
them. Moreover, people continue to change habitats at rates that are
likely to prevent species from adapting to them. Nevertheless, a few
species have traits that enable them to thrive in highly modified environ-
ments.
Many sets of species that use similar resources have members that
depend for their continued mutual existence on their tolerance of sub-
optimal conditions. Tolerant species cannot dominate the "best" habitat
patches, and intolerants depend on the best habitats for their survival.
When people change a habitat to produce novel conditions, the most
OCR for page 127
LOCAL AND REGIONAL INDICATORS
127
TABLE 5.1 Emlen's study of the effect of urbanization on the avian
species assemblage. Abundance went up 26-fold but diversity declined
from 21 to 15 species. Moreover, many of the local specialties were
replaced by exotics (house sparrow, starling) and widely distributed
commensals like Inca doves, mockingbirds, cardinals and house finches.
Individuals/100 acres
Species Desert City
Gambel's quail 0.3
White-winged dove 0.5 140
Mourning dove 1.9 30
Inca dove 230
Roadrunner 0.5
Black-chinned hummingbird 6
Gilded flicker 1.9
Gila woodpecker 0.3 14
Ash-throated flycatcher 0.8 2
Verdin 2.5 14
Cactus wren 6.8 2
Curve-billed thrasher 6.9 5
Bendire's thrasher 0.2
Mockingbird 0.3 45
Black-tailed gnatcatcher 1.6
Starling 35
Loggerhead shrike 0.1
Brown-headed cowbird 0.4 1
Hooded oriole 0.6
House sparrow 520
Cardinal 1 7
Pyrrhuloxia 0.6
House finch 0.3 170
Brown towhee 1.2
Black-throated sparrow 16.5
Rufus-winged sparrow 2.5
tolerant species are likely to succeed exuberantly, whereas the intolerant
ones become confined to nature reserves.
One reason why so many Old World species have moved in and
exploded in suburban environments to the detriment of natives may be
that they have had more time to adjust to humans. In addition, some
species transplanted to new continents simultaneously escape from their
predators. For example, an Australian native acacia that is not particu-
OCR for page 128
28
ECOLOGICAL INDICATORS FOR THE NATION
larly abundant in the western Australian kwongan where it evolved,
became a scourge in the similarly poor soils and Mediterranean climate of
the similarly diverse fynbos in the Southwestern Cape Province of South
Africa. Gypsy moth outbreaks are common in eastern North America,
but rare in these insects' native Europe. (Diamond [1997] used similar
ideas to understand the broad aspects of the distribution of human civili-
zations and the origin of technological advance.)
Thus, three factors contribute to the extraordinary abundance of a
few species in anthropogenic environments:
· Exotics may have had more time to adjust to humans.
· Exotics may have escaped many of their natural predators.
· Only a subset of native species (the tolerants) are preadapted to
"degraded" environments.
To evaluate the deficiency of diversity in an area of Tucson, one can-
not use the raw value of Di, species density, because it gives the city credit
for exotic species that merely follow human settlement, and for tolerant
natives that would thrive anywhere. An indicator that depreciates the
value of an area according to the proportion of its species that thrive in
anthropogenic habitats is needed. There is no shortage of such habitats,
and there is not likely to be in the foreseeable future.
One way to achieve this would be to recalculate Di, after excluding
the contributions of the tolerants and the exotics. However, because not
enough is currently known to identify tolerant species, the best that can
be done is to exclude the exotics, as was done for NAT IMP (see Chap-
ter 4~. Let
Gi = Si,n/CA '
where Sin is the number of native species at the site and cAZ gives the
number of species expected in a site of area A. (The coefficients c and z are
determined for the taxon in areas of the region free of exotics. Therefore,
cAZ amounts to an alternative expression for Sn.) Because Gi measures
native species density, it makes a better index of local diversity and gives
a truer picture of the value of a place in supporting diversity.
The complement of Gi is an indicator of true deficiency in diversity,
labeled U to signify unnaturalness:
Ui = [cAZ _ Si n] /cAZ.
Ui measures the proportion of native species expected at a site (of area A)
but not found there. Thus, Ui combines the change due to exotic species
OCR for page 129
LOCAL AND REGIONAL INDICATORS
129
with that caused by the overall loss of species. Because values of this
indicator of the deficiency of natural diversity change relatively slowly
with time, decadal monitoring is probably sufficient.
As an example, assuming that Emlen's data accurately reflect what to
expect of 100 acres of Sonoran desert, one can calculate Ui for 100 acres of
Tucson. In 100 acres, cAZ is 21 species. But Sin is 12. Thus, Ui is 9/21 or
0.43. Tucson's 100 acres are 43 percent deficient in natural species diver-
sity. Because everything gets scaled nonlinearly by the power law, that
deficiency is the same for any amount of area in the city.
As another example of an indicator of deficiency of natural diversity,
consider the proportion of nonnative fish species in a region, usually a
state, because most states have agencies that collect data on the distribu-
tion and abundances of fishes, especially game fishes. New techniques
are unlikely to change the ease of obtaining the necessary input data. The
proportion of nonnative species in the fauna of a state is an imprecise
measure because not all exotics are of equal importance. Deciding
whether to list an exotic can be tricky. Some anadromous species that
invade fresh waters only briefly for spawning are counted as exotics.
Other nonnative species, such as guppies in Oregon, survive only in very
specific environments, in this case, hot springs. Conversely, introduced
trout cannot survive the hot summers in most lowland waters of the
eastern United States, but they thrive in cool tailwaters below dams. Some
nonnative fishes are temporary survivors that live for only a few weeks or
months. Crude state lists do not distinguish such species from wide-
spread permanent residents, but the methods described above can adjust
an indicator so that very localized or temporary species do not count as
much as other exotics.
The proportion of nonnative fish species varies greatly among states,
from 0.02 (one [anadromous] species out of 56) in Alaska (Morrow 1980)
to .47 (36 out of 76) for Washington State (Wydoski and Whitney 1979~. In
relatively dry western states with low diversities of native species and
high proportions of dammed rivers, the indicator currently has values
greater than 0.2, and for many states the values are greater than 0.3. In the
wetter eastern states, which also have higher native fish diversities, values
are about 0.07 to 0.08. Values in all states are almost certain to increase
over time because established exotics are almost impossible to extermi-
nate, new introductions continue, and the quality of habitat for native
fishes continues to deteriorate in most states. Because most states have
agencies that collect data on the distributions and abundances of fishes,
especially game fishes, information is typically available by state rather
than by region or ecoregion. Often, however, the information is old or not
systematic, and so it is not always reliable, especially with respect to
diagnosis and distribution of nonnative species.
OCR for page 130
130
ECOLOGICAL INDICATORS FOR THE NATION
The Index of Biotic Integrity:
An Indicator of Species Diversity of Aquatic Ecosystems
Additive multimetric indicators have been developed and used to
compare the species diversity of aquatic systems with what should be in
those systems in the absence of human-caused perturbations (called appro-
priate diversity by the NRC [1994~. The most widely used multimetric
indicator is the Index of Biotic Integrity (IBI), which has been developed
and tested in a variety of aquatic ecosystems (Kerr et al. 1986, Karr and
Chu 1999~. The use of IBI requires general agreement about which organ-
isms indicate by their abundance or absence, poor or good ecological and
water characteristics. The IBI provides a method for quantifying those
qualitative assessments. The IBI is primarily a community-level rather
than an ecosystem indicator because it is based on taxonomic assem-
blages within specific phylogenetic groups and specific biogeographic
regions. The original IBI was developed for freshwater fish communities
in streams in the Midwest. Recently, similar indicators have been applied
to freshwater benthic macroinvertebrate communities in several regions
(and even to some terrestrial communities).
An IBI is calculated from a set of measures of distribution and relative
abundance of selected taxa. Each measure is assigned a numerical value-
an integer ranging from 0 to 6 based on the qualitative judgment of the
index developers. The final IBI, which is the sum of the individual scores
(usually 10 to 12), is unbounded but typically is between 0 and 60. Because
individual scores are discontinuous, statistical analysis of the additive
scores are generally inappropriate. More detail on the mechanisms for
developing multimetric indices was provided by Barbour et al. (1995~.
Typically, IBIs are developed for biogeographic systems such as
ecoregions where similar communities of organisms are expected. For
example, Ohio has developed an extensive set of IBIs that vary by
ecoregion, drainage area, and stream habitat (Ohio EPA 1987~. As men-
tioned above, the statistical properties of additive indices make it unreal-
istic to add or average scores across spatial scales to create a national
indicator (Norris 1995, Gerritsen 1995~. The only effective way to aggre-
gate measures into a multimetric indicator is already incorporated into
the regulatory policies of the U.S. EPA and state environmental agencies.
The Clean Water Act requires that attainment of water quality be reported
on the basis of the number of stream miles meeting the criteria. As IBI
scores have and will continue to be incorporated into state and federal
regulatory standards, attainment will be reported in relation to stream
miles assessed.
Representative terms from entire chapter:
regional indicators
