Indicator Species, Pages 437-450, John H Lawton and Kevin J Gaston.pdf

438



INDICATOR SPECIES



cator species to predict the diversity of other, unstudied

taxa, for scientific or conservation reasons, has proved

to be more contentious and much more difficult.



I. SPECIES AS INDICATORS OF THE

STATE OF THE ENVIRONMENT

There are three distinct uses of the term ‘‘indicator

species’’ in research in ecology and biodiversity. They

are a species, or group of species, that do the following:

1. Reflect the biotic or abiotic state of an environment

2. Reveal evidence for, or the impacts of, environmental change

3. Indicate the diversity of other species, taxa, or entire communities within an area

This article explains, provides examples of, and evaluates each of these uses of the term, focusing primarily on

terrestrial and freshwater ecosystems; broadly similar

conclusions apply to marine ecosystems, but marine

examples lie beyond the scope of the article. We pay

most attention to the third use of the term ‘‘indicator

species,’’ because this seems most appropriate for an

encyclopedia devoted to biodiversity. The most up-todate evaluation and review of indicator species in the

scientific literature is by McGeoch (1998). She concentrates on terrestrial insects as ‘‘bioindicators’’ (in all

three senses of the word) but the general principles

that she discusses extend to all ecosystems and organisms, not just to terrestrial insects.

Everybody knows that living organisms are sensitive

to the state of their environment. Pollution from human

activities kills many species and reduces the abundance

of others. These changes in abundance can be used to

assay the state of the environment.



A. An Example: Acid Deposition

Sulfur dioxide, produced by burning fossil fuel, particularly coal, enters the atmosphere and is eventually deposited on terrestrial and freshwater ecosystems via

three routes: (a) as tiny solid particles, (b) washed from

the air in rain or snow, or (c) as droplets formed in

clouds. Deposition often occurs hundreds of kilometers

from the source. Dissolved in water, sulfur dioxide

forms sulfuric acid, resulting in what is frequently referred to as ‘‘acid rain,’’ but because there are three

principal routes involved in its transfer to terrestrial



and freshwater ecosystems, it is more correctly called

‘‘acid deposition’’ (Erisman and Draaijers, 1995). Sulfur

dioxide is not the only source of acidification; oxides

of nitrogen, again produced by burning fossil fuel, are

also involved, but sulfur dioxide is the main agent of

acidification in most ecosystems.

In terrestrial ecosystems, this deposition kills lichens

and acidifies the soil, leading to changes in the vegetation. Lakes become progressively more acidic as deposition loads increase, until eventually they may become

virtually lifeless. A trained biologist, visiting for the first

time an area subject to acid deposition, will often be

able to deduce that the habitat is being polluted simply

by looking at the species that are present and those

that ought to be there but are not. Beautifully clear

Scandinavian lakes, lacking any fish or amphibians,

supporting few birds and a species-poor and taxonomically unusual invertebrate fauna, have been reduced to

this impoverished state by the transnational export of

sulfur dioxide from coal-burning power-stations in the

United Kingdom and elsewhere in Europe. Here, living

organisms act as powerful indicators of the state of the

environment and the damage being done to it by human

activities, often performed many hundreds of kilometers away.



B. Management of European Rivers

Because the species composition and richness of biological communities change as the environment changes,

we can use species as indicators of the state of the

environment for practical management purposes. The

techniques have been particularly well developed to

assess organic and inorganic pollution in European rivers, managed for recreation, fisheries, and drinking water. The advantages of using living organisms as indicators of water quality are that they avoid the need for

expensive chemical analyses, and, probably more important, organisms integrate the impacts of pollutants

over space and time, All chemical traces of a major

pollution incident may disappear from a river in a matter of hours as the pollution is flushed from the system.

Nonetheless, the biotic community may show evidence

of the damage for many months. It is extremely difficult,

and prohibitively expensive, for chemists to measure

all the organic and inorganic chemical pollutants entering a river, and it is certainly impossible for them to

work out what all the combined impacts of such a

cocktail might be. But living communities reflect the

integrated effects of all the compounds that find their

way deliberately and accidentally into watercourses,

and hence they act as sensitive indicators of the state



439



INDICATOR SPECIES



of the environment. A valuable source of further information on the use of living organisms to monitor the

environmental health of rivers and lakes is provided by

Rosenberg and Resh (1993).

Two widely used European examples of this approach are the German Saprobic Index, and RIVPACS

in the United Kingdom. RIVPACS is now used by the

Environment Agency to manage UK rivers. Both the

German and UK approaches require accurately identified samples to be taken of the organisms found along

sections of the river. RIVPACS focuses on invertebrates,

the German index on invertebrates, microbes, and

higher plants. Both rely on the fact that some species

are extremely tolerant of pollution (the aquatic larvae

of some chironomid midges, for instance), while others

are extremely sensitive, particularly to the low oxygen

levels produced by organic pollution (for instance, the

larvae of many mayflies). The species present in the

samples are given scores, depending on their known

tolerances, and the data from all species are combined

to produce a composite and very sensitive index of

pollution levels for any particular section of a river.



C. Widespread Application

Use of organisms to indicate the state of the environment is widespread, taxonomically and geographically,

for a wide range of environmental issues. Use of species

as indicators of the state of the environment is not

confined to freshwater, or to Europe and North

America. A wide variety of organisms has been suggested, or used, as indicators of human impacts. In

Europe, suites of plant, fungal, and insect species are

only found in, and hence are good indicators of, ancient

woodland; they are entirely absent from plantations,

even though these may be several hundred years old.

The use of lichens as sensitive indicators of air pollution

is well known, but organisms as different as mites and

geckos (agile, climbing lizards) have been used, or suggested, for similar purposes. Lichens have also been

used as indicators of fire history in Brazilian cerado (a

type of dry, scrubby forest), tiger beetles as indicators

of tropical forest degradation in Venezuela, and dayflying Lepidoptera (butterflies and moths) as indicators

of the state of seminatural grasslands for conservation

in Europe. Many other similar examples exist.



D. Interpretation Requires Care

In these and similar cases, considerable care is needed

before a species or group of species can be used as

reliable indicators of damaging (or beneficial) human



impacts on ecosystems. All populations of living organisms fluctuate over time and vary in abundance spatially, because of natural variations in the weather,

normal changes in the physical environment, and fluctuations in the abundances of natural enemies, competitors, and essential resources (food and shelter). Just

because one or more species is declining does not mean

that human impacts are to blame. In the case of lichens

and atmospheric pollution or freshwater invertebrates

and river quality, the links between anthropogenic pollutants and changes in the distributions and abundances

of organisms are thoroughly researched and well understood. But even quite major declines in some species

have proved exceptionally difficult to link to damage

to the environment caused by people.



E. Amphibian Decline

The so-called amphibian decline is a particularly dramatic example (Blaustein and Wake, 1995). In many

parts of the world, population biologists interested in

amphibians (frogs, toads, newts, salamanders, etc.)

have recently become alarmed by apparent major declines in the abundance, and the complete disappearance, of many species from areas where formerly they

were common, often in regions apparently remote from

human impacts. The declines are not happening everywhere, and the magnitude of many of those that have

been claimed is difficult to assess because of the lack

of long-term data prior to the supposed population

collapses; some of them may be perfectly natural. The

worrying aspects of the phenomenon are that while it

is apparently global in scope, the causal mechanism (or

mechanisms) remains obscure. It has been suggested,

for example, that the amphibian decline is indicative

of rising global levels of damaging ultraviolet light (UVB) caused by loss of the earth’s protective stratospheric

ozone layer. Amphibian eggs, exposed in shallow water,

and the adults with their thin wet skins may be particularly sensitive to UV-B, as are human sunbathers without sunblock. Others doubt the explanation. More recently a global pandemic has been implicated. But what

should suddenly trigger lethal outbreaks of disease in

amphibians is unclear.



F. Environmental Toxicology

In all the examples so far, the organisms being used as

actual or possible indicators of environmental health

have been in their natural environment. There is another related but quite separate way in which biologists

use the sensitivity of organisms to set environmental



440



INDICATOR SPECIES



standards, namely in the science of environmental toxicology, or ecotoxicology for short. In many areas of

human endeavor, the aim is to apply some beneficial

technology with minimum environmental damage.

Crop spraying with pesticides is a good example, and

so is the discharge of treated effluent from a factory.

Some environmentalists claim that these types of operations should not lead to any environmental contamination; factories should have zero discharges, and if we

must use pesticides, they should be targeted to reach

only the crop and the pest and not, for example, the

soil, nontarget organisms, or adjacent watercourses.

However, zero discharges or precision pesticides, if

they can be achieved at all, can often only be obtained

at great economic cost. The more pragmatic solution

is to ask whether there are minimal levels of discharge,

spray drift into watercourses, and so forth that cause

no detectable environmental damage. To provide answers to this admittedly difficult question, environmental toxicologists use a wide variety of laboratory

bioassays with standard organisms. Examples from

freshwater include the alga, Chlorella vulgaris, the water

flea, Daphnia magna, the amphipod shrimp, Gammarus

pulex, and the rainbow trout, Salmo gairdneri. The fundamental problem is to try and establish acceptable

levels of contamination. Defining ‘‘acceptable’’ obviously requires political as well as biological judgment.

However, traces of a compound in water, air, or soil

that cause no detectable changes in the performance

(growth, survival, or reproduction) of the test organisms are clearly more acceptable than doses that kill

50% of the population (so called LD50 levels). Basically,

the bioassays seek to set environmental standards for

levels of potential pollutants in soil, air, and freshwater,

using a range of standard laboratory organisms as indicators (Shaw and Chadwick, 1998), but there can be

no absolute standards about what is safe or acceptable.

The general trend in modern societies is for standards

to gradually tighten.



II. SPECIES AS INDICATORS OF

ENVIRONMENTAL CHANGE

If the amphibian decline (discussed in the previous

section) is real, it is an example of a group of organisms

acting as indicators not only of the state of the environment, but also as indicators of ongoing changes to the

global environment, albeit of an unknown nature. In

other words, given that species are sensitive to the condition of their environment, monitoring organisms not



only tells you about the current state of an environment,

but repeated monitoring can tell you about changes in

that environment. To act as indicators of change rather

than current environmental health, it is necessary to

have at least two sets of data on the particular indicator

species in question, taken in the same way, at the same

place(s), on two separate occasions. More frequent sampling allows greater confidence in the direction of apparent trends and the detection of more subtle environmental changes.



A. Not All Monitoring Is about

Environmental Degradation

Not all monitoring of species seeks to record environmental degradation. Increasingly after mining operations, for example, mine operators are required to restore spoil heaps and mine pits by sowing or planting

native vegetation. Monitoring selected groups of common animals on nearby undisturbed control sites and

on the restored land can give a good indication of the

recovery of the entire ecosystem and of the success of

the restoration project. For instance, when biologists

monitored ant assemblages on abandoned, replanted

bauxite mines in Australia, they found that the ants

provided a good indication of the recovery of these

ecosystems. Even after 14 years there were still differences between the ant communities found in the natural

Eucalyptus forest and the restored land.



B. Historical Records of Change

1. Lake Acidification

It may not always be necessary to sample in real time.

When anthropogenic acidification of lakes was first discovered, many people doubted that the phenomenon

was real. In particular, there was considerable opposition to the notion from the power-generating industry,

because solving the problem (by burning low-sulfur

coal, adding ‘‘scrubbers’’ to power station chimneys to

remove sulfur dioxide, or switching to natural gas) was

inevitably going to be expensive. After all, there were

few historic data on the state of the acidified lakes.

Perhaps they had always been that way?

Resolving the problem required knowledge of the

fact that lake phytoplankton (the tiny, unicellular plants

that float in the upper layers of lakes) are extremely

sensitive environmental indicators, because different

species grow best in very different conditions determined by nutrient status and pH (acidity). When algae

die, they sink to the bottom where their bodies and



INDICATOR SPECIES



characteristic pigments are buried and some are preserved (incipient fossils), particularly the resistant, silicious outer cases of a group called diatoms. An undisturbed core through the sediments records the history

of a lake’s phytoplankton, with the oldest flora at the

bottom. Cores showed unequivocally that many Scandinavian lakes that are acid now were not acid before the

Industrial Revolution; the oldest diatoms—species not

found in acid lakes—are gradually replaced in the sample column by acid-tolerant species. Diatoms are wonderfully sensitive indicators of environmental change

(Fig. 1).



2. Plants and Carbon Dioxide

Herbarium specimens (pressed plants collected for taxonomic purposes) and fossil leaves can also be used

as indicators of past environmental change. Another

consequence of the rapid rise in the burning of fossil fuel

since the Industrial Revolution has been an accelerating

rise in the concentration of atmospheric carbon dioxide,

one of the main agents of ‘‘global warming.’’ We will

deal with species as indicators of anthropogenic global

climate change (as it is more accurately known) later.

Here we want to focus on physiological and developmental responses within single species to rising carbon dioxide.



441



If plants are grown in a greenhouse under different

atmospheric carbon dioxide concentrations, from below the pre–Industrial Revolution levels of about 280

parts per million by volume (ppm), through what are

roughly present levels of 350 ppm, to levels that may

be reached by the end of the 21st century (700 ppm),

several interesting things happen. In particular, in the

present context, stomatal densities on the undersides

of the leaves decline. Stomata are the tiny pores in

the leaf surface through which plants take up carbon

dioxide (needed for photosynthesis), and through

which they lose water vapor. It has been known for a

long time that plants control the opening and closing of

stomata to optimize carbon dioxide uptake and reduce

water loss. More surprising, we now also know that

plants grown in high carbon dioxide have lower densities of stomata; something happens during leaf development to reduce the number of stomata. How and what

is currently unclear. Why is simple enough. In a high

carbon dioxide world, the plant needs fewer stomata

to take up the carbon dioxide it requires and hence can

satisfy the needs of photosynthesis and reduce water

loss by developing fewer pores in the leaves.

Now back to those herbarium specimens and fossil

leaves. If you look at 200-year-old (and very precious)

herbarium and modern specimens of the same species,



FIGURE 1 The pH history of Lilla Oăresjoăn, a 0.6 km2 lake in southwest Sweden. A core of the bottom sediments 3.5 m long

records the history of the lake extending back to 12600 BP, using the valves (‘‘shells’’) of diatoms preserved in the deposits.

Different species of diatoms have different pH preferences and can be classified accordingly. Acidobiontic species thrive in acid

waters; alkilophilous species prefer more alkaline conditions. Combining data from the remains of all species of diatoms allows

the pH history of the lake to be reconstructed. The lake has passed through four pH periods. A, an alkaline period after deglaciation.

B, a naturally more acidic period. C, a period with higher pH, which started at the same time as agricultural expansion in the

region, and D, a rapid, recent period of acidification. The post-1960 phase has no similarity with any of the previous periods.

From Renberg, I. (1990), Phil Trans. R. Soc. London B 327, 357–361.



442



INDICATOR SPECIES



FIGURE 2 Changes in stomatal densities of leaves as indicators of changes in atmospheric carbon dioxide concentrations. (A)

Percentage changes in stomatal densities on the lower surfaces of the leaves from eight species of trees and shrubs collected from

the English midlands and preserved as herbarium specimens. The oldest specimen was collected in 1750. As atmospheric CO2

has risen during and since the industrial revolution, so stomatal densities have fallen. From Woodward, F. I. (1987), Nature 327,

617–618. (B, C) Assuming that similar effects occur in all species of plants, fossil leaves can be used as indicators of atmospheric

CO2 concentrations extending back many millions of years. The CO2 content of the earth’s atmosphere appears to have fluctuated

markedly and apparently naturally during the past 400 million years. From Beerling, D. J., and Woodward, F. I. (1997), Bot. J

Linn. Soc. 124, 137–153.



sure enough, stomatal densities decline as global atmospheric carbon dioxide levels increase (Fig. 2A). The

same approach has recently been used to try and trace

atmospheric carbon dioxide levels throughout most of

the Phanerozoic, from the time when plants first colonized the land. Here the method is more contentious,

because different species of truly fossil plants with presumed similar growth forms have to be used in different

geological periods. Nevertheless, the pattern of apparent changes in global atmospheric carbon dioxide concentrations over hundreds of millions of years, revealed

by this method (Fig. 2B and C), are in reasonable



agreement with alternative, independent, and also contentious geochemical methods. Here is a really unusual

use of species as indicators of environmental change.



C. Species as Indicators

of Climate Change

1. The Sensitivity of Species to Climate:

Fossils Again

Current, rapidly rising concentrations of atmospheric

carbon dioxide are the primary cause of anthropogenic



INDICATOR SPECIES



global climate change. However, the earth’s climate has

always changed, naturally, with no intervention from

human beings. One of the ways we know this is through

the careful documentation of the types and distributions

of organisms in the fossil and subfossil record. The

science of paleoclimatology, which seeks to reconstruct

the history of earth’s climate, relies heavily on changes

in fossil and subfossil species assemblages to deduce

what the earth’s climate was like thousands or even

millions of years ago. To take one example, in the

modern world many types of corals occur exclusively

in tropical marine environments; it is a winning bet

that fossil corals of the same type indicate an ancient

tropical sea, even though the rocks bearing the fossils

may now lie in much colder parts of the world.

In more recent geological time, we can use changes

in the distributions and abundances of plants and animals to trace major changes in the earth’s climate during

the Holocene (the most recent geological past) and

Pleistocene glacials and interglacials. Plant remains preserved in packrat middens in dry air of the southeastern

United States attest a much wetter climate only a few

thousand years ago. Hippopotamus bones and teeth dug

up from under Trafalgar Square provide unequivocal

evidence of a much warmer London. Pollen grains preserved in peats and lake sediments record in exquisite

detail the march northward of European and North

American forests from the end of the last glaciation

12,000 years ago (Huntley and Birks, 1983). The forests

spread with remarkable speed (an average of about 200

m per year, but sometimes as fast as 2 km a year) to

achieve present distributions in the northern parts of

both continents from glacial refugia thousands of kilometres to the south (Fig. 3A and B). The information

is not won easily. It requires huge patience and great

skill to identify thousands upon thousands of pollen

grains extracted onto microscope slides. But once done,

the record reads like a speeded-up movie, as spruce,

oaks, white pine, hemlock, beech, and chestnut swept

north in successive waves through what is now the

United States and Canada; in the more species-poor

forests of Europe, pines were followed by birch, then

oak. These invasions are as dramatic as any in human

history, but they were silent and recorded only by pollen grains.



2. Contemporary Changes in

Species Distributions

Historical changes aside, there is now no doubt that

the world is currently warming quite rapidly. An upward trend in global annual mean surface temperatures



443



is apparent from about 1920, particularly over the last

two decades (from c. 1980); global mean surface temperatures in July 1998 were the highest ever recorded.

Do organisms act as indicators of these changes, perhaps, as with the freshwater species discussed earlier,

acting subtly to integrate several of the changes human’s

find difficult to comprehend in the bald statistics? Climate change does not simply involve warming; it involves changes in rainfall, extreme weather events

(droughts and storms), and even locally cooler conditions. All these complex changes should show up in

changes in the distributions and abundances of organisms.

They do. Species are proving to be extremely sensitive indicators of contemporary climate change, where

historical records allow decent reconstruction of former

and current distributions. Populations of Edith’s checkerspot butterfly Euphydryas editha are disappearing

from southern California and northern Mexico, at the

current southern end of its distribution, and from more

lowland sites; sites where previously recorded populations still exist are on average 2Њ further north than

sites where populations went extinct (Fig. 4). These

are exactly the changes we would expect in a warming

world. Twenty years ago in northwest Europe, little

egrets Egretta garzetta (small white herons) used to be

rare visitors from the Mediterranean. Now they are

breeding in northern France and southern England in

an astonishing expansion of range. Populations of many

other European birds, butterflies, and other organisms

are spreading north at the present time, as the climate warms.

Of course, none of this tells us whether the climate

change that is certainly happening is ‘‘natural’’—it

could have happened anyway and may have nothing

to do with anthropogenically produced greenhouse

gasses—or whether it is indeed due to human activities. Using species as indicators of climate change

tells us unequivocally that the earth’s climate is changing, but so does the mercury in the thermometer.

What neither tells us is why, and no end of work

on species as indicators will solve that dilemma. As

we have already seen, this situation is not unique to

climate change. It generally holds whenever we use

species as indicators of the state of the environment.

Indicator species can tell us whether an environment

is, or is not, changing. They do not tell us why the

changes are taking place. That almost always requires

additional detective work, although knowledge of an

organism’s biology will frequently provide valuable

clues. Three examples, using birds as indicators, illustrate the problem in more detail.



444



INDICATOR SPECIES



FIGURE 3 The migration of trees across North America and Europe after the end of the last glaciation, revealed by

pollen remains in lake sediments and peat. (A) The spread of oaks in North America, with radiocarbon ages in

thousands of years (contours) and the present range of the genus (shaded). From Davis, M. B. (1981) in Forest

Succession: Concepts and Applications (D. C. West, H. H. Shugart, and D. B. Botkin, Eds.), Springer-Verlag, New York.

(B) Estimates of the oversall rates of spread of trees on two continents, based on data of the type shown in part A.

From Williamson, M. (1996), Biological Invasions, Chapman & Hall, London.



D. Birds as Indicators of Large-Scale

Environmental Changes

Birds are widely used indicators, because in Europe,

North America, and other parts of the world where

there are large armies of amateur bird watchers their

populations and distributions have been recorded well

enough, for long enough, to reveal major environmental trends.



1. Peregrine Falcons and DDT

The catastrophic collapse of peregrine falcon Falco peregrinus populations throughout the northern hemisphere in the 1950s signaled widespread contamination

of the environment by chlorinated hydrocarbon insecticides, first DDT, then other compounds such as aldrin

and dieldrin. The total, and rapid, disappearance of

these dramatic birds signaled to ornithologists that

something was seriously wrong with the environment,



but what? It took a great deal of clever biological detective work (see Ratcliffe, 1980) to link the decline of

peregrine populations to the accumulation of these persistent pesticides up the food chain, resulting in eggshell thinning, reproductive failure, and (in extreme

cases) direct poisoning of adult birds. Although some

populations have now recovered, signaling a recovery

in environmental quality, the species is still missing

from many parts of its former range—some coastal

populations in England, for instance. Nobody knows

why.



2. Migratory Songbird Declines in

North America

In North America, considerable concern is currently

being expressed over widespread declines in summer

migrant birds, particularly warblers. Unlike the socalled amphibian decline, nobody questions the phenomenon; just like the amphibian decline, nobody re-



445



INDICATOR SPECIES



ally knows why it is happening. There is no doubting

the data; many species are indeed declining very

quickly, in as clear an indication as one wants that

something is wrong with the environment, but what?

Several possibilities exist, and they are unlikely to be

mutually exclusive. One explanation focuses on the

destruction of tropical forests in the birds’ wintering

areas. Another suggestion is that there are other unknown problems there or on the migration routes. A

third possibility is extensive habitat fragmentation and

urbanization in the breeding forests of the eastern seaboard. This human modification of the northeast forests

markedly increases nest losses of migrant songbirds to

jays, crows, cowbirds, and racoons, all species that

thrive in the slipstream of urban humans.



3. Declines in Formerly Common

Farmland Birds in Northwest Europe



FIGURE 4 The fate of 151 previously recorded populations of

Edith’s checkerspot butterfly, Euphydryas editha, in western North

America. The populations ranged from northern Mexico to southern

Canada and were visited by Camille Parmisan and other biologists

between 1992 and 1996. Populations that had disappeared because

of habitat degradation (e.g., loss of usable host plants) were omitted

from the analysis. Dividing the populations into five, evenly spaced

latitudinal bands between 30Њ N and 53Њ N (A) reveals that

significantly more southern populations have gone extinct than

northern populations; sites where previously recorded populations

still exist were, on average, 2Њ further north than sites where

populations were extinct. Extinctions were also higher at lower

altitudes (B) (n is the number of populations in each latitudinal

or altitudinal band). Both results are consistent with the effects

of global climate warming on the butterfly, leading to a northward

and upward shift in its geographical range. Reproduced, with the

permission of McMillan Journals Ltd, from Nature 282 (1996),

page 766.



In the intensively agricultural areas of northwest Europe—over the whole of lowland England, for example—a whole raft of formerly ‘‘common farmland birds’’

are also in steep decline (Tucker and Heath, 1994).

They include skylarks (Alauda arvensis), European tree

sparrows (Passer montanus), corn buntings (Milaria calaudra), gray partridges (Perdix perdix), and song

thrushes (Turdus philomelos). Here the problem is now

reasonably well understood, though many details remain unresolved. Modern farming is so efficient and

clean that there is little for the birds to eat. Weeds are

killed with herbicides, which remove both seeds and

rich sources of insects that feed on the weeds. The crop

itself is sprayed to remove insects and is harvested so

efficiently that few seeds are spilled on the way. Modern

farms are biodiversity deserts, an indication of the

power of people to squeeze nature to the margins while

apparently maintaining a green and pleasant land. If

present trends continue, skylarks will be rare birds in

Britain in 20 years.



III. SPECIES AS INDICATORS

OF BIODIVERSITY

A. The Nature of the Problem

Common sense suggests that the known losses of plants

and birds from European farmland will go hand-inhand with much more poorly documented declines in

many other, less familiar and cryptic taxa, from land

snails to glowworms, and hoverflies to harvest spiders.

In other words, changes in the distribution and abun-



446



INDICATOR SPECIES



dance of well-known groups should serve as broad indicators of the status of, and changes in, a much wider

sample of a region’s flora and fauna. The assumption

here is that birds (or other conspicuous species) might

serve as biodiversity indicators—that is, as surrogates

of overall biodiversity. But although it seems intuitively

reasonable to use familiar, well-studied, and easily censused groups as indicators of what is happening to many

other taxa, despite a great deal of research, the idea is

actually contentious.

Following (but slightly modifying) the work of

McGeoch (1998), we can define a biodiversity indicator

as a group of taxa (e.g., genus, tribe, family, or order,

or a selected group of species from a range of higher

taxa) whose diversity (e.g., overall species richness,

number of rare species, levels of endemism) reflects

that of other higher taxa in a habitat, group of habitats,

or geographic region. The idea is simple enough, and

if it can be shown to work, it is important because

biologists then have a relatively simple means of assessing overall biodiversity for purely scientific reasons,

for setting conservation priorities, or for monitoring

the effectiveness of conservation management.



B. Taxa That Have Been Suggested as

Indicators of Biodiversity

The groups of organisms whose richness has been evaluated most thoroughly in the greatest number of places

on earth are also the most familiar. The natural history

sections of bookshops are dominated (sometimes exclusively) by volumes on plants and birds. If insects figure

at all, butterflies will be on the top of the list, although

there are fascinating differences between nations. Japan

loves dragonflies. Birds, higher plants, butterflies, and

dragonflies are all groups that occur in most places in

the world but whose individual species are seldom so

widespread. In much of the world they are also groups

whose species are, relatively speaking, taxonomically

well known and stable, readily identifiable, and have

biologies that are well understood. They are easy to

find, inventory, and count, and they are reasonably, but

not overwhelmingly, diverse in any one place. These

are all desirable attributes of groups that mightbe used

as indicators of the diversity of many other, much less

well known taxa—that is, as indicators of the overall

biodiversity of a region.

Other groups have many of these same attributes

but have not gained the same popularity, perhaps because often they are not also large bodied or perceived

as being quite so attractive. The list of those that have

been advocated as useful biodiversity indicators at one



time or another is very long. It includes soil nematodes,

moths, beetles galore (tiger, carabid, dung, and buprestid, to name but four), termites, fish, frogs, and snakes.

Whatever the group, they must also have one further

attribute—namely, that they genuinely indicate levels

of biodiversity or at least some of the components of

primary interest. The fact that the scientific literature

contains suggestions for so many different possible indicators shows that there is little consensus on the matter.

Many have been called, but few are chosen. Why? There

are two, related, reasons. First, scientific knowledge on

the degree of coincidence in patterns of biodiversity

between different taxa is surprisingly poor. Second, as

knowledge improves, coincidence between many taxa

turns out to be much worse than people had imagined,

or indeed hoped, would be the case.



C. Knowledge Is Poor Because of the

Effort Required

Gathering information on the diversity of different

groups of organisms, even in one place, is enormously

time-consuming. Two examples illustrate the problem.

To map the presence and absence of breeding birds

(conspicuous, ‘‘easy’’ to find and to identify) in every

10 ϫ 10 km grid-square in Britain and Ireland (there are

3672 squares) took more than half a million individual

record cards, filled in by an army of amateur birdwatchers coordinated by professional ornithologists in

the British Trust for Ornithology (BTO). The task took

4 years and about 100,000 hours of fieldwork (Gibbons

et al., 1993). Now imagine the effort required to do the

same thing for all the other hundreds of different groups

of organisms found in this one small corner of Europe.

It has been done for a sample of taxa (we will return

to what these data show in a moment), but many groups

remain unmapped.

At a much smaller spatial scale in a tropical forest

in Cameroon, a group of biologists attempted to measure the impacts of forest disturbance on just eight

groups (birds, butterflies, flying beetles, canopy beetles,

canopy ants, leaf-litter ants, termites, and soil nematodes). The birds and the butterflies took 50 and 150

scientist-hours, respectively, to survey. But the effort

required climbed rapidly for smaller-bodied, more cryptic, less well-known groups—1600 hours for the beetles, 2000 for the termites, and 6000 for the nematodes

(Lawton et al., 1998). Despite the fact that this work

as a whole took about five scientist-years, inventories

for most groups that were surveyed were still only partial, and most taxa remained unexamined (fungi, higher

plants, spiders, soil mites, collembola, earthworms, liz-



INDICATOR SPECIES



ards, frogs, and mammals, to name some of the most

conspicuous gaps).

Given this background, it is hardly surprising that

biologists do not have a complete inventory of all the

species that occur even in a single, moderately sized

area (a field, small wood, or lake)—a so-called ATBI

(All Taxa Biodiversity Inventory) (Oliver and Beattie,

1996). A moments thought will also show that to use

one or two groups (for the sake of the argument, say

birds and butterflies) as indicators of the richness of

other taxa in fact requires several such areas to be investigated to properly test the hypothesis that high bird

diversity (or any other single group) reflects a high

diversity of many other groups.

Although progress has been made in this area over

the past decade, considerable work remains to be done.

Even in otherwise well-studied situations, many groups

remain to be examined. Hence, at the present time, and

effectively by default, some groups are being used as

indicators of biodiversity, even though we cannot show

categorically that the richness of one or more groups

of organisms truly reflects the overall, or even a major

portion of the overall, biodiversity of an area. As a result

there is little consensus about what a ‘‘good’’ indicator

group, or groups, might be, because there are too few

hard data, from a range of habitats and geographic

regions round the world, on which to draw firm conclusions. But as data slowly emerge, they are not encouraging for those who wish to use simple, single-taxon

indicators of biodiversity.



D. Indicator Reliability

Where knowledge exists, it suggests that single or small

numbers of taxa will usually be poor indicators of the

biodiversity of other groups.



1. Tropical versus Temperate and Other

Major Diversity Gradients

It would be wrong to think that there is no coincidence

between patterns of diversity in different groups of organisms. Of course there is. In the broadest terms, it is

axiomatic that most major terrestrial and freshwater

groups are more species rich in the tropics than in

temperate regions, at low elevations than at high ones,

in forests than in deserts, and on large land masses than

tiny islands. Whether you are a botanist, a bird-watcher,

or a bug hunter, to find the most species it is generally

advisable to head to hot and humid mainland tropics

with lots of trees. It is easy to assume that there must

therefore be reasonably good correlations between major diversity gradients for different groups. There can



447



be, but even at this scale often there are not. Penguin

diversity peaks in Antarctica, not the tropics, and there

are many other examples of similar ‘‘reverse diversity

gradients’’ that buck the average trend. On the eastern

side of North America, the diversity of breeding warblers increases from south to north—suggesting that

this conspicuous taxon, which is easy to identify (at

least the breeding males!) and to survey, is probably

highly unsuitable as an indicator of patterns of biodiversity in most other taxa (in which diversity typically

decreases from south to north).

In the case of breeding North American warblers,

we can spot the problem because we have enough information about the organisms involved. But the whole

point about indicator taxa for biodiversity is that typically we will not be armed with, and indeed should not

need, information about ‘‘other’’ groups; knowledge of

the indicator taxon should suffice and be reliable. The

evidence suggests otherwise.



2. Hot Spots

Major gradients in diversity aside, at similarly large

scales an indicator group might be used to identify local

geographic hot spots in the species richness of one or

more other groups (peaks in the landscape of species

richness) or to determine relative levels of richness in

those other groups (hot spots versus all spots) (Gaston

1996b; Reid 1998). At the continental scale, the procedure has frequently been found to fail on both counts

(Gaston 1996a), with mismatches between the occurrence of peaks in the richness of different groups being

commonplace.

Across the United States and southern Canada, hot

spots (local areas with unusually high diversity) overlap

partially between some pairs of taxa (trees, tiger beetles,

amphibians, reptiles, birds, and mammals), but the pattern is not a general one. Numbers of species in different

large grid cells for two groups are often significantly

positively correlated, for example, birds and tiger beetles or mammals and swallowtail butterflies. But these

correlations are frequently weak, of rather limited predictive value, and in some cases explained by latitudinal

gradients in diversity. In other words, although such

correlations may sometimes enable a very general impression of the patterns in richness of one group to be

obtained from the patterns in richness of another, their

predictive powers are low.

These conclusions seem to hold at finer resolutions

over more constrained areas. Thus, species-rich areas

for different taxa in Britain (birds with butterflies, dragonflies, etc.) frequently do not coincide at a scale of

10 ϫ 10 km squares (Pendergast et al., 1993) (Fig. 5).



448



INDICATOR SPECIES



FIGURE 5 Coincidence between hot spots for butterflies and up to seven other taxa in Britain.

Hot spots are unusually species-rich sites (here defined as the top 5 percentile in Britain). All of

the most species-rich localities in Britain for butterflies lie in southern England. Increasingly dark

shading indicates that butterfly hot spots coincide with hot spots for an increasing number of other

taxa. Note that many butterfly hot spots are not unusually rich in any other species (open circles)

and that only one locality (in southeast England, just in from the coast) is a hot spot for all eight

taxa in this particular survey. The other taxa are breeding birds, dragonflies, moths, mollusks,

aquatic higher plants, and liverworts (simple plants). Information from Prendergast et al. (1993),

with additional data and figure kindly provided by John Prendergast.



Hot spots in this study are not distributed randomly,

overlapping more often than expected by chance, but

still at a low level. Likewise, different taxa are species

poor or species rich in different areas of the Transvaal

region of South Africa. At even finer scales, within the

Cameroon forest mentioned earlier, disturbance impacted on the diversity of eight taxa in very different

ways. All declined drastically in completely cleared

areas, but intermediate levels of forest disturbance had

very different effects on the diversity of different groups.

As a result, changes in the diversity of one taxon could

not be used to predict changes in the diversity of any

other (Lawton et al., 1998). A summary of these and

related studies showing similar results is provided by

Gaston (1996a, 1996b) and by Pimm and Lawton

(1998).



3. A Commonsense Explanation

This lack of, or relatively feeble correlation between,

species rich-areas for different groups of organisms

makes the search for simple, robust, single-taxon indicators of overall biodiversity look increasingly like a



lost cause. With hindsight, perhaps this emerging result

is obvious (Reid, 1998). Major geographic gradients in

biodiversity aside, within particular geographic regions

or at smaller habitat scales, the conditions favoring one

group of species may be hostile to another. Mollusks

like it cool and wet, butterflies like it warm and sunny,

and high bird diversity is more likely in tall vegetation

than short vegetation, irrespective of weather. Commonsense natural history suggests that there is unlikely

to be a single indicator taxon able to predict the diversity

of all, or even a majority of others.



4. Rare Species and Endemic Species

Biologists and conservationists are often interested not

only in patterns of species richness but also in the

distribution of unusually rare species, or of endemic

species. Do sites with unusual numbers of rare species

frequently coincide across different taxa? Again, the

answer seems to be no, or only weakly (Pimm and

Lawton, 1998; Prendergast et al., 1993), for the reasons

just outlined.

Endemic species may be different (Bibby et al.,