Townsend C.R., Begon M., Harper J.L. Essentials of Ecology

Подождите немного. Документ загружается.

invertebrates in a trout stream are considerably more at risk of predation during

daylight hours. And these conclusions are all the more robust because they derive

both from the readily controlled conditions of a laboratory experiment and from

the more realistic, but more variable, circumstances of a ﬁeld experiment.

In the Taieri River in New Zealand, 198 sites were selected in a stratiﬁed

manner by choosing streams of similar dimensions at random in each of three

tributaries from each of eight subcatchments of the river. Care was taken not to

succumb to the temptation of choosing sites with easy access (near roads or bridges)

in case this biased the results. The sites were classiﬁed as containing: (i) no ﬁsh;

(ii) Galaxias only; (iii) trout only; or (iv) both Galaxias and trout. At every site

a variety of physical variables were measured (stream depth, ﬂow velocity,

phosphorus concentration in the stream water, percentage of the streambed

composed of gravel, etc.). A statistical procedure called multiple discriminant

functions analysis was then used to determine which physical variables, if any,

distinguished one type of site from another. Means and standard errors of these

key environmental variables are presented in Table 1.1.

Trout occurred almost invariably below waterfalls that were large enough

to prevent their upstream migration; they tended to occur at low elevations

because sites without waterfalls downstream tended to be at lower elevation. Sites

containing Galaxias (or with no ﬁsh) were always upstream of one or several

large waterfalls. The few sites that contained both trout and Galaxias were below

waterfalls, at intermediate elevations, and in sites with cobble beds; the unstable

nature of these beds may have promoted coexistence (at low densities) of the two

species. This descriptive study at the population level therefore takes advantage

of a ‘natural’ experiment (streams that happen to contain trout or Galaxias) to

determine the effect of the introduction of trout. The most probable reason for

the restriction of populations of Galaxias to sites upstream of waterfalls, which

cannot be climbed by trout, is direct predation by trout on the native ﬁsh below

the waterfalls (a single small trout in a laboratory aquarium has been recorded

consuming 135 Galaxias fry in a day).

Chapter 1 Ecology and how to do it

the population level – brown trout

and the distribution of native ﬁsh

Table 1.1

Means and, in brackets, standard errors for important discriminating variables for ﬁsh assemblage classes

in 198 sites in the Taieri River. In particular, compare the ‘Galaxias only’ and ‘brown trout only’ classes.

Galaxias are found on their own if there are large waterfalls downstream of the site (and at relatively high

elevations where the stream bed has an intermediate representation of cobbles). Brown trout, on the other

hand, generally occur where there are no downstream waterfalls (at slightly lower elevations and with a

bed composition similar to the Galaxias class).

FROM TOWNSEND & CROWL, 1991

VARIABLES

NUMBER OF ELEVATION % OF THE BED

NUMBER WATERFALLS (m ABOVE COMPOSED

SITE TYPE OF SITES DOWNSTREAM SEA LEVEL) OF COBBLES

Brown trout only 71 0.42 (0.05) 324 (28) 18.9 (2.1)

Galaxias only 64 12.3 (2.05) 567 (29) 22.1 (2.8)

No ﬁsh 54 4.37 (0.64) 339 (31) 15.8 (2.3)

Trout + Galaxias 9 0.0 (0) 481 (53) 46.7 (8.5)

9781405156585_4_001.qxd 11/5/07 14:40 Page 19

That an exotic predator such as trout has direct effects on Galaxias distribu-

tion or mayﬂy behavior is not surprising. However, we can ask whether these

changes have community consequences that cascade through to other species. In

the relatively species-poor stream communities in the south of New Zealand, the

plants are mainly algae that grow on the streambed. These are grazed by various

insect larvae, which in turn are prey to predatory invertebrates and ﬁsh. As we

have seen, trout have replaced Galaxias in many of these streams. An experiment

involving artiﬁcial ﬂow-through channels (several meters long, with mesh ends

to prevent escape of ﬁsh but to allow invertebrates to colonize naturally) placed

into a real stream was used to determine whether trout affect the stream food

web differently from the displaced Galaxias. Three treatments were established

(no ﬁsh, Galaxias present, and trout present, at naturally occurring densities) in

each of several randomized blocks located in a stretch of a stream with each block

separated by more than 50 m. Algae and invertebrates were allowed to colonize

for 12 days before introducing the ﬁsh. After a further 12 days, invertebrates and

algae were sampled (Figure 1.8).

A signiﬁcant effect of trout reducing invertebrate biomass was evident

(P = 0.026), but the presence of Galaxias did not depress invertebrate biomass

from the no-ﬁsh control. Algal biomass, perhaps not surprisingly then, achieved

its highest values in the trout treatment (P = 0.02). It is clear that trout do have

a more pronounced effect than Galaxias on invertebrate grazers and, thus, on

algal biomass. The indirect effect of trout on algae occurs partly through a reduc-

tion in invertebrate density, but also because trout restrict the grazing behavior

of the invertebrates that are present (see Figure 1.7b).

The sequence of studies above provided the impetus for a detailed energetics

investigation of two neighboring tributaries of the Taieri River (with very similar

physicochemical conditions), one being occupied by just trout and the other

(because of a waterfall downstream) containing only Galaxias. No other ﬁsh were

present in either stream. The hypothesis under examination was that the rate at

which radiation energy was captured through photosynthesis by the algae would

be greater in the trout stream because there would be fewer invertebrates and

thus a lower rate of consumption of algae. Indeed, annual net ‘primary’ production

(the rate of production of plant, in this case algal, biomass) was six times greater

in the trout stream than in the Galaxias stream (Figure 1.9).

Part I Introduction

the community – brown trout

cause a cascade of effects

the ecosystem – trout and

energy ﬂow

Invertebrate biomass (g m

–2

)

Algal biomass (µg cm

–2

)

Fish predation regime Fish predation regime

NG T N G T

4(a) (b)

Figure 1.8

(a) Total invertebrate biomass and (b) algal biomass (chlorophyll a)

(± SE) for an experiment performed in summer in a small New

Zealand stream. In experimental replicates where trout are present,

grazing invertebrates are rarer and graze less; thus, algal biomass

is highest. G, Galaxias present; N, no ﬁsh; T, trout present.

AFTER FLECKER & TOWNSEND, 1994

9781405156585_4_001.qxd 11/5/07 14:40 Page 20

Moreover, the primary consumers (invertebrates that eat algae) produced new

biomass in the trout stream at about 1.5 times the rate in the Galaxias stream,

while trout themselves produced new biomass at roughly nine times the rate that

Galaxias do (Figure 1.9).

Thus, the algae, invertebrates and ﬁsh are all ‘more productive’ in the trout

stream than in the Galaxias stream; but Galaxias consume only about 18% of

available prey production each year (compared to virtually 100% consumption

by trout); while the grazing invertebrates consume about 75% of primary pro-

duction in the Galaxias stream (compared to only about 21% in the trout stream)

(Figure 1.9). Thus, the initial hypothesis appears to be conﬁrmed: it is strong

control by trout of the invertebrates that releases algae to produce and accumulate

biomass at a fast rate.

A further ecosystem consequence ensues: in the trout stream, the higher

primary production is associated with a faster rate of uptake by algae of plant

nutrients (nitrate, ammonium, phosphate) from the ﬂowing stream water (Simon

et al., 2004).

This series of studies, therefore, illustrates some of the variety of ways in which

ecological investigations may be pursued, and both the range of levels in the

biological hierarchy that ecology spans and the way in which studies at differ-

ent levels may complement one another. While it is necessary to be cautious

when interpreting the results of an unreplicated study (only one trout and one

Galaxias stream in the ‘ecosystem study’), the conclusion that a trophic cascade

is responsible for the patterns observed at the ecosystem level can be made with

some conﬁdence because of the variety of other corroborative studies conducted

at the individual, population and community levels. Although brown trout are

exotic invaders in New Zealand, and they have far-reaching effects on the ecology

of native ecosystems, they are now considered a valuable part of the fauna,

particularly by anglers, and generate millions of dollars for the nation. Many

other invaders have dramatic negative economic impacts (Box 1.5).

Chapter 1 Ecology and how to do it

Figure 1.9

Annual estimates for ‘production’ of biomass at one trophic level,

and the ‘demand’ for that biomass (the amount consumed) at the

next trophic level, for (a) primary producers (algae), (b) invertebrates

(which consume algae), and (c) ﬁsh (which consume invertebrates).

Estimates are for a trout stream and a Galaxias stream. In the

former, production at all trophic levels is higher, but because the

trout consume essentially all of the annual invertebrate production

(b), the invertebrates consume only 21% of primary production (a).

In the Galaxias stream, these ﬁsh consume only 18% of invertebrate

production, ‘allowing’ the invertebrates to consume the majority

(75%) of annual primary production.

Production/demand (g AFDM

–1

–2

)

300

250

200

150

100

1.5

0.5

Galaxias Trout

Algae

Galaxias Trout

Invertebrates

Galaxias Trout

Fish

(a) (b) (c)

Production Demand

AFTER HURYN, 1998

9781405156585_4_001.qxd 11/5/07 14:41 Page 21

Part I Introduction

1.5 TOPICAL ECONCERNS

1.5 Topical ECOncerns

A recent analysis concluded that tens of thousands

of invading exotic species in the United States cause

economic losses totaling $137 billion each year

(Pimentel et al., 2000). Table 1.2 breaks down the total

into a variety of taxonomic groups.

Let us consider a few invaders with particularly

dramatic consequences. The yellow star thistle

(Centaurea solstitalis) now dominates more than

4 million hectares in California, resulting in the total

loss of once productive grassland. Rats are estimated

to destroy $19 billion of stored grains nationwide per

year, as well as causing ﬁres (by gnawing electric

wires), polluting foodstuffs, spreading diseases and

preying on native species. Introduced carp reduce

water quality by increasing turbidity, while 44 native

ﬁsh are threatened or endangered by ﬁsh invaders.

The red ﬁre ant (Solenopsis invicta) kills poultry,

lizards, snakes and ground-nesting birds; in Texas

alone, its estimated damage to livestock, wildlife and

public health is put at about $300 million per year, and

a further $200 million is spent on control. The zebra

mussel (Dreissena polymorpha), which arrived in

Michigan’s Lake St. Clair in ballast water released

from ships from Europe, has reached most aquatic

habitats in the eastern United States, and is expected

to spread nationwide in the next 20 years. The large

populations that develop threaten native mussels and

other fauna, not only by reducing food and oxygen

availability but by physically smothering them. The

mussels also invade and clog water intake pipes,

so that millions of dollars need to be spent clearing

them from water ﬁltration and hydroelectric generat-

ing plants. Overall, pests of crop plants, including

weeds, insects and pathogens, engender the biggest

economic costs. However, imported human disease

organisms, particularly HIV and inﬂuenza viruses, cost

$6.5 billion to treat and result in 40,000 deaths per

year. (See Pimentel et al., 2000, for further details and

references.)

Globalization has been the prevalent economic

ideology in recent times. Globalization of the biota,

in which successful invaders are moved around

the world, often driving local species extinct, can be

expected to lead to a general homogenization of the

world’s biota. [Lövei (1997) has colorfully referred to

this as ‘McDonaldization’ of the biosphere.] Does

biotic homogenization matter? Why?

Invasions and homogenization of the biota: does it matter?

NA, not available.

AFTER PIMENTEL ET AL., 2000

NUMBER OF LOSS AND CONTROL TOTAL

TYPE OF ORGANISM INVADERS MAJOR CULPRITS DAMAGE COSTS COSTS

Plants 5,000 Crop weeds 24.4 9.7 34.1

Mammals 20 Rats and cats 37.2 NA 37.2

Birds 97 Pigeons 1.9 NA 1.9

Reptiles and amphibians 53 Brown tree snake 0.001 0.005 10.006

Fishes 138 Grass carp, etc. 1.0 NA 1.0

Arthropods 4,500 Crop pests 17.6 2.4 20.0

Mollusks 88 Asian clams 1.2 0.1 1.3

Microbes (pathogens) >20,000 Crop pathogens 32.1 9.1 41.2

Table 1.2

Estimated annual costs (billions of dollars) associated with invaders in the United States.

9781405156585_4_001.qxd 11/5/07 14:41 Page 22

1.3.2 Successions on old ﬁelds in Minnesota:

a study in time and space

‘Ecological succession’ is a concept that must be familiar to many who have

simply taken a walk in open country – the idea that a newly created habitat, or

one in which a disturbance has created an opening, will be inhabited, in turn,

by a variety of species appearing and disappearing in some recognizably repeat-

able sequence. Widespread familiarity with the idea, however, does not mean

that we understand fully the processes that drive or ﬁne-tune successions; yet

developing such understanding is important not just because succession is one

of the fundamental forces structuring ecological communities, but also because

human disturbance of natural communities has become ever more frequent and

profound. We need to know how communities may respond to, and hopefully

recover from, such disturbance, and how we may aid that recovery.

One particular focus for the study of succession has been the old agricultural ﬁelds

of the eastern USA, abandoned as farmers moved west in search of ‘fresh ﬁelds and

pastures new’. One such site is now the Cedar Creek Natural History Area, roughly

50 km north of Minneapolis, Minnesota. The area was ﬁrst settled by Europeans in

Chapter 1 Ecology and how to do it

Red ﬁre ants, Solenopsis.

Yellow star thistle, Centaurea solstitialis.

Zebra mussels, Dreissena polymorpha.

9781405156585_4_001.qxd 11/5/07 14:41 Page 23

1856 and was initially subject to logging. Clearing for cultivation then began about

1885, and land was ﬁrst cultivated between 1900 and 1910. Now there are agri-

cultural ﬁelds that are still under cultivation and others that have been abandoned

at various times since the mid-1920s. Cultivation led to depletion of nitrogen

from soils that already were naturally poor in this important plant nutrient.

In the ﬁrst place, studies at Cedar Creek illustrate the value of ‘natural

experiments’. To understand the successional sequence of plants that occur in ﬁelds

in the years following abandonment we could plan an artiﬁcial manipulation,

under our control, in which a number of ﬁelds currently under cultivation were

‘forcibly’ abandoned and the communities in them sampled repeatedly into the

future. (We would need a number of ﬁelds because any single ﬁeld might be

atypical, whereas several would allow us to calculate mean values for, say, ‘number

of new species per year’, and place conﬁdence intervals around those means.) But

the results of this experiment would take decades to accumulate. The natural

experiment alternative, therefore, was to use the fact that records already exist of

when many of the old ﬁelds were abandoned. This is what Tilman and his team

did. Thus, Figure 1.10 illustrates data from a group of 22 old ﬁelds surveyed

in 1983, having been abandoned at various times between 1927 and 1982 (i.e.

between 1 and 56 years previously). Interpreted cautiously, these can be treated

as 22 ‘snapshots’ of the continuous process of succession in old ﬁelds at Cedar

Creek in general, even though each ﬁeld was itself only surveyed once.

A number of the shifting balances during succession are clear from the ﬁgure

as statistically signiﬁcant trends. Over the 56 years, the cover of ‘invader’ species

(mostly agricultural weeds) decreased (Figure 1.10a) while the cover of species

from nearby prairies increased (Figure 1.10b): the natives reclaimed their land.

Of more general applicability, the cover of annual species decreased over time,

while the cover of perennial species increased (Figure 1.10c, d). Annual species

(those that complete a whole generation from seed to adult through to seeds again

within a year) tend to be good at increasing in abundance rapidly in relatively

empty habitats (the early stages of succession); whereas perennials (those that live

for several or many years and may not reproduce in their early years) are slower

to establish but more persistent once they do.

On the other hand, natural experiments like this, while frequently suggestive

and stimulating (and too good an opportunity to miss), usually only generate

correlations. They may therefore fail to establish what actually causes the observed

patterns. In the present case, we can see the problem by noting, ﬁrst, that ﬁeld

age is itself strongly correlated with nitrogen concentration in the soil – perhaps

the single most important plant nutrient (Figure 1.10e). The question therefore

arises: are the correlations in Figure 1.10a–d the result of an effect of ﬁeld age

itself? Or is the causal agent nitrogen, with which age is correlated?

Manipulative ﬁeld experiments can be used to help support – or refute – what so

far is no more than a plausible explanation based on correlation. It seems to follow

from the proposed explanation (time matters) that nitrogen itself has little role to

play in driving these successions, and that manipulating nitrogen should do little to

alter the species sequences that these ﬁelds have followed. To test this, Tilman’s team

selected a pair of ﬁelds (one abandoned for 46 years and the other for 14 years) and,

over a 10-year period starting in 1982, subjected six replicate 4 m × 4 m plots in each

ﬁeld to one of two treatments: nitrogen added at rates of either 1 or 17 g m

−2

−1

(Inouye & Tilman, 1995). Two questions in particular were being asked.

Part I Introduction

. . . in generating correlations

artiﬁcial experiments:

the search for causation

the use of natural

experiments...

9781405156585_4_001.qxd 11/5/07 14:41 Page 24

1 Do patches receiving different supply rates of nitrogen become less similar

in species composition over time?

2 Do patches receiving similar supply rates of nitrogen become more similar

in species composition over time?

The answer to the ﬁrst question was clear: plots within a ﬁeld were initially

similar to one another but, 10 years later, plots receiving different amounts of

nitrogen had diverged in species composition – and the greater the difference in

nitrogen input, the greater the divergence (Inouye & Tilman, 1995).

The answer to the second question is illustrated in Figure 1.11. At the start

of the experiment, the ﬁeld abandoned for 46 was very different in species com-

position to the one only abandoned for 14 years. But 10 years later, plots within

the two ﬁelds that had been subjected to similar rates of nitrogen input had

become remarkably similar (Figure 1.11).

Thus, this experiment tends to refute the simplicity of our proposed explanation.

Time itself is not the only cause of successional changes in species composition

of these old ﬁelds. Differences in available nitrogen cause successions to diverge;

similarities cause them to converge much more quickly than they would other-

wise do. Time (= opportunity to colonize) and nitrogen are clearly intimately

Chapter 1 Ecology and how to do it

Percent cover

(b) Prairie species

(a) Invader species

(d) Perennial species

(e) Soil nitrogen

Field age (years)

1002030

Field age (years)

40 50 60 1002030

Field age (years)

40 50 60

100 2030405060 100 2030405060

Soil nitrogen (mg kg

–1

)

0 102030405060

200

400

600

800

1000

Figure 1.10

Twenty-two ﬁelds at different stages in

an old-ﬁeld succession were surveyed

to generate the following trends with

successional stage (ﬁeld age): (a) invader

species decreased, (b) native prairie

species increased, (c) annual species

decreased, (d) perennial species

increased, and (e) soil nitrogen content

increased. The best ﬁt lines (see Box 1.2)

are highly signiﬁcant in every case

(P < 0.01).

AFTER INOUYE ET AL., 1987

9781405156585_4_001.qxd 11/5/07 14:41 Page 25

intertwined and further experiments will be required to disentangle their web of

cause and effect – just one of many unanswered ecological questions.

Finally, experimental manipulations over extended periods like these may

also provide important insights into the possible effects of more chronic human

disturbances to natural communities. The lower rate of nitrogen addition in the

experiment (1 g of nitrogen m

−2

−1

) was similar to that experienced in many

parts of the world as a result of increased atmospheric deposition of inorganic

nitrogen (mainly derived from the burning of fossil fuels). Even these low levels

apparently led to convergence of previously dissimilar communities over a 10-year

period (Figure 1.11b). Experiments like this are crucial in helping us to predict

the effects of pollutants, a point that is taken further in the next example.

1.3.3 Hubbard Brook: a long-term commitment of

large-scale signiﬁcance

The Cedar Creek study took advantage of a temporal pattern (a succession that

takes decades to run its course) being reﬂected more or less accurately by a

pattern in space (ﬁelds abandoned for different periods). The spatial pattern has

the advantage that it could be studied within the time-bite of most research pro-

jects (3–5 years). It would have been better still to follow the ecological pattern

through time but rather few researchers or institutions have risen to the challenge

of designing research programs that last for decades.

A notable exception has been the work of Likens and associates at the Hubbard

Brook Experimental Forest, an area of temperate deciduous forest drained by small

streams in the White Mountains of New Hampshire in the USA. The researchers

were pioneers with no precedents to follow. They decided to think big, and their

work has shown the value of large-scale studies and long-term data records. The

study commenced in 1963 and continues to the present. In the second edition of

their classic book Biogeochemistry of a Forested Ecosystem, Likens and Bormann

(1995) make poignant reference to three of their original collaborators who had

died since the study began. Long term indeed.

Part I Introduction

Similarity

1.0

0.8

0.6

0.4

0.2

(a)

(b)

1981 1983 1986 1987 1989 1991 1993 19931981 1983 1986 1987 1989 1991

YearYear

Figure 1.11

Results from an experiment in which plots within two old ﬁelds from Figure 1.10 were given artiﬁcial

nitrogen addition treatments starting in 1982: one of the ﬁelds had been abandoned for 46 years and the

other for 14 years. (a) Between 1982 and 1992, plots receiving 17 g of nitrogen m

−2

−1

in the two ﬁelds

became increasingly similar in composition. The similarity index measures the extent to which the species

composition in the pair of ﬁelds is similar – identical compositions produce a similarity index of 1, entirely

different compositions produce a similarity of 0. (b) Like (a) but with only 1 g of nitrogen m

−2

−1

Note in this case that there was still convergence in species composition between the two ﬁelds but to

a lesser extent. In both cases the best ﬁt lines are highly signiﬁcant.

AFTER INOUYE & TILMAN, 1995

insight into the effects of

nitrogen pollution

9781405156585_4_001.qxd 11/5/07 14:41 Page 26

The research team developed an approach called ‘the small watershed technique’

to measure the input and output of chemicals from individual catchment areas

in the landscape. Because many chemical losses from terrestrial communities are

channeled through streams, a comparison of the chemistry of stream water with

that of incoming precipitation can reveal a lot about the differential uptake and

cycling of chemical elements by the terrestrial biota. The same study can reveal

much about the sources and concentrations of chemicals in the stream water,

which in turn may inﬂuence the productivity of stream algae and the distribution

and abundance of stream animals.

The catchment area (or watershed) – the extent of terrestrial environment

drained by a particular stream – was taken as the unit of study because of the

role that streams play in chemical export from the land. Six small catchments

were deﬁned and their outﬂows were monitored (Figure 1.12). A network of

precipitation gauges recorded the incoming amounts of rain, sleet and snow.

Chemical analyses of precipitation and stream water made it possible to calculate

the amounts of various chemical elements entering and leaving the system. In

most cases, the output of chemicals in streamﬂow was greater than their input from

rain, sleet and snow (Table 1.3). The source of the excess chemicals was weather-

ing of parent rock and soil, estimated at about 70 g m

−2

−1

. The exception was

nitrogen; less was exported in stream water than was added to the catchment

in precipitation and by ﬁxation of atmospheric nitrogen by microorganisms in

the soil.

Likens had the brilliant idea of performing a large-scale experiment in which

all the trees were felled in one of Hubbard Brook’s six catchments. In terms of

experimental design, statistical purists might argue the study was ﬂawed because

Chapter 1 Ecology and how to do it

COURTESY OF GENE LIKENS

Figure 1.12

The Hubbard Brook experimental forest. Note the experimental

stream catchment from which all trees were removed – extending

from the top left toward the center of the photograph.

the catchment area as a unit

of study

insights from a large-scale ﬁeld

experiment

9781405156585_4_001.qxd 11/5/07 14:41 Page 27

it was unreplicated. However, the scale of the undertaking rather precluded

replication. In any case, it was the asking of a dramatically new question that

made this study a classic rather than elegant statistical design.

Within a few months of felling all the trees in the drainage basin, the con-

sequences were evident in the stream water. The overall export of dissolved

inorganic substances from the disturbed catchment rose to 13 times the normal

rate (Figure 1.13). Two phenomena were responsible. First, the enormous

reduction in transpiring surfaces (leaves) led to 40% more precipitation passing

through the ground water to be discharged to the streams, and this increased

outﬂow caused greater rates of leaching of chemicals and weathering of rock and

soil. Second, and more signiﬁcantly, deforestation effectively broke the link

between decomposition and nutrient uptake. In the spring, when the deciduous

trees would normally have started production and taken up inorganic nutrients

released by decomposer activity, these were instead available to be leached in the

drainage water.

Likens knew from the beginning that the rain and snow at Hubbard Brook

were quite acid but it was some years before the widespread nature of acid rain in

North America became clear. In fact, Hubbard Brook is more than 100 km from

the nearest urban industrial area, yet precipitation and stream water were both

markedly acid as a result of atmospheric pollution from fossil fuels. The long-

term records kept so meticulously since 1963 at Hubbard Brook have proved

invaluable in monitoring progress in the war against acid rain and its long-term

consequences. The value of such records of stream water concentrations can be

seen for hydrogen, sulfate and nitrate, three ions associated with acid rain (which

in simple terms is a mixture of dilute nitric and sulphuric acids; sulphuric acid is

the dominant acid in the eastern USA). There have been statistically signiﬁcant

declines in average annual concentrations of H

and SO

2−

since 1964/65, and

also of NO

−

, thought the latter is subject to much greater year to year variation

(Figure 1.14). Of note, however, is the fact that the results for shorter periods sug-

gest quite different trends. Consider the hydrogen ion graph where three periods

of 4 years are highlighted in different colors. The ﬁrst suggests an increasing

trend, the second no change and the third a decreasing trend. In fact, no statistic-

ally signiﬁcant, long-term trend was established until nearly two decades of data

had been amassed (Likens, 1989).

Part I Introduction

Table 1.3

Annual chemical budgets for forested catchment areas at Hubbard Brook (kg ha

−1

). Inputs are

for dissolved materials in precipitation or in dryfall (gases or associated with particles falling from the

atmosphere). Outputs are losses in stream water as dissolved material plus particulate organic material in

the streamﬂow. The source of the excess chemicals (where outputs exceeded inputs) was weathering of

parent rock and soil. The exception was nitrogen (as ammonium or nitrate ions) – less was exported than

arrived in precipitation because of nitrogen uptake in the forest.

*Net change is positive when the catchment gains matter and negative when it loses it.

AFTER LIKENS ET AL., 1971

−−

Input 2.7 16.3 38.3 1.1 2.6 0.7 1.5

Output 0.4 8.7 48.6 1.7 11.8 2.9 6.9

Net change* +2.3 +7.6 −10.3 −0.6 −9.2 −2.2 −5.4

for statistically signiﬁcant trends

to become evident, many years

of data may be required

9781405156585_4_001.qxd 11/5/07 14:41 Page 28