|
How Air
Pollution Can Change the Response of
Plants to
Other Environmental Stresses
J N B
Bell
Professor of
Environmental Pollution,
Centre for
Environmental Policy/Division of Biology
Imperial College
London, Silwood Park campus, Ascot
Berkshire, SL5 7PY,
UK.
It has been known since
the 17th century that air pollution can cause serious damage
to vegetation, causing both visible blemishes on foliage and/or
reductions in growth and crop yield. Research into this topic started in
the late 19th century, initially in Germany, but followed
rapidly by the UK, USA, Canada and elsewhere. Initially interest was
centred on the traditional pollutants in the form of coal smoke and SO2,
later it was realised that there were other phytotoxic pollutants
present in the atmosphere. Thus in the period after World War II
interest shifted to photochemical oxidants, particularly O3,
formed as secondary pollutants resulting from complex atmospheric
chemical reactions involving nitrogen oxides and volatile organic
compounds under conditions of high temperature and sunlight. The
recognition of O3 as a widespread pollutant, particularly in
rural areas, occurred first in California, then elsewhere in North
America, followed by Europe and Japan in the 1970s and much more
recently in some parts of the developing world, although in the latter
case the severity of the problem is largely unrecognised. Subsequently
other pollutants became recognised as of major importance, notably NO2
and NH3 and their products. Widespread problems are not
restricted to O3, because acid rain and total nitrogen
deposition are known to cause large scale problems of acidification and
eutrophication, respectively.
The best example of an
attempt to estimate the economic impacts of air pollution on crops was
the National Crop Loss Assessment Network (NCLAN) carried out in the USA
in the 1980s, followed by a similar programme in Europe. This
concentrated mainly on O3 and predicted a loss of $3 billion
per annum for the 10 principal widely grown crops in the USA. It should
be noted that this programme concentrated on the impacts of pollution on
crop yield. However, there are a numerous other stresses, both abiotic
and biotic, which cause enormous losses to agriculture, up to an order
of magnitude higher than the estimates of NCLAN. Thus if the response of
plants to these other stresses could be modified by air pollution, then
the economic consequences could be substantial, including both negative
and positive effects. In this article a brief overview will be given of
the evidence for air pollutants impacting on these other stresses.
Abiotic stresses
There are a wide range
of abiotic stresses which impact vegetation, including frost/low
temperature, high temperature, drought, salinity, wind and mineral
deficiency/toxicity. The best researched areas of interactions of
abiotic stresses with air pollution are for frost and drought, with, to
a much lesser extent, salinity.
Frost
There is abundant
evidence that air pollutants can sensitise plants to frost damage,
particularly in the case of SO2 and O3. Numerous
controlled experiments have demonstrated that fumigated plants which are
subsequently exposed to low temperatures show much greater frost injury
than those that were previously exposed to charcoal- filtered clean air.
Much interest in this issue was shown in both eastern North American and
Europe during the 1980s, when major research programmes examined the
possible role of O3 in .forest decline.. This followed
observations that tree health showed particularly severe deterioration
following harsh winters. It was demonstrated conclusively that a
summertime fumigation with O3 resulted in increased sensitivity to
subsequent frost stress, including visible injury and delayed hardening
in a number of tree species. Likewise it is now well known that heather
in European heathlands, impacted by high levels of N deposition is
sensitive to frost damage, leading to gaps in the canopy which permit
the invasion of nitrophilous grasses and deterioration of these habitats
of high conservation value.
Similar effects have
also been demonstrated for herbaceous species, including grasses,
cereals and clovers. In the UK the potential for SO2 to
exacerbate frost injury has been shown in both laboratory and field
studies. In the latter case, of particular interest are observations at
two open-air fumigation systems, where cereals were subjected to SO2
in fields where different plots received different concentrations of SO2
from computer controlled arrays of pipes. In such a situation the crop
is subjected to all the local natural stresses, both abiotic and biotic.
It was recorded in both systems that following periods of cold weather
injury was significantly greater in the plots which had been fumigated
with SO2, with an indication of a dose/response relationship.
My own research group has demonstrated a similar phenomenon, but this
time caused by ambient pollution. In February 1986 we were carrying out
an overwinter experiment in which four cultivars of red clover were
grown in open-top chambers, ventilated with ambient or clean charcoal-
filtered air, or else in outside chamberless plots. This month was
exceptionally severe for the south of England, as the temperature fell
below zero for four weeks. At the end of this period severe low
temperature injury was seen on all the experimental plants, this being
similar in the outside plots and the ambient air chambers, but
significantly reduced in all cultivars in the charcoal-filtered
chambers. Thus ambient air pollution had sensitised the clovers to frost
injury. The pollutant(s) involved could not be identified unequivocally.
O3 can be dismissed as it is not elevated under winter
conditions, but it is known that large areas of Western Europe,
including the UK were blanketed in a cloud of high SO2
concentrations at the time, although the role of NO2 cannot
be precluded.
Drought
It is generally held
that drought reduces air pollution injury by reducing stomatal
conductance, thereby inhibiting the uptake of pollutant gases into the
leaf. However, this appears to be a complex situation, about which it is
difficult to generalise, not the least because air pollutants have been
demonstrated both to open and close stomata, depending on species,
pollutant concentration and prevailing environmental conditions. In the
case of soya beans in the USA and trees in the UK it has been shown that
the nature of the O3/drought interaction reverses, depending
on pollutant concentration. of particular interest is a study in the UK,
which showed that pretreatment with a SO2+NO2
mixture predisposed a grass species to drought stress. The grass was
fumigated with a range of pollutant concentrations or else kept in clean
air, before all plants being transferred to the latter treatment, with
half being kept well-watered and the other half droughted. The watered
plants showed no effect of pollution pretreatment, but the droughted
plants showed a concentration-dependant reduction in growth, when
pretreated with SO2+NO2. As in the case of frost
injury, there has been considerable interest in the role of O3/drought
interactions in contribution to forest decline, in view of the latter
apparently becoming worse following very dry summers. However, it is
difficult to elucidate the respective roles of these two stresses in
view of the fact that very dry summers always coincide with particularly
high O3 levels.
Biotic stresses
As in the case of
abiotic stresses, their biotic equivalents are responsible for massive
crop losses worldwide. These can take many forms, the principal being
fungal pathogens, herbivorous insect pests, viruses, bacteria and
nematodes. The interactions of air pollutants on all of these have been
investigated, with emphasis on fungal interactions.
Fungal
Pathogens
Observations have been
made for many years that the incidence of a range of fungal pathogens
appear to be related, either positively or negatively, to elevated
levels of air pollutants, most of these being in the vicinity of point
sources. In general it appears that biotrophs (fungi not readily
cultured in the absence of the host) were reduced in pathogenicity,
while non-biotrophs showed a more mixed response. It was postulated that
the former was a reflection of the intimate connection between the
metabolism of the pathogen and the host, with pollutant-induced
impairment of the latter resulting in damage to the fungus. In the case
of stimulation it was suggested that injury to the host's surface
provided courts of infection for the pathogen. Many studies have been
carried out on pathogen/pollutant interactions, although only a
relatively small number of the vast range of potential combinations have
been investigated. Thus only a selection will be given here,
concentrating on the results of my own research group. One particularly
interesting study of ours involved investigation of the natural
infection of barley, growing in one of the open-air SO2
fumigation systems mentioned earlier. This showed that in three
successive growing seasons infection by powdery mildew was stimulated by
SO2, but the reverse occurred for leaf blotch. A controlled
fumigation with the same concentrations of SO2 in chambers
confirmed the causal nature of the field system observations. The use of
economic models for infection/crop yield reduction of these
host/pathogen systems indicated that the combined net effect of them
would be negative at all concentrations employed. Of particular interest
here is the fact that powdery mildew is classed as a biotroph,
indicating that the suggestion that biotrophs are reduced in infection
by air pollution is certainly not a universal phenomenon.
There are two
interesting cases where fungal pathogen incidence has been related to
air pollution levels to the extent that they are considered to be (and
in one case has been) employed as biomonitors. Both of these have been
investigated by my group. The first of these is black spot of roses a
ubiquitous disease of rose leaves. In the past when SO2
levels were very high in many parts of the UK, this disease was reported
as absent from the more polluted areas. A study in the mid-1960s showed
that its incidence was negatively related to the prevailing SO2
concentration in a study of nurseries nationwide. The causality of this
relationship was confirmed by controlled fumigation experiments.
Subsequently with the
massive fall in concentrations, anecdotal evidence indicated that it had
reinvaded locations where it was formerly absent. A study by a member of
my group involved revisiting the locations of the 1960s survey and
showed clearly that reinvasion had occurred everywhere and that there
was no indication of any relationship with the residual SO2
levels. The second disease was rather more controversial and studies
have produced intriguing results. This was the tarspot of sycamore,
which takes the form of large black lesions on the tree's leaves. A
field study in cities of northern England in the mid-1970s showed a
negative relationship between a calculated tarspot index (TSI: number of
spots per leaf unit area) and SO2 concentrations, which
showed a similar cut- off level for total elimination as in the earlier
black spot research. Thus it was postulated that TSI was a good
bioindication of prevailing SO2 concentrations and it was
incorporated into lichen biomonitor surveys in parts of northern England
in the 1970s. However, a survey in the mid-1980s in Edinburgh failed to
find the same relationship, and it was suggested that the earlier
findings were an artefact viz. that in the more polluted areas towards
the centre of cities the activities of street cleaners and park keepers
increasingly removed the only source of inoculum in the form of spores
over wintering on dead leaves on the ground. I used this example some
years ago in a lecture I gave on a short course on environmental science
for lawyers, as an exemplary lesson in the dangers of attributing
causality to observed relationships. It was entitled .Correlation and
Causality: Park Keepers or Pollution.! However, I had to eat my words!
In the late 1980s some of the last controls on SO2 emissions
from domestic coal-burning were imposed around colliery villages in
north-east England. A colleague from the local University of Newcastle
upon Tyne observed that with falling SO2 concentrations, not
only did black spot of roses show reinvasion, but so also did tarspot of
sycamore! Intrigued by this we studied the relationship between TSI and
air pollution in the London area and a region of northern England
(around the city of Sheffield). In the case of the Sheffield area
tarspot had been recorded as absent in the mid-1970s in the more
polluted places, in a biomonitoring study. When we resurveyed the area
in 1999 the disease was prevalent in effectively all sites, and showed
no relationship to the contemporary (much lower) SO2
concentrations. Interestingly there was one site where it was absent: in
the very middle of Sheffield, where a single tree was surrounded by
concrete paving and there was no possibility of spores surviving over
winter to reinfect. Thus there was overwhelming evidence that SO2
and TSI were negatively related and that the latter was potentially a
good biomonitor. However, as part of the same study, the TSI of sycamore
trees was measured along a transect of 50 km from a rural area into
central London. It was found that the disease disappeared within London.
Initially it was postulated that its relative short spore dispersal
distance might have delayed reinvasion. However, this was refuted by a
field study where saplings from a clean area were exposed along the same
transect, together with inoculum in the form of infected dead leaves
placed alongside. Again the fungus failed to reinfect within London.
There would appear to be only two possible explanations: differences in
climate along the transect and/or increasing NO2
concentrations towards central London. The former can be dismissed as
the changes in climatic parameters between the rural area and central
London are far less than those around the UK, where the disease is
ubiquitous (including Edinburgh). The most likely explanation is that
the fungus is sensitive to NO2, a pollutant which has its
highest levels in the UK in London. This would suggest the potential for
TSI to be used as a biomonitor for NO2.
Insects
As in the case of fungal
pathogens, for many years there have been field observations of changes
in herbivorous pest (both sucking and chewing) infestation of vegetation
in polluted areas. These have largely shown increases in incidence, but
there are some opposite cases, particularly at very high pollutant
concentrations. These observations have been made around point, diffuse
and line scources (in the form of busy roads). Various studies have
examined the causality and underlying mechanisms of these effects,
employing field, filtration, fumigation and survey techniques. Some of
the work of my research group on this topic will now be summarised.
Our initial work started
as a result of discussions between myself and a colleague. The latter
was an entomologist, interested in the relationship between host plant
chemistry and herbivorous insect performance. My colleague was well
aware that subtle changes in the amino acid composition of a plant could
have major impacts on the performance of insects feeding on it. I was
aware (from the literature) that air pollution could influence plant
amino acid composition. Thus, ipso facto, we established the
hypothesis that air pollution could alter pest infestations via changes
in plant chemistry. We had no funding to test this hypothesis, but then
in 1982 a gift arrived from Germany in the form of a student who wanted
to carry out his Diplom Arbeit with us, and we suggested this topic.
Probably one of the best decisions we ever made! As in other work, we
developed a programme to verify the causality of field observations. For
many years colleagues had worked on models to predict the outbreak of
the serious sucking pest, black bean aphid, on crops in summer, by
studying the number of eggs on its over winter host in different parts
of the country. Consistently these models under predicted the outbreaks
in areas impacted by the pollution plume downwind to the east of London.
Thus we decided to investigate the impacts of SO2 and NO2
on this aphid. We fumigated the bean for short periods withSO2
and NO2 and then put the plants in clean air. Weighed aphids
were caged on the leaves of fumigated plants and then reweighed three
days later. It was demonstrated that the growth rate of aphids feeding
on previously fumigated plants was increased above that of those on
clean air controls. A parallel study examined the growth rate of aphids
feeding on artificial diets, mimicking bean phloem sap, and showed no
effect. Thus our hypothesis was confirmed that air pollution could
change pest performance via effects on the host plant and most of our
subsequent research has suggested that this is mediated via plant
chemistry, particularly shifts in amino acid composition. Subsequently
we have shown that this phenomenon is widely distributed across a wide
range of pest/host plant systems. More confirmation of the relationship
was determined by three further parts of the same investigation: a
filtration study in central London showed the same effects, as did a
transect exercise in which plants were exposed along a declining
gradient of air pollution out from central London; finally an
examination of aphid trap data around the UK showed that most species
occurred at greater levels in areas with higher SO2
pollution. The magnitude of the response of these insects which have a
very short life cycle, indicates that they could readily escape the
control of their natural enemies in the form of predators and parasites.
Subsequent research has confirmed this, including measurements of
infestation by aphids on barley in one of the field SO2
fumigation systems discussed earlier. Ozone has also been shown to
affect aphid performance, mediated via the host plant. Our own work has
shown in chamber filtration experiments in a rural area that the
pollution climate of a relatively low O3 summer produced a
substantial increase in colony size of black bean aphids feeding on
beans. Other work has shown that O3 levels only just above
the natural background can produce substantial increases in aphid growth
rates. In the case of O3 there is evidence that there may be
complex interactions with duration and temperature of exposure.
Conclusion
The subject of air
pollution impacts on plant response to natural stresses remains
relatively little studied. Yet it has potentially very significant
implications for crop production not least in developing countries. It
has so far been totally ignored in the establishment of air quality
standards and guidelines, with the single exception of the United
Nations Economic Commission for Europe's incorporation of extreme low
temperatures into the critical levels for forests and natural
vegetation, which are lowered by 25% in such circumstances. My
conclusion is that the world is missing the opportunity to address a
really serious threat to food security, notably in developing countries.
Everybody understands the importance of frost, drought, insect pests and
pathogens. Few understand the direct impacts of air pollution on crops.
And fewer still understand the potential for the first of these to be
impacted by the second!
Biblography
Davison, A.W. &
Barnes, J.D. (2002) Air pollutant-biotic stress interactions. In: Bell,
J.N.B. & Treshow, M. (eds.) . 2nd edition. John Wiley & son,
Chichester. Pp 359-377.
Flückiger, W., Braun,
S. & Hiltbrunner, E. (2002) Effects of air pollutants on biotic
stresses. In: Bell, J.N.B. & Treshow, M. (eds.) . 2nd
edition. John Wiley & son, Chichester. Pp 379-406. |
