Previous PageTable Of ContentsNext Page


Roger M. Gifford

CSIRO Division Plant Industry. GPO Box 1600. Canberra, ACT 2601


In considering agriculture, forestry and natural ecosystems, we are asking in the first instance what the effect of the global build-up of atmospheric carbon dioxide and other trace gases on vegetation will be. And if the consequential “greenhouse effect” warming causes average temperatures to continue to increase, by say 30oC in the next half century, how will such warming and any associated changes in climate patterns affect primary production?

The increase in concentration of the greenhouse effect gases like carbon dioxide (CO2) are smooth and continuous, and therefore relatively predictable.

Their first order effects should, in principle, be relatively predictable too.

The climate changes that might ensue are statistical and rather unpredictable especially for any one location or region. It is not even possible to say during a period of unusual weather, such as that which we have experienced this year, whether it is due to the progressive change in atmospheric composition or is just part of the normal distribution of variability. That can only be discerned years afterwards.

Both the smooth and continuous changes and the postulated statistical climatic changes would, if continued, certainly have significant, possibly massive, impacts on primary production from the land, on hydrology, on natural ecosystems, and on the landscape in general. But the different character of smooth and of statistical changes renders evaluation of integrated impacts very difficult indeed. It is meaningless to contemplate statistical possibilities alone without evaluating their impact in the context of much more certain continuous changes.

There are other difficulties. One is that there appear to be potential benefits to be gained from the changes as well as the much discussed disadvantages. Much of the potential benefit derives from the smooth continuous relative predictable change. Many of the possible costs would be associated with the speculative statistical climatic changes.

Another problem is time-scale. While, for purposes of providing a focus for discussion, today’s conditions are often compared with postulated conditions associated with twice the amount of the offending trace gases, change does not stop there. While a small change soon may well be considered desirable, a much bigger change later in the same direction could be problematic - catastrophic for some - but maybe a century away.

Finally, among these difficulties there is the problem of big surprises that we fail to predict. The opening up of the ozone hole over the Antarctic at the end of the long Antarctic “night” was a surprise. But it may not persist. There could be other surprises. For example, Australia is

located on the edge of the monsoonal and cyclonic belts and is influenced strongly by the (as yet) unpredictable El Nino-Southern Oscillation phenomenon responsible for the big drought of the early 1980s and the wet period now. The timing, tracking, frequency and intensity of all these meteorological phenomena could well be influenced in unpredictable ways by a general warming. This could bring big surprises.

How do we balance the unknown risks of such surprises against any social costs of foregoing fossil fuel consumption or of any potential agricultural benefits of a warmer, wetter, high CO2 world in the next few decades?

The potential scope of this topic is very broad. For example, considering crop production alone, we could consider impacts on:

• rate of plant growth

• length of the growing season

• special developmental effects like vernalisation (chilling)

• weed competition

• pest and disease incidence and management

• optimal geographic boundaries for specific crops

• soil erosion and salinisation effects

• system management-Strategies under continuous change

• crop breeding requirements

It all gets very complicated. Even if we knew exactly what environmental changes will occur over what time scale, we have a big job to resolve their impacts on all aspects.

Primary Atmospheric Changes

Global changes are occurring in the composition of both the lower atmosphere (the troposphere) and the upper atmosphere (the stratosphere). In the troposphere the most important changes are the rapidly increasing concentration of the gases carbon dioxide, chlorofluorocarbons, methane, nitrous oxide and, to some extent, ozone. These changes are well documented and there is no room for doubt about the validity of the measured increases (Pearman, 1987). In the stratosphere increase in CFCs and halons and decrease in ozone is the subject of concern, but there is more uncertainty about the long term ozone changes in the stratosphere.

Table 1. Concentrations and rates of increase of the main greenhouse effect gases.


Concentration (parts per billion)

Rate of increase
(% per year)


Carbon dioxide

350 000



1 700





Nitrous oxide






Carbon dioxide increase is by far the biggest cause for concern. Of the offending trace gases, it is highest in concentration (Table 1) and increasing at the most rapid rate in absolute terms (Figure 1). In the table the concentrations are expressed in parts per billion. They are too small to express conveniently in terms of percent - that is in parts per hundred.

Fjgure 1. The history of global atmospheric carbon dioxide concentration based on Antarctic measurements. Solid line: instrumental monitoring of the atmosphere. Dashed line: measurements of air bubbles trapped in Antarctic ice-cores (ppmv: parts per million by volume). (Derived from Pearman, 1987).

Each of these gases contributes to the greenhouse effect because each absorbs radiation in different wavebands of the solar spectrum. Carbon dioxide contributes more than half the total greenhouse effect warming. The other ones, however, are more effective greenhouse effect gases than you might expect froni their relatively low concentration.

Until the Montreal agreement to reduce manufacture of chlorofluorocarbons, it was expected that CFCs, growing at 5-6% per year would be contributing about half as much greenhouse effect as CO alone would over the next several decades. But now, assuming the Montreal agreement is adhered to or strengthened, CFCs are expected to start to decline in relative importance as greenhouse effect gases.

However, there is no growing concern that methane could ultimately rival CO2 as a greenhouse effect as if the growth in its concentration keeps on at 1.2% p.a. compared with the 0.46% p.a. for CO2.

Agriculture has a special involvement here because it is probably a major contributor to methane emissions. Methane is given off, for example, in the belches of ruminant animals and by rice paddies. Cattle feeding on coarse cellulosic vegetation,as opposed to starchy grains, emit prodigious amounts of methane. This source needs to be carefully evaluated in comparison to other sources like release from coal mines, landfills and leakage from natural gas distribution systems.

Insofar as their direct impacts on vegetation goes, the chlorofluorocarbons, methane, nitrous oxide and carbon monoxide are at much too low concentration to have any direct g1obal effect on plants or animals. Ozone, which is increasing in the lower atmosphere, is toxic to plants and in some regions of the USA where it can be particularly abundant it is now at concentrations that are probably reducing the rate of plant growth somewhat (Heck et al. 1983). In Australia, ozone is probably not at sufficiently high concentration to have significant direct effects on vegetation in general. It seems unlikely to present much problem in the foreseeable future.

The big direct atmospheric impact on plants is from CO2 . Carbon dioxide is a plant fertiliser. Doubling the CO2 concentration to 700 parts per million typically increases growth and yield of experimental plants by 30-40% (Warrick and Gifford, 1986). There is much species variation in responsiveness to CO2 enrichment, however. Stimulation is larger for plants growing with water deficiency (Gifford, 1979) or at high temperatures (Idso and Kimball, 1989). But CO2 is probably rather less stimulatory for plants suffering mineral deficiency. Much work is needed to build a sound basis for prediction including for non-domesticated species and for mixed populations of species growing in competition. Higher carbon dioxide concentration also has the potential to reduce the amount of water used by plants. Whether these potential agricultural advantages are expressed in the field will depend on how we manage the agricultural system in the face of continuous gradual environmental change.

Turning to the stratosphere, we still do not know whether we have a potential global problem with reduced ozone levels letting increased ultraviolet radiation down to ground level. If ultraviolet were to increase, we would expect some, but not all, plant species to suffer somewhat reduced growth rates. However, short tern experiments with UV effects on plant performance have been misleading because plants exposed continuously to high UV in the field have been found to develop protective defences. For animals the risk is more frequent eye problems and incidence of skin cancer on exposed parts of the body, like the nose. As far as global stratospheric ozone is concerned, monitoring has shown a 2-3% decline in concentration over the last decade at non-polar latitudes. There is no clear evidence yet of actual problems from increased UV radiation over agricultural lands.

In summary, as long as we persist in abolishing CFCs over the next 2 or 3 decades, there seems little likelihood of significant direct impacts of UV change on Australian agriculture. Of course, there is always the risk of surprises and the CFCs already released will remain airborne for many decades. So a monitoring and analysis programme of stratospheric chemistry and dynamics of ozone must be sustained along with a research effort into biological impact of extra UV radiation.

Climatic Repercussions

Let us move on now to the climatic changes that may be being caused by the atmospheric composition change. The “greenhouse effect” is the predominant effect here, but changes in vegetation brought about by the “CO2 fertilising effect” are also expected to be involved in climate change.

Understanding how vegetation, and changes in vegetation, affect climate is one of the topics that is most exercising modellers of global climate at present. It is thought to play a significant part in climate determination but cannot yet be quantified properly. So for now we can only contemplate the repercussions of the greenhouse effect that involves warming at ground level and cooling of the stratosphere.

The reality of this warming tendency is almost without doubt. All models predict that a doubling CO2 concentration would cause a global average temperature increase of 1.50C to 4.50C. Also certain is that, associated with any temperature increase, there will be a humidity increase and a rainfall increase of about 10% on a global averaged basis. This is not only an integral part of the “greenhouse effect” theory, but is also born out by the historic and paleaoclimatic record. For example, the annual quantity of water discharged from the world’s major rivers was analysed recently in relation to temperature (Probst and Tardy, 1989). It was demonstrated that runoff was positively related to annual average temperature for 45 major river basins of the world over the last 80 years. This means that as temperature increases, the associated intensification of the hydrologic cycle increases water input to catchments more than higher temperatures increase evaporation from them. While convincing evidence that we are truly moving into a warmer, more humid and wetter, as well as CO2 enriched, world may still take a decade or so to emerge, it is looking very probable. However, that is where the high level of certainty ends. When it comes to regional details, details about the seasonal distribution of changed temperature and rainfall, and details about frequency of extreme events like heat waves, droughts, floods, and cyclones, our forecasting ability is abysmal. All we can be sure of is that, as the world warms up, climatic patterns are bound to change to some degree. Unfortunately there is likely to be places that become drier despite the overall wetter world.

Impacts On Plant Productivity

To illustrate the kind of impact on vegetation that might be occurring I have five examples. First, consider the continental scale. Global data on annual growth of different ecosystems in relation to annual rainfall and annual mean temperature has enabled a general model relating these three things for natural vegetation to be formulated. Pittock and Nix (1986) applied this model using present day rainfall and temperature data for Australia and to the climatic situation expected with twice as much atmospheric CO2 That is, they assumed that temperature increased by 1OC in the north of Australia grading to 4OC increase in Tasmania. Summer rainfall increased, winter rainfall decreased.

Figure 2. Percentage change, according to a simple mathematical model, in the net primary productivity of vegetation relative to the present for a climate scenario roughly equivalent to a doubling of atmospheric carbon dioxide concentration. It excludes the direct CO2 fertilising effect.

(Derived from Pittock and Nix, 1985).

Figure 2 shows the modelled result of those climatic changes alone. Annual vegetation growth increased over almost all the country except for the south-west coast of Western Australia which had a small decline in modelled vegetation productivity. The maximum percent increase, in excess of 40%, was in the Great Sandy Desert of Western Australia. This model did not take account of the direct CO2 fertilising effect which in my research experience could cause up to a 30-40% further increase in annual growth of the so-called C plant species that dominate temperate vegetation. This does not take account, however, of any change in fire frequency or intensity, pest and disease incidence, or the long term impact of any change in soil erosion and salinisation, all of which could modify the conclusions substantially. This exercise was a crude one. Much refinement is required. Species composition of the vegetation, for example, is bound to change with “greenhouse” boosted conditions.

While adapted perennial vegetation generally experiences increased growth with warmer temperatures, the yield of cereal crops seems likely to suffer reduction with warmer conditions - all else equal. The reason for this is that warm conditions hasten plant development rate and the attainment of maturity, so there is less time to fill the grain. Stapper examined this temperature effect for irrigated wheat at Griffith, NSW and for dryland wheat at Wagga Wagga, NSW, using a comprehensive and successful wheat yield simulation model (SIMTAG). A 3OC temperature increase alone reduced modelled grain yield of irrigated wheat by 28% (Figure 3). Such a yield reduction would be approximately cancelled by a CO2 doubling that would cause the warmer temperature.

Percent yield reduction of irrigated (Griffith, NSW) and dryland (Wagga Wagga, NSW) wheat as a function of temperature increase above today’s conditions. Results of a simulation model (from Gifford, 1987).

The dryland wheat result was different. While total crop biomass was reduced 18% by a 3OC warmer season, the proportion of biomass present as grain increased sharply because the acceleration in development meant that grain filling occurred at a wetter time of year. So, on balance, temperature alone had no effect on grain yield. With twice the amount of CO2 in the atmosphere and more rainfall, we can therefore expect about a 20-40% increase in grain yield of that non-irrigated wheat near Wagga.

The third example is of wheat production on the Darling Downs in south-east Queensland. The simulation model used in this study by McKeon et al. (1987) not only predicted yield, but also predicted soil loss resulting from an increase of water runoff from the field. The model predicts yield and soil loss under present conditions very well. They simulated the effect of a 20C temperature increase, combined with 20% less winter rainfall and 30% increased summer rainfall, but again did not include the CO2 fertilising effect. As Table 2 shows, the greenhouse scenario increased yield 23%. If the CO2 fertiliser effect had been included, yield could have increased by about 50-60% above today’s yield. The model also predicted an increase in the already serious soil loss, from 32 t/ha/yr to an astounding 95 t/ha/yr. The large soil loss occurred because of the unsustainable practice of bare fallow during the summer Months. The prospect of climate change makes change of cropping practice even more urgent on the Darling Downs according to that analysis. In principle, however, moister conditions and higher plant productivity should make the job of soil conservation easier if it is managed correctly. We must not miss the opportunity to face that challenge.

Table 2. Simulation of wheat yield, water runoff and soil loss from the Darling Downs, Queensland, with temperature and rainfall changed according to a possible scenario of climate under doubled greenhouse-effect gas concentrations (from McKeon et al., 1987).


Crop yield (t/ha/yr)

Rainfall runoff

Soil loss (t/ha/yr)

With 1912—1984 climate




1912—84 climate +2 0C




1912—84 climate +30% summer rainfall and -20% winter rainfall




1912-84 with both temperature and rainfall changes




Besides the risk of further acceleration of soil erosion from bare fallow, there are other risks from climatic change which could undermine the potential production increases unless managed correctly. These risks include the speculative possibilities of greater season-to-season weather variability, worsened pest and disease incidence and weed competition, increased wild-fire frequency and intensity, changes in the frequences and intensities of windstorms, droughts, floods, frosts and heavy rain. The probability of increase or decrease in these risks cannot yet be assessed. So it is essential that management strategies be devised which will protect against such hazardous eventualities so that the potential increase in productivity from the warmer, wetter, flare humid, CO2 enriched greenhouse-like conditions is not lost.

A more clear cut negative effect on crop production is likely to occur for production from species for which a winter chilling response of reproductive development occurs. The mild winter of 1988 apparently led to poor stone fruit yields in several orchard areas because the cumulative chilling requirement for successful fruiting was not met. This is a non-linear effect of warmer conditions. That is, the size of the yield reduction is out of all proportion to the size of the average winter temperature increase. A one degree average temperature increase can have a huge effect in marginal areas for, say, peach production. Since orchards are established for long periods it would be prudent for any orchardist to study his situation very carefully before choosing a variety or species for a new orchard planting.

Besides such disadvantages of the global change, even the potential advantages could turn into disadvantages if the pace of change is faster than the land-management community can modify farming practices and organisational arrangements to match the continuously changing conditions. Furthermore, if the entire vast fossil fuel resources of the world were eventually burned, the climatic consequences would doubtless be generally catastrophic involving a 6-90C increase in global average temperature. Every effort should therefore be made in Australia and elsewhere to reduce the global rate of emissions of the greenhouse effect gases to as low a rate as possible as soon as possible. However, since Australia has little control over that and the world is inevitably committed to significant carbon dioxide increase and climate change over the next few decades, we must learn to be flexible, to forecast change accurately and adapt.


1. Gifford, R.M. (1979). Growth and yield of CO2 - enriched wheat under water limited conditions. Aust.J.Plant Physiol. 6367-378.

2. Gifford, R.M. (1987). Direct effects of higher carbon dioxide concentrations on vegetation. In “Greenhouse: planning for climate change”, ed. G.I. Pearnian. Melbourne, CSIRO, pp506-545.

3. Heck, W.H., Adams, R.M., Cure, W.W., Heagle, A.S., Heggestad, H.E., Kohnut, R.J., Kress, L.W., Rawlings, 3.0. and Taylor, O.C. (1983). A reassessment of crop loss from ozone. Environ.Sci. and Technol. 17:574a-581a.

4. Idso, S.B. and Kimball, B.A. (1989). Growth response of carrot and radish to CO2 enrichment. Environ, and Exp. Botany 29:135-139.

5. McKeon, G.M., Howden, S.M., Silburn, D.M., Carter, 3.0., Clewett, 3.F., Hammer, G.L., Johnston, Lloyd P.L., Mott, J.J., Walker, B., Weston, E.J. and Wilcocks, J.R. (1987). The effect of climate change on crop and pastoral production in Queensland. In “Greenhouse: planning for climate change”, ed. G.I. Pearman. Melbourne, CSIRO pp 546-563.

6. Pearman, G.I. (1987). Greenhouse gases: evidence for atmospheric changes and anthropogenic causes. In “Greenhouse: planning for climate change”, ed. G.I. Pearman. Melbourne, (‘SIRO, pp3-2l.

7. Pittock, A.B. and Nix, H. .A. (1985). The effect of changing climate on Australian biomass production. Climatic Change 8:243-255.

8. Probst, .J.L. and Tardy, Y. (1989). Global runoff fluctuations during the last 80 years in relation to world temperature change. American J. Science 289:267—285.

9. Warrick, R.A. and Gifford, R.M. (1986). CO2, climatic change and agriculture:

10. assessing the response of food crops to the direct effects of increased CO and climate change. ln “The greenhouse effect, climatic change, and ecosystems”, eds B. Bolin, B.R. Doos,. Jager and R.A. Warrick. Chichester, Wiley and Sons, pp393-474.

Previous PageTop Of PageNext Page