Elements of Change 1995: Session I Nijs

In the context of global climatic change, Nijs and colleagues have undertaken a series of experiments aimed at increasing understanding of the effects of increased carbon dioxide and temperature on grassland vegetation. Goals include a better assessment of the risks and opportunities for agricultural production as well as improved knowledge of ecosystem functioning. Their focus is on grasslands in natural and agroecosystems because grasslands make up 50% of agricultural land in Belgium, constitute a major carbon pool of the global carbon cycle, are a stabilizing factor in soils, and are a major pool of biodiversity. The scale of focus in this work ranges from the sub-cellular to the ecosystem level. The work is primarily experimental but also includes some modeling.

The first approach, in 1986, was to set up ecosystems in enclosed sunlit chambers and use these to estimate carbon balance over minutes, days, weeks, seasons, and years. The enclosures were contained in a greenhouse. Control temperatures were designed to recreate ambient conditions in Antwerp and CO₂ levels were ambient or enriched. A switching system enabled the researchers to sample the different chambers. They obtained daily profiles of carbon flux density and CO₂ exchange rates, and evapotranspiration rates for the whole system. Drought stress experiments were also being conducted to see if plants grown in elevated CO₂ conditions are more susceptible to prolonged drought conditions.

Leaf thickness increased by CO₂ enrichment could reduce the dose of UV-B plants receive in sensitive areas. Conversely, ecosystems exposed to higher temperatures typically yield thinner leaves, which could make them more susceptible to UV-B.

In a more recent system, built in 1991, the simulations of climatic change involve exposing test plots to:

Free air CO₂ enrichment experiments (FACE) are now gaining in popularity over enclosed systems. In such a system, emission points for CO₂-enriched air flood the vegetation under study. There are grassland FACE experiments underway (Zürich, Switzerland), but one key problem with FACE is that air temperature cannot be controlled. In a newly developed technique at the University of Antwerp, infrared lamps are used to raise ambient canopy temperatures. They can then be compared to controls using thermocouples shielded from sunlight, yielding a 2.5°C rise. Infrared thermometers would be even better.

A new project will study interactions between UV-B rise due to ozone depletion, CO₂ enrichment, and temperature rise, in an effort to uncover synergisms that might emerge. One possible interaction mechanism is that leaf thickness increased by CO₂ enrichment could alter the dose of UV-B plants receive in sensitive areas. The thickening caused by CO₂ enhancement is either an increase in density or in number of cell layers. Conversely, ecosystems exposed to higher temperatures typically yield thinner leaves, which could make them more susceptible to UV-B.

A key question in recent experiments was "Is N availability a limiting factor for the response to CO₂?" This is relevant because N inputs are high in agriculture and low in natural systems. The answer, from these experiments on cool-temperate grasses, may have relevance to current European Union thinking that there is too much land in agricultural use and that this is causing undue environmental stress (such as increase in environmental nitrates). It could influence policy towards altering heavily used agricultural land into more extensively managed lands, depending on their future CO₂ response.

Further, there are questions about how complex the ecosystem needs to be in order to get accurate results. Plants grow naturally in complex stands so it doesn't make sense to eliminate competition in the experiments. Also, plants' reactions may change with different neighbors. A series of experiments focusing on two species interactions is presently being conducted at the University of Antwerp (UA) to derive general principles on how CO₂ and temperature influence resource acquisition (nutrients, light, space, and water). First results show that modification of resource capture and its dynamics can potentially explain shifts in species composition within grasslands.

Experiments on temporal variability reveal that CO₂ and temperature effects change during the growing season. In early Spring, for example, high temperature increased productivity; this response then reversed in summer when it severely depressed productivity.

Within a group of six species, examined at UA in a screening experiment, the response to either CO₂ or temperature was generally the same in sign (whether positive or negative) for a great number of plant or stand features. This demonstrates that there is a functional type for cool-temperate species, in the sense that they all respond in basically the same way to the stimuli. Figure 11.1 summarizes the response of the major characteristics to elevated CO₂ and elevated temperature. For many of those characteristics, CO₂ and temperature effects were opposite.

Experiments on temporal variability reveal that CO₂ and temperature effects change during the growing season. In early Spring, for example, high temperature increased productivity; this response then reversed in summer when it severely depressed productivity. This emphasizes the need to examine results over the full growing season.

CO₂ and temperature both affect assimilation capacity which is in turn highly determined by the leaf nitrogen status. Examining the relationship between photosynthetic productivity and nitrogen provides a connection to a number of issues, such as eutrophication, excess fertilization, atmospheric nitrogen deposition, and the control function of nitrogen in the plant. It further allows an assessment of whether limited nutrient availability can reduce the response to CO₂.

The procedure to approach assimilation used here was based on a separation of the biophysical and biochemical components of photosynthesis. This was achieved by combining several models including the model of Farquhar (1980) and the model of Ball (1987). This allows an investigation of whether the photosynthesis-nitrogen relationship is universal for different species and different growth conditions which would allow the modeling of carbon uptake as a function of the nitrogen status of the ecosystem.

Both long-term exposure to elevated CO₂ or increased air temperature reduced the photosynthetic efficiency of leaf nitrogen.

Results from several species show that this was not the case; both long-term exposure to elevated CO₂ or increased air temperature reduced the photosynthetic efficiency of leaf nitrogen. The actual leaf nitrogen efficiency in elevated CO₂ increased though, due to the direct CO₂ effect, which compensated for the long-term acclimation. It was also shown that exposure to the global change treatments used here induced an internal reorganization within the leaf in the sense that the investment of nitrogen in different processes was altered to maintain optimal functioning. Prior to further extrapolations, the model results were compared with independently measured data, which assured a sufficient predictive capability.

Subsequent model results provided an answer to the principle question of this project, namely, whether nutrient deficient ecosystems will still be able to respond to elevated CO₂. Figure 11.2 shows that the model predicts no attenuation in the response to CO₂ when the leaf nutrient status deteriorates, even up to the point of complete N-deficiency. This result was confirmed by harvests of both above-ground and total stand dry matter in different species. The response to warming, on the other hand, does depend on nutrient availability. In a nutrient-rich environment, the effect of +4°C is predominantly negative, but this is reversed to a positive effect when the N status of the leaf becomes more deficient. Long-term modeling from other studies (Parton, 1995) suggests that some of the effects found here would decline, but not disappear, over time.

The results from this study of grassland ecosystems suggest that non-CO₂ greenhouse gas emissions should be preferentially reduced. While the direct fertilization effects of CO₂ will compensate for its negative warming effects on productivity, methane, NOX and CFCs have no such compensating effects.

The experimental and modeling results produced by this study provide an underpinning for agricultural management policy and land use planning. They predict consequences of certain land use decisions, particularly with respect to fertilizer inputs. Damage from high temperature, for example, can predominantly be expected in agroecosystems while natural or extensively managed ecosystems can better withstand future temperature increases which will make them become relatively more important as carbon reservoirs. There is little potential for manipulation in this however as a lower N input into grasslands would reduce productivity more than it would alleviate the temperature stress.

Concerning the seasonal dynamics of productivity, the results indicate that the difference between Spring and Summer productivity will increase in a future climate. In addition to this, an increased sensitivity to extreme weather events ( e. g. high temperatures) is expected to further exacerbate this trend in seasonal dynamics.

With respect to greenhouse gas reduction strategies, the results from this study of grass land ecosystems suggest that non-CO₂ greenhouse gas emissions should be preferentially reduced. While the direct fertilization effects of CO₂ will compensate for its negative warming effects on productivity, methane, NOX and CFCs have no such compensating effects. A full analysis of this however, will have to include a similar investigation for the other major ecosystem types of the world, as well as the specific emission reduction costs of each of these atmospheric components.