The Jasper Ridge Global Change Experiment
1992-1997 Open-top Chamber Experiment
The primary objectives of this experiment were to study the effects of increased atmospheric CO2 on carbon and water budgets (how much enters, is kept in place, and leaves), the nitrogen cycle, and species composition. All of these factors interact in an ecosystem and are influenced by CO2 concentration.
Experimental Design and Facilities
The open-top chamber experimental design involved two separate facilities at Jasper Ridge. Field chamber facilities were located on a ridge-top in an area where a sandstone soil and serpentine soil abut. The MECCA (MicroEcosystems for Climate Change Analysis) facilities were located approximately 1.5 kilometers from the field site. The MECCA facilities consisted of 20 open chamber enclosures. Each chamber contained 24-36 microcosms of the local grassland communities. These microcosms consisted of large, vertical sections of PVC pipe (the sections were one meter tall and either .2 or .4 meters in diameter) that were filled with soil, and seeded with species representative of the natural ecosystems. This MECCA design made a high number of treatment replicates possible. Both the field chambers and the MECCA chambers were supplied with either ambient air or CO2 enriched air through a system of blowers, ducts, and manifolds. Both these facilities were almost completely dismantled at the end of the program. See the figures for design details.
The chamber walls were constructed of polyethylene to allow light transmission. The field chambers were open on top to allow precipitation to enter. Each MECCA chamber was covered with a screen that distributed ambient rainfall uniformly inside the chamber.
These facilities were designed to explore long-term ecosystem-scale responses to elevated CO2. The field facility design allowed study of adjacent, natural serpentine and sandstone ecosystems. This design allowed comparing responses of the nutrient-limited serpentine grassland with those of the moderately fertile sandstone grassland. These two ecosystems were also simulated in the MECCA chambers╣ microcosms.
In the field and MECCA chambers, CO2 was delivered at ambient or twice ambient levels. Nutrient levels and water inputs were also manipulated in some of the MECCA chambers╣ microcosms.
The overall design concept was for the field site to both test hypotheses in a reasonably natural setting and provide a reference point for evaluating MECCA site experiments. In the artificial ecosystems created at the MECCA site, experiments were performed which involved many combinations of CO2, water, nutrients, soils, and plant species.
Overall, the Mecca experiments and the field plots showed similar sensitivities to nutrient and CO2 changes.
The adjustment of plant water budgets played an important role in several responses observed in the experiment. Higher levels of atmospheric CO2 can cause plants to partially close leaf pores (stomata) through which CO2 enters leaves for photosynthesis and water vapor escapes to the atmosphere. This partial closing of the stomata under elevated CO2 can allow the plants to acquire more CO2 with less water loss. Lower plant water loss (higher water use efficiency) can result in increased soil moisture.
Aboveground production (biomass) in both the field and MECCA sites increased under conditions of elevated CO2 and ambient nutrients when production for the entire growing season was measured. However, the increase at both sites was a consequence of increased aboveground biomass of species requiring late-season soil moisture. It appears that the increased production of late season annuals was not a direct effect of increased CO2 but an indirect effect due to the greater water use efficiency of the plant community earlier in the season induced by elevated CO2. This greater water use efficiency earlier in the season increased late-season soil moisture.
Early and mid-season annuals did not achieve higher productivity under higher CO2 when aboveground biomass was measured in May. These earlier developing annuals have developmental transitions, such as flowering, that are largely regulated by photoperiod and/or cumulative heat. These developmental switches make them less capable of extending their growing period significantly to take advantage of additional water availability later in the season. In this Mediterranean climate, water availability is not normally a prolonged limiting resource for early and mid-season annuals during their relatively inflexible growing season that usually ends in April-May at Jasper Ridge. Under elevated CO2, active transpiration of late annuals continued longer into the summer drought, indicating a shift in seasonality. Generally, the experiment showed that elevated CO2 induced growth increases in plants with longer or flexible lifespans such as tarweeds (Hemizonia).
In the field plots, under elevated CO2, soil moisture increased significantly in sandstone plots. Soil moisture changes were small and inconsistent in serpentine plots. The sandstone community produces more plant canopy than the serpentine community. The denser canopy probably contributed to soil moisture retention. The porous and rocky nature of serpentine soil probably contributed to its moisture loss.
Increased soil moisture (and probably increased root exudates) increased soil microbial biomass which, in turn, influenced nutrient competition between microbes and plants. In general, it was concluded that belowground CO2 induced changes that may play a larger role in ecosystem function than aboveground changes.
The response of litter decomposition to increases in CO2 concentration can affect the global carbon cycle. Some of the CO2-induced mechanisms that could affect litter and the carbon cycle are: (1) changes in the amount of litter (2) changes in litter chemistry of individual species, (3) shifts in plant species, (4) altered plant structure affecting the allocation of litter between root, soil surface, and aboveground locations, (5) increases in soil moisture, and (6) changes in microbial activity.
Litter decomposition was studied in an experiment beginning in 1994 involving sandstone ecosystems at both the field and MECCA sites. Litter for the experiment consisted of fully senescent shoot tissue (aging material that is no longer photosynthesizing) from annual grasses and legumes. The litter was from plants grown under combinations of the following treatments: ambient air, CO2 enriched air at twice ambient, unfertilized, and fertilized. The harvested litter was segregated by treatment and placed in litter bags. During the following growing season, litter bags were placed in both microcosms and field plots in varying positions: in the soil, on the soil surface, and suspended 5 cm above the soil surface. The microcosms containing the litter bags were exposed to the same treatments as were used to produce the litter. Litter bags were positioned in November and collected for analysis the following July. Measurements included soil moisture, litter chemistry, litter mass loss, and nitrogen dynamics. The following conclusions were reached. Little change in litter chemistry was found under varying CO2 conditions when species composition of the litter was held constant. CO2 conditions during plant growth did not appear to affect litter mass loss. Although wetter soils due to elevated CO2 can lead to accelerated decomposition rates, in this experiment litter mass loss was not affected by any treatment. It is likely that shifts in species composition of communities, such as stimulation of legume growth, will have more important effects on litter chemistry than direct CO2 concentration effects.
The microcosm experiments studied the success of non-native grasses under elevated CO2 and elevated nutrient levels. Higher CO2 concentration did not appear to increase productivity of these grasses in serpentine soil. However, nutrient addition did increase productivity of non-native grasses in sandstone soil. These findings suggest that soil fertility affects the success of non-native grasses. Nutrient addition in the form of nitrogen deposition from air pollution may increase the dominance of non-native grasses.
An important conclusion from the Open-top Chamber Experiment related to carbon storage. It has been thought that some ecosystems might respond to higher levels of CO2 by producing more biomass (fixing more carbon) above and below ground, thereby storing some excess carbon from the atmosphere. This was not found to be the case in the two grassland ecosystems studied. Overall, in these experiments, under higher CO2, more carbon was fixed during the growing season and more was released by system respiration. There was an increase in carbon flowing through the system but no evidence of increased carbon storage.
Experimental Design Issues
The use of chambers and blowers in this experiment had several impacts on the ecosystems studied. The chambers inhibited insect access to the plants, thereby affecting pollination and herbivory. The use of blowers to mix CO2 additions and deliver air to the plants somewhat accelerated drying out of plants. Finally, the chambers inhibited the entrance of wind blown rain into the plant community, which, in turn, affected the ratio of standing to lying litter (dried plant material). The physical position of litter affects the litter decomposition rate, which plays an important role in carbon storage and the carbon cycle.
Although these effects do not invalidate the experiment╣s conclusions, the Jasper Ridge Global Change Experiment that began in 1997 uses a design that eliminates these effects.
ReferencesChiariello, NR, and Field, CB (1996) Annual grassland responses to elevated CO2 in long-term community microcosms. In: Community, Population and Evolutionary Responses to Elevated Carbon dioxide Concentration, pp. 139-175. Academic Press, San Diego. Dukes, JS, and Field, CB (2000) Diverse mechanisms for CO2 effects on grassland litter decomposition. Global Change Biology 6, pp. 145-154. Field, CB, Chapin, FS, III, Chiariello, NR, Holland, EA, and Mooney, HA (1996) The Jasper Ridge CO2 experiment: Design and Motivation. In: Carbon Dioxide and Terrestrial Ecosystems, pp. 121-145. Academic Press, San Diego.
Field, CB, Lund, CP, Chiariello, NR, and Mortimer, BE (1997) CO2 effects on the water budget of grassland microcosm communities. In: Global Change Biology 3, pp. 197-206.