Limited effect of increased atmospheric CO2 concentration on ombrotrophic bog vegetation
Author for correspondence: Marcel R. Hoosbeek Tel: +31 317 484109 Fax: +31 317 482419 Email:marcel.hoosbeek@ bodeco.beng.wag-ur.nl
- • Boreal and subarctic peatlands contain 20–30% of the world’s soil organic carbon, and if growing, they constitute sinks for atmospheric CO2. We hypothesized that even in the nutrient-poor bog environment, elevated CO2 would stimulate Sphagnum growth more than vascular plant growth, thereby improving Sphagnum’s competitive strength and enhancing carbon (C) sequestration.
- • Free-air carbon dioxide enrichment (FACE) experiments took place on predominantly ombrotrophic peatbog-lawns in Finland (FI), Sweden (SW), The Netherlands (NL), and Switzerland (CH).
- • After 3 yr of treatment, increased CO2 concentration (560 ppm on volume basis) had no significant effect on Sphagnum or vascular plant biomass at either site.
- • This research suggests that, just as with other nutrient-poor ecosystems, increased atmospheric CO2 concentrations will have a limited effect on bog ecosystems.
Since the industrial revolution, the atmospheric CO2 concentration has been increasing exponentially. However, the rate of increase of atmospheric CO2-C during the 1980s (3.2 ± 0.2 Gt C yr−1) is smaller than expected when considering burning of fossil C (5.5 ± 0.5 Gt C yr−1 in the 1980s), which is in part because of feedbacks involving terrestrial ecosystems (CO2 fertilization: 1.0 ± 0.5 Gt C yr−1) (Schimel, 1995). About 60 Gt of CO2-C is cycled annually along the terrestrial photosynthesis-decomposition pathway, 20 times more than the annual net addition to the atmosphere (Goudriaan, 1992). So, even small changes in net primary productivity or in decomposition of soil organic carbon (SOC) could significantly influence the net increase of atmospheric CO2. Because elevated CO2 may strongly influence net primary productivity, species abundance and community composition (Mooney et al., 1999) and the chemical and physical composition of plant material, and therefore the decomposability of plant litter, strong feedbacks to the pools of SOC are expected. In this regard, northern peatlands are particularly important, because they contain 20–30% of the world’s SOC (Gorham, 1991), and if growing, they constitute a sink for atmospheric CO2 (Van Breemen, 1995).
Through their anatomical, morphological, physiological and organo-chemical properties, Sphagnum species create waterlogged, nutrient-poor, acidic conditions, thereby outcompeting most vascular plants (Van Breemen, 1995). We hypothesized that even in the nutrient-poor bog environment, elevated CO2 will stimulate plant growth (Silvola, 1985; Jauhiainen et al., 1992, 1994). We expected growth of Sphagnum to be stimulated more than that of vascular plants because increased growth lowers N concentration, which affects vascular plants much more than Sphagnum, because growth of Sphagnum is less nutrient limited than that of co-occurring vascular plants (Sveinbjörnsson & Oechel, 1992; Jauhiainen et al., 1998). Increased peat growth (due to lower decomposition rates of Sphagnum vs. vascular plant material), which positively feeds back to Sphagnum itself further depresses vascular plants. This sequence of events forms a strong positive feedback of increased sequestration of CO2-C.
The Bog Ecosystem Research Initiative (BERI) was funded by the European Commission and comprised nine research groups in five European countries. In each country, experimental field sites were installed on ombrotrophic bogs dominated by a mixture of peat mosses (Sphagnum spp.) and vascular plants. The hypothesis was tested with free-air CO2 enrichment (FACE) technology during the growing seasons of 1996–98.
Materials and Methods
The FACE experiments took place on predominantly ombrotrophic peatbog-lawns in eastern Finland (FI), southern Sweden (SW), The Netherlands (NL) and Switzerland (CH). The fifth BERI site in north-west England is not included in this study because Sphagnum biomass data were not available. Locations, climate summaries and atmospheric N deposition rates are provided in Table 1. The experiments in FI, SW and CH took place in situ, but in NL large peat cores (1.1 m diameter and 0.6 m depth) were collected in frozen condition and transferred to containers near Wageningen. The containers (same size as peat cores) were buried to a depth of 0.5 m in a grass lawn. An overview of the moss and vascular plant species composition of the field sites is provided in Table 2.
Table 1. Location, climate summary, and atmospheric N deposition
|FI||Salmisuo, Ilomantsi, eastern Finland (62°47′ N, 30°56′ E)|| −12||16|| 600|| 183||0.4|
|SW||Kopparåsmyren, southern Sweden (57°8′ N, 14°30′ E)|| −2||16|| 800|| 90–100||0.8|
|NL||Peat from Dwingeloo (52°49′ N, 6°25′ E) transplanted to Wageningen (51°99′ N, 5°70′ E)|| 2||17|| 805||< 10||3.9|
|CH||La Chaux-des-Breuleux, Swiss Jura (47°13′ N, 7°3′ E)|| −5||15||1390|| 80–120||1.8|
Table 2. Overview of the dominant moss and vascular plant species in 1996
|FI||Sphagnum balticum||61||Eriophorum vaginatum||14|
| ||Sphagnum papillosum||33||Vaccinium oxycoccos|| 4|
| ||Sphagnum magellanicum|| 4||Andromeda polifolia|| 2|
| || || ||Scheuchzeria palustris|| 2|
|SW||Sphagnum magellanicum||69||Eriophorum angustifolium|| 8|
| ||Sphagnum papillosum||13||Vaccinium oxycoccos|| 3|
| ||Sphagnum balticum|| 7||Drosera rotundifolia|| 3|
| ||Sphagnum rubellum|| 7||Calluna vulgaris|| 2|
| || || ||Andromeda polifolia|| 2|
|NL||Sphagnum magellanicum||97||Vaccinium oxycoccos||19|
| ||Sphagnum papillosum|| 1||Erica tetralix|| 9|
| || || ||Eriophorum angustifolium|| 4|
| || || ||Drosera rotundifolia|| 2|
| || || ||Calluna vulgaris|| 1|
|CH||Sphagnum fallax||62||Carex nigra|| 3|
| ||Polytrichum strictum||37||Vaccinium oxycoccos|| 3|
| || || ||Eriophorum vaginatum|| 2|
At each site 10 FACE rings (Miglietta et al., 2001) with a diameter of 1 m (in NL 1.1 m) were randomly laid out on the bog surface. In five rings the atmospheric CO2 concentration was kept at ambient levels (about 360 ppm on volume basis) while in the other rings the CO2 concentration was maintained at 560 ppm for 24 h d−1. The elevated CO2 rings were located at a distance of at least 6 m from ambient air rings to prevent CO2 pollution. To avoid edge effects, no data were collected in the outer 15 cm of each plot.
Blowers next to each FACE ring supplied ambient air or CO2-enriched air to circular tubes resting on the bog surface on which 72 small venting pipes were mounted. The venting pipes had small holes directed towards the centre of the ring at 6 and 12 cm height above the moss surface. Air was sampled in the middle of elevated CO2 rings at 7.5 cm above the moss surface and analysed for CO2 with an infrared gas analyser. Based on the measured CO2 concentration and wind speed, the CO2 supply was adjusted automatically via a PC and mass flow controllers to maintain the target concentration of 560 ppm. During the winter months the FACE system was turned off because of minimal biological activity.
Calibration and test experiments were conducted to optimize and evaluate the performance of the FACE equipment. System performance was considered adequate when the 1-min average CO2 concentrations measured at the centre of the ring stayed within 20% of the preset target for more than 80% of the time. The temporal performance of the BERI FACE systems largely met these quality criteria. When calculated over the entire season, and averaged over the five replicated rings, the percentage of time in which 1-min average CO2 concentrations deviated less than 20% from the preset 560 ppm target was 95% in FI, 96% in SW, 98% in NL and 96% in CH (Miglietta et al., 2001).
Plant cover was measured nondestructively using the point-quadrat method (Jonasson, 1988). The point-quadrat covered a permanently marked area of 25 by 37.5 cm and consisted of 150 points with a grid size of 2.5 cm. At each of the 150 points a needle was lowered and vegetation hits were recorded. Sphagnum cover was measured at the beginning of the experiment (May or June 1996) and thereafter at the end of each growing season. Vascular plant cover was determined each year at ‘peak biomass’.
Sphagnum length growth was measured according to a modified cranked wire method (Clymo, 1970). In each ring four stainless steel wires with a small brush at the lower end were inserted into the peat and fixed at 2 cm below the surface. The distance between the moss surface and the top of the wire was measured each month during the growing seasons. At the end of the experiment Sphagnum was harvested at each wire in 7.5 cm diameter columns for conversion of length to biomass. The bulk density of the moss cover was determined for the sections 0–1 cm (capitula) and 1–3 cm below the moss surface by drying at 70°C to constant weight.
Above ground vascular plant biomass was determined at the end of the experiment by clipping off all vascular plants at the moss surface in a 37 × 25 cm2 area per ring. All species were sorted into current year and older parts and dried at 70°C to constant weight. Below ground vascular biomass was determined in three 10-cm diameter and 30-cm deep columns per ring. After sorting out, all fractions were dried at 70°C to constant weight.
Sphagnum biomass production under ambient CO2 ranged from 190 ± 43 in CH to 310 ± 33 g m−2 yr−1 in NL (Table 3). We did not find significant increased CO2 concentration treatment effects on Sphagnum biomass at either site (using ANOVA with site and treatment as factors). Below ground vascular biomass was about four times larger than above ground vascular biomass in FI and SW (Table 4). In NL and CH this difference was smaller. The above ground response to increased CO2 concentration was positive in FI, NL and CH, but negative in SW, while the below ground CO2 response was positive at all sites. However, none of these responses is statistically significant. Because neither Sphagnum nor vascular plant biomass production was affected by increased CO2 concentration in this 3-yr experiment, the hypothesis stating that elevated CO2 would stimulate the growth of Sphagnum more than that of vascular plants was rejected.
Table 3. Sphagnum dry weight production (g m−2 years−1) after 3 yr of exposure to either ambient or increased CO2 concentrations in FACE rings (n = 5; n.s., not significant at P ≤ 0.05; SE, standard error)
|FI||256 ± 23||273 ± 38||1.1||n.s.|
|SW||211 ± 25||187 ± 19||0.9||n.s.|
|NL||310 ± 33||377 ± 19||1.2||n.s.|
|CH||190 ± 43||162 ± 16||0.9||n.s.|
Table 4. Vascular plant above and below ground biomass (g m−2) in 1998 after 3 years of exposure to either ambient or increased CO2 concentrations in FACE rings (n = 5; n.s. = not significant at P ≤ 0.05)
|FI||a 73 ± 9||a 85 ± 11||1.2||n.s.|
| ||b 272 ± 55||b 300 ± 64||1.1||n.s.|
|SW||a 67 ± 7||a 52 ± 12||0.8||n.s.|
| ||b 297 ± 61||b 383 ± 23||1.3||n.s.|
|NL||a 227 ± 29||a 286 ± 60||1.3||n.s.|
| ||b 350 ± 71||b 413 ± 70||1.2||n.s.|
|CH||a 122 ± 14||a 124 ± 24||1.0||n.s.|
| ||b 248 ± 59||b 311 ± 77||1.3||n.s.|
The BERI FACE experiment has been the only CO2 enrichment study in bog ecosystems. Direct comparison with similar nutrient-poor wetland ecosystems is therefore not possible. Other nutrient-poor ecosystems, like the Arctic tundra and Alpine grassland, also showed no or small increases of biomass under elevated CO2 (Körner, 1996). According to Oechel & Vourlitis (1994) nutrient availability in these nutrient-poor ecosystems must increase before plants are able to benefit from CO2 fertilization.
The effect of elevated CO2 on a nutrient-rich wetland ecosystem was investigated by Drake (1992), who used open top chambers on a Chesapeake Bay marsh. In a mixed community of the C3 sedge Scirpus olneyi and the C4 grasses Spartina patens and Distichlis spicata, the biomass of the C3 component increased over 100%, which was accompanied by a decrease in the biomass of the C4 component of the community. In general, nutrient-rich ecosystems show an increase in biomass under elevated CO2 (Körner, 1996).
Because of their anatomical, morphological, physiological and organo-chemical properties and the ability of Sphagnum species to create waterlogged, nutrient-poor, acidic conditions, thereby outcompeting most vascular plants, we hypothesized that under nutrient-poor conditions Sphagnum species would still be able to benefit from elevated CO2, thereby improving their competitive strength. However, this research suggests that, just as with other nutrient-poor ecosystems, elevated CO2 will have a limited effect on bog ecosystems.
The European Commission funded this research (ENV4-CT95–0028). The Swiss part of the project was funded by the Swiss Federal Office for Education and Science (project 95.0415).