You are on page 1of 6

Oecologia (1997) 111:249±254

Ó Springer-Verlag 1997

Johann G Zaller á John A. Arnone III

Activity of surface-casting earthworms in a calcareous grassland under elevated atmospheric CO2

Received: 3 November 1996 / Accepted: 11 January 1997

Abstract Earthworms make up the dominant fraction of the biomass of soil animals in most temperate grasslands and have important e€ects on the structure and function of these ecosystems. We hypothesized that the e€ects of elevated atmospheric CO2 on soil moisture and plant biomass production would increase earthworm activity, expressed as surface cast production. Using a screen-aided CO2 control facility (open top and open bottom rings), eight 1.2-m2 grassland plots in Switzerland have been maintained since March 1994 at ambient CO2 concentrations (350 ll CO2 l)1) and eight at elevated CO2 (610 ll CO2 l)1). Cumulative earthworm surface cast production measured 40 times over 1 year (April 1995±April 1996) in plots treated with elevated CO2 (2206 g dry mass m)2 year)1) was 35% greater (P<0.05) than that measured in plant communities maintained at ambient CO2 (1633 g dry mass m)2 year)1). At these rates of surface cast production, worms would require about 100 years to egest the equivalent of the amount of soil now found in the Ah horizon (top 15 cm) under current ambient CO2 concentrations, and 75 years under elevated CO2. Elevated atmospheric CO2 had no in¯uence on the seasonality of earthworm activity. Cumulative surface cast production measured over the 7-week period immediately following the 6week summer dry period in 1995 (no surface casting) was positively correlated (P<0.05) with the mean soil water content calculated over this dry and subsequent wetter period, when viewed across all treatments. However, no correlations were observed with soil temperature or with annual aboveground plant biomass productivity. No CO2-related di€erences were observed in total nitrogen (Ntot) and organic carbon (Corg) concentration of surface casts, although concentrations of both elements varied seasonally. The CO2-induced

increase in earthworm surface casting activity corresponded to a 30% increase of the amount of Ntot (8.9 mg N m)2 vs. 6.9 mg N m)2) and Corg (126 mg C m)2 vs. 94 mg C m)2) egested by the worms in one year. Thus, our results demonstrate an important indirect stimulatory e€ect of elevated atmospheric CO2 on earthworm activity which may have profound e€ects on ecosystem function and plant community structure in the long term. Key words Carbon dioxide enrichment á Cast production á Cast C and N á Lumbricidae á Producer-consumer interactions

Introduction
Ever since the late 1800s, with the pioneering work of Hensen (1877) and Darwin (1881), earthworms have been known for the important e€ects they have on the chemistry and physical structure of soils (e.g. Sto È ckli 1928; Edwards and Lofty 1972; Satchell 1983; Lee 1985; Edwards and Bohlen 1996). Earthworms speed up the decomposition of plant litter and mineralization of soil organic matter by reducing the size of detritus particles for microorganisms (Swift et al. 1979) and by mineralizing nutrients directly in their guts (Barley and Jennings 1959; Sharpley and Syers 1977; Syers et al. 1979). Earthworms can also strongly in¯uence the density, diversity and activity of soil microorganisms (e.g. Parle 1963; Scheu 1987; Daniel and Anderson 1992; Brown 1995). Consequently, the productivity of plant communities has been shown to be enhanced by the presence of earthworms (Hopp and Slater 1948; Waters 1955; Stockdill 1982), and data from microcosm studies suggest that earthworms may also signi®cantly in¯uence plant population dynamics and plant community structure (e.g. Thompson et al. 1993). Since earthworms represent the dominant fraction of the biomass of soil animals in most temperate grasslands (Lee 1985; Curry 1994), they thus have profound e€ects on the structure

J. G. Zaller (&) á J. A. Arnone III Botanisches Institut, Universita È t Basel, Scho È nbeinstrasse 6, CH-4056 Basel, Switzerland Fax: +41 61 267 35 04; e-mail: zaller@ubaclu.unibas.ch

250

and function of these ecosystems. In calcareous grasslands in Switzerland between 1.5 and 10 t of earthworms live in each hectare of soil (Sto È ckli 1958). More than 70% of these earthworms are surface casting species which can produce between 20 and 40 t (dry mass) of nutrient-rich casts per hectare in a single year (e.g. Glasstetter 1991; Zu È rcher 1994). Despite the important role that earthworms play in many terrestrial ecosystems, we know nothing about how rising atmospheric CO2 may a€ect their activity and how altered earthworm activity may feed back on ecosystem processes such as carbon (C) and nitrogen (N) cycling. Plant CO2 studies which have included animals such as herbivores have revealed that an understanding of interactions between plants and animals is essential for the reliable prediction of the responses of natural ecosystems to rising atmospheric CO2 (e.g. Fajer et al. 1989; Bazzaz 1990; Johnson and Lincoln 1990; Arnone et al. 1995). The two most likely e€ects that rising atmospheric CO2 will have on earthworms are (1) increases in C supply via stimulated above- and belowground plant productivity (cf. Strain and Cure 1985; Rogers et al. 1994), including litter production (Ko È rner and Arnone 1992; Owensby et al. 1994; Arnone and Ko È rner 1995; Navas et al. 1995), and (2) increased soil water contents (Owensby et al. 1993; Jackson et al. 1994; P. Niklaus, unpublished work). The latter e€ect appears to result from decreased stomatal conductance (e.g. Morison 1987; Jackson et al. 1994; Lauber and Ko È rner 1997) and leaf-level transpiration, leading to decreased canopy evapotranspiration under elevated CO2 (e.g. Stocker et al. 1997). Because earthworms are sensitive to changes in soil moisture (Evans and Guild 1947; Zicsi 1959; Gerard 1967) and because they feed mainly on plant detritus (e.g. Waters 1955; Piearce 1978), we hypothesized that elevated CO2 would stimulate earthworm activity (i.e. surface cast production), and consequently the amounts of soil, soil carbon and soil nitrogen egested by earthworms. We tested these hypotheses over 1 year in a calcareous grassland in the second growing season under elevated CO2 treatment.

(Ogermann et al. 1994). Depth of the Ah-horizon in the experimental plots averages 15 cm. Experimental design Eight 1.2-m2 experimental plots have been maintained since March 1994 at current ambient CO2 concentrations (``A'': 350 ll CO2 l)1) and eight at elevated CO2 (``E'': 610 ll CO2 l)1) using a screenaided CO2 control facility (SACC, open top and open bottom rings; Leadley et al. 1997) as a part of an ongoing, long-term study at this site (Leadley and Ko È rner 1996). Eight plots have been left unscreened (``C''). All plots are arranged in a randomized complete block design with eight blocks. CO2 treatments have been maintained round the clock over the whole year except for the period from 15 December 1995 to 6 March 1996 when air temperatures were below freezing or when the ground was covered with snow. SACC screens enabled free movement of most above- and belowground fauna into and out of each plot. Further details about the experimental design can be found in Leadley and Ko È rner (1996). Precipitation and soil temperature (5 cm depth) were continuously measured near the experimental plots (M. Jaeggi, unpublished work). We also measured soil temperature in each plot periodically using a portable electronic soil thermometer (5 cm depth) and correlated these with the continuous measurements to reconstruct continuous soil temperatures for each experimental plot. Water content over the top 10 cm of soil in all SACC plots and four unscreened plots was also continuously monitored using time-domain re¯ectometry. Earthworm surface cast production and plant biomass Surface cast production was measured on average every 9 days from 5 April 1995 to 12 April 1996 on a permanent 30 cm ´ 30 cm sampling area in each of the experimental plots. At each sampling, we recorded in the ®eld the number and fresh mass of newly produced casts using an electronic pocket balance. After weighing, each cast was returned to its original position. We then slightly deformed each cast to facilitate the identi®cation of newly produced casts at the next sampling date. At least two cast subsamples from each plot were taken at each sampling date to calculate fresh mass/dry mass ratios (80°C, 24 h) and for the determination of organic C (Corg) and total N contents (Ntot, see below). All cast data are expressed on an oven-dried basis per unit land area (m2). Vegetation in each plot was harvested down to a height of 5 cm (typical mowing height) in June and October 1995 and dried at 80°C for 48 h. Plant dry mass is expressed per unit land area. C and N content of casts Total C and N content of dried and pulverized cast subsamples was measured with a CHN analyzer (LECO 1000, St. Joseph, Mich., USA). Inorganic C content (Cinorg) of casts was measured using a hydrochloric acid digestion (cf. Nelson and Sommers 1982) and a carbonate carbon analyser (LECO CC-1000, St. Joseph, Mich., USA). The Corg content of casts was calculated by subtracting the Cinorg content from the total C measured. According to the cast production data, we distinguished six earthworm activity periods over the year and pooled all cast samples from each plot for chemical analysis. Statistical analyses The CO2 treatment e€ect on the rate of surface cast production (mass and number), cumulative surface cast production, and on C and N concentrations in casts over time were analyzed using repeated measurement ANOVAs (Sokal and Rohlf 1981; SYSTAT 1992). ANOVAs included two CO2 levels (A and E plots) and block as independent variables. The block factor was deleted from the

Materials and methods
Study site The calcareous grassland studied here is located on a 20° southwest-facing slope at an elevation from 500 to 530 m near the village of Nenzlingen in the Jura Mountains of Switzerland (47°27¢ N, 7°34¢ E). Precipitation from April 1995 to April 1996 was 1060 mm with a mean annual air temperature of 8.7°C (M. Jaeggi, unpublished work). Up to 1993 this grassland had been used for hundreds of years as a non-intensively managed pasture, mainly for cattle. It is no longer grazed but is mown twice a year in early summer and autumn. Among the approximately 100 vascular plant species found on this site, the grass Bromus erectus is the dominant species (Zoller 1954; Leadley and Ko È rner 1996). Soils are classi®ed as a transition Rendzina (pH is about 6.5, bulk density of the top soil 1.1 g cm)3), with a well developed, stone-free, loamy topsoil and a rapid transition at 10±15 cm depth to the underlying scree material

251 model if no statistical signi®cance was indicated (i.e. P > 0.05). We used unpaired t-tests to identify signi®cant (P O 0.05) CO2 e€ects on C and N concentrations of casts at individual sampling times. To assess the relationships between surface cast production and soil temperature, soil water content (WC) and above-ground plant biomass productivity, simple linear regression models were ®tted using data from A and E plots, as well as data from unscreened plots. We regressed the mean soil temperature calculated over a 13week period in 1995, which included 6 weeks without surface casting and the following 7-week casting period, against the cumulative cast production over the 7-week casting period. We did the same for soil WC. We also regressed total annual above-ground plant biomass production measured over the 1995 growing season against the cumulative surface cast production from April 1995 to April 1996. ANOVAs were used to test for the signi®cance of regression coecients of these linear models (Sokal and Rohlf 1981).

Surface cast production and plant biomass We distinguished six periods of surface casting activity over the year, each interrupted by a period of inactivity (Fig. 1b). Rates of cast production (g m)2 day)1 and number of casts m)2 day)1) ¯uctuated over the year but with similar patterns observed in plots maintainted at ambient and elevated CO2 (Fig. 1b, c). Production rates were highest when soil temperatures at 5 cm depth were about 17°C and when soils were moist (35% dry mass). Such periods occurred usually after one to several days of rains (Fig. 1a, b), and lowest when soils were warm and dry (above 25°C, water content 24% dry mass). Rates of surface cast production expressed in g m)2 day)1 in communities maintained at elevated CO2 were up to 6 times higher than those measured in communities at ambient CO2 (P<0.05, Fig. 1b). However, differences were generally smaller and occurred less frequently when production was expressed in number of casts m2 day)1 (Fig. 1c). Cumulative surface cast production after 1 year (373 days) was 35% greater …P ˆ 0:016† in communities treated with elevated CO2 (2206 g m)2) than in those treated with ambient CO2 (1633 g m)2, Fig. 2a). Cumulative cast production measured over the same period in the unscreened plots was similar to that observed in elevated CO2 plots (2290 g m)2, data not shown). Cumulative surface cast production measured over the 7week period (period 3 in Fig. 2a) immediately following the 6-week summer dry period (no surface casting) in 1995 was positively correlated (P<0.05) with the mean soil water content calculated over the entire 13-week

Results
Climate and soil conditions Precipitation and soil temperatures did not di€er between A and E plots (Fig. 1a) but mean annual soil water content was 10% greater (P<0.01) in elevated CO2 plots (33% dry mass) than in ambient CO2 plots (30% dry mass) (P. Niklaus, personal communication). Mean soil WC measured in E plots (26.8 ‹ 0.6% dry mass) over the 6-week-long summer dry period in 1995 was signi®cantly (P<0.05) greater than that measured in the A plots over the same period (23.3 ‹ 0.8% dry mass). When averaged over the entire 13-week dry (6 weeks) and subsequent wetter period (7 weeks), soil WC in E plots was 29.7 ‹ 0.7% and that in A plots was 27.0% ‹ 0.8% (P<0.05).
Fig. 1 a Daily sums of precipitation at the study site (®lled bars) and mean soil temperatures in experimental plots (5 cm depth, averaged across treatments), b masses and c numbers of surface casts produced in a calcareous grassland exposed to either ambient (350 ll l)1) or elevated (610 ll l)1) atmospheric CO2 (means ‹ SE, n = 8). Earthworm activity periods denoted by circled numbers

252 Fig. 2 a Cumulative surface cast production (dry mass) over 1 year b cast Ntot content (mg N g)1) and c cast Corg (mg C g)1) content averaged over each of the six activity periods, denoted by circled numbers (each bar represents mean ‹ SE, n = 8 plots) produced in a calcareous grassland maintained at either ambient (350 ll l)1) or elevated (610 ll l)1) atmospheric CO2

Fig. 3 Cumulative surface cast production (dry mass) over the 6-week period following the 7-week summer inactivity period (see Fig. 2a) plotted as a function of: a the mean soil temperature at 5 cm depth and b mean soil water content over the entire 13 weeks. c Total annual surface cast production as a function of above-ground plant biomass production in a calcareous grassland maintained at either ambient (350 ll l)1) or elevated (610 ll l)1) atmospheric CO2. P values are for signi®cance of regressions; n.s. P>0.05

period, when viewed across all treatments (Fig. 3b, P ˆ 0:03). However, the same analysis using soil temperature showed no signi®cant correlation (Fig. 3a), nor was total annual cumulative surface cast production signi®cantly correlated with annual aboveground plant biomass production (Fig. 3c). Aboveground plant biomass production (June plus October 1995 harvests) was not signi®cantly di€erent in plots treated with elevated CO2 (367 ‹ 28.3 g m)2) than in plots maintained at ambient CO2 (306.4 ‹ 23.2 g m)2; P. Leadley, personal communication). Cast Ntot and Corg Neither the Ntot nor the Corg concentrations of casts were a€ected by CO2 treatment in any of the six earthworm activity periods (Fig. 2b, c). However, concentrations did vary seasonally (P<0.01) with the highest values measured in the fall and early winter …Ntot ˆ 5:5 mg N gÀ1 ; Corg ˆ 80 mg C gÀ1 † and the low-

est in the spring …Ntot ˆ 3:9 mg N gÀ1 ; Corg ˆ 51:4 mg C gÀ1 †: However, because cumulative cast production under elevated CO2 summed over the year was 35% greater than that in plots treated with ambient CO2, so was the total cumulative land-area-based amount of Ntot and Corg (28% greater at high CO2 by year's end, P<0.05). After 1 year, the cumulative amount of N (Ntot) which was egested by earthworms in surface casts was 6.9 g N m)2 at ambient CO2 and 8.9 g N m)2 at elevated CO2 (P<0.05). Over the same period cumulative Corg egested was 93.5 g C m)2 at ambient CO2 and 125.8 g C m)2 at elevated CO2 (P<0.05).

Discussion
CO2-induced stimulation of soil turnover by earthworms In 1 year, earthworms produced a mass of surface casts equivalent to 1.0% of the top 15 cm of surface soil at

253

ambient CO2 and 1.3% at elevated CO2, assuming a soil bulk density of 1.1 g cm)3 (e.g. Ogermann et al. 1994). At these rates of surface cast production, worms in the ecosystems maintained at current ambient CO2 concentrations would require about 100 years to egest the equivalent amount of soil as that now present in the Ah horizon (top 15 cm). At elevated CO2 they would need 75 years. Subsurface cast production was apparent on the outer surface of the minirhizotron tube installed in each plot, no discernable di€erences between ambient and elevated CO2 were evident on any of the sampling dates (J. Arnone, unpublished work). Annual surface cast production measured in our study (mass basis) averaged across all treatments (about 20 t cast dry mass ha)1 year)1) was near the bottom of the range reported for temperate grasslands (e.g. Sto È ckli 1928; Evans and  1982; Glasstetter 1991; Zu Guild 1947; Bouche È rcher 1994). There are several possible explanations for this. 1. Most data are from managed ecosystems where cast production is often stimulated by fertilizer application (Lee 1985; Edwards and Bohlen 1996). 2. Di€erences in climatic conditions when cast production is measured can lead to widely di€ering estimates of cast production at the same site (Sto È ckli 1928; Gerard 1967). 3. Finally, earthworm surface cast production is usually measured by removing the casts produced from the soil surface. This opens burrow exits and can actually stimulate cast production and lead to overestimation of production (e.g. Darwin 1881). We removed surface casts only for weighing in the ®eld but replaced them immediately in their original position. Seasonal oscillations in surface cast production were not a€ected by CO2 treatments and corresponded to ®ndings from previous studies conducted in the grasslands of the Jura Mountains (Glasstetter 1991; Zu È rcher 1994) and in grasslands in other temperate regions (Evans and Guild 1948; Gerard 1967; Nordstro È m 1975; Nowak 1975; James 1991). In spite of the long history of earthworm research, factors which determine earthworm cast production are not well understood (Lee 1985). Certainly soil moisture a€ects cast production, but it is unclear whether this is primarily due to its direct e€ects on earthworm populations (Bolton and Phillipson 1976; Daniel et al. 1996) or whether it is indirectly due to moisture-induced changes in plant biomass and litter production which a€ects worm food supply (Scheu 1987). In the grasslands we studied, it appears that soil moisture during dry periods a€ects worm activity directly (compare Fig. 3b with Fig. 3c), which would explain the increased earthworm activity at high CO2. CO2-induced stimulation of Ntot and Corg cycling by earthworms Despite the lack of di€erences in the concentrations of Ntot and Corg in surface casts between ambient and

elevated CO2, CO2-induced changes of the cumulative amounts of these elements deposited on the soil surface in one year by earthworms, in relation to the total amounts present in the top 15 cm of soil, were substantial. For example, at elevated CO2 1.6% of the Ntot pool of the topsoil was egested annually by earthworms on the soil surface, compared to 1.3% at ambient CO2 (soil bulk density 1.1 g cm)3, soil Ntot ˆ 0:33%, Ogermann et al. 1994). Similarly, earthworms in ecosystems maintained at high CO2 egested 2.0% of the Corg present in the top 15 cm of soil in surface casts annually, compared to 1.5% at ambient CO2 (soil Corg ˆ 3:8%, Ogermann et al. 1994). Although we found no CO2-induced changes of Ntot and Corg concentrations in surface casts their seasonal increase in the autumn indicates a functional relationship between earthworm activity and timing of periods of increased above- and below-ground plant litter production (Baker and Garwood 1959; Garwood 1967; Kretzschmar 1983). At this point we do not know whether the amount of available nutrients in worm casts was signi®cantly a€ected by elevated CO2. Thus, the results of our study clearly demonstrate that elevated atmospheric CO2 can have important indirect stimulatory e€ects on the activity of surface casting earthworms, which in the long term must in¯uence plant community responses to rising atmospheric CO2. Finally, our results demonstrate the need to include interactions among organisms when attempting to predict the e€ects of rising atmospheric CO2 on terrestrial plant communities and ecosystems.
Acknowledgements We gratefully acknowledge the use of the CO2 enrichment facility provided by the CO2 -team of the University of Basel (Swiss Priority Program on the Environment). M. Jaeggi (Geographical Institute, Basel) supplied meteorological data. P. Jordan helped with statistics. P. Bohlen, S. Ha È ttenschwiler, Ch. Ko È rner, P. Leadley and P. Niklaus provided valuable comments on previous versions of this manuscript. This work was made possible by grants from the Swiss National Science Foundation to J.A. Arnone III (NF 3100±042401.94/1) and Ch. Ko È rner (NF-SPPU 5001-035214).

References
Arnone JA III, Ko È rner Ch (1995) Soil and biomass carbon pools in model-communities of tropical plants under elevated CO2. Oecologia 104:61±71 Arnone JA III, Zaller JG, Ziegler C, Zandt H, Ko È rner Ch (1995) Leaf quality and insect herbivory in model tropical plant communities after long-term exposure to elevated atmospheric CO2. Oecologia 104:72±78 Baker HK, Garwood EA (1959) Studies on the root development of herbage plants. IV. Seasonal changes in the root and stubble weights of various leys. J Br Grassland Soc 14:99±104 Barley KP, Jennings AC (1959) Earthworms and soil fertility. III. The in¯uence of earthworms on the availability of nitrogen. Aust J Agric Res 10:364±370 Bazzaz FA (1990) The response of natural ecosystems to the rising global CO2 levels. Annu Rev Ecol Syst 21:167±196 Bolton PJ, Phillipson J (1976) Burrowing, feeding, egestion and energy budgets of Allolobophora rosea (Savigny) (Lumbricidae). Oecologia 23:225±245

254  MB (1982) Ecosyste Á me prairial. 4.3 Un exemple d'activite  Bouche animale: le ro à le des lombriciens. Acta Oecol Oecol Gen 3:127±154 Brown GG (1995) How do earthworms a€ect micro¯oral and faunal community diversity? Plant Soil 170:209±231 Curry JP (1994) Grassland invertebrates. Ecology, in¯uence on soil fertility and e€ects on plant growth. Chapman and Hall, London Daniel O, Anderson JM (1992) Microbial biomass and activity in contrasting soil materials after passage through the gut of the earthworm Lumbricus rubellus Ho€meister. Soil Biol Biochem 24:465±470 Daniel O, Kohli L, Schuler B, Zeyer J (1996) Surface cast production by the earthworm Aporrectodea nocturna in a pre-alpine meadow. Biol Fert Soils 22:171±178 Darwin C (1881) The formation of vegetable mould through the action of worms. John Murray, London Edwards CA, Bohlen PJ (1996) Biology and ecology of earthworms, 3rd edn. Chapman and Hall, London Edwards CA, Lofty JR (1972) Biology of earthworms. Chapman and Hall, London Evans AC, Guild WJMcL (1947) Studies on the relationships between earthworms and soil fertility. I. Biological studies in the ®eld. Appl Biol 34:307±329 Evans AC, Guild WJ (1948) Studies on the relationships between earthworms and soil fertility. V. Field populations. Ann Appl Biol 35:485±493 Fajer ED, Bowers MD, Bazzaz FA (1989) The e€ects of enriched carbon dioxide atmospheres on plant-insect herbivore interactions. Science 243:1198±1200 Garwood EA (1967) Seasonal variation in appearance and growth of grass roots. J Br Grassland Soc 22:121±130 Gerard BM (1967) Factors a€ecting earthworms in pastures. J Anim Ecol 36:235±252 Glasstetter M (1991) Die Bodenfauna und ihre Beziehungen zum Na È hrsto€haushalt in Geosystemen des Tafel- und Faltenjura (Nordwestschweiz). Physiogeographica 15:1±231 Hensen V (1877) Die Ta È tigkeit des Regenwurms fu È r die Fruchtbarkeit des Erdbodens. Z Wiss Zool 28:361 Hopp H, Slater CS (1948) In¯uence of earthworms on soil productivity. Soil Sci 66:421±428 Jackson RB, Sala OE, Field CB, Mooney HA (1994) CO2 alters water use, carbon gain, and yield for the dominant species in a natural grassland. Oecologia 98:257±262 James SW (1991) Soil, nitrogen, phosphorus, and organic matter processing by earthworms in tallgrass prairie. Ecology 72:2101± 2109 Johnson RH, Lincoln DE (1990) Sagebrush and grasshopper responses to atmospheric carbon dioxide concentration. Oecologia 84:103±110 Ko È rner Ch, Arnone JA III (1992) Responses to elevated carbon dioxide in arti®cial tropical ecosystems. Science 257:1672±1675 Kretzschmar A (1983) Soil transport as a homeostatic mechanism for stabilizing the earthworm environment. In: Satchell JE (ed) Earthworm ecology. From Darwin to vermiculture. Chapman and Hall, London, New York, pp 59±66 Lauber W, Ko È rner Ch (1997) In situ stomatal responses to longterm CO2 enrichment in calcareous grassland plants. Acta Oecol, in press Leadley PW, Ko È rner Ch (1996) E€ects of elevated CO2 on plant species dominance in a highly diverse calcareous grassland. In: Ko È rner Ch, Bazzaz FA (eds) Carbon dioxide, populations, and communities, Academic Press, San Diego, pp 159±175 Leadley PW, Niklaus P, Stocker R, Ko È rner Ch (1997) Screen aided CO2 control (SACC): a middle ground between FACE and open top chamber. Acta Oecol, in press Lee KE (1985) Earthworms. Their ecology and relationships with soils and land use. Academic Press, Syndey Morison JIL (1987) Intercellular CO2 concentration and stomatal response to CO2. In: Zeiger E, Farquhar GD, Cowan IR (eds) Stomatal function. Stanford University Press, Stanford, pp 229±251 Navas ML, Guillerm JL, Fabreguettes J, Roy J (1995) The in¯uence of elevated CO2 on community structure, biomass and carbon balance of Mediterranean old-®eld microcosms. Global Change Biol 1:325±335 Nelson DW, Sommers LE (1982) Total carbon, organic carbon, and organic matter. In: Page AL, Miller RH, Keeney DR (eds) Methods of soil analysis, part 2. Chemical and microbiological properties, 2nd edn Am Soc Apron, Soil Sci Soc Am Publisher, Medisen, USA. pp 539±579 Nordstro È m S (1975) Seasonal activity of lumbricids in Southern Sweden. Oikos 26:307±315 Nowak E (1975) Population density of earthworms and some elements of their production in several grassland environments. Ekol Pol 23:459±491 Ogermann P, Spycher B, Schaub D, Sollberger R (1994) Die Landschaftsstruktur im Raum Nenzlingen ± geoo È kologisch gesehen. Regio Basil 35:91±100 Owensby CE, Coyne PI , Ham JM, Auen LM, Knapp AK (1993) Biomass production in a tallgrass prairie ecosystem exposed to ambient and elevated levels of CO2. Ecol Appl 3:644±653 Owensby CE, Auen LM , Coyne PI (1994) Biomass production in a nitrogen-fertilized, tallgrass prairie ecosystem exposed to ambient and elevated levels of CO2. Plant Soil 165:105±113 Parle JN (1963) A microbiological study on earthworm casts. J Gen Microbiol 31:13±22 Piearce TG (1978) Gut contents of some lumbricid earthworms. Pedobiologia 18:153±157 Rogers HH, Runion GB , Krupa SV (1994) Plant responses to atmospheric CO2 enrichment with emphasis on roots and the rhizosphere. Environ Pollut 83:155±189 Satchell JE (ed) (1983) Earthworm ecology. From Darwin to vermiculture. Chapman and Hall, London Scheu S (1987) The role of substrate-feeding earthworms (Lumbricidae) for bioturbation in a beechwood soil. Oecologia 72:192±196 Sharpley AN, Syers JK (1977) Seasonal variation in casting activity and in the amount and release to solution of phosphorus forms in earthworm casts. Soil Biol Biochem 9:227±231 Sokal RR, Rohlf FJ (1981) Biometry. The principles and practice of statistics in biological research, 2nd edn. Freeman, New York Stockdill SMJ (1982) E€ects of introduced earthworms on the productivity of New Zealand pastures. Pedobiologia 24:29±35 Stocker R, Leadley PW, Ko È rner Ch (1997) Carbon and water ¯uxes in a calcareous grassland under elevated CO2. Funct Ecol, in press Sto È ckli A (1928) Studien u È ber den Ein¯uss des Regenwurmes auf die Bescha€enheit des Bodens. Landwirt Jb Schweiz 42:1±121 Sto È ckli A (1958) Die Regenwurmarten in landwirtschaftlich genutzten Bo È den des schweizerischen Mittellandes. Landwirt Jb Schweiz 72:699±725 Strain BR, Cure JD (1985) Direct e€ects of increasing carbon dioxide on vegetation. Carbon dioxide research, State of the art. US Department of Energy, Washington Swift MJ, Heal OW, Anderson JM (1979) Decomposition in terrestrial ecosystems. Blackwell, Oxford Syers JK, Sharpley AN, Keeney DR (1979) Cycling of nitrogen by surface-casting earthworms in a pasture ecosystem. Soil Biol Biochem 11:181±185 SYSTAT (1992) Statistics, version 5.2 Systat, Evanston Thompson L, Thomas CD, Radley JMA, Williamson S, Lawton JH (1993) The e€ect of earthworms and snails in a simple plant community. Oecologia 95:171±178 Waters RAS (1955) Numbers and weights of earthworms under a highly productive pasture. NZ J Sci Technol 36:516±525 Zicsi A (1959) Faunistisch-systematische und o È kologische Studien u È ber die Regenwu È rmer Ungarns II. Acta Zool Acad Hung 5:401±447 Zoller H (1954) Die Typen der Bromus erectus-Wiesen des Schweizer Juras Beitra È ge zur geobotanischen Landesaufnahme der Schweiz 33: 1±309 Zu È rcher S (1994) Die Losungsproduktion der Lumbriciden als Indikator fu È t. Master's Thesis, È r die bodenbiologische Aktivita Geographisches Institut, Universita È t Basel