ARTICLE IN PRESS
Soil Biology & Biochemistry 39 (2007) 2058–2069 www.elsevier.com/locate/soilbio
Earthworm activity as a determinant for N2O emission from crop residue Elena Rizhiyaa,d, Chiara Bertoraa,c, Petra C.J. van Vlietb, Peter J. Kuikmana, Jack H. Fabera, Jan Willem van Groenigena,b, a Alterra, Wageningen University and Research Centre, P.O. Box 47, 6700 AA Wageningen, The Netherlands Department of Soil Quality, Wageningen University and Research centre, P.O. Box 47, 6700 AA Wageningen, The Netherlands c University of Turin, Department of Agronomy, Forest and Land Management, Italy d Agrophysical Research Institute, Grazhdanskii pr. 14, 195220 St. Petersburg, Russia
b
Received 30 October 2006; received in revised form 1 March 2007; accepted 8 March 2007 Available online 10 April 2007
Abstract Earthworm activity may have an effect on nitrous oxide (N2O) emissions from crop residue. However, the importance of this effect and its main controlling variables are largely unknown. The main objective of this study was to determine under which conditions and to what extent earthworm activity impacts N2O emissions from grass residue. For this purpose we initiated a 90-day (experiment I) and a 50-day (experiment II) laboratory mesocosm experiment using a Typic Fluvaquent pasture soil with silt loam texture. In all treatments, residue was applied, and emissions of N2O and carbon dioxide (CO2) were measured. In experiment I the residue was applied on top of the soil surface and we tested (a) the effects of the anecic earthworm species Aporrectodea longa (Ude) vs. the epigeic species Lumbricus rubellus (Hoffmeister) and (b) interactions between earthworm activity and bulk density (1.06 vs. 1.61 g cm3). In experiment II we tested the effect of L. rubellus after residue was artificially incorporated in the soil. In experiment I, N2O emissions in the presence of earthworms significantly increased from 55.7 to 789.1 mg N2O-N kg1 soil (L. rubellus; po0.001) or to 227.2 mg N2O-N kg1 soil (A. longa; po0.05). This effect was not dependent on bulk density. However, if the residue was incorporated into the soil (experiment II) the earthworm effect disappeared and emissions were higher (1064.2 mg N2O-N kg1 soil). At the end of the experiment and after removal of earthworms, a drying/wetting and freezing/thawing cycle resulted in significantly higher emissions of N2O and CO2 from soil with prior presence of L. rubellus. Soil with prior presence of L. rubellus also had higher potential denitrification. We conclude that the main effect of earthworm activity on N2O emissions is through mixing residue into the soil, switching residue decomposition from an aerobic and low denitrification pathway to one with significant denitrification and N2O production. Furthermore, A. longa activity resulted in more stable soil organic matter than L. rubellus. r 2007 Elsevier Ltd. All rights reserved. Keywords: Greenhouse gases; Nitrous oxide; Emission factor; Earthworm ecological groups; Residue application; Bulk density
1. Introduction The return of plants or plant parts to the soil in agroecosystems may lead to considerable nitrous oxide (N2O) fluxes. This may occur when crop residues remain in Corresponding author. Alterra, Wageningen University and Research Centre, P.O. Box 47, 6700 AA Wageningen, The Netherlands. Tel.: +31 317 474784; fax: +31 317 419000. E-mail address:
[email protected] (J.W. van Groenigen).
0038-0717/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.soilbio.2007.03.008
the field after harvest (Velthof et al., 2002), during renovation of managed grassland by plowing of the sod and/or application of herbicide (Pinto et al., 2004; Vellinga et al., 2004), and during incorporation of cover crops. The global N2O emission from crop residue has been estimated at 0.4 Tg N per year, using the IPCC default emission factor of 1.25% of applied residue N emitted as N2O (Mosier et al., 1998). However, this default emission factor is based on relatively few experimental studies (Novoa and Tejeda, 2006). Recent experiments showed that the emission factor for crop residues can vary considerably
ARTICLE IN PRESS E. Rizhiya et al. / Soil Biology & Biochemistry 39 (2007) 2058–2069
with residue quality, particularly the carbon/nitrogen (C/N) ratio and the amount of mineralizable N (Velthof et al., 2002). Generally, higher emissions follow incorporation of residue with lower C/N ratios (deCatanzaro and Beauchamp, 1985; Baggs et al., 2000). Besides quality and quantity of the applied residue, N2O emissions are affected by soil parameters controlling the rate of decomposition and denitrification, such as water content, temperature and aeration (Firestone, 1982; Stott et al., 1986). Available organic C after incorporation of residue is a major substrate for denitrifier activity and therefore also for N2O emission (Beauchamp, 1997). However, high rates of soluble organic C may also decrease the ratio of N2O/N2 evolving from denitrification (Firestone, 1982; Paul and Beauchamp, 1989). Although biotic N2O emission from crop residue is ultimately the result of microbial activity, it has been established that soil macrobiota are instrumental in stimulating microbial N2O production. In recent years, especially the role of earthworm activity in mineralization of residue-N has been stressed (Cortez et al., 2000; Wardle, 2002; Swift et al., 2004; Postma-Blaauw et al., 2006). Earthworms increase soil N availability and -cycling by stimulating the transfer of N from plant material to inorganic forms that can be utilized by microorganisms and plants (Lavelle, 1983; Bohlen and Edwards, 1995; Cortez et al., 2000). Also, the earthworm gut is a nearoptimal environment for N2O production in terms of microflora, anaerobicity and concentration of mineral N and available C. As a result, N2O emissions from the ingested soil are elevated as compared to the surrounding soil matrix (Karsten and Drake, 1997; Horn et al., 2003). Earthworms also affect the overall physical properties of soil such as porosity and bulk density through bioturbation, casts deposition on the soil surface and litter incorporation into the soil (Parkin and Berry, 1994, 1999). These earthworm-mediated physical properties may in turn affect N2O emissions by changing gas diffusion rates (of N2O out of the soil and O2 into the soil) and soil moisture conditions (Granli and Bøckman, 1994; Ball et al., 1999; Van Groenigen et al., 2005b). In terms of greenhouse gas budget of the soil, earthworms are also considered to contribute to sequestration of C in the soil through stimulating stabilization of soil organic matter in stable micro-aggregates rich in organic C. These micro-aggregates are predominantly formed in macro-aggregates originating from earthworm casts (Guggenberger et al., 1996; Bossuyt et al., 2005; Pulleman et al., 2005). The effect of earthworm activity on greenhouse gas emissions from soils is therefore crucial but intricate—they appear to contribute to both decreased carbon dioxide (CO2) fluxes and increased N2O fluxes. The net contribution of earthworms to greenhouse gas emissions from soils in terms of global warming potential has not yet been assessed. Using morphological, behavioral and life-history criteria, earthworms have been classified into ‘ecological groups’
2059
or ‘strategies’ (Bouche´, 1977; Perel, 1977; Lee, 1985; Hendrix and Bohlen, 2002). Some of these criteria, particularly food choice, depth distribution and burrowing behavior, are thought to be relevant to earthworm functioning in decomposition and soil amelioration. Species with an epigeic strategy live in the upper soil horizons and litter layers with high organic residue content. They ingest large amounts of fresh residue and generally build only superficial and unstable horizontal burrows. Anecic species create permanent vertical burrows and come to the surface to feed on partially decomposed residue, mixing the residue deeper into the soil. Endogeic species forage below the surface, ingest large quantities of soil high in organic matter and build mostly horizontal burrows. Several studies have demonstrated differences between earthworm species that could be linked to these strategies (Shaw and Pawluk, 1986a, b; Brun et al., 1987). With respect to crop residue decomposition, mineralization of N has been shown to be enhanced by epigeic and anecic species, but not by an endogeic species (Postma-Blaauw et al., 2006). However, N2O emissions have not yet been quantified in relation to earthworm species characteristics. Despite a large body of literature on the impact of earthworms on C and N cycling in soils, there is hardly any direct data on the mechanism through which they affect N2O emissions from crop residue and on the soil properties that may control such a mechanism. For agricultural systems, several soil management decisions may interact with a possible earthworm effect. Particularly, mixing the residue into the soil by plowing may affect soil physical properties to a higher degree than earthworm activity. Both decreased bulk density due to plowing and increased bulk density due to trafficking have been shown to strongly affect N2O emissions (Hansen et al., 1993; Ball et al., 1999; Van Groenigen et al., 2005a). Therefore, the main objective of the present study was to quantify how and under what conditions earthworm activity changes N2O emissions from crop residue in agricultural systems. In two mesocosm studies we quantified earthworm effects on N2O emissions from crop residue (i) as mediated by two common earthworm species representing different ecological groups; (ii) as related to bulk density; and (iii) as affected by residue application method. We used plant residue with a relatively low C/N ratio, as N2O emissions from such residues are considered to be the most important (Baggs et al., 2000). 2. Materials and methods 2.1. Experimental set-up We studied the influence of earthworm activity on N2O and CO2 emissions from crop residue in two laboratory mesocosm experiments during 90 days (experiment I) and 50 days (experiment II). The treatments for both experiments are summarized in Table 1. In experiment I, crop residue was applied on top and treatments differed with
ARTICLE IN PRESS E. Rizhiya et al. / Soil Biology & Biochemistry 39 (2007) 2058–2069
2060
Table 1 Treatments included in the two mesocosm experiments Experiment
Treatment
Bulk density (g cm3)
Residue applieda
Earthworm presenceb
I
T T T T T T T T
1.0 1.1 1.2 2.0 2.1 2.2 2.3 2.4
1.06 1.06 1.06 1.61 1.61 1.61 1.61 1.61
— T T — T T T T
— — AL — — AL LR AL+LR
II
T 3.0 T 3.1 T 3.2
1.30 1.30 1.30
— M M
— — LR
a
— ¼ no residue; T ¼ residue on top of the soil; M ¼ residue mixed into the soil. — ¼ no earthworms; AL ¼ A. longa (anecic); LR ¼ L. rubellus (epigeic); AL+LR ¼ A. longa+L. rubellus.
b
respect to earthworm species and bulk density. In experiment II, crop residue was mixed into the soil and treatments differed with respect to earthworm presence. The soil was collected from a field on the former experimental farm ‘‘De Kandelaar’’, in Marknesse, located in the Noord-Oost Polder of the Netherlands (521 430 N, 51 520 E; 1 m elevation). The soil can be classified as a loamy Typic fluvaquent (Soil Survey Staff, 1998). The soil mineral fraction was composed of 29% sand, 54% silt and 17% clay. The bulk density of the plowing layer in the field was 1.61 g cm3. At the time of sampling (22nd February 2006), the soil had been under pasture for 3 years (Bertora et al., 2007). After collection the soil from the plowing layer (0–30 cm) was air-dried at 201 C for 1 week and subsequently sieved through a 10 mm sieve. Total C and N contents were 32.8 and 2.0 g kg1, respectively; pH (in CaCl2 extract) was 7.5. The basic setup of mesocosms was similar to a previous study (Bertora et al., 2007). The mesocosms consisted of 6.1 l polypropylene buckets that were filled with 4 kg of dry soil mixed with 0.250 l water kg1 soil. Based upon previous experiments, this corresponded to the optimal moisture level for earthworm activity (Bertora et al., 2007). Subsequently, the mesocosms were pre-incubated during 7 days until the initial CO2 fluxes had decreased to background levels. Grass was collected from a field near Alterra on 24th April 2006, chopped to pieces of 1–3 cm and subsequently oven-dried at 40 1C for 7 days. The residue contained 35 g kg1 N and 422 g kg1 C, with a corresponding C/N ratio of 12.1. Residue and earthworms were applied on 11th May and 19th June for experiments I and II, respectively. In both experiments, residue was applied to the soil at a rate of 2.50 g dry matter kg1 soil, or 87.5 mg N kg1 soil. Residue was either left on top (experiment I) or mixed into the soil (experiment II). In experiment II, the blank treatment was also mixed but no residue was added. Three different bulk densities (1.06,
1.30 and 1.61 g cm3) were manually established in the two experiments (Table 1) by pushing with the downside of another bucket on the soil surface until the required volume was reached. In experiment I, the residue was applied on top after the different bulk densities were established. Two species of earthworm from different ecological groups were included in our mesocosm studies: the anecic species Aporrectodea longa (Ude); and the epigeic species Lumbricus rubellus (Hoffmeister). Both species were collected from park areas in Wageningen (NL) on 20th February 2006 (A. longa) and 28th April 2006 (L. rubellus). Earthworms were kept in the pasture soil with grass residue at 16 1C until the start of the respective experiments. Two days prior to the experiment and before weighing, earthworms were kept on wet filter paper in order to empty their guts. After residue application, earthworms were added to the respective mesocosms in densities based on previously reported values from Dutch pastures (Didden, 2001). Four individuals of A. longa (both adults and large juveniles, on average totaling 3.2 g, or 100 individuals m2) and/or six (adult) individuals of L. rubellus (on average totaling 7.8 and 9.0 g for experiments I and II, respectively, or 150 individuals m2) were applied per mesocosm. After applying residue and earthworms, mesocosms were kept closed by a black polyethylene cover that allowed gaseous exchange with the outside air, retarded water evaporation and prevented earthworms from escaping. The mesocosms were set up under controlled conditions in a climate room with constant temperature (16 1C) and air humidity (65%). The soil moisture content in the mesocosms was gravimetrically adjusted every 2–4 days for each individual mesocosm, always directly after a flux measurement session. The average evaporation from the mesocosms was 0.9 g kg1 soil d1. Both experiments were set up as a randomized block design with 5 replicates. The total number of mesocosms was therefore 40 for experiment I and 15 for experiment II.
ARTICLE IN PRESS E. Rizhiya et al. / Soil Biology & Biochemistry 39 (2007) 2058–2069
2.2. Gaseous emissions Fluxes of N2O and CO2 from the mesocosms were measured 2–3 times a week for both experiments, resulting in 33 measurements in 90 days for experiment I and 24 measurements in 50 days for experiment II. The flux measurement protocol largely followed that of a previous study (Bertora et al., 2007). At least half an hour prior to flux measurements, the cover was removed from the mesocosms in order to eliminate any possible N2O or CO2 gradient inside the mesocosm headspace. During this period, the lights remained switched on in order to prevent earthworms from escaping. Each mesocosm was subsequently closed with a polypropylene lid equipped with two rubber septa. At the start of the experiment, when fluxes were relatively high, it was closed for a period of approximately 30 min. After 30 days, the closing time was increased to approximately 50 min. The N2O increase in the mesocosm headspace was measured using a photo-acoustic infrared gas analyzer (Innova 1312), using two teflon tubes equipped with a soda-lime filter to minimize interference by CO2 (Velthof et al., 2002). Fluxes were calculated assuming a linear increase of the concentration of N2O over time during the closing of the lid. This was checked occasionally for each treatment during the experiment. Values were corrected for ambient N2O concentration and for mixing of the gas sample with the previous measurement in the internal volume of the gas monitor. CO2 emissions were measured using the same gas analyzer and a similar setup, but without the soda-lime filter and after a separate closing period (about 30 min) of the mesocosms. Cumulative emissions for both N2O and CO2 were calculated assuming linear changes between subsequent measurements (Kool et al., 2006).
2.3. Post-incubation analyses On 9th August, 90 days after the start of experiment I and 50 days after the start of experiment II, the studies were terminated. Subsequently, soil cores were collected from both experiments for measuring potential denitrification. Two or three (depending on the height of soil in the mesocosm) stainless steel ring samplers (5 cm diameter, 5 cm height) were attached to each other in order to sample the whole profile of the mesocosm. The rings were subsequently separated from each other, and placed in a PVC container with a volume of 0.8 l. We applied approximately 200 mg NO3-N kg1 dry soil as a KNO3 solution to the container, bringing the soil to near saturation. The boxes were subsequently closed with airtight PVC lids with two septa and were flushed with N2 for 5 min. Finally, we replaced 8% of the gas phase of the containers with acetylene (Van Beek et al., 2004; Bertora et al., 2007). N2O build-up inside the closed containers was measured after 24 and 48 h. Denitrification
2061
rates were calculated from the increase in headspace N2O concentration between 24 and 48 h (Van Beek et al., 2004). + Mineral N content (NO 3 and NH4 ), dissolved organic nitrogen (DON) and pH were measured on 9th August for both experiments. A subsample of approximately 200 g was dried and sieved through a 5-mm sieve. The soil was extracted with a 0.01 M CaCl2 solution and concentrations of the N species and pH were subsequently determined using standard methodology (Kool et al., 2006). Also on 9th August, earthworms were collected from the mesocosms and weighed. In order to obtain information about the stability of soil organic matter in the respective treatments, a cycle of wetting/drying and freezing/thawing treatments was started on subsamples, without earthworms present. The soil from all treatments was gently broken along its natural planes of weakness (Kong et al., 2005) and passed through an 8 mm sieve. For all mesocosms from both experiments, 500 g soil was put in 1 l glass jars (55 jars in total). The jars were set up in a randomized block design for each of the two experiments. On 13th August, all samples were wetted with 100 ml water and this moisture level was maintained until 18th August, after which lids were removed and the samples were allowed to air dry until 26th August. Subsequently, the samples were frozen overnight at 12 1C and thawed on 27th August at 16 1C. During the whole period from 10th August through 2nd September, N2O and CO2 emissions from the jars were measured 14 times using the same procedure as described above, using Teflon-fitted lids equipped with septa. 2.4. Data analysis The significance of the effects of earthworm species and bulk density was quantified using analysis of variance (ANOVA). Blank treatments (without residue) were not included in the statistical analysis. The analyzed variables for the two mesocosm experiments were cumulative N2O and CO2 emissions, as well as mineral N, DON, and total dissolved N at the end of the experiments. For the post-incubation analyses, cumulative N2O and CO2 emissions as well as potential denitrification were analyzed. The cumulative N2O flux data needed to be log-transformed before statistical analysis. For experiment I, the effect of earthworm species was tested as a two-way block design with presence of A. longa and L. rubellus as independent factors (treatments T2.1, T2.2, T2.3 and T2.4—Table 1); the effect of earthworm presence vs. bulk density was tested as a two-way block design with presence of A. longa and bulk density as independent factors (treatments T1.1, T1.2, T2.1 and T2.2). For experiment II, the effect of earthworm presence was tested as a oneway block design with presence of L. rubellus as factor (T3.1 and T3.2). Significance of differences in earthworm biomass between start and end of the experiment was tested using a paired one-sided t-test for the individual treatments.
ARTICLE IN PRESS 2062
E. Rizhiya et al. / Soil Biology & Biochemistry 39 (2007) 2058–2069
3. Results 3.1. Earthworm survival rates In experiment I, all residue had disappeared from the soil surface in the mesocosms containing L. rubellus (T2.3 and T2.4). Most residue had disappeared in the presence of A. longa (T1.2 and T2.2). In experiment II, no fresh residue mixed into the soil could be determined in the presence of L. rubellus (T3.2). In the residue treatments without earthworms (T1.1, T2.1 and T3.1) partly decomposed residue could still be clearly distinguished. At the end of the mesocosm experiments, all earthworms (L. rubellus) in experiment II had survived, and the average weight had significantly decreased from 1.5 to 1.1 g individual1 (po0.05; paired one-sided t-test). In experiment I, two individuals of L. rubellus had died in two different mesocosms, both in the presence of A. longa (T2.4; Table 1). In the mesocosms with one earthworm species, the average weight of L. rubellus had significantly decreased from 1.3 to 1.1 g individual1 (po0.01). The average weight of A. longa significantly increased from 0.8 to 1.7 g individual1 (po0.001). 3.2. Gas fluxes With respect to earthworm species and presence, cumulative N2O emissions in residue treatments within experiment I with a bulk density of 1.61 g cm3 varied between 56 and 789 mg N kg1 soil. Cumulative emissions were lowest in the treatment without earthworms (T2.1) and highest for the treatment with L. rubellus (T2.3) (Fig. 1a), corresponding to 0.23 and 3.18 kg N2O-N ha1, respectively (based upon the field bulk density of 1.61 g cm3 and a 25 cm plowing layer). The simultaneous presence of both earthworm species resulted in lower emissions than the presence of L. rubellus alone. Presence of L. rubellus and A. longa, as well as their interaction, significantly affected N2O fluxes (po0.001, po0.05, and po0.001, respectively). In the absence of earthworms, soil with higher bulk density (T2.1) had significantly (po0.01) higher N2O emissions than soil with low bulk density (T1.1), raising emissions from 20 to 56 mg N2O-N kg1 soil (Fig. 1b). Presence of A. longa significantly (po0.05) increased these emissions to 287 mg N kg1 soil for low bulk density (T1.2) and 227 mg N kg1 soil for high bulk density (T2.2; Fig. 1b). In experiment II, no significant effect of earthworm presence could be detected, with cumulative N2O emissions of T3.1 (no earthworms) and T3.2 (earthworms) averaging 1013 mg N kg1 soil (Fig. 1c). CO2 emissions in experiment I were significantly (po0.001) affected by presence of L. rubellus, resulting in the highest emissions for T 2.3 at high bulk density (881 mg C kg1 soil). Presence of A. longa did not have a significant impact (Fig. 2a). As with N2O fluxes, bulk density and presence of A. longa had a significant (po0.01 in both cases) impact on CO2 emissions, with higher bulk
density and earthworm presence both resulting in higher emissions (Fig. 2b). In experiment II, earthworm activity did not have a significant impact on cumulative CO2. Cumulative fluxes for T3.1 and T3.2 averaged 518 mg C kg1 soil (Fig. 2c). Fig. 3 summarizes the results of both experiments in terms of N2O-N emitted as percentage of the applied residue-N, after correction for background fluxes. As it was not possible to establish treatments with earthworms but without residue (due to starvation of the earthworms), all treatments were corrected for the background treatments T1.0, T2.0 resp. T3.0). When the residue was left on top (experiment I), emissions totaled 0.02% and 0.05% of the applied residue-N for bulk density 1.06 and 1.61 g cm3. In the presence of L. rubellus emission increased to a maximum of 0.89%. With incorporated residue and a bulk density of 1.30 g cm3 emissions were not significantly different due to earthworm presence (1.22%). 3.3. Mineral N and total dissolved N At the start of the experiment, nitrate concentration in 1 the soil for all treatments was 20.8 mg NO 3 -N kg . At the end of the experiment, nitrate content in experiment I was significantly lower at higher bulk density than at lower bulk density (po0.001), but earthworm species did not make a significant difference (Table 2). Highest NO 3 concentrations were found in T3.2 (experiment II), but no significant effect of earthworm presence was detected. Ammonium content was low and not significantly different between treatments, with the exception of a small bulk density effect (po0.05; Table 2). DON content was not significantly different for any of the treatments in both experiments (not depicted). The significant effect of L. rubellus presence in experiment I on total dissolved N largely reflected differences in NO 3 content. 3.4. Post-incubation analyses The wetting/drying and freezing/thawing cycles after the mesocosm studies for experiment I resulted in cumulative N2O and CO2 emissions that were highest in soil with former presence of L. rubellus (Table 3). This effect was significant for both fluxes (po0.001 and po0.05, respectively). In experiment II, no significant effect of epigeic earthworm presence could be detected for N2O or CO2 fluxes (Table 3). However, potential denitrification in soils with prior presence of L. rubellus (T3.2) significantly (po0.05) increased from 33.9 to 88.7 mg N h1 kg1 (Table 3). 4. Discussion 4.1. N2O emissions from crop residue In this study, earthworm presence significantly increased N2O emissions from 55.7 up to 789.1 mg N2O-N kg1 soil
ARTICLE IN PRESS E. Rizhiya et al. / Soil Biology & Biochemistry 39 (2007) 2058–2069
2063
N2O cumulative emissions (µg N2O-N kg-1 dry soil)
1000 Level of significance: L. rubellus < 0.001 A. longa < 0.05 L. rubellus x A. longa < 0.001
800
Residue + L. rubellus Residue + L. rubellus + A. longa Residue + A. longa
600
Residue Blank 400
200
0
350 Level of significance: A. longa < 0.05 Bulk density < 0.01 A. longa x Bulk density < 0.01
300 250
Residue, 1.06 g cm-3 + A. longa Residue, 1.61 g cm-3 + A. longa Residue, 1.61 g cm-3
200
Residue,1.06 g cm-3 Blank,1.61 g cm-3
150
Blank, 1.06 g cm-3
100 50 0
1400 1200
Residue + L. rubellus Residue Blank
1000 800 600 400 Level of significance: L. rubellus n.s.
200 0 0 -200
10
20
30
40
50
60
70
80
90
days
Fig. 1. Cumulative N2O emissions for the two mesocosm experiments: (a) the effect of earthworm species when residue is applied on top in experiment I; (b) interaction between bulk density and earthworm presence when residue is applied on top in experiment I; (c) effect of earthworm presence when residue is mixed into the soil in experiment II. Significance of the different factors are listed (blank treatments not included in the statistical tests). Error bars denote standard errors (n ¼ 5).
(Fig. 1a). However, this effect was not observed when the residue was artificially mixed into the soil (Figs. 1a and 3). This strongly suggests that the main effect of earthworms
on N2O emissions from crop residue is through incorporation of residue into the soil. Direct emissions from the earthworms themselves were measured following Bertora
ARTICLE IN PRESS E. Rizhiya et al. / Soil Biology & Biochemistry 39 (2007) 2058–2069
2064
900
CO2 cumulative emissions (mgCO2-C kg-1 dry soil)
800
Level of significance: L. rubellus < 0.001 A. longa n.s. L. rubellus x A.longa < 0.05
700 600
Residue + L. rubellus Residue + L. rubellus + A. longa Residue + A. longa Residue Blank
500 400 300 200 100 0
800 700
Level of significance: A. longa < 0.01 Bulk density < 0.01 A. longa x Bulk density n.s.
600
Residue,1.61 g cm-3 + A. longa Residue, 1.61 g cm-3 Blank, 1.61 g cm-3
500
Residue, 1.06 g cm-3 + A. longa Residue, 1.06 g cm-3
400
Blank, 1.06 g cm-3 300 200 100 0
600 Level of significance: L. rubellus n.s.
500
Residue + L. rubellus Residue Blank
400 300 200 100 0 0
10
20
30
40
50
60
70
80
90
days Fig. 2. Cumulative CO2 emissions for the two mesocosm experiments: (a) the effect of earthworm species when residue is applied on top in experiment I; (b) interactions between bulk density and earthworm presence when residue is applied on top in experiment I; (c) effect of earthworm presence when residue is mixed into the soil in experiment II. Significance of the different factors are listed (blank treatments not included in the statistical tests). Error bars denote standard errors (n ¼ 5).
et al. (2007), but never exceeded 2.6 ng N2O-N h1 g1 fresh weight, which was often 2 orders of magnitude lower than the observed difference in N2O emissions during the first
half of experiment I, when the main effect of earthworm presence was established. Even assuming the highest published emission rate from L. rubellus in the literature
ARTICLE IN PRESS E. Rizhiya et al. / Soil Biology & Biochemistry 39 (2007) 2058–2069
1.61 g cm-3
1.06g cm-3
Residue + L. rubellus
Residue + A. longa
Residue
Residue
Experiment II (residue in corporated)
Residue + L. rubellus + A. longa
Residue + L. rubellus
Residue + A. longa
Experiment I (residue on top)
1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 Residue
residue-derived N2O emissions (% of applied residue-N)
of 25 ng N2O-N h1 g fresh weight1 (Matthies et al., 1999), and assuming that all N2O produced by earthworms was emitted from the soil (which is certainly an overestimation since a fraction will be reduced to N2), only approximately 8% of the observed differences during that period could be explained by direct emissions from the earthworms.
1.30 g cm-3
Fig. 3. N2O emissions for the different treatments in experiment I and II expressed as % of residue-N. Emissions in the presence of earthworms are corrected for the background without earthworms, as a treatment without residue but with earthworms was not possible due to starvation of the worms. Error bars denote standard errors (n ¼ 5).
2065
Moreover, in experiment II no increase in N2O emissions in the presence of L. rubellus was observed, despite the fact that a higher earthworm biomass was applied than in experiment I. Incorporating residue into the soil may affect N2O emissions through denitrification or nitrification in several ways; the residue can (i) change the amount of mineral N through mineralization or temporary immobilization; (ii) supply easily available C as an energy source; and (iii) decrease aerobicity due to decomposition activity. In our study, NO 3 does not appear to have been limited; 1 concentrations were 20.8 mg NO at the start of 3 -N kg the experiment, and remained equal or increased during the experiment in the residue treatments (Table 2). Furthermore, in the 1.61 g cm3 treatments, anaerobic (reduced) soil features were visually present in all treatments, irrespective of residue or earthworm presence. Therefore, the most likely explanation for the earthworm/mixing effect on N2O is an increased supply of easily available C for denitrification. Although differences in CO2 emissions were relatively small for all residue treatments (Fig. 2a), in experiment I decomposition in the absence of earthworms was located on top of the soil, whereas earthworms relocated the decomposing residue into the soil. This implied a shift from decomposition under aerobic conditions towards decomposition under increasingly anaerobic
Table 2 Average values and standard errors (n ¼ 5) for mineral N and total dissolved N (Nts) at the end of the mesocosm experiments Experiment
I
Source of variation Earthworm speciesb
Treatmenta
T T T T T T T T
1.0 1.1 1.2 2.0 2.1 2.2 2.3 2.4
NO3 (mg N kg1)
NH4 (mg N kg1)
Nts (mg N kg1)
Average
St. err.
Average
St. err.
Average
St. err.
51.6 68.2 71.9 6.4 22.6 22.3 34.4 37.4
0.7 1.9 3.1 0.6 4.1 6.5 5.1 8.1
1.1 1.1 1.2 1.9 1.2 1.3 1.3 1.3
0.0 0.0 0.1 0.2 0.0 0.1 0.0 0.1
55.6 72.3 76.0 13.5 28.0 27.6 40.0 42.7
0.7 2.1 3.1 0.4 3.9 5.8 4.8 7.7
AL LR AL LR
ns ns ns
ns ns ns
ns * ns
Bulk densityb
AL B AL B
ns *** ns
ns * ns
ns ns ns
II
T 3.0 T 3.1 T 3.2
51.1 62.6 79.4
LR
ns
Source of variation Earthworm presencec
4.1 3.7 4.1
* ¼ o 0.05; ** ¼ o 0.01; ***o 0.001; ns ¼ not significant. a Treatment codes are described in Table 1. b AL ¼ A. longa (anecic); LR ¼ L. rubellus (epigeic); B ¼ bulk density. c Presence of L. rubellus (epigeic).
1.0 1.2 1.3 ns
0.0 0.0 0.1
55.7 68.3 84.4 ns
4.1 4.0 4.0
ARTICLE IN PRESS E. Rizhiya et al. / Soil Biology & Biochemistry 39 (2007) 2058–2069
2066
Table 3 Properties of the soil after the two mesocosm studies (averages with standard errors; n ¼ 5); cumulative N2O and CO2 emissions after wetting/drying and freezing/thawing cycles; potential denitrification Experiment
I
Source of variation Earthworm speciesb
Treatmenta
T T T T T T T T
1.0 1.1 1.2 2.0 2.1 2.2 2.3 2.4
Cum. N2O emission (mg N2O-N kg1 soil)
Cum. CO2 emission (mg CO2-C kg1 soil)
Pot. denitrification (mg N h1 kg1)
Average
St. err.
Average
St. err.
Average
St. err.
2.9 0.2 2.8 33.4 -2.3 -5.7 83.8 65.1
11.7 1.9 3.1 16.0 6.0 4.3 21.5 47.8
26.6 28.9 27.3 29.9 33.2 30.9 43.3 35.7
6.9 6.7 3.4 4.2 3.0 4.8 4.6 3.0
9.3 10.3 46.7 31.3 22.7 19.4 14.0 30.4
2.4 0.9 23.0 2.8 5.6 4.1 1.9 9.5
AL LR AL LR
ns *** ns
ns * ns
ns ns ns
Bulk densityb
AL B AL B
ns ns ns
ns ns ns
ns ns ns
II
T 3.0 T 3.1 T 3.2
6.1 18.0 73.3
LR
ns
Source of variation Earthworm presencec
2.3 8.8 15.3
28.2 35.9 37.0 ns
4.4 5.2 4.2
20.2 33.9 88.7
1.5 7.4 39.6
*
* ¼ o 0.05; ** ¼ o 0.01; ***o 0.001; ns ¼ not significant. a treatments codes are listed in Table 1. b AL ¼ A. longa (anecic); LR ¼ L. rubellus (epigeic); B ¼ bulk density. c Presence of L. rubellus (epigeic).
conditions due to O2 diffusion limitations in the soil. In the presence of NO 3 this would probably result in an increase in denitrification. In experiment II this was achieved in both treatments by the artificially mixing of residue into the soil. After 90 days, cumulative N2O emissions were highest in the presence of A. longa at a bulk density of 1.06 g cm3 (Fig. 1b). In the absence of earthworms, N2O emissions increased at higher bulk density (Figs. 1b and 3). This corresponds to earlier work on the effect of bulk density on N2O emissions, and probably reflects the effect of higher anaerobicity at higher bulk density (e.g. Hansen et al., 1993; Van Groenigen et al., 2005a). The low NO 3 concentration in the control treatment with a high bulk density (6.4 mg N kg1; T2.0), as compared to the lowdensity control (51.6 mg N kg1; T1.0) is an indication of higher denitrification rates at high bulk density (Table 2). The increasing N2O emission in the presence of A. longa at 1.06 g cm3 (Fig. 2b; T1.2) after day 50, as compared to 1.61 g cm3 (T2.2), is possibly related to shifts in the N2O/ N2 ratio of denitrification. At the end of the experiment, T1.2 had both higher NO 3 concentrations (Table 2) and lower rates of C respiration (Fig. 2b). Both trends tend to increase the N2O/N2 ratio (Granli and Bøckman, 1994).
4.2. Earthworm activity, species, and interactions The observed 18-fold increase in N2O emissions in the presence of earthworms may likely represent the upper limit of this effect, as earthworm density was relatively high and soil conditions were near-optimal for earthworm activity. Based upon the surface area of the mesocosms, the earthworm densities in our experiments are equivalent to 100 individuals m2 for A. longa, and 150 individuals m2 for L. rubellus. These densities are within normal ranges for Dutch pasture systems, but rather high for arable soils, especially for epigeic species such as L. rubellus (Didden, 2001). However, earthworm densities of up to 350 individuals m2 have been reported in arable soils (Bohlen et al., 1995). Although no-tillage systems resulted in highest earthworm densities (House and Parmelee, 1985b), arable soils that were chisel-plowed only had marginally lower densities of earthworms (Bohlen et al., 1995). Therefore, we consider the earthworm densities in our experiment to be relatively high but ecologically relevant. The weight range during the experiments (increasing from 0.8 to 1.7 g individual1) for A. longa was consistent with previously published values (Schmidt, 1999;
ARTICLE IN PRESS E. Rizhiya et al. / Soil Biology & Biochemistry 39 (2007) 2058–2069
Baker and Whitby, 2003), as was the weight range of L. terrestris (1.1–1.5 g individual1) (Postma et al., 2006). The soil moisture content that we established in our experiments was near optimal for earthworm activity, based upon an earlier mesocosm study with the same soil (Bertora et al., 2007). In that study, earthworm activity decreased with moisture contents, with A. longa reverting to diapause as a survival strategy at 125 g water kg1 (Lee, 1985; Bertora et al., 2007). As most decomposition of surface-applied residue in a system with such a low soil moisture content would proceed through an aerobic pathway, this would lead to N2O emissions closer to those of our residue treatments without earthworms. Our 18-fold increase in N2O emissions therefore probably reflected a maximum that may not always be achieved under field conditions. With respect to different ecological groups of earthworms, it is intrinsically difficult to separate effects of earthworm density and species composition, as we did not change these variables independently in our experimental setup. For example, the presence of L. rubellus increased emissions of both N2O and CO2 as compared to A. longa (Figs. 1a and 2a). However, it is not immediately clear whether this was an indication of a ‘species effect’ based on feeding- and burrowing strategy, or whether it is a ‘density effect,’ reflecting the higher density of L. rubellus in our mesocosms. In our study, cumulative N2O emissions for the treatment with both earthworm species (T2.4; Fig. 1a) were intermediate between the two single species treatments (T2.2 and T2.3). This suggests competition between the two species for the residue, with L. rubellus having a higher activity-related N2O emission than A. longa. A possible explanation for this difference may be that A. longa, as an anecic species, incorporates residue deeper into the soil (Bouche´, 1977). N2O produced at larger depths takes longer to diffuse upwards to the soil surface, and therefore is more likely to be reduced to N2 before it is emitted (Baggs et al., 2000). Since the depth of our mesocosms was limited, emissions in the field due to A. longa activity may be lower than those measured in our study if significant N2O reduction takes place. For CO2, this pattern is different, with presence of L. rubellus leading to increased emissions irrespective of A. longa presence. Combined with the different growth trends for the two species (weight loss for L. rubellus; weight gain for A. longa) this suggests competition for a limited foodsource or differences in food quality demands, with A. longa having a competitive advantage over L. rubellus. N2O and CO2 fluxes during the freezing/thawing and drying/wetting cycles after removal of the earthworms also suggest a ‘species effect’ (Table 3). Whereas initial presence of A. longa (T2.2) did not induce emissions different from the control residue treatment, L. rubellus (T2.3) significantly increased fluxes of both N2O and CO2. Moreover, in experiment II prior presence of L. rubellus (T3.2) increased potential denitrification from 33.9 to 88.7 mg N h1 kg1. Since potential denitrification as we defined it primarily reflects quality and quantity of substrate-C for denitrification,
2067
this suggests that L. rubellus activity does not lead to stabilization of residue-C in the soil. This would be in line with previous results indicating that epigeic species, as opposed to anecic and endogeic species, do not contribute to soil aggregate formation and organic matter stabilization (Six et al., 2004a, and references therein). It remains to be determined what the effect of endogeic earthworm activity would be on N2O and CO2 emissions from crop residue applied to the soil. The literature suggests that endogeic species are most instrumental in stabilizing soil organic matter, especially in combination with anecic species (Six et al., 2004a). This may ultimately result in more stable soil organic matter, i.e. lower emissions of N2O and CO2 after physical disturbance. However, as endogeic species do not compete with epigeic and anecic species for fresh residue as a food source, initial N2O emissions from crop residue through epigeic and anecic species may not decrease much. 4.3. Implications for soil management Emissions of N2O from residue-N in our experiments ranged from 0.02% to 1.22% of the applied residue, with highest emissions occurring when residue was mixed into the soil by earthworms or artificially. This effect was hypothesized to be related to C availability in the soil vs. at the soil surface. Therefore, this raises the possibility of reducing N2O emissions by delaying incorporation of crop residue until most decomposition of easily available C has been completed. However, this hypothesis should be tested under field conditions. Especially, potential trade-offs between direct emissions (N2O from the soil) and indirect emissions (N2O from volatilized NH3 from plant residue or from surface water NO 3 due to increased runoff) should be taken into account (Mosier et al., 1998). Our establishment of different bulk densities by pressing may have led to slightly higher bulk densities near the surface, and thereby to relatively more stress for L. rubellus than for A. longa. In the field, this effect may be absent or even reversed, although the presence of the earthworm effect at different bulk densities in our experiment suggests that this effect might not be very strong. It should also be stressed that the reported effects in this paper apply to crop residues with a low C/N ratio, such as incorporation of cover crops and green manure (Lahti and Kuikman, 2003), vegetable residues (Velthof et al., 2002) and renewal of grassland in spring (Vellinga et al., 2004). Residues with much higher C/ N ratios will probably lead to different (and much lower) emission patterns due to temporary immobilization of mineral N (Baggs et al., 2000). For residue with low C/N ratios, delayed plowing may have an added benefit in that delayed mineralization of residue-N may reduce leaching during the winter, and optimize uptake of N in the succeeding cropping season (Lahti and Kuikman, 2003). The decreasing temperatures that are normally associated with delayed incorporation of crop residues in autumn were not simulated in this study. However, as N2O emissions
ARTICLE IN PRESS 2068
E. Rizhiya et al. / Soil Biology & Biochemistry 39 (2007) 2058–2069
generally decrease at lower temperature (Granli and Bøckman, 1994) we do not expect this effect to change the possible benefits of later incorporation on N2O emissions. It has been widely reported that zero tillage leads to increases in earthworm populations (e.g. House and Parmelee, 1985b; Mackay and Kladivko, 1985) as well as to prolonged periods of enhanced N2O emissions (Six et al., 2004b, and references therein). Although not all major crops in these systems have the low C/N ratios used in this paper, these systems often include either cover crops or other leguminous crops with relatively low C/N ratios (e.g. House and Parmelee, 1985a). Our results suggest that there may be a functional link between the increase in earthworm population and elevated N2O emissions. Increases in earthworm density due to zero-tillage management are often dominated by species with epigeic characteristics, which we found to increase N2O emissions considerably. Different ecological traits in earthworms may be therefore be controlling both increased C stabilization and increased N2O emissions in agricultural ecosystems. 5. Conclusions We conclude that earthworm activity has the potential to increase N2O emissions from crop residues up to 18-fold; that the earthworm effect is largely independent of bulk density; and that earthworm species specifically impact N2O emissions and residue stabilization in soil organic matter. However, earthworm-mediated emissions of N2O mostly resulted from residue incorporation into the soil, and disappeared when plowing of residue into the soil was simulated. Our results suggest that, irrespective of earthworm activity, farmers may decrease direct N2O emissions from crop residues with a relatively low C/N ratio by leaving it on top for a few weeks before plowing it into the soil. However, field studies should confirm this effect, and possible trade-offs to other (indirect) emissions of N2O should be taken into consideration before this can be recommended. Acknowledgements Elena Rizhiya’s stay at Alterra was funded through an IAC scholarship from the Dutch Ministry of Agriculture, Nature and Food Quality. The authors would like to thank Eduard Hummelink, Tamas Salanki, Willem Menkveld and Dorien Kool for assistance with field and laboratory work. We are grateful to Gerard Velthof for discussions on the setup and results of the experiment. References Baggs, E.M., Rees, R.M., Smith, K.A., Vinten, A.J.A., 2000. Nitrous oxide emission from soils after incorporating crop residues. Soil Use and Management 16, 82–87. Baker, G.H., Whitby, W.A., 2003. Soil pH preferences and the influences of soil type and temperature on the survival and growth of Aporrectodea longa (Lumbricidae). Pedobiologia 47, 745–753.
Ball, B.C., Parker, J.P., Scott, A., 1999. Soil and residue management effects on cropping conditions and nitrous oxide fluxes under controlled traffic in Scotland 2. Nitrous oxide, soil N status and weather. Soil and Tillage Research 52, 191–201. Beauchamp, E.G., 1997. Nitrous oxide emission from agricultural soils. Canadian Journal of Soil Science 77, 113–123. Bertora, C., Van Vliet, P.C.J., Hummelink, E.W.J., Van Groenigen, J.W., 2007. Do earthworms increase N2O emissions in ploughed grassland? Soil Biology & Biochemistry 39, 632–640. Bohlen, P.J., Edwards, C.A., 1995. Earthworm effects on N dynamics and soil respiration in microcosms receiving organic and inorganic nutrients. Soil Biology & Biochemistry 27, 341–348. Bohlen, P.J., Edwards, W.M., Edwards, C.A., 1995. Earthworm community structure and diversity in experimental agricultural watersheds in northeastern Ohio. Plant and Soil 170, 233–239. Bossuyt, H., Six, J., Hendrix, P.F., 2005. Protection of soil carbon by microaggregates within earthworm casts. Soil Biology & Biochemistry 37, 251–258. Bouche´, M.B., 1977. Strate´gies lombriciennes. In: Lohm, U., Persson, E. (Eds.), Soil Organisms as Components of Ecosystems. Swedish Natural Science Research Council, Stockholm, pp. 122–132. Brun, J.J., Cluzeau, D., Trehen, P., Bouche, M.B., 1987. Biostimulation— prospects and limitations of biological improvement of soils by stimulation or introduction of earthworms. Revue d’Ecologie et de Biologie du Sol 24, 685–701. Cortez, J., Billes, G., Bouche´, M.B., 2000. Effect of climate, soil type and earthworm activity on nitrogen transfer from a nitrogen-15-labelled decomposing material under field conditions. Biology and Fertility of Soils 30, 318–327. deCatanzaro, J.B., Beauchamp, E.G., 1985. The effect of some carbon substrates on denitrification rates and carbon utilization in soil. Biology and Fertility of Soils 1, 183–187. Didden, W.A.M., 2001. Earthworm communities in grasslands and horticultural soils. Biology and Fertility of Soils 33, 111–117. Firestone, M.K., 1982. Biological denitrification. In: Stevenson, F.J. (Ed.), Nitrogen in Agricultural Soils. American Society of Agronomy, Madison, pp. 289–326. Granli, T., Bøckman, O.C., 1994. Nitrous oxide from agriculture. Norwegian Journal of Agricultural Sciences (suppl.) 12, 1–128. Guggenberger, G., Thomas, R.J., Zech, W., 1996. Soil organic matter within earthworm casts of an anecic-endogeic tropical pasture community, Colombia. Applied Soil Ecology 3, 263–274. Hansen, S., Mæhlum, J.E., Bakken, L.R., 1993. N2O and CH4 fluxes in soil influenced by fertilization and tractor traffic. Soil Biology & Biochemistry 25, 621–630. Hendrix, P.F., Bohlen, P.J., 2002. Exotic earthworm invasions in North America: ecological and policy implications. Bioscience 52, 801–811. Horn, M.A., Schramm, A., Drake, H.L., 2003. The earthworm gut: an ideal habitat for ingested N2O-producing microorganisms. Applied and Environmental Microbiology 69, 1662–1669. House, G.J., Parmelee, R.W., 1985a. Comparison of soil arthropods and earthworms from conventional and no-tillage agroecosystems. Soil and Tillage Research 5, 351–360. House, G.J., Parmelee, R.W., 1985b. Comparison of soil arthropods and earthworms from conventional and no-tillage agroecosystems. Soil & Tillage Research 5, 351–360. Karsten, G.R., Drake, H.L., 1997. Denitrifying bacteria in the earthworm gastrointestinal tract and in vivo emission of nitrous oxide (N2O) by earthworms. Applied and Environmental Microbiology 63, 1878–1882. Kong, A.Y.Y., Six, J., Bryant, D.C., Denison, R.F., Van Kessel, C., 2005. The relationship between carbon input, aggregation, and soil organic carbon stabilization in sustainable cropping systems. Soil Science Society of America Journal 69, 1078–1085. Kool, D.M., Hoffland, E., Hummelink, E.W.J., Van Groenigen, J.W., 2006. Increased hippuric acid content of urine can reduce soil N2O fluxes. Soil Biology & Biochemistry 38, 1021–1027. Lahti, T., Kuikman, P.J., 2003. The effect of delaying autumn incorporation of green manure crop on N mineralization and spring wheat
ARTICLE IN PRESS E. Rizhiya et al. / Soil Biology & Biochemistry 39 (2007) 2058–2069 (Triticum aestivum L.) performance. Nutrient Cycling in Agroecosystems 65, 265–280. Lavelle, P., 1983. The structure of earthworm communities. In: Satchell, J.W. (Ed.), Earthworm Ecology: From Darwin to Vermiculture. Chapman & Hall, London, pp. 449–465. Lee, K.E., 1985. Earthworms: Their Ecology and Relationship with Soils and Land Use. Academic Press, Sydney, Australia. Mackay, A.D., Kladivko, E.J., 1985. Earthworms and rate of breakdown of soybean and maize residues in soil. Soil Biology & Biochemistry 17, 851–857. Matthies, C., GrieXhammer, A., Schmittroth, M., Drake, H.L., 1999. Evidence for involvement of gut-associated denitrifying bacteria in emission of nitrous oxide (N2O) by earthworms obtained from garden and forest soils. Applied and Environmental Microbiology 65, 3599–3604. Mosier, A.R., Kroeze, C., Nevison, C., Oenema, O., Seitzinger, S.P., Van Cleemput, O., 1998. Closing the global N2O budget: nitrous oxide emissions through the agricultural nitrogen cycle. OECD/IPCC/IEA phase II development of IPCC guidelines for national greenhouse gas inventory methodology. Nutrient Cycling in Agroecosystems 52, 225–248. Novoa, R.S.A., Tejeda, H.R., 2006. Evaluation of the N2O emissions from N in plant residues as affected by environmental and management factors. Nutrient Cycling in Agroecosystems 75, 29–46. Parkin, T.B., Berry, E.C., 1994. Nitrogen transformations associated with earthworm casts. Soil Biology & Biochemistry 26, 1233–1238. Parkin, T.B., Berry, E.C., 1999. Microbial nitrogen transformations in earthworm burrows. Soil Biology & Biochemistry 31, 1765–1771. Paul, J.W., Beauchamp, E.G., 1989. Denitrification and fermentation in plant-residue-amended soil. Biology and Fertility of Soils 7, 303–309. Perel, T.S., 1977. Differences in lumbricid organization connected with ecological properties. In: Lohm, U., Persson, E. (Eds.), Soil Organisms as Components of Ecosystems. Swedish Natural Science Research Council, Stockholm, pp. 122–132. Pinto, M., Merino, P., Del Prado, A., Estavillo, J.M., Yamulki, S., Gebauer, G., Pietrzak, S., Lauf, J., Oenema, O., 2004. Increased emissions of nitric oxide and nitrous oxide following tillage of a perennial pasture. Nutrient Cycling in Agroecosystems 70, 13–22. Postma-Blaauw, M.B., Bloem, J., Faber, J.H., Van Groenigen, J.W., De Goede, R.G.M., Brussaard, L., 2006. Earthworm species composition affects the soil bacterial community and net nitrogen mineralization. Pedobiologia 50, 243–256. Postma, J.F., Keijzers, R.C.M., Faber, J.H., Belfroid, A.C., Stroomberg, G.J., Den Besten, P.J., 2006. Environmental risk assessment on aquatic–terrestrial gradients in a freshwater tidal area. Aquatic Ecosystem Health & Management 9, 339–349. Pulleman, M.M., Six, J., Uyl, A., Marinissen, J.C.Y., Jongmans, A.G., 2005. Earthworms and management affect organic matter incorporation
2069
and microaggregate formation in agricultural soils. Applied Soil Ecology 29, 1–15. Schmidt, O., 1999. Intrapopulation variation in carbon and nitrogen stable isotope ratios in the earthworm Aporrectodea longa. Ecological Research 14, 317–328. Shaw, C., Pawluk, S., 1986a. Fecal microbiology of Octolasion tyrtaeum, Aporrectodea turgida and Lumbricus terrestris and its relation to the carbon budgets of 3 artificial soils. Pedobiologia 29, 377–389. Shaw, C., Pawluk, S., 1986b. The development of soil structure by Octolasion tyrtaeum, Aporrectodea turgida and Lumbricus terrestris in parent materials belonging to different textural classes. Pedobiologia 29, 327–339. Six, J., Bossuyt, H., DeGryze, S., Denef, K., 2004a. A history of research on the link between (micro)aggregates, soil biota, and soil organic matter dynamics. Soil and Tillage Research 79, 7–31. Six, J., Ogle, S.M., Breidt, F.J., Conant, R.T., Mosier, A.R., Paustian, K., 2004b. The potential to mitigate global warming with no-tillage management is only realized when practised in the long term. Global Change Biology 10, 155–160. Soil Survey Staff, 1998. Keys to Soil Taxonomy. United States Department of Agriculture, Washington, 326p. Stott, D.E., Elliott, L.F., Papendick, R.I., Campbell, G.S., 1986. Lowtemperature or low water potential effects on the microbial decomposition of wheat residue. Soil Biology & Biochemistry 18, 577–582. Swift, M.J., Izac, A.M.N., van Noordwijk, M., 2004. Biodiversity and ecosystem services in agricultural landscapes—are we asking the right questions? Agriculture, Ecosystems and Environment 104, 113–134. Van Beek, C.L., Hummelink, E.W.J., Velthof, G.L., Oenema, O., 2004. Denitrification rates in relation to groundwater level in a peat soil under grassland. Biology and Fertility of Soils 39, 329–336. Van Groenigen, J.W., Kuikman, P.J., De Groot, W.J.M., Velthof, G.L., 2005a. Nitrous oxide emission from urine-treated soil as influenced by urine composition and soil physical conditions. Soil Biology & Biochemistry 37, 463–473. Van Groenigen, J.W., Velthof, G.L., Van der Bolt, F.J.E., Vos, A., Kuikman, P.J., 2005b. Seasonal variation in N2O emissions from urine patches: effects of urine concentration, soil compaction and dung. Plant and Soil 273, 15–27. Vellinga, T.V., Van den Pol—van Dasselaar, A., Kuikman, P.J., 2004. The impact of grassland ploughing on CO2 and N2O emissions in the Netherlands. Nutrient Cycling in Agroecosystems 70, 33–45. Velthof, G.L., Kuikman, P., Oenema, O., 2002. Nitrous oxide emission from soils amended with crop residues. Nutrient Cycling in Agroecosystems 62, 249–261. Wardle, D.A., 2002. Communities and Ecosystems: Linking the Aboveground and Belowground Components. Princeton University Press, Oxford.
