- Email: [email protected]
Competition and cocoon consumption by the earthworm Aporrectodea longa
Applied Soil Ecology 10 (1998) 127±136
Competition and cocoon consumption by the earthworm Aporrectodea longa P.R. Dalbya,b,*, G.H. Bakera,c, S.E. Smitha,b b
a CRC Soil and Land Management, PMB1, Glen Osmond, 5064 SA, Australia Department of Soil Science, Waite Campus, University of Adelaide, PMB1, Glen Osmond, 5064 SA, Australia c CSIRO Division of Entomology, PMB2, Glen Osmond, 5064 SA, Australia
Received 5 August 1997; accepted 20 December 1997
Abstract We investigated interactions between the deep burrowing earthworm Aporrectodea longa and three other common pasture species (A. caliginosa, A. trapezoides and Microscolex dubius) and the roles these other species would have on reducing the ability of A. longa to colonise agricultural land in the high rainfall zone (>600 mm) of southern Australia. Experiments were conducted in pots in the laboratory or glasshouse and in cages in the ®eld. In most experiments, ®eld soil was used and in some experiments, an arti®cial soil of commercially available garden loam mixed with bentonite clay was used. To determine whether competition occurred, the growth, survival and reproduction of earthworms was compared between single-species and mixed-species treatments. Competition between A. longa and A. caliginosa was weak and no competitive effects were measured between A. longa and A. trapezoides. The presence of A. longa signi®cantly reduced the reproductive output of M. dubius, possibly by removing its food source and habitat or consuming its cocoons. This is the ®rst reported evidence that individuals of one earthworm species may consume the cocoons of another. The establishment of A. longa into pastures in southern Australia is unlikely to be impeded by the presence of earthworm species already established. The spread of A. longa in this region will not signi®cantly reduce populations of A. trapezoides and A. caliginosa, but is likely to decrease populations of M. dubius signi®cantly. # 1998 Elsevier Science B.V. Keywords: Competition; Earthworm; Cocoon; Predation; Aporrectodea longa; Aporrectodea caliginosa; Aporrectodea trapezoides; Microscolex dubius
1. Introduction Aporrectodea longa (Ude) (Lumbricidae) is an earthworm species native to Europe which is being *Corresponding author. Present address: Department of Soil Science, Waite Campus, University of Adelaide, PMB1, Glen Osmond, 5064 SA, Australia. Tel.: +61 8 8303 6518; fax: +61 8 8303 6511; e-mail: [email protected]. 0929-1393/98/$19.00 # 1998 Elsevier Science B.V. All rights reserved. PII S0929-1393(98)00031-6
considered for introduction into the high-rainfall permanent pastures of southern Australia to increase plant production (Baker et al., 1994). If a species of earthworm is to be introduced into an area where it currently does not occur, its potential to compete with species already established in that area needs to be evaluated. Competition occurs when the growth or reproductive output of one or more organism(s) is reduced, due
128
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
to another organism either consuming a shared resource or physically or chemically blocking access to a resource (Nicholson, 1954). Such competitive interactions have been demonstrated across a wide range of organisms and ecosystems (Schoener, 1983; Gurevitch et al., 1992). However, very little is known about species interactions in earthworm communities (Curry, 1994) apart from a few pot experiments (Abbott, 1980; Dalby et al., 1998) and a ®eld survey in Mexico (Fragosa, 1993). Abbott and Fragosa produced evidence suggesting that competition occurred between different species of earthworms. Dalby et al., 1996 found no evidence to suggest that A. longa would compete with a native megascolecid, if introduced to adjacent pastures. However, not enough is known about competitive interactions between earthworms for predictions to be made on whether any two earthworm species are likely to compete with each other. In Tasmania, both A. caliginosa and A. longa have higher population densities in areas where both species coexist compared to areas where only one species is present (Temple-Smith et al., 1993). Furthermore, these two species are often found together in permanent pastures in Tasmania (R. Garnsey, personal communication) and New Zealand (J. Springett, personal communication). Whether this is because of positive interactions between the two species, or a sharing of similar habitat preferences, is not known. The most widespread species of earthworms found in high-rainfall pastures in southern Australia are Aporrectodea trapezoides (DugeÂs) (Lumbricidae) and Aporrectodea caliginosa (Savigny) (Lumbricidae) (Baker et al., 1994). In the Mount Lofty Ranges, South Australia, the area in which this study was based, Microscolex dubius (Fletcher) (Acanthodrilidae) is also widespread, although at low densities (Baker et al., 1992a, b). We report on a series of experiments that aimed to determine whether there is competition between A. longa and three other species: A. caliginosa, A. trapezoides and M. dubius. The null hypothesis in the ®rst three experiments was that adding a second species to or increasing the density of the ®rst species would have no effect on the survival or growth of the ®rst species (i.e. there was no competitive effect). In one of these experiments, A. longa was shown to reduce the number of cocoons of M. dubius. To determine whether the mechanism of cocoon reduc-
tion was cocoon consumption, the ®nal two experiments investigated the ability of A. longa to consume earthworm cocoons or arti®cial spheres of similar size to cocoons. 2. Materials and methods 2.1. Earthworms Individuals of the species A. longa and A. caliginosa were collected from a pasture soil near Smithton in northern Tasmania, Australia. Individuals of A. trapezoides and M. dubius were collected from the Waite Campus of the University of Adelaide, South Australia. They were stored temporarily in a mixture of soil and sphagnum moss in the dark at 158C before being weighed and placed into pots of soil or in ®eld cages (see below). The average weight of individual earthworms at the beginning of an experiment was not signi®cantly different between treatments within a species. At the end of the experiment, earthworms were handsorted from the soil and weighed. Earthworms (except for M. dubius) were pre-treated and weighed at the beginning and end of experiments. Pre-treatment consisted of incubation in 50 ml screwtop, plastic specimen containers containing moist ®lter paper (Whatman no. 24) for 24 h (A. caliginosa, A. trapezoides) or 72 h (A. longa) to void their intestinal contents and standardise the water content of their tissue (Dalby et al., 1996). This treatment was not used on M. dubius, because it kills most individuals (Dalby, unpublished data). 2.2. Soil Soils were collected from pastures at Springmount and Woodside in the Mount Lofty Ranges. The site at Springmount is on a slope facing north (138) and receives an annual average rainfall of 800 mm, with the majority of the rain falling between June and October (Baker et al., 1992a). The soil is Bleached, Eutrophic, Brown Chromosol; clay loamy (Isbell, 1995), with a pH (CaCl2) of 4.9% and 5.1% carbon. The vegetation is a mix of grasses, clover and broadleaf weeds and the paddock is used primarily for sheep grazing and occasionally for cattle grazing. The earthworm population is composed of A. trapezoides (200
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
and 400 individuals mÿ2) and A. caliginosa (150±350 individuals mÿ2, Baker (1998), unpublished data). Both species are active and are most abundant between June and October in the top 10 cm of the soil. For the rest of the year they are found quiescent further down the soil pro®le (Baker et al., 1992a). The Woodside soil was taken from an area of gently sloping land (48), facing east which has an average annual rainfall of 800 mm, with the majority of the rainfall falling between April and October. The soil is Malanic-Mottled, Eutrophic, Brown Chromosol; loamy (Isbell, 1995) with a pH (CaCl2) of 4.6% and 3.0% carbon (Merry, unpublished). The vegetation is dominated by grasses and the land used for cattle grazing. The resident population of earthworms include A. rosea, A. trapezoides, A. caliginosa and M. dubius (Baker, 1998, unpublished data). An ``arti®cial'' soil was made by mixing commercially available, garden-loam which contained low levels of organic carbon (0.18%) with a bentonite clay (0.12% organic carbon) taken from a deposit on ``Arumpo'' station in south-western New South Wales. Clay was added because pure loam has a poor water holding capacity and the range of soil moistures between ®eld capacity and wilting point is very low. The loam was sieved (5 mm mesh) to remove stones. 2.2.1. Field experiment ± interactions between A. longa , A. caliginosa and A. trapezoides The aim of this experiment was to determine whether there were signi®cant competitive or mutualistic interactions between A. longa and either A. caliginosa or A. trapezoides by measuring changes in the growth or survival of each species in single species and mixed species treatments. There were seven treatments with nine replicates each:
A. longa added at 7 cageÿ1, A. longa added at 14 cageÿ1, A.longa (7 cageÿ1) plus A. trapezoides (30 cageÿ1), A. longa (7 cageÿ1) plus A. caliginosa (30 cageÿ1), A. caliginosa (30 cageÿ1), A. trapezoides at (30 cageÿ1), Control (no earthworms).
Cages were constructed from 30 cm diameter PVC plastic pipe, cut into 20 cm lengths. The lengths of pipe were inserted 15 cm into the soil in early spring
129
(September±October) while the soil was moist. These were removed with the soil column intact in the dry summer period (December±February), when the majority of earthworms had moved down the soil pro®le to aestivate. Fine, nylon mesh was secured over the base before the cages were placed back in the soil. These cages could then be used in the following growing season (autumn±winter) with minimal contamination by local populations, although there was still some contamination from cocoons hatching and small juveniles which had not burrowed down past 15 cm. Further details of the method are outlined in (Baker et al., 1996). Earthworms were added to cages at Springmount in late June 1995. The average starting weight for individuals of A. trapezoides was 288 mg, for A. caliginosa 289 mg, and for A. longa 961 mg. The percentages of the starting population that had a clitellum (i.e. were mature adults) were 15% for A. longa, 29% for A. caliginosa and 1.3% for A. trapezoides. The number of clitellate earthworms in each treatment was similar for A. longa (1-way ANOVA, F0.13, df3, P>0.05), and for A. caliginosa and A. trapezoides (1-way ANOVA, F<0.57, df2, P>0.05). The experiment was terminated in mid-October 1995 after 15 weeks. A multiple regression was used to analyse the data, which took into account the effects of the spatial distribution of the cores, the high death rate of earthworms in some cages, and the presence of contaminants ± i.e. earthworms which had hatched in the cages from cocoons produced in the previous season. 2.2.2. Pot experiment ± interaction between A. longa and A. caliginosa The aim of this experiment was to determine if there were signi®cant competitive or mutualistic interactions between A. longa and A. caliginosa by measuring changes in the growth or survival of each species in single species and mixed species treatments. There were ®ve treatments with 10 replicates each:
A. caliginosa added at 6 worms potÿ1, A. caliginosa added at 12 worms potÿ1, A. longa added at 6 worms potÿ1, A. longa added at 12 worms potÿ1, A. longa and A. caliginosa together, each added at 6 potÿ1.
130
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
The earthworms were added to 4 l pots (20 cm diameter, 20 cm depth) containing 2.85 kg of (ovendried) soil which was collected from Springmount to a depth of 15 cm, air-dried then sieved through a 5 mm mesh sieve. The pots were kept in a constant temperature water bath at 158C (18C). The water potential was kept above ÿ10 kPa by weekly watering to constant weight. The mean starting weights of individual A. longa and A. caliginosa were 1282 and 361 mg, respectively; A. longa is naturally a larger earthworm than A. caliginosa. The earthworms were stored for four weeks after collection before being added to the pots in early July, 1993. The experiment was terminated 10 weeks after the addition of earthworms to the soil. 2.2.3. Pot experiment ± interactions between A. longa and M. dubius The aim of this experiment was to investigate whether A. longa had a signi®cant in¯uence on the survival, growth or reproduction of M. dubius. Soil columns were collected from Woodside by inserting 15 cm diameter, 20 cm long PVC cores to a depth of 15 cm and removing them with the soil column intact. The soil was watered to a water potential of ÿ23 kPa. The soil columns were kept in a waterbath at 158C in a glasshouse with air temperatures ranging from 98C to 328C. A square of ®ne, nylon mesh was secured over the top of the core and a 10 mm diameter hole cut out to allow the pasture plants (mostly ryegrass) to grow through. A cap was placed on the base to stop earthworms from escaping. Two individuals of M. dubius were added to each pot. Half of the pots had four A. longa added which is equivalent to the density of A. longa in a previous ®eld experiment at Woodside (Dalby, unpublished data). That experiment had shown that A. longa reduced the resident populations of M. dubius. The density of added M. dubius was set at 2 potÿ1 because this is equivalent to the average density of M. dubius found at Woodside (100 mÿ2). There were eight replicates for each treatment. Sheep dung was collected from holding pens, dried and ground through a 1 mm mesh sieve. 8.85 g (dry weight) of this sheep dung was mixed with 18 g of reverse osmosis (RO) water and added to the surface of each pot as a food source for M. dubius.
The average individual weight of M. dubius individuals was 191 mg in the treatment with A. longa and 189 mg in the treatment without A. longa. The average individual weight of A. longa was 1074 mg in the treatment with added M. dubius and 1080 mg in the treatment without M. dubius. Earthworms of this size were chosen because larger earthworms (1550 mg) lost weight under these conditions (Dalby, unpublished data). All M. dubius and 28% of the four A. longa in each pot were clitellate. The pots of soil were watered to constant weight three times a week to maintain the water potential between ÿ10 and ÿ25 kPa. Ryegrass (the dominant pasture plant species in this soil) was allowed to germinate from seed reserves. The shoots of the ryegrass were cut after two weeks and every week thereafter to simulate grazing. The experiment was terminated after 10 weeks. Litter was removed from the surface by hand, separated from soil by ¯otation in a saturated sodium chloride solution (360 g lÿ1), rinsed with RO water, dried (608C for 24 h) and weighed. Cocoons were wet sieved from the soil, after handsorting for earthworms, through a 1.6 mm sieve which was ®ne enough to catch the smallest cocoons of M. dubius (2.6 mm long and 1.8 mm wide). 2.2.4. Can A. longa consume earthworm cocoons or objects the same size as cocoons? Our ®rst experiment to test this question was designed to determine if A. longa can consume cocoons of M. dubius and whether large worms are better able to do so than small worms. Cocoons of M. dubius were collected from a site on the Waite campus immediately before the start of the experiment by wet sieving soil through a 1.6 mm sieve and removing the cocoons with forceps. Individuals of A. longa were separated into two size classes, large (av. wt. 3981 mg) and small (av. wt. 719 mg). A control (no earthworms) was included to ensure that cocoons were not disappearing for reasons other than the presence of A. longa. There were six replicates for each treatment. Twenty ®ve grams (dry weight) of the ``arti®cial'' soil (see above) was added to 50 ml plastic specimencontainers and adjusted to a moisture content of 15% (g gÿ1) which corresponded to a water potential of ÿ10 kPa. Another 10g of ``arti®cial'' soil which contained 7% sheep dung and moistened to ÿ10 kPa
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
(23% g gÿ1) was then added on top. Five grams of the sheep dung:soil mix was added ®rst, followed by ®ve cocoons which were placed evenly onto the surface and then the remaining 5 g of dung:soil mix was added on top. One individual of A. longa was added to each container. Earthworm casts were collected from the surface every day for 48 days, and sieved through a 0.5 mm sieve to extract cocoons. After 48 days, earthworms were handsorted from the soil and weighed. The soil remaining in the containers after 48 days was wet-sieved through a 1.6 mm mesh to extract cocoons of M. dubius. Our second experiment was designed to determine the preference of A. longa for small plastic spheres of two sizes; 1.5 and 2.5 mm diameter, which are of similar size to M. dubius cocoons (1.8 mm diameter). Plastic spheres were used because cocoons of M. dubius are dif®cult to collect due to their small size and low density in the soil (4.5 cocoons kgÿ1). The three treatments were large earthworms (fresh weight3000±3200 mg), medium-sized earthworms (fresh weight1500±1800 mg) and small earthworms (fresh weight800±900 mg), with three replicates for each treatment. Treatments with different earthworm sizes were included to determine whether the size of an earthworm restricted its ability to consume beads. Plastic specimen containers (50 ml) with screw top lids were ®lled with 40 g (dry weight) of ``arti®cial'' soil which contained 7% (g gÿ1) sheep dung and was moistened to a moisture content of 23% (g gÿ1). The small plastic beads were added to the soil at 2.5 beads gÿ1 soil for each size class (2500 kgÿ1 dry soil). One individual of A. longa was added to each container and the containers were kept at a constant temperature of 158C (18C). Earthworm casts were collected after 7 and 14 days, by sorting through the soil and picking them out with forceps. Any beads adhering to the surface of the casts were removed and discarded because they might not have passed through the intestines of the earthworms. The casts were then sieved through a 0.5 mm sieve to separate the beads. 3. Analyses Data were analysed using analysis of variance and differences between individual means were deter-
131
mined using Tukey's test except where a Bartlett's test of equal variances indicated a signi®cant difference between variances of the means. Under such circumstances a Kruskal±Wallis test was used. Ratios were transformed using arcsin-squareroot (x0.5) and counts were transformed using log(x1). 4. Results 4.1. Field experiment ± interactions between A. longa, A. caliginosa and A. trapezoides The average level of survival in the ®eld experiment was 58.5% for A. longa, 80.7% for A. caliginosa and 31.6% for A. trapezoides. The number of earthworms collected at the end of the experiment was not highly correlated to the number added at the start due to contamination by other individuals (especially A. caliginosa) and a high number of deaths in a few cages. An analysis of variance between original treatments was therefore an invalid method of statistical analysis, because it could not be certain that replicate cages within ``treatments'' had a similar number of individuals of the same species. To overcome this, a multiple regression was used to analyse the relationship between numbers of individuals of each species with numbers of individuals of other species and its spatial relationship with other cages. There was a signi®cant negative relationship between the number of A. caliginosa and the number of A. longa handsorted from the cages at the end of the experiment (Multiple regression, df2, F3.44, P<0.05). For every extra 7.0 A. longa found in the treatment with A. caliginosa at the end of the experiment there was an average of 1.4 less A. caliginosa. Because of the problems in identifying original treatments, an analysis of the effect of interactions on growth was not done. 4.2. Pot experiment 1 ± interaction between A. longa and A. caliginosa Overall survival of A. caliginosa and A. longa was greater than 95%, with no signi®cant differences in survival between treatments for either A. caliginosa (Kruskal±Wallis, H2.00, P>0.05) or A. longa (1-way ANOVA, df2, F0.00, P>0.05).
132
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
4.3. Pot experiment ± effect of A. longa on M. dubius
Both species of earthworms lost weight over the period of the experiment (Figs. 1 and 2). Individuals of A. caliginosa lost signi®cantly more weight (Kruskal±Wallis, H13.72, P<0.05) in the treatment with A. longa than they did in the single species treatments. A. longa lost signi®cantly more weight (1-way ANOVA, df2, F6.63, P<0.05) in the treatment with A. caliginosa and in the high density, single species treatment than in the low density, single species treatment.
Few M. dubius (<0.25 potÿ1) were recovered at the end of the experiment, although other residential species were recovered from the soil including A. trapezoides, Aporrectodea rosea (Savigny) and A. caliginosa. The average number of residents was 4.01 potÿ1. Signi®cantly more M. dubius cocoons were sieved from the treatments with no A. longa added (1.69) compared to the treatments with A. longa added (0.81) (2-way ANOVA, df1, F4.47, P<0.05). All individuals of A. longa were recovered from the treatments with M. dubius and 94% were recovered from the treatment with A. longa alone. A. longa increased in weight by 23.2% over the period of the experiment, representing an average growth rate of 6.63 mg indÿ1 wkÿ1. There were no signi®cant differences between treatments in terms of weight change (1-way ANOVA, df1, F1.52, P>0.05). The mean number of A. longa cocoons sieved from the soil was 1.1 potÿ1 on average for the pots which had A. longa added, with no signi®cant differences between treatments (1-way ANOVA. df1, F0.59, P>0.05). Pots with A. longa present had signi®cantly less (1-way ANOVA, df3, F5.06, P<0.05) litter (plant litteradded dung) remaining on the surface of the soil compared to pots without A. longa after 10 weeks (Fig. 3).
Fig. 2. Weight change of A. longa in pots containing Springmount soil kept at 158C (18C) for 10 weeks. Vertical bars represent standard errors of the means, n10. Bars with same letters are not significantly different at the 5% level. L6A. longa at a density of 6 worms potÿ1. L12A. longa at a density of 12 worms potÿ1. CLA. caliginosa and A. longa at a density of 6 worms potÿ1 each.
Fig. 3. Plant litter and sheep dung remaining on the surface of pots at 158C (18C) containing intact Woodside soil after 10 weeks. Vertical bars represent standard errors of the means, n10. Bars with same letters are not significantly different at the 5% level. MM. dubius, MLA. longaM. dubius.
Fig. 1. Weight change of A. caliginosa in pots containing Springmount soil kept at 158C (18C) for 10 weeks. Vertical bars represent standard errors of the means, n10. Bars with same letters are not significantly different at the 5% level. C6A. caliginosa at a density of 6 worms potÿ1. C12A. caliginosa at a density of 12 worms potÿ1. CLA. caliginosa and A. longa at a density of 6 worms potÿ1 each.
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
133
4.4. Cocoon consumption experiment
5. Discussion
All earthworms in the cocoon consumption experiment survived over the 48 day period and all lost weight. The weight loss averaged 8.9% for large A. longa and 16.0% for small A. longa. There were no signi®cant differences between large and small A. longa in terms of the percentage weight-loss (1-way ANOVA, df1, F2.85, P>0.05). No cocoons were found in surface-collected casts. All cocoons could be accounted for in the control treatment, although one cocoon had hatched. Not all cocoons could be accounted for in the treatments with A. longa added, with both earthworm treatments losing approximately one cocoon out of ®ve.
The results of this study demonstrate that earthworm species compete with each other. This is the ®rst documented evidence of competition between earthworm species under ®eld conditions, and along with Abbott (1980) is the only evidence that earthworms compete with each other. There are two mechanisms for competitive interactions suggested by the data; competition for food and habitat resources and cocoon consumption. Competition for food and resources (scramble competition; Nicholson, 1954) occurs when two species feed on the same resources which are scarce for at least one of the species. In the case of interactions between A. longa and M. dubius, the resources were litter and dung. Although A. longa also had access to other resources in the soil, M. dubius (an epigeic species) was totally reliant on litter and dung as a food source and for shelter. By burying most of the litter and dung, A. longa removed food resources and shelter. This was a likely contributing factor to the reduced cocoon production by M. dubius in the presence of A. longa. For A. longa and A. caliginosa on the other hand, determining what might be the scarce resources is more dif®cult to de®ne because the food resources are more intimately mixed with mineral soil. Both species have similar habitat preferences, so there is a high chance that they use and compete for similar resources. However, attempts in the past to categorise food resources of earthworms have been largely unsuccessful. In studies of earthworm diets, many of the prominent food groups that are described are poorly identi®ed in terms of their source, chemical composition and likely digestibility (Piearce, 1978; Judas, 1992; Gunn and Cherrett, 1993). Although enzyme studies suggest that earthworms are able to digest a wide range of organic compounds (Tracey, 1951; Nielsen, 1963; Parle, 1963a, b; UrbaÂsÏek, 1990; UrbaÂsÏek and Pizl, 1991; Zhang et al., 1993), the lack of competition between A. trapezoides and A. longa suggest that at least some species utilise different fractions of the soil organic matter pool. An increased understanding is required of which components of soil organic matter are accessed by earthworms to further understand the mechanisms of interactions between earthworms which feed in the soil layers.
4.5. Plastic spheres experiment The average quantity of casts collected for each individual earthworm was 2.5 g for the small earthworms (3.0 g gÿ1 starting weight of earthworm), 4.8 g for the medium-sized earthworms (2.8 g gÿ1 starting weight of earthworm) and 7.8 g for the large earthworms (2.5 g gÿ1 starting weight of earthworm). Both small and large beads were collected from the casts of every individual earthworm. The density of the beads in the casts was lower than the density found in the soil, especially for the larger beads (Table 1). The density of beads gÿ1 cast material was signi®cantly higher for smaller beads than for larger beads (2-way ANOVA, df1, F6.59, P<0.05) but there were no signi®cant differences in bead density between the different size classes of earthworms (2-way ANOVA, df2, F1.21, P>0.05). Table 1 Density of beads found in the casts of A. longa of three different size classes which had been placed in soil mixed with sheep dung for 7 days at 158C (18C) Size class of A. longa (g)
Density of beads (beads gÿ1 cast material) 1.5 mm
2.5 mm
0.75±1.0 1.5±2.5 3.0±4.0
1.96 (0.30) 1.88 (0.76) 1.12 (0.19)
0.76 (0.28) 1.06 (0.12) 0.66 (0.06)
The soil had small (1.5 mm diameter) and large (2.5 mm diameter) beads mixed throughout at 2.5 beads gÿ1 for each bead size. Numbers in brackets are standard errors of the mean, n3.
134
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
Table 2 Consumption rates of soil by different species of lumbricid earthworms Species
Consumption rate (mg dayÿ1 gÿ1 earthworm)
Reference
A. caliginosa
362±2353 3750±4730 280±420 2630±4190 1130±1270 1920±3010 80±460 0.43±2.55 242±713 70±180 6.5±16.9
Curry et al. (1995) Martin (1982) Barley (1959) Martin (1982) Bolton and Phillipson (1976) Martin (1982) Shipitalo et al. (1988) Dickschen and Topp (1987) Curry et al. (1995) Shipitalo et al. (1988) Heine and Larink (1993)
A. trapezoides A. rosea L. rubellus L. terrestris
The results presented suggest that A. longa can cause the destruction of M. dubius cocoons. If this reduction in cocoon numbers was due solely to A. longa consuming the cocoons and that individual earthworms came across the cocoons by chance and did not actively search for them, then it would be expected that the earthworms would have to consume 1320 mg soil g earthworm tissueÿ1 dayÿ1 (Appendix A). This estimate is within the range of consumption rates measured for other lumbricid species of earthworms (Table 2), so it is theoretically possible that the reduction in M. dubius cocoons in the pot experiment can be accounted for by direct consumption by A. longa. If the cocoons were disproportionately distributed near the surface of the soil and if A. longa is feeding more often near the soil surface, then the consumption rate of soil by A. longa that would be required to destroy the cocoons of M. dubius would be less. Predation of eggs or young of other species as a mechanism for competition has been shown previously in birds (Hakkarainen and Korpimaeki, 1996; Spooner et al., 1996) and ®sh (Garner, 1996). In these cases, eggs or young were a high quality food resource which were exploited by a competing species, which had the effect of increasing its competitive advantage. It is not surprising that earthworms would choose to consume cocoons which are high in energy and protein and would make an excellent food source. It is also possible that earthworms consume cocoons and eggs of other soil fauna such as snails and slugs; probably only chemical toxicity or physical size would inhibit this behaviour. Individuals of A. longa
larger than 800 mg consumed spheres as large as 2.5 mm in diameter. It is therefore conceivable that A. longa of this size would be able to consume the cocoons of other species if their cocoons were smaller than 2.5 mm. It is also possible that A. longa could consume cocoons larger than 2.5 mm diameter. Another large earthworm, Lumbricus terrestris (Lumbricidae) (L) is able to consume seeds up to 3 mm in diameter and 6 mm in length (Piearce et al., 1994). However, although A. longa may also be physically able to consume objects of this size, the data presented here shows they avoid larger beads. Shumway and Koide (1994) have similarly found that earthworms prefer to consume smaller objects (0.5±1.0 mm). Therefore these earthworm species are unlikely to have a signi®cant impact on the numbers of larger cocoons in soil. Inter-speci®c competition within the spatial scale of a cage may not translate into competitive exclusion at larger, ®eld scale. There is a greater opportunity for competitive avoidance at a larger scale as heterogeneity of habitat types and resources increases. Numerous theoretical models have shown that habitat subdivision can allow two species, a fugitive species and a superior competitor, to coexist as a stable metapopulation (Armstrong, 1976; Hastings, 1980; Schmida and Ellner, 1984). Such metapopulation models are dif®cult to apply to earthworm populations in agricultural ®elds because it is dif®cult to measure boundaries of any one population, let alone numerous, neighbouring populations. Inoculation programs offer a unique opportunity to study the dispersal and colonising behaviour of earthworms and to study whether
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
some species of earthworms outcompete other species, or reduce their population densities when they invade an area. 5.1. Implications for introducing A. longa Although A. longa and A. caliginosa compete against each other, the interaction appears weak. The results presented here suggest that on an average, 10 A. longa are required to reduce the numbers of A. caliginosa by two individuals in a population. This means that addition of 10 A. longa to a soil volume that supports a population of 10 A. caliginosa, would reduce the population of A. caliginosa to eight but the total earthworm population would increase to 18. Furthermore, the opportunities for individuals to avoid competition at the scale of a paddock are much greater than those in a 30 cm diameter cage, and so the negative effects of one on the other is likely to be overstated in these experiments. Therefore it would be expected that total earthworm activity would be increased by adding A. longa to a soil already colonised by A. caliginosa. Adding A. longa to communities of A. caliginosa has in fact been shown by others to increase plant productivity and improve soil conditions to a greater extent than having just A. caliginosa alone (Syers and Springett, 1983; Springett, 1985; Baker, 1998). The weak competitive interaction demonstrated between A. longa and A. caliginosa in the pot and ®eld experiments suggests that the positive association between A. longa and A. caliginosa found in Tasmania (Temple-Smith et al., 1993) is due to similar habitat preferences rather than a positive interaction between the two species. It is quite likely that A. longa will reduce the numbers of M. dubius where the two coexist. The effect that M. dubius can have on nutrient cycling and pasture production is unknown and therefore the impact on nutrient cycling of A. longa reducing M. dubius populations in pastures cannot be predicted accurately. Competition between earthworm species was demonstrated in the short-term experiments outlined in this paper. Competition between A. longa and A. caliginosa was weak and there was no demonstrated competition between A. longa and A. trapezoides. In the ®eld, the ability of earthworm species to avoid competition is likely to be greater than in
135
experimental cages or pots. It seems likely therefore that introducing A. longa to soils that contain populations of either A. caliginosa or A. trapezoides would not be limited by competitive interactions. A. longa reduced the number of cocoons of M. dubius. This may be as a result of A. longa burying surface litter thus depriving M. dubius of habitat and/or food and A. longa consuming cocoons of M. dubius. Acknowledgements This research was funded by the CRC for Soil and Land Management. The authors would like to especially thank J. Rye for the use of the Springmount ®eld site and Dr. Ray Correll for his assistance with statistical analyses. Appendix R M
T N B C
cocoon reduction due to presence of A. longa (0.48 (0.81 out of 1.69)) soil mass (2.0 kg, assuming M. dubius cocoons were spread through the soil column, although they are more likely to be found in the top layer as M. dubius is epigeic in habit) time (70 days) number of 8 A. longa cageÿ1 (8) average earthworm biomass (1300 mg) consumption rate to consume 48% of soil over 70 days ((RM)/(NBT))
References Abbott, I., 1980. Do earthworms compete for food? Soil Biol. Biochem. 12, 523±530. Armstrong, R.A., 1976. Fugitive species: experiments with fungi and some theoretical considerations. Ecology 57, 953±963. Baker, G.H., 1998. Recognising and responding to the influences of agriculture and other land-use practices on soil fauna in Australia. Appl. Soil Ecol. 9, 303±310. Baker, G.H., Barrett, V.J., Grey-Gardener, R., Buckerfield, J.C., 1992a. The life history and abundance of the introduced earthworms Aporrectodea trapezoides and A. caliginosa (Annelida: Lumbricidae) in pasture soils in the Mount Lofty Ranges. South Australia. Aust. J. Ecol. 17, 177±188. Baker, G.H., Buckerfield, J., Grey-Gardner, R., Merry, R., Doube, B., 1992b. The abundance and diversity of earthworms in
136
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
pasture soils in the Fleurieu Peninsular. South Australia. Soil Biol. Biochem. 12, 1389±1395. Baker, G.H., Carter, P.J., Barrett, V.J., Kilpin, G.P., Buckerfield, J.C., Dalby, P.R., 1994. The introduction and management of earthworms to improve soil structure and fertility in southeastern Australia. In: Pankhurst, C.E., Doube, B.M., Gupta, V.V.S.R., Grace, P.R. (Eds.), Soil Biota: Management in Sustainable Farming Systems. CSIRO, Australia, pp. 42±49. Baker, G.H., Barrett, V.J., Carter, P.J., Woods, J.P., 1996. A method for caging earthworms for use in field experiments. Soil Biol. Biochem. 28, 331±339. Barley, K.P., 1959. The influence of earthworms on soil fertility II. Consumption of soil and organic matter by the earthworm Allolobophora caliginosa (Savigny). Aust. J. Agric. Res. 10, 179±185. Bolton, P.J., Phillipson, J., 1976. Burrowing, feeding, egestion and energy budgets of Allolobophora rosea (Savigny) (Lumbricidae). Oecologia 23, 225±245. Curry, J.P., 1994. Grassland Invertebrates: Ecology, Influence on Soil Fertility and Effects on Plant Growth. Chapman&Hall, London. Curry, J.P., Byrne, D., Boyle, K.E., 1995. The earthworm population of a winter cereal field and its effects on soil and nitrogen turnover. Biol. Fertility Soils 19, 166±172. Dalby, P.R., Baker, G.H., Smith, S.E., 1996. ``Filter paper method'' to remove soil from earthworm intestines and to standardise the water content of earthworm tissue. Soil Biol. Biochem. 28, 685±687. Dalby, P.R., Baker, G.H., Smith, S.E., 1998. Potential impact of an introduced lumbricid on a native woodland in South Australia. Appl. Soil Ecol., 9, 351±354. Dickschen, F., Topp, W., 1987. Feeding activities and assimilation efficiencies of Lumbricus rubellus (Lumbricidae) on a plant only diet. Pedobiologia 30, 31±37. Fragosa, C., 1993. Les peuplements de vers de terre dans l'est et sud'est de Mexique. Ph.D. thesis. University of Paris 6. Paris, France. Garner, P., 1996. Microhabitat use and diet of 0 cyprinid fishes in a lentic, regulated reach of the River Great Ouse England. J. Fish Biol. 48, 367±382. Gunn, A., Cherrett, J.W., 1993. The exploitation of food resources by soil meso- and macroinvertebrates. Pedobiologia 37, 303± 320. Gurevitch, J., Morrow, L.L., Wallace, A., Walsh, J.S., 1992. A meta-analysis of competition in field experiments. American Naturalist 140, 539±572. Hakkarainen, H., Korpimaeki, E., 1996. Competitive and predatory interactions among raptors: an observational and experimental study. Ecology 77, 1134±1142. Hastings, A., 1980. Disturbance, coexistence, history and competition for space. Theoret. Popul. Biol. 18, 363±373. Heine, O., Larink, O., 1993. Food and cast analyses as a parameter of turn-over of materials by earthworms (Lumbricus terrestris L.). Pedobiologia 37, 245±256. Isbell, R.F., 1995. Australian soil and land survey handbook series, vol. 4. CSIRO Division of Soils, Townsville.
Judas, M., 1992. Gut content analysis of earthworms (Lumbricidae) in a beechwood. Soil Biol. Biochem. 24, 1413±1417. Martin, N.A., 1982. The interaction between organic matter in soil and the burrowing activity of three species of earthworms (Oligochaeta: Lumbricidae). Pedobiologia 24, 185±190. Nicholson, A.J., 1954. Experimental demonstrations of balance in populations.. Nature 173, 862. Nielsen, R.L., 1963. Effects of soil minerals on earthworms. New Zealand J. Agric. 83, 433±435. Parle, J.N., 1963a. A microbiological study of earthworm casts. J. Gen. Microbiol. 31, 13±22. Parle, J.N., 1963b. Microorganisms in the intestines of worms. J. Gen. Microbiol. 31, 1±11. Piearce, T.G., 1978. Gut contents of some lumbricid earthworms. Pedobiologia 18, 153±157. Piearce, T.G., Roggero, N., Tipping, R., 1994. Earthworms and seeds. J. Biol. Educ 28, 195±202. Schmida, A., Ellner, S.P., 1984. Coexistence in plants with similar niches. Vegetatio 58, 29±55. Schoener, T.W., 1983. Field experiments on interspecific competition. Am. Naturalist 122, 240±285. Shipitalo, M.J., Protz, R., Tomlin, A.D., 1988. Effect of diet on the feeding and casting activity of Lumbricus terrestris and L. rubellus in laboratory culture. Soil Biol. Biochem. 29, 233±237. Shumway, D.L., Koide, R.T., 1994. Seed preferences of Lumbricus terrestris L.. Appl. Soil Ecol. 1, 11±15. Spooner, A., Pribil, S., Picman, J., 1996. Why do Gray Catbirds destroy eggs in nests of other birds? Experimental tests of alternative hypotheses. Can. J. Zool. 74, 1688±1695. Springett, J.A., 1985. Effect of introducing Allolobophora longa on root distribution and some soil properties in New Zealand pastures. In: Fitter, A.H. (Ed.) Ecological Interactions in Soil. Blackwell, Oxford, pp. 399±405. Syers, J.K., Springett, J.A., 1983. Earthworm ecology in grassland soils. In: Satchell, J.E. (Ed.) Earthworm Ecology. Chapman and Hall, London, pp 67±83. Temple-Smith, M.G., Kingston, T.J., Furlonge, T.L., Garnsey, R.B., 1993. The effect of introduction of the earthworms Aporrectodea caliginosa and Aporrectodea longa on pasture production and turf mat degradation in Tasmania. In: Temple-Smith, M.G., Pinkard, T. (Eds). The Role of Earthworms in Agriculture and Management: Report of a National Workshop. Department of Primary Industry and Fisheries, Tasmania, Technical Report 1/96. Tracey, M.V., 1951. Cellulase and chitinase of earthworms. Nature 167, 776±777. UrbaÂsÏek, F., 1990. Cellulase activity in the gut of some earthworms. Revue d'Ecologie et Biologie du Sol 27, 21±28. UrbaÂsÏek, F., Pizl, V., 1991. Activity of digestive enzymes in the gut of five earthworm species (Oligochaeta; Lumbricidae). Revue d'Ecologie et Biologie du Sol 28, 461±468. Zhang, B.G., Rouland, C., Lattaud, C., Lavelle, P., 1993. Activity and origin of digestive enzymes in gut of the tropical earthworm Pontoscolex corethrurus. Eur. J. Soil Biol. 29, 7±11.
Competition and cocoon consumption by the earthworm Aporrectodea longa P.R. Dalbya,b,*, G.H. Bakera,c, S.E. Smitha,b b
a CRC Soil and Land Management, PMB1, Glen Osmond, 5064 SA, Australia Department of Soil Science, Waite Campus, University of Adelaide, PMB1, Glen Osmond, 5064 SA, Australia c CSIRO Division of Entomology, PMB2, Glen Osmond, 5064 SA, Australia
Received 5 August 1997; accepted 20 December 1997
Abstract We investigated interactions between the deep burrowing earthworm Aporrectodea longa and three other common pasture species (A. caliginosa, A. trapezoides and Microscolex dubius) and the roles these other species would have on reducing the ability of A. longa to colonise agricultural land in the high rainfall zone (>600 mm) of southern Australia. Experiments were conducted in pots in the laboratory or glasshouse and in cages in the ®eld. In most experiments, ®eld soil was used and in some experiments, an arti®cial soil of commercially available garden loam mixed with bentonite clay was used. To determine whether competition occurred, the growth, survival and reproduction of earthworms was compared between single-species and mixed-species treatments. Competition between A. longa and A. caliginosa was weak and no competitive effects were measured between A. longa and A. trapezoides. The presence of A. longa signi®cantly reduced the reproductive output of M. dubius, possibly by removing its food source and habitat or consuming its cocoons. This is the ®rst reported evidence that individuals of one earthworm species may consume the cocoons of another. The establishment of A. longa into pastures in southern Australia is unlikely to be impeded by the presence of earthworm species already established. The spread of A. longa in this region will not signi®cantly reduce populations of A. trapezoides and A. caliginosa, but is likely to decrease populations of M. dubius signi®cantly. # 1998 Elsevier Science B.V. Keywords: Competition; Earthworm; Cocoon; Predation; Aporrectodea longa; Aporrectodea caliginosa; Aporrectodea trapezoides; Microscolex dubius
1. Introduction Aporrectodea longa (Ude) (Lumbricidae) is an earthworm species native to Europe which is being *Corresponding author. Present address: Department of Soil Science, Waite Campus, University of Adelaide, PMB1, Glen Osmond, 5064 SA, Australia. Tel.: +61 8 8303 6518; fax: +61 8 8303 6511; e-mail: [email protected]. 0929-1393/98/$19.00 # 1998 Elsevier Science B.V. All rights reserved. PII S0929-1393(98)00031-6
considered for introduction into the high-rainfall permanent pastures of southern Australia to increase plant production (Baker et al., 1994). If a species of earthworm is to be introduced into an area where it currently does not occur, its potential to compete with species already established in that area needs to be evaluated. Competition occurs when the growth or reproductive output of one or more organism(s) is reduced, due
128
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
to another organism either consuming a shared resource or physically or chemically blocking access to a resource (Nicholson, 1954). Such competitive interactions have been demonstrated across a wide range of organisms and ecosystems (Schoener, 1983; Gurevitch et al., 1992). However, very little is known about species interactions in earthworm communities (Curry, 1994) apart from a few pot experiments (Abbott, 1980; Dalby et al., 1998) and a ®eld survey in Mexico (Fragosa, 1993). Abbott and Fragosa produced evidence suggesting that competition occurred between different species of earthworms. Dalby et al., 1996 found no evidence to suggest that A. longa would compete with a native megascolecid, if introduced to adjacent pastures. However, not enough is known about competitive interactions between earthworms for predictions to be made on whether any two earthworm species are likely to compete with each other. In Tasmania, both A. caliginosa and A. longa have higher population densities in areas where both species coexist compared to areas where only one species is present (Temple-Smith et al., 1993). Furthermore, these two species are often found together in permanent pastures in Tasmania (R. Garnsey, personal communication) and New Zealand (J. Springett, personal communication). Whether this is because of positive interactions between the two species, or a sharing of similar habitat preferences, is not known. The most widespread species of earthworms found in high-rainfall pastures in southern Australia are Aporrectodea trapezoides (DugeÂs) (Lumbricidae) and Aporrectodea caliginosa (Savigny) (Lumbricidae) (Baker et al., 1994). In the Mount Lofty Ranges, South Australia, the area in which this study was based, Microscolex dubius (Fletcher) (Acanthodrilidae) is also widespread, although at low densities (Baker et al., 1992a, b). We report on a series of experiments that aimed to determine whether there is competition between A. longa and three other species: A. caliginosa, A. trapezoides and M. dubius. The null hypothesis in the ®rst three experiments was that adding a second species to or increasing the density of the ®rst species would have no effect on the survival or growth of the ®rst species (i.e. there was no competitive effect). In one of these experiments, A. longa was shown to reduce the number of cocoons of M. dubius. To determine whether the mechanism of cocoon reduc-
tion was cocoon consumption, the ®nal two experiments investigated the ability of A. longa to consume earthworm cocoons or arti®cial spheres of similar size to cocoons. 2. Materials and methods 2.1. Earthworms Individuals of the species A. longa and A. caliginosa were collected from a pasture soil near Smithton in northern Tasmania, Australia. Individuals of A. trapezoides and M. dubius were collected from the Waite Campus of the University of Adelaide, South Australia. They were stored temporarily in a mixture of soil and sphagnum moss in the dark at 158C before being weighed and placed into pots of soil or in ®eld cages (see below). The average weight of individual earthworms at the beginning of an experiment was not signi®cantly different between treatments within a species. At the end of the experiment, earthworms were handsorted from the soil and weighed. Earthworms (except for M. dubius) were pre-treated and weighed at the beginning and end of experiments. Pre-treatment consisted of incubation in 50 ml screwtop, plastic specimen containers containing moist ®lter paper (Whatman no. 24) for 24 h (A. caliginosa, A. trapezoides) or 72 h (A. longa) to void their intestinal contents and standardise the water content of their tissue (Dalby et al., 1996). This treatment was not used on M. dubius, because it kills most individuals (Dalby, unpublished data). 2.2. Soil Soils were collected from pastures at Springmount and Woodside in the Mount Lofty Ranges. The site at Springmount is on a slope facing north (138) and receives an annual average rainfall of 800 mm, with the majority of the rain falling between June and October (Baker et al., 1992a). The soil is Bleached, Eutrophic, Brown Chromosol; clay loamy (Isbell, 1995), with a pH (CaCl2) of 4.9% and 5.1% carbon. The vegetation is a mix of grasses, clover and broadleaf weeds and the paddock is used primarily for sheep grazing and occasionally for cattle grazing. The earthworm population is composed of A. trapezoides (200
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
and 400 individuals mÿ2) and A. caliginosa (150±350 individuals mÿ2, Baker (1998), unpublished data). Both species are active and are most abundant between June and October in the top 10 cm of the soil. For the rest of the year they are found quiescent further down the soil pro®le (Baker et al., 1992a). The Woodside soil was taken from an area of gently sloping land (48), facing east which has an average annual rainfall of 800 mm, with the majority of the rainfall falling between April and October. The soil is Malanic-Mottled, Eutrophic, Brown Chromosol; loamy (Isbell, 1995) with a pH (CaCl2) of 4.6% and 3.0% carbon (Merry, unpublished). The vegetation is dominated by grasses and the land used for cattle grazing. The resident population of earthworms include A. rosea, A. trapezoides, A. caliginosa and M. dubius (Baker, 1998, unpublished data). An ``arti®cial'' soil was made by mixing commercially available, garden-loam which contained low levels of organic carbon (0.18%) with a bentonite clay (0.12% organic carbon) taken from a deposit on ``Arumpo'' station in south-western New South Wales. Clay was added because pure loam has a poor water holding capacity and the range of soil moistures between ®eld capacity and wilting point is very low. The loam was sieved (5 mm mesh) to remove stones. 2.2.1. Field experiment ± interactions between A. longa , A. caliginosa and A. trapezoides The aim of this experiment was to determine whether there were signi®cant competitive or mutualistic interactions between A. longa and either A. caliginosa or A. trapezoides by measuring changes in the growth or survival of each species in single species and mixed species treatments. There were seven treatments with nine replicates each:
A. longa added at 7 cageÿ1, A. longa added at 14 cageÿ1, A.longa (7 cageÿ1) plus A. trapezoides (30 cageÿ1), A. longa (7 cageÿ1) plus A. caliginosa (30 cageÿ1), A. caliginosa (30 cageÿ1), A. trapezoides at (30 cageÿ1), Control (no earthworms).
Cages were constructed from 30 cm diameter PVC plastic pipe, cut into 20 cm lengths. The lengths of pipe were inserted 15 cm into the soil in early spring
129
(September±October) while the soil was moist. These were removed with the soil column intact in the dry summer period (December±February), when the majority of earthworms had moved down the soil pro®le to aestivate. Fine, nylon mesh was secured over the base before the cages were placed back in the soil. These cages could then be used in the following growing season (autumn±winter) with minimal contamination by local populations, although there was still some contamination from cocoons hatching and small juveniles which had not burrowed down past 15 cm. Further details of the method are outlined in (Baker et al., 1996). Earthworms were added to cages at Springmount in late June 1995. The average starting weight for individuals of A. trapezoides was 288 mg, for A. caliginosa 289 mg, and for A. longa 961 mg. The percentages of the starting population that had a clitellum (i.e. were mature adults) were 15% for A. longa, 29% for A. caliginosa and 1.3% for A. trapezoides. The number of clitellate earthworms in each treatment was similar for A. longa (1-way ANOVA, F0.13, df3, P>0.05), and for A. caliginosa and A. trapezoides (1-way ANOVA, F<0.57, df2, P>0.05). The experiment was terminated in mid-October 1995 after 15 weeks. A multiple regression was used to analyse the data, which took into account the effects of the spatial distribution of the cores, the high death rate of earthworms in some cages, and the presence of contaminants ± i.e. earthworms which had hatched in the cages from cocoons produced in the previous season. 2.2.2. Pot experiment ± interaction between A. longa and A. caliginosa The aim of this experiment was to determine if there were signi®cant competitive or mutualistic interactions between A. longa and A. caliginosa by measuring changes in the growth or survival of each species in single species and mixed species treatments. There were ®ve treatments with 10 replicates each:
A. caliginosa added at 6 worms potÿ1, A. caliginosa added at 12 worms potÿ1, A. longa added at 6 worms potÿ1, A. longa added at 12 worms potÿ1, A. longa and A. caliginosa together, each added at 6 potÿ1.
130
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
The earthworms were added to 4 l pots (20 cm diameter, 20 cm depth) containing 2.85 kg of (ovendried) soil which was collected from Springmount to a depth of 15 cm, air-dried then sieved through a 5 mm mesh sieve. The pots were kept in a constant temperature water bath at 158C (18C). The water potential was kept above ÿ10 kPa by weekly watering to constant weight. The mean starting weights of individual A. longa and A. caliginosa were 1282 and 361 mg, respectively; A. longa is naturally a larger earthworm than A. caliginosa. The earthworms were stored for four weeks after collection before being added to the pots in early July, 1993. The experiment was terminated 10 weeks after the addition of earthworms to the soil. 2.2.3. Pot experiment ± interactions between A. longa and M. dubius The aim of this experiment was to investigate whether A. longa had a signi®cant in¯uence on the survival, growth or reproduction of M. dubius. Soil columns were collected from Woodside by inserting 15 cm diameter, 20 cm long PVC cores to a depth of 15 cm and removing them with the soil column intact. The soil was watered to a water potential of ÿ23 kPa. The soil columns were kept in a waterbath at 158C in a glasshouse with air temperatures ranging from 98C to 328C. A square of ®ne, nylon mesh was secured over the top of the core and a 10 mm diameter hole cut out to allow the pasture plants (mostly ryegrass) to grow through. A cap was placed on the base to stop earthworms from escaping. Two individuals of M. dubius were added to each pot. Half of the pots had four A. longa added which is equivalent to the density of A. longa in a previous ®eld experiment at Woodside (Dalby, unpublished data). That experiment had shown that A. longa reduced the resident populations of M. dubius. The density of added M. dubius was set at 2 potÿ1 because this is equivalent to the average density of M. dubius found at Woodside (100 mÿ2). There were eight replicates for each treatment. Sheep dung was collected from holding pens, dried and ground through a 1 mm mesh sieve. 8.85 g (dry weight) of this sheep dung was mixed with 18 g of reverse osmosis (RO) water and added to the surface of each pot as a food source for M. dubius.
The average individual weight of M. dubius individuals was 191 mg in the treatment with A. longa and 189 mg in the treatment without A. longa. The average individual weight of A. longa was 1074 mg in the treatment with added M. dubius and 1080 mg in the treatment without M. dubius. Earthworms of this size were chosen because larger earthworms (1550 mg) lost weight under these conditions (Dalby, unpublished data). All M. dubius and 28% of the four A. longa in each pot were clitellate. The pots of soil were watered to constant weight three times a week to maintain the water potential between ÿ10 and ÿ25 kPa. Ryegrass (the dominant pasture plant species in this soil) was allowed to germinate from seed reserves. The shoots of the ryegrass were cut after two weeks and every week thereafter to simulate grazing. The experiment was terminated after 10 weeks. Litter was removed from the surface by hand, separated from soil by ¯otation in a saturated sodium chloride solution (360 g lÿ1), rinsed with RO water, dried (608C for 24 h) and weighed. Cocoons were wet sieved from the soil, after handsorting for earthworms, through a 1.6 mm sieve which was ®ne enough to catch the smallest cocoons of M. dubius (2.6 mm long and 1.8 mm wide). 2.2.4. Can A. longa consume earthworm cocoons or objects the same size as cocoons? Our ®rst experiment to test this question was designed to determine if A. longa can consume cocoons of M. dubius and whether large worms are better able to do so than small worms. Cocoons of M. dubius were collected from a site on the Waite campus immediately before the start of the experiment by wet sieving soil through a 1.6 mm sieve and removing the cocoons with forceps. Individuals of A. longa were separated into two size classes, large (av. wt. 3981 mg) and small (av. wt. 719 mg). A control (no earthworms) was included to ensure that cocoons were not disappearing for reasons other than the presence of A. longa. There were six replicates for each treatment. Twenty ®ve grams (dry weight) of the ``arti®cial'' soil (see above) was added to 50 ml plastic specimencontainers and adjusted to a moisture content of 15% (g gÿ1) which corresponded to a water potential of ÿ10 kPa. Another 10g of ``arti®cial'' soil which contained 7% sheep dung and moistened to ÿ10 kPa
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
(23% g gÿ1) was then added on top. Five grams of the sheep dung:soil mix was added ®rst, followed by ®ve cocoons which were placed evenly onto the surface and then the remaining 5 g of dung:soil mix was added on top. One individual of A. longa was added to each container. Earthworm casts were collected from the surface every day for 48 days, and sieved through a 0.5 mm sieve to extract cocoons. After 48 days, earthworms were handsorted from the soil and weighed. The soil remaining in the containers after 48 days was wet-sieved through a 1.6 mm mesh to extract cocoons of M. dubius. Our second experiment was designed to determine the preference of A. longa for small plastic spheres of two sizes; 1.5 and 2.5 mm diameter, which are of similar size to M. dubius cocoons (1.8 mm diameter). Plastic spheres were used because cocoons of M. dubius are dif®cult to collect due to their small size and low density in the soil (4.5 cocoons kgÿ1). The three treatments were large earthworms (fresh weight3000±3200 mg), medium-sized earthworms (fresh weight1500±1800 mg) and small earthworms (fresh weight800±900 mg), with three replicates for each treatment. Treatments with different earthworm sizes were included to determine whether the size of an earthworm restricted its ability to consume beads. Plastic specimen containers (50 ml) with screw top lids were ®lled with 40 g (dry weight) of ``arti®cial'' soil which contained 7% (g gÿ1) sheep dung and was moistened to a moisture content of 23% (g gÿ1). The small plastic beads were added to the soil at 2.5 beads gÿ1 soil for each size class (2500 kgÿ1 dry soil). One individual of A. longa was added to each container and the containers were kept at a constant temperature of 158C (18C). Earthworm casts were collected after 7 and 14 days, by sorting through the soil and picking them out with forceps. Any beads adhering to the surface of the casts were removed and discarded because they might not have passed through the intestines of the earthworms. The casts were then sieved through a 0.5 mm sieve to separate the beads. 3. Analyses Data were analysed using analysis of variance and differences between individual means were deter-
131
mined using Tukey's test except where a Bartlett's test of equal variances indicated a signi®cant difference between variances of the means. Under such circumstances a Kruskal±Wallis test was used. Ratios were transformed using arcsin-squareroot (x0.5) and counts were transformed using log(x1). 4. Results 4.1. Field experiment ± interactions between A. longa, A. caliginosa and A. trapezoides The average level of survival in the ®eld experiment was 58.5% for A. longa, 80.7% for A. caliginosa and 31.6% for A. trapezoides. The number of earthworms collected at the end of the experiment was not highly correlated to the number added at the start due to contamination by other individuals (especially A. caliginosa) and a high number of deaths in a few cages. An analysis of variance between original treatments was therefore an invalid method of statistical analysis, because it could not be certain that replicate cages within ``treatments'' had a similar number of individuals of the same species. To overcome this, a multiple regression was used to analyse the relationship between numbers of individuals of each species with numbers of individuals of other species and its spatial relationship with other cages. There was a signi®cant negative relationship between the number of A. caliginosa and the number of A. longa handsorted from the cages at the end of the experiment (Multiple regression, df2, F3.44, P<0.05). For every extra 7.0 A. longa found in the treatment with A. caliginosa at the end of the experiment there was an average of 1.4 less A. caliginosa. Because of the problems in identifying original treatments, an analysis of the effect of interactions on growth was not done. 4.2. Pot experiment 1 ± interaction between A. longa and A. caliginosa Overall survival of A. caliginosa and A. longa was greater than 95%, with no signi®cant differences in survival between treatments for either A. caliginosa (Kruskal±Wallis, H2.00, P>0.05) or A. longa (1-way ANOVA, df2, F0.00, P>0.05).
132
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
4.3. Pot experiment ± effect of A. longa on M. dubius
Both species of earthworms lost weight over the period of the experiment (Figs. 1 and 2). Individuals of A. caliginosa lost signi®cantly more weight (Kruskal±Wallis, H13.72, P<0.05) in the treatment with A. longa than they did in the single species treatments. A. longa lost signi®cantly more weight (1-way ANOVA, df2, F6.63, P<0.05) in the treatment with A. caliginosa and in the high density, single species treatment than in the low density, single species treatment.
Few M. dubius (<0.25 potÿ1) were recovered at the end of the experiment, although other residential species were recovered from the soil including A. trapezoides, Aporrectodea rosea (Savigny) and A. caliginosa. The average number of residents was 4.01 potÿ1. Signi®cantly more M. dubius cocoons were sieved from the treatments with no A. longa added (1.69) compared to the treatments with A. longa added (0.81) (2-way ANOVA, df1, F4.47, P<0.05). All individuals of A. longa were recovered from the treatments with M. dubius and 94% were recovered from the treatment with A. longa alone. A. longa increased in weight by 23.2% over the period of the experiment, representing an average growth rate of 6.63 mg indÿ1 wkÿ1. There were no signi®cant differences between treatments in terms of weight change (1-way ANOVA, df1, F1.52, P>0.05). The mean number of A. longa cocoons sieved from the soil was 1.1 potÿ1 on average for the pots which had A. longa added, with no signi®cant differences between treatments (1-way ANOVA. df1, F0.59, P>0.05). Pots with A. longa present had signi®cantly less (1-way ANOVA, df3, F5.06, P<0.05) litter (plant litteradded dung) remaining on the surface of the soil compared to pots without A. longa after 10 weeks (Fig. 3).
Fig. 2. Weight change of A. longa in pots containing Springmount soil kept at 158C (18C) for 10 weeks. Vertical bars represent standard errors of the means, n10. Bars with same letters are not significantly different at the 5% level. L6A. longa at a density of 6 worms potÿ1. L12A. longa at a density of 12 worms potÿ1. CLA. caliginosa and A. longa at a density of 6 worms potÿ1 each.
Fig. 3. Plant litter and sheep dung remaining on the surface of pots at 158C (18C) containing intact Woodside soil after 10 weeks. Vertical bars represent standard errors of the means, n10. Bars with same letters are not significantly different at the 5% level. MM. dubius, MLA. longaM. dubius.
Fig. 1. Weight change of A. caliginosa in pots containing Springmount soil kept at 158C (18C) for 10 weeks. Vertical bars represent standard errors of the means, n10. Bars with same letters are not significantly different at the 5% level. C6A. caliginosa at a density of 6 worms potÿ1. C12A. caliginosa at a density of 12 worms potÿ1. CLA. caliginosa and A. longa at a density of 6 worms potÿ1 each.
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
133
4.4. Cocoon consumption experiment
5. Discussion
All earthworms in the cocoon consumption experiment survived over the 48 day period and all lost weight. The weight loss averaged 8.9% for large A. longa and 16.0% for small A. longa. There were no signi®cant differences between large and small A. longa in terms of the percentage weight-loss (1-way ANOVA, df1, F2.85, P>0.05). No cocoons were found in surface-collected casts. All cocoons could be accounted for in the control treatment, although one cocoon had hatched. Not all cocoons could be accounted for in the treatments with A. longa added, with both earthworm treatments losing approximately one cocoon out of ®ve.
The results of this study demonstrate that earthworm species compete with each other. This is the ®rst documented evidence of competition between earthworm species under ®eld conditions, and along with Abbott (1980) is the only evidence that earthworms compete with each other. There are two mechanisms for competitive interactions suggested by the data; competition for food and habitat resources and cocoon consumption. Competition for food and resources (scramble competition; Nicholson, 1954) occurs when two species feed on the same resources which are scarce for at least one of the species. In the case of interactions between A. longa and M. dubius, the resources were litter and dung. Although A. longa also had access to other resources in the soil, M. dubius (an epigeic species) was totally reliant on litter and dung as a food source and for shelter. By burying most of the litter and dung, A. longa removed food resources and shelter. This was a likely contributing factor to the reduced cocoon production by M. dubius in the presence of A. longa. For A. longa and A. caliginosa on the other hand, determining what might be the scarce resources is more dif®cult to de®ne because the food resources are more intimately mixed with mineral soil. Both species have similar habitat preferences, so there is a high chance that they use and compete for similar resources. However, attempts in the past to categorise food resources of earthworms have been largely unsuccessful. In studies of earthworm diets, many of the prominent food groups that are described are poorly identi®ed in terms of their source, chemical composition and likely digestibility (Piearce, 1978; Judas, 1992; Gunn and Cherrett, 1993). Although enzyme studies suggest that earthworms are able to digest a wide range of organic compounds (Tracey, 1951; Nielsen, 1963; Parle, 1963a, b; UrbaÂsÏek, 1990; UrbaÂsÏek and Pizl, 1991; Zhang et al., 1993), the lack of competition between A. trapezoides and A. longa suggest that at least some species utilise different fractions of the soil organic matter pool. An increased understanding is required of which components of soil organic matter are accessed by earthworms to further understand the mechanisms of interactions between earthworms which feed in the soil layers.
4.5. Plastic spheres experiment The average quantity of casts collected for each individual earthworm was 2.5 g for the small earthworms (3.0 g gÿ1 starting weight of earthworm), 4.8 g for the medium-sized earthworms (2.8 g gÿ1 starting weight of earthworm) and 7.8 g for the large earthworms (2.5 g gÿ1 starting weight of earthworm). Both small and large beads were collected from the casts of every individual earthworm. The density of the beads in the casts was lower than the density found in the soil, especially for the larger beads (Table 1). The density of beads gÿ1 cast material was signi®cantly higher for smaller beads than for larger beads (2-way ANOVA, df1, F6.59, P<0.05) but there were no signi®cant differences in bead density between the different size classes of earthworms (2-way ANOVA, df2, F1.21, P>0.05). Table 1 Density of beads found in the casts of A. longa of three different size classes which had been placed in soil mixed with sheep dung for 7 days at 158C (18C) Size class of A. longa (g)
Density of beads (beads gÿ1 cast material) 1.5 mm
2.5 mm
0.75±1.0 1.5±2.5 3.0±4.0
1.96 (0.30) 1.88 (0.76) 1.12 (0.19)
0.76 (0.28) 1.06 (0.12) 0.66 (0.06)
The soil had small (1.5 mm diameter) and large (2.5 mm diameter) beads mixed throughout at 2.5 beads gÿ1 for each bead size. Numbers in brackets are standard errors of the mean, n3.
134
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
Table 2 Consumption rates of soil by different species of lumbricid earthworms Species
Consumption rate (mg dayÿ1 gÿ1 earthworm)
Reference
A. caliginosa
362±2353 3750±4730 280±420 2630±4190 1130±1270 1920±3010 80±460 0.43±2.55 242±713 70±180 6.5±16.9
Curry et al. (1995) Martin (1982) Barley (1959) Martin (1982) Bolton and Phillipson (1976) Martin (1982) Shipitalo et al. (1988) Dickschen and Topp (1987) Curry et al. (1995) Shipitalo et al. (1988) Heine and Larink (1993)
A. trapezoides A. rosea L. rubellus L. terrestris
The results presented suggest that A. longa can cause the destruction of M. dubius cocoons. If this reduction in cocoon numbers was due solely to A. longa consuming the cocoons and that individual earthworms came across the cocoons by chance and did not actively search for them, then it would be expected that the earthworms would have to consume 1320 mg soil g earthworm tissueÿ1 dayÿ1 (Appendix A). This estimate is within the range of consumption rates measured for other lumbricid species of earthworms (Table 2), so it is theoretically possible that the reduction in M. dubius cocoons in the pot experiment can be accounted for by direct consumption by A. longa. If the cocoons were disproportionately distributed near the surface of the soil and if A. longa is feeding more often near the soil surface, then the consumption rate of soil by A. longa that would be required to destroy the cocoons of M. dubius would be less. Predation of eggs or young of other species as a mechanism for competition has been shown previously in birds (Hakkarainen and Korpimaeki, 1996; Spooner et al., 1996) and ®sh (Garner, 1996). In these cases, eggs or young were a high quality food resource which were exploited by a competing species, which had the effect of increasing its competitive advantage. It is not surprising that earthworms would choose to consume cocoons which are high in energy and protein and would make an excellent food source. It is also possible that earthworms consume cocoons and eggs of other soil fauna such as snails and slugs; probably only chemical toxicity or physical size would inhibit this behaviour. Individuals of A. longa
larger than 800 mg consumed spheres as large as 2.5 mm in diameter. It is therefore conceivable that A. longa of this size would be able to consume the cocoons of other species if their cocoons were smaller than 2.5 mm. It is also possible that A. longa could consume cocoons larger than 2.5 mm diameter. Another large earthworm, Lumbricus terrestris (Lumbricidae) (L) is able to consume seeds up to 3 mm in diameter and 6 mm in length (Piearce et al., 1994). However, although A. longa may also be physically able to consume objects of this size, the data presented here shows they avoid larger beads. Shumway and Koide (1994) have similarly found that earthworms prefer to consume smaller objects (0.5±1.0 mm). Therefore these earthworm species are unlikely to have a signi®cant impact on the numbers of larger cocoons in soil. Inter-speci®c competition within the spatial scale of a cage may not translate into competitive exclusion at larger, ®eld scale. There is a greater opportunity for competitive avoidance at a larger scale as heterogeneity of habitat types and resources increases. Numerous theoretical models have shown that habitat subdivision can allow two species, a fugitive species and a superior competitor, to coexist as a stable metapopulation (Armstrong, 1976; Hastings, 1980; Schmida and Ellner, 1984). Such metapopulation models are dif®cult to apply to earthworm populations in agricultural ®elds because it is dif®cult to measure boundaries of any one population, let alone numerous, neighbouring populations. Inoculation programs offer a unique opportunity to study the dispersal and colonising behaviour of earthworms and to study whether
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
some species of earthworms outcompete other species, or reduce their population densities when they invade an area. 5.1. Implications for introducing A. longa Although A. longa and A. caliginosa compete against each other, the interaction appears weak. The results presented here suggest that on an average, 10 A. longa are required to reduce the numbers of A. caliginosa by two individuals in a population. This means that addition of 10 A. longa to a soil volume that supports a population of 10 A. caliginosa, would reduce the population of A. caliginosa to eight but the total earthworm population would increase to 18. Furthermore, the opportunities for individuals to avoid competition at the scale of a paddock are much greater than those in a 30 cm diameter cage, and so the negative effects of one on the other is likely to be overstated in these experiments. Therefore it would be expected that total earthworm activity would be increased by adding A. longa to a soil already colonised by A. caliginosa. Adding A. longa to communities of A. caliginosa has in fact been shown by others to increase plant productivity and improve soil conditions to a greater extent than having just A. caliginosa alone (Syers and Springett, 1983; Springett, 1985; Baker, 1998). The weak competitive interaction demonstrated between A. longa and A. caliginosa in the pot and ®eld experiments suggests that the positive association between A. longa and A. caliginosa found in Tasmania (Temple-Smith et al., 1993) is due to similar habitat preferences rather than a positive interaction between the two species. It is quite likely that A. longa will reduce the numbers of M. dubius where the two coexist. The effect that M. dubius can have on nutrient cycling and pasture production is unknown and therefore the impact on nutrient cycling of A. longa reducing M. dubius populations in pastures cannot be predicted accurately. Competition between earthworm species was demonstrated in the short-term experiments outlined in this paper. Competition between A. longa and A. caliginosa was weak and there was no demonstrated competition between A. longa and A. trapezoides. In the ®eld, the ability of earthworm species to avoid competition is likely to be greater than in
135
experimental cages or pots. It seems likely therefore that introducing A. longa to soils that contain populations of either A. caliginosa or A. trapezoides would not be limited by competitive interactions. A. longa reduced the number of cocoons of M. dubius. This may be as a result of A. longa burying surface litter thus depriving M. dubius of habitat and/or food and A. longa consuming cocoons of M. dubius. Acknowledgements This research was funded by the CRC for Soil and Land Management. The authors would like to especially thank J. Rye for the use of the Springmount ®eld site and Dr. Ray Correll for his assistance with statistical analyses. Appendix R M
T N B C
cocoon reduction due to presence of A. longa (0.48 (0.81 out of 1.69)) soil mass (2.0 kg, assuming M. dubius cocoons were spread through the soil column, although they are more likely to be found in the top layer as M. dubius is epigeic in habit) time (70 days) number of 8 A. longa cageÿ1 (8) average earthworm biomass (1300 mg) consumption rate to consume 48% of soil over 70 days ((RM)/(NBT))
References Abbott, I., 1980. Do earthworms compete for food? Soil Biol. Biochem. 12, 523±530. Armstrong, R.A., 1976. Fugitive species: experiments with fungi and some theoretical considerations. Ecology 57, 953±963. Baker, G.H., 1998. Recognising and responding to the influences of agriculture and other land-use practices on soil fauna in Australia. Appl. Soil Ecol. 9, 303±310. Baker, G.H., Barrett, V.J., Grey-Gardener, R., Buckerfield, J.C., 1992a. The life history and abundance of the introduced earthworms Aporrectodea trapezoides and A. caliginosa (Annelida: Lumbricidae) in pasture soils in the Mount Lofty Ranges. South Australia. Aust. J. Ecol. 17, 177±188. Baker, G.H., Buckerfield, J., Grey-Gardner, R., Merry, R., Doube, B., 1992b. The abundance and diversity of earthworms in
136
P.R. Dalby et al. / Applied Soil Ecology 10 (1998) 127±136
pasture soils in the Fleurieu Peninsular. South Australia. Soil Biol. Biochem. 12, 1389±1395. Baker, G.H., Carter, P.J., Barrett, V.J., Kilpin, G.P., Buckerfield, J.C., Dalby, P.R., 1994. The introduction and management of earthworms to improve soil structure and fertility in southeastern Australia. In: Pankhurst, C.E., Doube, B.M., Gupta, V.V.S.R., Grace, P.R. (Eds.), Soil Biota: Management in Sustainable Farming Systems. CSIRO, Australia, pp. 42±49. Baker, G.H., Barrett, V.J., Carter, P.J., Woods, J.P., 1996. A method for caging earthworms for use in field experiments. Soil Biol. Biochem. 28, 331±339. Barley, K.P., 1959. The influence of earthworms on soil fertility II. Consumption of soil and organic matter by the earthworm Allolobophora caliginosa (Savigny). Aust. J. Agric. Res. 10, 179±185. Bolton, P.J., Phillipson, J., 1976. Burrowing, feeding, egestion and energy budgets of Allolobophora rosea (Savigny) (Lumbricidae). Oecologia 23, 225±245. Curry, J.P., 1994. Grassland Invertebrates: Ecology, Influence on Soil Fertility and Effects on Plant Growth. Chapman&Hall, London. Curry, J.P., Byrne, D., Boyle, K.E., 1995. The earthworm population of a winter cereal field and its effects on soil and nitrogen turnover. Biol. Fertility Soils 19, 166±172. Dalby, P.R., Baker, G.H., Smith, S.E., 1996. ``Filter paper method'' to remove soil from earthworm intestines and to standardise the water content of earthworm tissue. Soil Biol. Biochem. 28, 685±687. Dalby, P.R., Baker, G.H., Smith, S.E., 1998. Potential impact of an introduced lumbricid on a native woodland in South Australia. Appl. Soil Ecol., 9, 351±354. Dickschen, F., Topp, W., 1987. Feeding activities and assimilation efficiencies of Lumbricus rubellus (Lumbricidae) on a plant only diet. Pedobiologia 30, 31±37. Fragosa, C., 1993. Les peuplements de vers de terre dans l'est et sud'est de Mexique. Ph.D. thesis. University of Paris 6. Paris, France. Garner, P., 1996. Microhabitat use and diet of 0 cyprinid fishes in a lentic, regulated reach of the River Great Ouse England. J. Fish Biol. 48, 367±382. Gunn, A., Cherrett, J.W., 1993. The exploitation of food resources by soil meso- and macroinvertebrates. Pedobiologia 37, 303± 320. Gurevitch, J., Morrow, L.L., Wallace, A., Walsh, J.S., 1992. A meta-analysis of competition in field experiments. American Naturalist 140, 539±572. Hakkarainen, H., Korpimaeki, E., 1996. Competitive and predatory interactions among raptors: an observational and experimental study. Ecology 77, 1134±1142. Hastings, A., 1980. Disturbance, coexistence, history and competition for space. Theoret. Popul. Biol. 18, 363±373. Heine, O., Larink, O., 1993. Food and cast analyses as a parameter of turn-over of materials by earthworms (Lumbricus terrestris L.). Pedobiologia 37, 245±256. Isbell, R.F., 1995. Australian soil and land survey handbook series, vol. 4. CSIRO Division of Soils, Townsville.
Judas, M., 1992. Gut content analysis of earthworms (Lumbricidae) in a beechwood. Soil Biol. Biochem. 24, 1413±1417. Martin, N.A., 1982. The interaction between organic matter in soil and the burrowing activity of three species of earthworms (Oligochaeta: Lumbricidae). Pedobiologia 24, 185±190. Nicholson, A.J., 1954. Experimental demonstrations of balance in populations.. Nature 173, 862. Nielsen, R.L., 1963. Effects of soil minerals on earthworms. New Zealand J. Agric. 83, 433±435. Parle, J.N., 1963a. A microbiological study of earthworm casts. J. Gen. Microbiol. 31, 13±22. Parle, J.N., 1963b. Microorganisms in the intestines of worms. J. Gen. Microbiol. 31, 1±11. Piearce, T.G., 1978. Gut contents of some lumbricid earthworms. Pedobiologia 18, 153±157. Piearce, T.G., Roggero, N., Tipping, R., 1994. Earthworms and seeds. J. Biol. Educ 28, 195±202. Schmida, A., Ellner, S.P., 1984. Coexistence in plants with similar niches. Vegetatio 58, 29±55. Schoener, T.W., 1983. Field experiments on interspecific competition. Am. Naturalist 122, 240±285. Shipitalo, M.J., Protz, R., Tomlin, A.D., 1988. Effect of diet on the feeding and casting activity of Lumbricus terrestris and L. rubellus in laboratory culture. Soil Biol. Biochem. 29, 233±237. Shumway, D.L., Koide, R.T., 1994. Seed preferences of Lumbricus terrestris L.. Appl. Soil Ecol. 1, 11±15. Spooner, A., Pribil, S., Picman, J., 1996. Why do Gray Catbirds destroy eggs in nests of other birds? Experimental tests of alternative hypotheses. Can. J. Zool. 74, 1688±1695. Springett, J.A., 1985. Effect of introducing Allolobophora longa on root distribution and some soil properties in New Zealand pastures. In: Fitter, A.H. (Ed.) Ecological Interactions in Soil. Blackwell, Oxford, pp. 399±405. Syers, J.K., Springett, J.A., 1983. Earthworm ecology in grassland soils. In: Satchell, J.E. (Ed.) Earthworm Ecology. Chapman and Hall, London, pp 67±83. Temple-Smith, M.G., Kingston, T.J., Furlonge, T.L., Garnsey, R.B., 1993. The effect of introduction of the earthworms Aporrectodea caliginosa and Aporrectodea longa on pasture production and turf mat degradation in Tasmania. In: Temple-Smith, M.G., Pinkard, T. (Eds). The Role of Earthworms in Agriculture and Management: Report of a National Workshop. Department of Primary Industry and Fisheries, Tasmania, Technical Report 1/96. Tracey, M.V., 1951. Cellulase and chitinase of earthworms. Nature 167, 776±777. UrbaÂsÏek, F., 1990. Cellulase activity in the gut of some earthworms. Revue d'Ecologie et Biologie du Sol 27, 21±28. UrbaÂsÏek, F., Pizl, V., 1991. Activity of digestive enzymes in the gut of five earthworm species (Oligochaeta; Lumbricidae). Revue d'Ecologie et Biologie du Sol 28, 461±468. Zhang, B.G., Rouland, C., Lattaud, C., Lavelle, P., 1993. Activity and origin of digestive enzymes in gut of the tropical earthworm Pontoscolex corethrurus. Eur. J. Soil Biol. 29, 7±11.