Earthworm populations in a cool and wet district as
affected by tractor traf®c and fertilisation
Sissel Hansen
a,*, Freddy Engelstad
baNorwegian Centre for Ecological Agriculture, Tingvoll Gard, N-6630 Tingvoll, Norway bCentre for Soil and Environmental Research, N-1432 AÊ s, Norway
Received 21 September 1998; received in revised form 3 June 1999; accepted 8 June 1999
Abstract
In a ®eld trial (1985±1996) on a sandy loam soil the effects of tractor traf®c and fertilisation on earthworm populations were investigated. The tractor traf®c treatments compared normal farm practice (`normal') with reduced tractor traf®c (`low').
In the fertilisation treatments, three different cattle-manure management methods were compared. The cattle manure was applied as diluted slurry, as aerated slurry, or as separated manure with the composted faeces being applied to tilled land, while the urine was applied to ley. The different cattle manure management methods were compared at two different application levels. In addition, there were an unfertilised control and a higher application level with diluted slurry and mineral fertiliser treatments.
The endogeic species, Aporrectodea caliginosa and A. rosea constituted 67% and 7% of the population density, respectively, and 78% and 3% of the mass in the period 1993±1996. The epigeic species,Lumbricus rubellus, constituted 25% of the population density and 19% of the mass.
The population density, but not the species composition, was strongly affected by tractor traf®c. In the years 1987±1989, there was an average of 160 earthworms/m2 after normal, and 680 earthworms/m2 after low, tractor traf®c. During the
experimental period, the earthworm population decreased in treatments with low tractor traf®c. Important reasons for the reduction in earthworm population are likely to be soil acidi®cation and reduced manure application. In 1995, the differences in population density and mass between treatments with normal, and low, tractor traf®c were no longer statistically signi®cant (140 and 250 earthworms/m2, respectively). In 1996, when there was no tractor traf®c or fertilisation carried out, the corresponding population densities were 200 and 280 earthworms/m2.
The population density and mass of earthworms were highest in manured soil. Increasing the amount of slurry applied was associated with a higher earthworm population in some years and a lower population in others. Increasing slurry application was most favourable for endogeic species and most destructive for epigeic species.
Neither the average total population density nor mass of earthworms was affected by whether the manure was treated as slurry, or separated. However, more earthworms were found after addition of solid manure than of urine. Aerated slurry did not result in a higher earthworm population than diluted slurry.#1999 Elsevier Science B.V. All rights reserved.
Keywords:Acidi®cation; Aerated slurry; Cattle manure; Mineral fertiliser; Organic farming; Soil compaction Applied Soil Ecology 13 (1999) 237±250
*Corresponding author. Tel.: +47-7153-1342; fax: +47-7153-1339
E-mail adress:sissel.hansen@norsok.no (S. Hansen)
1. Introduction
Earthworms play an important role in the turnover of organic matter in soil and in building and main-taining a good soil structure (Lavelle, 1988). They, therefore, are essential for improved utilisation of added organic matter and, thus, for plant growth, especially in an extensive agricultural system, such as organic farming, which is based on nutrient release from turnover of organic matter.
Three ecological classes of earthworms have been identi®ed (BoucheÂ, 1977; Lee, 1985). Epigeic species, such asLumbricus rubellus, live near the surface and feed on surface litter. Anecic species (deep burrowing species), such asAporrectodea longaand Lumbricus terrestris also feed on surface litter, at least partially, but pull the organic material into vertical channels in the soil. Endogeic species, such as A. caliginosa,A. roseaandAllolobophora chlorotica feed on mineral and humus particles within the soil and move more horizontally in the soil.
Only a few investigations on earthworm species in agricultural soils have been carried out in Norway (Haraldsen and Engelstad, 1994; Haraldsen et al., 1994a, b). It is, therefore, of interest to examine what earthworm species are found in Norwegian agricul-tural soils under different conditions.
In Surnadal (Northwest Norway), a high population density and large mass of earthworms were found in an organic dairy farm from 1987 to 1989 (Hansen, 1996). Both, the population density and mass of earth-worms were strongly reduced by tractor traf®c in the loamy sand soil of this farm. A corresponding reduc-tion in earthworm populareduc-tions as a result of soil compaction has been found in many investigations (Aritajat et al., 1977b; BostroÈm, 1986; Pizl, 1992; SoÈchtig and Larink, 1992). However, when traf®c is removed, earthworm populations seem to recover quite quickly (Aritajat et al., 1977a; Haraldsen et al., 1994a). To our knowledge, no previous investigation has followed the effects of tractor traf®c on earthworm populations over many years in the same ®eld. There-fore, in this paper we present data on the long-term effects of tractor traf®c on earthworm populations in an organic dairy farm in Surnadal and the development of earthworms after the tractor traf®c ended.
Additions of organic matter in the form of farmyard manure have been found to increase earthworm
popu-lations under favourable soil conditions (Curry, 1976; Marshall, 1977; Andersen, 1979; Lofs-Holmin, 1983; Hansen, 1996). On the other hand, both cattle slurry and urine have been found to be transiently toxic to earthworms as a result of ammonia, benzoic acid and sodium sulphide content (Curry, 1976). The overall effects of long-term cattle slurry applications and treatments are, therefore, of interest. Aeration of slurry has been proposed as a strategy for, among other things, decreasing the toxicity of the slurry to earth-worms (Nebiker, 1974; Bauchhenss, 1981), but this does not always seem to succeed (Bieri and Besson, 1987; Hansen, 1996).
The ®ndings reported by Hansen (1996) are fol-lowed up in the present investigation by measurements of the long-term effects of manure management on earthworm species. We report the effects of ten years of tractor traf®c and fertilisation treatments on the population density, mass and species composition of earthworms in a loamy sand soil in a cool, wet district. The tractor traf®c treatments compared normal farm practice with reduced tractor traf®c. In the fertilisation treatments, three different cattle manure-management methods were compared. The cattle manure was applied as diluted slurry, as aerated slurry, or as separated manure with solid composted manure applied on tilled land and urine on ley. The manure was added at different levels and was compared with unfertilised and mineral fertiliser treatments.
2. Material and methods
2.1. The experimental site
The ®eld experiment was established in 1985 in Surnadal, Norway (638000N, 88880E) on an organic dairy farm. The previous crop rotation had a high proportion of ley, the manure was evenly distributed over the whole farm with small amounts applied each year, no chemical nitrogen fertilisers or pesticides were used, and the farmer had tried to avoid soil compaction as much as possible.
1991 to 1996 in Fig. 1. The soil was mostly covered with snow during the winter and frost never occurred for a long period when there was no snow layer.
The soil was a naturally drained loamy sand devel-oped on ¯uvial deposits (Typic Udorthents, USDA System of soil classi®cation). The average bulk den-sity at 7±11 cm depth was 1.25 g/cm3, with a pore volume of 52% (SE0.3) in 1985.
The soil is moderately rich in organic matter (organic carbon 2.1±2.2%, organic nitrogen 0.15± 0.17%) and low in potassium and magnesium.
2.2. Experimental design
A split-plot factorial design with two replicates was used, with tractor traf®c treatment on the main plots (288 m2) and fertilisation on the subplots (2.88 m2; sample area for yields 1.56.5 m2).
The trials were run for two ®ve-year periods. Han-sen (1996) preHan-sented details on crops, fertilisation, tractor traf®c, experimental design, climate, and sta-tistical analysis in the ®rst crop rotation (1985±1990). The second crop rotation period (1991±1996) fol-lowed the same principles as the ®rst, with some adjustments. The experimental site was the same throughout the investigation period and there was only one crop each year, as determined by the crop rotation. The crop rotation was: Year 1, green fodder con-sisting of rape (Brassica napus), barley (Hordeum vulgare), peas (Pisum arvense) and ®eld beans (Vicia faba); Year 2, barley with ley undersown; Years 3±5, ley containing timothy (Phleum pratense), red clover
(Trifolium pratense), white clover (T. repens) and alsike clover (T. hybridum). In 1990 (oats) and 1996 (ley), no fertilisation or tractor traf®c treatments were applied in order to test residual effects of previous treatments. In the present paper 1996 is called the after effect year.
The soil was ploughed in spring of 1985, 1986, 1990, 1991, and 1992.
In the ®rst crop rotation (Hansen, 1996) there were two ®eld trials, whereas in the second one, there was only one. The data presented here are only from the trial that was continued.
2.3. Tractor traffic
There were two tractor traf®c treatments: normal and low. In both these treatments, a tractor was used for soil tillage (ploughing, two harrowings, rolling) and sowing. These passes were made across the experimental plots with tractors similar to those used for the normal tractor traf®c treatment. Cereal harvest-ing was done with a lightweight experimental combine harvester. Green fodder and ley were cut with a two-wheel reaper and raked out of the ®eld by hand. This was the only traf®c in the treatment with low tractor traf®c.
In the normal tractor traf®c treatment, there was one additional pass of a tractor wheel-by-wheel once each spring, starting in 1986, and two passes shortly after each harvest. This was done all over the experimental plots in the normal tractor traf®c treatment, which means that each spot was covered ®ve times. Five Fig. 1. Monthly average temperatures (8C) and precipitation (mm), 1991±1996.
passes each year are comparable to normal dairy farm practice. From 1985±1989, a three-tonne tractor (2.5 tonnes on rear wheels, in¯ation pressure 150 kPa, tire width 32 cm) was used. From 1991 to 1995, the tractor weight was four tonnes. The tractor had dual settings and a total tyre width of 140 cm at the rear wheels (in¯ation pressure of 57 kPa). In front there were low-pressure tyres with a total width of 100 cm.
2.4. Fertilisation
The fertilisation levels (FL) one, two and three were intended to represent three different levels of farming intensity, with FL3 being the most intensive. FL1 corresponded to an organic farm where only the manure from the farm0s own cattle was used; at FL2, 35% more manure was added, while FL3 corre-sponded to the nutrients applied in conventional agri-cultural practice (Hansen, 1996). In addition, there was an unfertilised control (UNF).
There were three manure management methods: stored diluted slurry (DS), aerated slurry (AS), or separated manure [SM, solid composted manure on tilled land (1985, 1986, 1991, 1992) and diluted liquid on ley (1987±1989, 1993±1995)]. These methods were compared at FL1 and FL2.
DS was slurry, diluted with water to twice its original volume, shortly before use. The slurries were stored for four-to-six months prior to application in spring and after the ®rst harvest.
AS was cattle slurry aerated with a submersible pump. In the ®rst rotation (1985±1989), the aeration was done in a manure cellar (Hansen, 1996). In the second rotation, the slurry was aerated in a closed cylindrical container (radius and height2 m) with an ABS AGRO 75-4, 7.5 kW pump. The aeration started shortly after the container was ®lled. In the beginning, the pump was run continuously, but when the temperature reached 308C (after 5 days) the slurry was aerated for 10 min, fourteen times per day, for six weeks. The temperature was 28±328C in 1993 and 1994, and 19±258C in 1995. The AS used in spring was stored for only a few days after aeration, and the AS used after the ®rst herbage cut was stored for ca. two months.
SM was separated by free drainage in the stable in 1985 and from 1991 to 1996. From 1986 to 1989 a Key Dollar slurry separator was used (Morken, 1987) and
the liquid applied to ley from 1987 to 1989 came from the slurry separator. From 1993 to 1995, the liquid was diluted urine.
At FL3, DS was the only manure management. In addition, mineral fertiliser (18-3-15 from HYDROÐ 17.6% N as ammonium nitrate, 2.6% P and 14.6% K) was used either alone (NPK) or as DS plus NPK (DSN), where DS corresponded to FL1. NPK and DSN followed the fertiliser guidelines of conventional agriculture. In 1995 the NPK-fertilised treatments, but only those, were limed because of low soil pH. Manure samples were frozen (ÿ208C) for analyses, following the method described by (Horwitz, 1980).
Slurries were spread using cans (10 l) with a spray-ing plate, compost by dung fork, and mineral fertiliser by hand, spreading lengthways and crosswise. On ley, fertilisation was mostly done in cold and rainy weather.
At FL1 and FL2, the amounts of manure applied in the second crop rotation were only 70% of the amounts applied in the ®rst crop rotation period. This was in order to adjust the manure level to the common practice on organic farms in the area. Hansen (1996) gives the amounts of nitrogen, phosphorus and potas-sium applied in the ®rst crop rotation. The total amounts of manure, dry matter and nitrogen applied in the second crop rotation are presented in Table 1.
2.5. Sampling and measurements
Earthworms were sampled in ley years, after the second ley cut in 1987 (9 September) and after the ®rst cuts in 1988, 1989, 1993, 1995 and 1996 (end of June to beginning of July). One soil sample from a 10 20-cm2area was taken from each plot (0±15 cm depth in 1987, and 0±20 cm depth in 1988±1996). The earth-worms were hand sorted, washed, counted and weighed (fresh weight 1987±1989; after storage 1993±1996). At a depth of 15±20 cm, very few earth-worms were found in this shallow soil. In the whole investigation no anecic species were found. Therefore, there are likely to be very few earthworms below 20 cm. From 1993 on, the sampled earthworms were killed in 96% ethanol with 0.7% diethylphthalate, and stored either in 96% ethanol (1993 and 1995) or in 4% formalin (1996).
Table 1
Total amounts of manure (TM), manure dry matter (DM), total nitrogen (total N) and ammonium nitrogen (NH4-N) added per ha at different fertilisation levels and treatments
Year TA (tonne/ha) DM (kg/ha) Total N (kg/ha) NH4-N (kg/ha)
1991 1992 1993 1994 1995 total 1991 1992 1993 1994 1995 total 1991 1992 1993 1994 1995 total 1991 1992 1993 1994 1995 total
FL1 (mean of DS,AS,SM) 25 16 33 33 33 140 3112 1509 1635 1635 1326 9217 112 45 70 70 81 378 29 9 39 39 47 164 FL2 (mean of DS,AS,SM) 34 21 44 44 44 187 4164 2018 2183 2232 1773 12370 150 60 108 79 108 505 39 13 64 50 63 228
DS, mean of FL1 and FL2 35 28 59 59 59 239 2208 1247 2879 2925 2797 12056 95 40 116 88 117 455 58 17 64 53 64 256 AS, mean of FL1 and FL2 18 17 25 26 25 111 1373 1254 2195 2312 1179 8313 71 40 81 71 90 352 40 15 37 32 47 171 SM, mean of FL1 and FL2 36 11 31 31 31 140 7334 2789 653 618 673 12066 226 78 72 49 76 501 4 1 54 45 54 158
FL3, DS 70 44 80 140 140 475 4416 1936 3930 7015 6711 24008 189 62 128 210 281 870 116 26 64 126 154 487
FL3, NPK 137 83 120 209 176 725 73 44 63 110 93 383
FL3, DSN 1884 1065 2460 2500 2390 10299 160 68 125 209 203 765 91 33 64 116 109 413
species was carried out on stored material according to Sims and Gerard (1985). The presence of clitellum and/or tubercula pubertatis was used to classify the earthworms as adults or juveniles.
Soil physical determinations were made with low and normal tractor traf®c in selected fertilisation treatments. Soil porosity was determined in undis-turbed soil samples (7±11 cm depth) taken by means of four 100-cm3cylinders in autumn each year from 1985 to 1989 and in 1991, 1993, 1995 and 1996. Particle density, used to calculate material and pore volume, was determined according to De Boodt et al. (1967). Penetration resistance was measured in 1995 and 1996 with a fully automated penetrometer that followed the ASAE S13.2 standard (ASAE, 1987).
Soil samples (0±20 cm) for determinations of pH (H2O), ammonium-acetate lactate-soluble calcium,
phosphorous and potassium (Ca-AL, P-AL, K-AL, EÂgner et al., 1960), acid soluble potassium (K-HNO3, Reitemeier et al., 1951) and ignition loss, were
taken from all plots after the last harvest in autumn in 1985, 1986, 1987, 1988, 1989, 1990, 1993, 1995 and 1996. The organic carbon and total nitrogen concen-trations were determined in the soil samples from 1995 and 1996 with a Perkin±Elmer 2400 analyser.
2.6. Statistical analyses and calculations
Tractor traf®c and the interactions between tractor traf®c and fertilisation were tested with analyses of variance (ANOVA). The mean square difference of the interaction effect (replicatecompaction) was used as an error term to test the effect of compaction. Contrasts and least signi®cant differences were used to separate the effects of the different fertilisation treatments. Correlations of population density and mass of earthworms with soil pH, Ca-AL and ignition loss, organic carbon and total nitrogen in soil, soil porosity, amount of organic matter applied and ley yield, were tested with the Pearson correlation coef®cient, r. This analysis was carried out, both between years and within years and for low and normal tractor traf®c treatments separately. Pooled data from the ten fertilisation treatments within main plots were used for calculating between-year correlation coef®cients (n6), while analyses within each year were based on data for individual fertilisation treatments (n20).
3. Results
3.1. Earthworm populations
From 1987 to 1989, the population densities ranged from 50 to 1200 earthworms/m2, with a mean of 42030/m2 (SE); earthworm mass ranged from 8 to 360 g/m2, with a mean of 13010 g/m2(Table 2). There was a close correlation between the population density and mass of earthworms (r0.87,p<0.001). From the ®rst crop rotation (1987 to 1989) to the second (1993 to 1996), the population density and mass of earthworms decreased. The average total population density in the latter period was 26020/m2 and the average mass was 704 g/ m2. From 1993 to 1996, the earthworm species were determined.Aporrectodea caliginosaconstituted 67% of the population density and 78% of the mass. The corresponding ®gures were 25% and 19% for Lum-bricus rubellus, and 7% and 3% for Apporrectodea rosea. In addition, small numbers of Dendrodrilus rubidusandAllolobophora chloroticawere present.
The endogeic species Aporrectodea caliginosa,A. rosea and Allolobophora chlorotica dominated in 1993 and 1995, with 90% and 80% of the population density, respectively. In the aftereffect year (1996), when no cattle slurry was spread, the population densities of the epigeic species, L. rubellus and D. rubidus, increased and endogeic species decreased so that the epigeic species constituted 50% of the total.
3.2. Effect of tractor traffic
The population density responded strongly to trac-tor traf®c (p<0.001). An average of 500 earthworms/ m2was found with low tractor traf®c, compared with 180 earthworms/m2with normal tractor traf®c. The corresponding values for earthworm mass were 140 and 60 g/m2. The difference in earthworm populations between normal and low tractor traf®c treatments, was greatest at the beginning of the study period. In 1987 there was an average of 720 earthworms/m2after low, and 180/m2after normal, tractor traf®c. During the study period, the earthworm populations decreased in treatments with low tractor traf®c (Fig. 2). In 1993, the difference was still statistically signi®cant (p
The lower population density in the treatment with heaviest tractor traf®c coincided with denser soil and lower porosity (Fig. 3). Since in¯ation pressure in tractor wheels was lower in the second crop rotation,
the soil was less dense in the normal tractor traf®c treatment in the second crop rotation than in the ®rst one (Fig. 3). There was a close correlation between the volume of pores and the population density in 1988 Table 2
Earthworm densities (numbers/m2) and mass (g/m2) in subplots receiving different fertilisation treatments
Fertilisation treatments Mean 1987±1989 1993 1995 1996
Earthworm numbers No./m2 SE
Control 45080 27070 6030 130 40
FL1 (mean of DS,AS,SM) 47060 29030 17040 190 30
FL2 (mean of DS,AS,SM) 43050 41060 18030 270 39
DS, mean of FL1 and FL2 41070 33060 22050 330 40
AS, mean of FL1 and FL2 45070 30050 20040 210 40
SM, mean of FL1 and FL2 48070 42070 11030 150 40
DS, FL3 36090 620120 9050 510110
NPK, FL3 29050 21030 22050 140 50
DSN, FL3 430104 430150 29020 190 40
Mean 42030 36030 17020 23030
Earthworm mass(g/m2SE)
Control 9020 6020 20 10 3010
FL1 (mean of DS, AS, SM) 15020 8010 4010 6010
FL2 (mean of DS, AS, SM) 14020 10020 5010 7010
DS, mean of FL1 and FL2 13020 9020 6020 8010
AS, mean of FL1 and FL2 13020 8010 4010 6010
SM, mean of FL1 and FL2 17020 10020 3010 4010
DS, FL3 10020 20020 5030 10010
NPK, FL3 6020 5010 8030 4020
DSN, FL3 13030 11030 5010 5020
Mean 13010 10010 5010 6010
Note: In 1996 no fertilisation was carried out. The data are presented as means of normal and low tractor traffic treatments. The fertilisation treatments are given in Table 1.
Fig. 2. Population density (numbers/m2) and mass (g/m2) of earthworms during the period 1987±1996 in plots with normal, and low, tractor traffic. The data are presented as means of all fertilisation treatments,n20. Vertical lines show SE.
(r0.74,p <0.001), whereas in 1995 there was no signi®cant correlation. In 1995, the mean penetrom-eter resistance (1±20 cm soil depth) was 0.61 C.I. MPa in the low tractor traf®c treatment and 0.89 in the normal (p <0.02). The soil was also harder in the aftereffect year 1996, in the treatments with normal tractor traf®c (0.72 vs. 0.5 C.I.MPa,p<0.02).
The earthworms were heavier in soil with normal than with low compaction (p<0.02). This was mainly due to heavier juvenile earthworms. Tractor traf®c did not have any signi®cant affect on species composition.
3.3. Effect of fertilisation treatments
3.3.1. Effect of mineral fertiliser
In total, there were fewer earthworms in soil ferti-lised with mineral fertiliser only (NPK-fertiferti-lised) than in manured soil (Table 2,p<0.002), but the liming of
NPK-fertiliser treatments in spring 1995 led to an immediate increase in earthworm populations in these treatments (Fig. 4).
3.3.2. Effect of increased manure application
Increasing the manure application from FL1 to FL2 increased the population density (p <0.02) and the mass (p<0.01) of endogeic but not of epigeic species (Fig. 5). Therefore, for all earthworms together, the increases in the population density and mass from FL1 to FL2 were not statistically signi®cant (Table 3). A further increase in the manure application in the diluted slurry management (DS) increased both, the total population density and mass of earthworms in 1988, 1993 and 1996. This increase was due to an increased endogeic population in 1993 (Fig. 6). In 1989 and 1995, there was a tendency towards fewer earthworms with increasing manure level.
3.3.3. Cattle slurry vs. separated manure application
The total population density of earthworms was not signi®cantly affected by whether the manure was treated as slurry or separated with solid composted manure applied to tilled land and urine to ley. Since most of the organic matter under separated manure treatments was added with the solid manure in tillage years (1985, 1986, 1991, 1992), there was less food from the added manure for the earthworms late in the ley period. In the years following the application of the solid manure (1987, 1988, 1993), there was a tendency towards more earthworms where separated manure was applied than in the plots that received slurry, but signi®cantly fewer earthworms late in the second crop rotation period (1995, 1996) (Table 2 and 3). From 1993 to 1996, in the separated manure treatment, there was a large decrease in endogeic species (Fig. 7), but not in the epigeic species.
Fig. 3. Soil porosity (% v/v) in the period 1985±1996 in plots with normal, and low, tractor traffic. The data are presented as means of fertilisation treatments,n10. Vertical lines show SE.
Table 3
The probability level (p) for significant differences in population density and mass of earthworms between fertilisation level one and two (FL1, FL2), diluted and aerated slurry (DS, AS) and separated manure vs. slurry (SM-DS, AS), with the fertilisation treatments given in Table 1
Population density Mass
1987 1988 1989 1993 1995 1996 1987 1988 1989 1993 1995 1996
Contrast DS±AS 0.4 0.9 0.6 0.7 0.6 0.05 0.9 0.6 1 0.5 0.3 0.3
3.3.4. Aerated vs. diluted cattle slurry
Whether the slurry was aerated or diluted did not signi®cantly affect either the total population density or mass of earthworms in the treatment years (1987± 1995, Table 3). In 1993 and 1995, there was a
sig-ni®cantly lower mass of epigeic (p<0.05) but not of endogeic species of earthworms with aerated slurry than with diluted slurry (Fig. 7). Less organic matter was added with aerated slurry than with diluted slurry. From 1991 to 1995, the mean addition of organic Fig. 4. Earthworm mass (g/m2) and content of readily available calcium (Ca-AL) in soil without fertilisation (control), with NPK-fertilisation and with diluted slurry applied at fertilisation levels one, two and three (DS1, DS2 and DS3), 1987±1996. The earthworm mass is given for normal, and low, tractor traffic treatments and the content of Ca-AL only for the low tractor traffic treatment,n2. Vertical lines show SE. The fertilisation treatments are given in Table 1.
matter in the FL1 and FL2 diluted slurry management was 240 g/m2(dry matter basis) vs. 170 g/m2in the
aerated management. This is a result of the breakdown of organic matter during composting of the aerated slurry.
3.4. Changes and correlations during the treatment period
During the treatment period there was a decline in earthworm populations with low tractor traf®c in all fertilisation treatments, except high additions of diluted slurry (DS3) (Fig. 4).
Simultaneously with this decline, acidi®cation of soil in the research ®eld occurred and both, the soil pH and Ca-AL were quite low at the end of the experiment (Fig. 4). In treatments with low tractor traf®c, the mass of earthworms was closely correlated with soil pH (r0.95,p<0,004)andCa-AL (r0.96,p<0.003), but not with ignition loss. There was also a close correlation with herbage yields (r0.92, p <0.02), which declined during the study period. With normal tractor traf®c there were no signi®cant correlations. Fig. 5. Mass of endogeic and epigeic earthworm species (g/m2) at
fertilisation levels one and two (FL1 and FL2) in the treatment years 1993±1995 and the aftereffect year 1996 with normal, and low, tractor traffic. The data are presented as means of the manure treatments AS, DS and SM (aerated or diluted slurry or separated manure),n6. The fertilisation treatments are given in Table 1.
Fig. 6. Mass of endogeic and epigeic earthworm species (g/m2) without fertilisation (control) and with diluted slurry applied at fertilisation levels one, two and three (DS1, DS2 and DS3) in the treatment years 1993 and 1995, and the aftereffect year 1996. The data are presented as means of normal and low tractor traffic,
n4. The fertilisation treatments are given in Table 1.
3.5. Residual effects of previous treatments
Correlation analysis showed that the amount of dry matter added with manure in 1991±1995 was the environmental factor that was most closely correlated with the total population density (r0.61,
p <0.001) and mass (r0.49, p <0.002) of earth-worms in the aftereffect year (1996). The main reason for this was the close correlation between the previously added manure and the population density and mass of juvenile earthworms (r0.61,
p <0.001 and r0.50, p <0.001). Previously added manure was also correlated with the population density of A. caliginosa (r0.44, p <0.005) and
L. rubellus (r0.54, p <0.001). Mass was not sig-ni®cantly correlated with previously added manure forA. caliginosa, but forL. rubellus, it was (r0.50,
p<0.001).
The total population density of adult earthworms was not signi®cantly correlated with previously added manure, but was correlated with soil pH (r0.33, p <0.05). For A. rosea, the population density of earthworms was correlated with soil pH (r0.39, p <0.05), but not with previously added manure. The total population densities of A. caliginosaand L. rubellus were not correlated with soil pH.
Neither the mass nor the population density of any component of the earthworm population was corre-lated with ignition loss, organic carbon or total nitro-gen in 1996.
4. Discussion
4.1. Earthworm populations
The number of earthworms found in the present ®eld trial are relatively high compared with other Nordic sites (Andersen, 1979; Andersen et al., 1983; BostroÈm, 1986; Christensen et al., 1987, Nuu-tinen, 1992; Haraldsen et al., 1994a, b). All the present species are commonly found in Norwegian soil (Har-aldsen et al., 1994b). Their ecological traits ensure their survival in spite of regular ploughing. A combi-nation of agricultural practice and occasional water-logging of the soil may cause the absence of anecic species.
4.2. Effects of tractor traffic
The destructive effects of tractor traf®c on earth-worms are not surprising, and agree with the result of other investigations (e.g. BostroÈm, 1986; SoÈchtig and Larink, 1992).
The very destructive effects of tractor traf®c on earthworms and soil porosity in the present investiga-tion probably occurred because, in this cool and wet climate, the soil often is moist and thus vulnerable to damage. In addition, the structure of this loamy sand easily deteriorates (Hansen et al., 1993). The tractor traf®c also led to large yield declines (Hansen, 1996). We expected tractor traf®c to be more harmful to epigeic species than to endogeic species, as found by Pizl (1992), since earthworms near the surface are more vulnerable to direct pressure from the tractor wheels. This could not be con®rmed in the present investigation, however. One reason might be the high density and penetration resistance in the soil with normal tractor traf®c in the present investigation. In dense soil, the endogeic earthworms have to eat their way through the soil and cannot push it aside as easily as they can in looser soil (Dexter, 1978). As the tractor traf®c treatment lowered the soil porosity most in the ®rst crop rotation period (Fig. 3), the very harmful effect of tractor traf®c on endogeic earthworms found in the second crop rotation period could be a conse-quence of the ®rst crop rotation period.
The lack of a statistically detectable effect of pre-vious tractor traf®c on the earthworm population, a year after the last tractor traf®c (1996), is in accor-dance with the results of Aritajat et al. (1977a) and Haraldsen et al. (1994a).
4.3. Effect of fertilisation treatments
4.3.1. Effect of mineral fertiliser
Our ®nding of fewer earthworms in soil fertilised with mineral fertiliser only (NPK-fertilised), than in manured soil agrees with the results of other investi-gations (Lofs-Holmin, 1983; Haraldsen et al., 1994a). The direct adverse effects of mineral fertilisers on earthworm populations seem to be most pronounced in light soils and with the nitrogen applied in acidify-ing compounds (Lofs-Holmin, 1983; Ma et al., 1990). In the present investigation, the soil is light (sandy loam) and the nitrogen fertiliser has an acidifying
effect (NH4NO3). The likelihood of a direct adverse
effect of NPK-fertiliser on earthworms is indicated by the fact that, in the ®rst part of the investigation period (1987±1993), there was a tendency for fewer earth-worms in NPK-fertilised than in unfertilised soil (Table 2). However, this tendency was signi®cant only in 1989.
Some investigators has found a positive long-term effect of adding mineral nitrogen (Edwards and Lofty, 1982; BostroÈm, 1986). This is likely to be the result of increased crop growth and, thus, more organic mate-rial added to the soil. Liming of the soil in NPK-fertilised treatments may explain why there were no differences between earthworm populations in NPK-fertilised and manured treatments in 1995.
4.3.2. Effect of increased manure application
The signi®cant correlation between the amount of organic matter added with manure and the population density of earthworms in the aftereffect year (1996) shows the importance of manure as a source of food for earthworms, as previously observed, for example, by Uhlen (1953); Edwards and Lofty (1977); Lofs-Holmin (1983) and Standen (1984).
The reason for a tendency towards fewer earth-worms with increasing manure level in 1989 and 1995 may have been a toxic effect of cattle slurry on earthworms, as observed by (Curry, 1976). The fact that many dead earthworms were observed on the soil surface shortly after the slurry was spread in those years and the large increase in epigeic species in the aftereffect year 1996, supports this hypothesis.
4.3.3. Cattle slurry vs. separated manure application
A decrease in the earthworm population late in the ley periods in the separated manure treatment, when only the liquid fraction was being applied, can par-tially be explained by the fact that the effect of solid manure (FYM) does not last very long (Lofs-Holmin, 1983). This seems most evident for the endogeic species (Fig. 7). Because endogeic species feed mostly on mineral and humus particles within the soil, whereas epigeic species feed on plant residues and microorganisms at the surface, this is reasonable.
4.3.4. Aerated vs. diluted slurry
The main reason we did not get a positive effect of aerating the slurry on earthworm density, is likely to
be reduced availability of food for the earthworms in the aerated slurry treatment compared with the diluted slurry. This hypothesis is supported by the fact that, in the aftereffect year (1996), there were fewer earth-worms (p <0.05) and less earthworm mass in the aerated that in the diluted slurry treatment. Haraldsen et al. (1994b) also found a tendency towards fewer earthworms in the aftereffect year in the treatments previously manured with aerated compared with diluted slurry. This is consistent with the ®eld results of Bieri and Besson (1987).
In addition, it seems that we did not achieve reduced toxicity of the cattle slurry as a result of aeration as Nebiker (1974) and Bauchhenss (1981) did following more intense aeration which reduces the ammonia content; rather, a tendency to the opposite. This is indicated by the fact that, in plots where aerated slurry was applied, there was a lower mass of epigeic species than in plots with diluted slurry.
4.4. Reasons for the decline in earthworm
populations in treatments with low tractor traffic
The close correlations between population density and mass and soil pH and Ca-AL that were observed with the low tractor traf®c treatments suggest that acidi®cation of the soil is an important reason for the reduction in earthworm populations in these treat-ments. Earthworm species differ in acid tolerance, and the tolerance seems to vary with environmental conditions. Lofs-Holmin (1986) found thatA. caligi-nosa, A. rosea and L. rubellus, which were the dominant species in the present investigation, all were quite acid-tolerant. On the other hand, Lofs-Holmin (1983) found increased population density ofA. cali-ginosa in a sandy soil when the pH was increased from 5 to 6.6. Satchell (1955), Piearce (1972) and NordstroÈm and Rundgren (1974) also found positive correlations between pH and mass ofA. caliginosa, as did Standen (1984) for the total earthworm mass in a meadow.
(1985±1990), may also be an important reason for the earthworm decline.
Changing weather conditions theoretically could have been a reason for the decline in earthworm populations. As the soil was covered with snow during cold periods, no drought occurred during the investi-gation period and the earthworm sampling (except 1987) was done mostly at the same time each year, this is not likely to be a major factor in the earthworm decline. Neither should the crop rotation be important, as the samplings were done in ley in both the crop rotations with the rotations being very similar.
5. Conclusions
When other conditions were favourable for earth-worms, the near absence of tractor traf®c led to a high population density of earthworms in this loamy sand. The highest population density and mass of earth-worms were found in the manured soil. Increasing the amount of cattle slurry and urine decreased the popu-lation density and mass of earthworms transiently in some years, but not all. Aerated slurry application did not result in a higher earthworm population compared with diluted slurry.
Both, tractor traf®c and soil acidi®cation were found to depress the earthworm populations in this soil. The dominant earthworm species were the endo-geic speciesAporrectodea caliginosaand the epigeic speciesLumbricus rubellus.
Acknowledgements
We would like to thank K. Myhr for suggesting the earthworm investigations, J. Andersen for penetrom-eter measurements, B.R. Ertvaag for his assistance during the research period, A.-K. Lùes and V. Nuu-tinen for valuable comments and Professor W. Lock-eretz for valuable comments and language correction.
References
Andersen, C., 1979. The influence of farmyard manure and slurry on the earthworm population (Lumbricidae) in arable soil. In: Dindal, D.L., (Ed.) Soil Biology as Related to Land Use
Practices. Proc. 5 Int. Coll. Soil Zool. in Syracuse, US, pp. 325± 335.
Andersen, C., Eiland, F., Vinther, F.P., 1983. ékologiske under-sùgelser af jordbundens mikroflora og fauna i dyrkningssyste-mer med reducert jordbehandling, vaÊrbyg og efterafgrùde. Tidsskr. Planteavl 87, 257±296.
Aritajat, U., Madge, D.S., Gooderham, P.T., 1977a. The effects of compaction of agricultural soils on soil fauna. Pedobiologia 17, 262±282.
Aritajat, U., Madge, D.S., Gooderham, P.T., 1977b. The effects of compaction of agricultural soils on soil fauna 2. Laboratory investigations. Pedobiologia 17, 283±291.
ASAE, 1987. ASAE Standards 1987. 34th en. 2950 Niles Rd. St. Joseph, MI, USA.
Bauchhenss, J., 1981. Wirkung beluÈfteter und unbeluÈfteter GuÈlle auf die Regenwurmfauna. In: Bericht uÈber die 7. Arbeitstagung Fragen der GuÈllerei vom 29. Sept bis 2. Okt. 1981 in Gumpenstein. A-8952 Irdning, pp. 739±748.
Bieri, M., Besson, J.M., 1987. Der Einfluss unterschiedlich aufbereiteter GuÈllen auf die Regenwurmpopulation einer Kunstwiese. Schweiz. Landw. Fo. 26, 499±506.
BostroÈm, U., 1986. The effect of soil compaction on earthworms (Lumbricidae) in a heavy clay soil. Swed. J. Agric. Res. 16, 137±141.
BoucheÂ, M.B., 1977. StrateÁgies lombriciens. In: Lohm, U., Person, T., (Eds.), Soil Organisms as Components of Ecosystems, Stockholm, pp. 122±132.
Christensen, O., Daugbjerg, P., Hinge, J., Jensen, J.P., Sigurda-dottir, H., 1987. Effekten af dyrkningspraksis paÊ regnorme og deres mulige rolle som bioindikatorer. Tidsskr. Planteavl 91, 15±32.
Curry, J.P., 1976. Some effects of animal manures on earthworms in grassland. Pedobiologia 16, 425±438.
De Boodt, M., De Leenheer, L., Frese, L., Low, A.J., Perkamp, P.K., 1967. Determination of the particle density according to the alcohol-method. In: West-European Methods for Soil-Structure Determinations. The West-European Working Group on Soil Structure of the ISSS (en.), The State Faculty of Agricultural Science, Ghent.
Dexter, A.R., 1978. Tunnelling in soils by earthworms. Soil Biol. Biochem. 10, 447±449.
Edwards, C.A., Lofty, J.R., 1977. Biology of Earthworms, second en. Chapman and Hall, London, pp. 333.
Edwards, C.A., Lofty, J.R., 1982. Nitrogenous fertilizers and earthworm populations in agricultural soils. Soil Biol. Bio-chem. 14, 515±521.
EÂgner, H., Riehm, H., Domingo, W.R., 1960. Untersuchungen uÈber die chemische Boden-Analyse als Grundlage fuÈr die Beurtei-lung des NaÈhrstoffzustandes der Boden, Kungl.. Lantbruk-shoÈgskolans annaler 26, 199±215.
Hansen, S., 1996. Effects of manure treatment and soil compaction on plant production of a dairy farm system converting to organic farming practice. Agric. Ecosyst. Environ. 56, 173± 186.
Hansen, S., Maehlum, J.E., Bakken, L.R., 1993. N2O and CH4 fluxes in soil influenced by fertilization and tractor traffic. Soil Biol. Biochem. 25, 621±630.
Haraldsen, T.K., Engelstad, F., 1994. Meitemark i Nord-Norge. Norden, pp. 24±25.
Haraldsen, T.K., Larsen, M.A., Myhr, K., 1994a. Effects of cattle slurry and soil compaction on the earthworm population in a silty clay loam soil in central Norway. Norw. J. Agric. Sci. 8, 231±241.
Haraldsen, T.K., Sveistrup, T.E., Engelstad, F., 1994b. Jordegenskaper og meitemark i leirjord ved omlegging til ùkologiske dyrkings-systemer i Norge. Norsk Landbr. Forsk. Suppl. 17, 1±42. Horwitz, W., 1980. Official Methods of Analyses, 13th en.
Association of Official Analytical Chemists, Washington, DC. Lavelle, P., 1988. Earthworm activities and the soil system. Biol.
Fertil. Soils 6, 237±251.
Lee, K.E., 1985. Earthworms, Their Ecology and Relationships with Soils and Land Use. Academic Press, Sydney, pp. 411. Lofs-Holmin, A., 1983. Influence of agricultural practices on
earthworms (Lumbricidae). Acta Agric. Scand. 33, 225±234. Lofs-Holmin, A., 1986. Occurence of eleven earthworm species
(Lumbricidae) in permanent pastures in relation to soil-pH. Swed. J. Agric. Res. 16, 161±165.
Ma, W.C., Brussaard, L., de Ridder, J.A., 1990. Long-term effects of nitrogenous fertilizers on grassland earthworms (Oligochae-ta: Lumbricidae): their relation to soil acidification. Agric. Ecosyst. Environ. 30, 71±80.
Marshall, V.G., 1977. Effects of manures and fertilizers on soil fauna: a review. Special Publication, Commonwealth Bureau of Soils No. 3, Commonwealth Agricultural Bureaux, Slough, England, pp. 79.
Morken, J., 1987. Forsùk med Key Dollar Model 150 gjùdselse-parator. LTI-trykk, Landbruksteknisk Institutt, N-1432 AÊ s, 51, pp. 1±7.
Nebiker, H., 1974. Neues Verfahren zur aufbereitung von FluÈssigduÈnger. Schweiz. Landw. Monatsh. 2, 1±32.
NordstroÈm, S., Rundgren, S., 1974. Environmental factors and lumbricid associations in southern Sweden. Pedobiologia 14, 1±27.
Nuutinen, V., 1992. Earthworm community response to tillage and residue management on different soil types in southern Finland. Soil Tillage Res. 23, 221±239.
Piearce, T.G., 1972. Acid intolerant and ubiquitous Lumbricidae in selected habiatas in North Wales. J. Anim. Ecol. 41, 397±410. Pizl, V., 1992. Effect of soil compaction on earthworms (Lumbricidae) in apple orchard soil. Soil Biol. Biochem. 24, 1573±1575.
Reitemeier, R.F., Brown, I.C., Holmes, R.S., 1951. Release of native and fixed nonexchangeable potassium of soils containing hydromica. U.S. Dept. Agric. Techn. Bull. No. 1049. Satchell, J.E., 1955. Some aspects of earthworm ecology. In:
McKevan, D.K. (Ed.) Soil Zoology. Butterworths, London, pp. 180±201.
Sims, R.W., Gerard, B.M., 1985. Earthworms. Synopses of the British Fauna (New Series), The Linnean Society of London, London, pp. 157.
Standen, V., 1984. Production and diversity of enchytraeids, earthworms and plants in fertilized hay meadow plots. J. Appl. Ecol. 21, 293±312.
SoÈchtig, W., Larink, O., 1992. Effect of soil compaction on activity and biomass of endogeic lumbricids in arable soils. Soil Biol. Biochem. 24, 1595±1599.
