Electronic Journal of Polish Agricultural Universities (EJPAU) founded by all Polish Agriculture Universities presents original papers and review articles relevant to all aspects of agricultural sciences. It is target for persons working both in science and industry,regulatory agencies or teaching in agricultural sector. Covered by IFIS Publishing (Food Science and Technology Abstracts), ELSEVIER Science - Food Science and Technology Program, CAS USA (Chemical Abstracts), CABI Publishing UK and ALPSP (Association of Learned and Professional Society Publisher - full membership). Presented in the Master List of Thomson ISI.
Volume 8
Issue 3
Ulrich W. , Czarnecki A. , Paprzycka I. 2005. EARTHWORM ACTIVITY IN SEMI-NATURAL AND FARMLAND SOILS, EJPAU 8(3), #13.
Available Online: http://www.ejpau.media.pl/volume8/issue3/art-13.html


Werner Ulrich1, Adam Czarnecki2, Iwona Paprzycka2
1 Department of Animal Ecology, Nicolaus Copernicus University in Torun, Poland
2 Laboratory for Landscape Planning and Ecology, Nicolaus Copernicus University in Torun, Poland



Earthworms are keystone species in the process of soil formation. The present paper reports on comparative studies of earthworm activity at a semi-natural park site and a typical farmland in Northern Poland. Of the six species found, Aporrectodea caliginosa, A. rosea, Allolobophora chlorotica, and Lumbricus terrestris occurred at all sites. The epigeic species L. casteneus and D. octaedra were missing at the farmland sites. Least species rich was the farmland site with the most condensed soil where only A. caliginosa, A. chlorotica, and L. terrestris were present. The farmlands showed reduced earthworm activities as measured by cast production and macropore density with respect to the natural park site. Cast production in the farmlands appeared to be independent of soil compactness and relative humidity. Soil compactness above 1.7 Mg · m-2 hampered earthworm activity. The total cast weight was inversely correlated with soil pH. The two most alkalic sites had significantly lower total cast weights per m2 than the park site.

Key words: Lumbricidae, agriculture, earthworm casts, macropores, soil compactness.


Ever since Darwin [7], the influence of earthworms (Lumbricidae) on soil formation has been in the centre of attention of soil biologists. In temperate and tropical soils earthworms are regarded as keystone species and play a major role in the decomposition of organic material [9,22] and soil formation [12]. Their activity explains at least some part of the 'enigma of soil animal species diversity' [1,2,21]. Earthworm diversity and activity depend not only on soil chemistry and composition, but also on climatic and biotic factors [18]. An additional factor in anthropogenic soils, like farmlands, is the degree of anthropopressure [10].

Structure and fertility of farmland soils depend to a major extent on the type of farming. Reduced contents of organic materials [8] and intense farming cause soil compaction [4]. Further, Bakken et al. [3] emphasized that enhanced soil compaction is one of the major factors that induce subsequent degradation of farmland soils. Soil compaction alters the chemical and physical soil properties in a direction that hampers earthworm development and activity. Therefore an initial anthropogenic induced altering of soil properties might induce a cascade of effects. This leads to the hypothesis that anthropogenic influences on soil conditions might be a threshold process. Below a certain threshold human activity is counterbalanced by self-regulatory mechanisms that invoke the soil macrofauna. Above the threshold the alteration of conditions itself results in reduced macrofaunal (earthworm) activities and intensifies the negative properties. A spiral into an undesired direction starts.

To verify this hypothesis one has to know to what extent earthworms are involved in the process of soil formation in the agricultural landscape and above what level of soil compactness their activity is suppressed. The first step towards this knowledge is to establish the abundance of earthworms in farmlands of different soil types and to infer the dependence of earthworm abundance and function depending on human induced changes of soil properties.

The present study aims to describe the occurrence of earthworms in farmland soils analyzing the production and chemical composition of casts, and the amount of soil macropores in soils of different human impact.


In 1998 and 1999 two different sites were studied at the Agricultural University Field Stations at Koniczynka near Toruñ (central northern Poland). The first habitat was an about 100-year-old park (P) without any recent and past farming activities. The second study site was a typical farmland with three-year crop rotation (sugar beet, winter wheat, spring wheat). On this farmland there were studied four sites (T, G, Z, ZZ) of 24 x 50 m2 each with different degrees of soil compaction, humidity and pH (Table 1). The park was characterized by the least condensed soil, the highest amounts of organic C and N, a higher C/N ratio, and the lowest pH (Fig. 1). Instead mean soil humidity was quite similar across the sites. All four farmland soils were alkalic. To level out the possible effect of different weather conditions during the sampling, the following results are based on total yearly counts.

Table 1. Basic soil parameters of the five study sites (means of 1998 and 1999)


Soil density

Humidity, %

C org, %

N org, %






































P - park, T - typical field, G - moderately condensed soil, Z - condensed soil, ZZ - highly condensed soil

Earthworm cast was sampled from March to October 1998 and 1999 according to the method of Norgove et al. [16]. In 1998 the farmland sites were planted with spring wheat, in 1999 - with winter wheat. 10 randomly chosen plots per site (in the park within the phytosociologically most representative part), 0.5 x 0.5 m2 each, were screened by hand up to 30 cm deep (in total 320 probes per site and per year). The cast was sampled once a week. In one week after heavy rains two samples were taken. From these data the total yearly time span of cast deposition was inferred. The cast weight (g·m-2) was determined after 24h drying at 105°C. Organic C contents were determined with the method of Tiurin, N-contents with the Kjeldahl method.

Fig. 1. Park site

Soil macropores of the mull layer (0.3 x 0.3 m2) were analyzed according to the methods of McKenzi & Dexter [15] and Lightart et al. [14]. In line with common praxis [9,13], circular pores between 1 and 10 mm long with smooth walls were counted as being made by earthworms.

There were used non-parametric Kruskal-Wallis ANOVA by ranks and the U-test to infer differences across the five sites. Spearman's rank order correlation served to infer pairwise dependencies. All statistical analyses were carried out with the Statistica 5 package (Statsoft 1997).

Evenness was measured with the index of Simpson [E = n/(n-1)(1-∑p2)] with n being the sample size as this index is least dependent on the sample size [17].


Six species of earthworms were found at the study sites (Table 2). Aporrectodea caliginosa, A. rosea, Allolobophora chlorotica, and Lumbricus terrestris occurred at all sites. At T, G, and Z the epigeic species L. casteneus and D. octaedra were missing. It was ZZ which was least species rich; A. caliginosa, A. chlorotica, and L. terrestris were present only. Evenness was highest in the park (E = 0.49 ± 0.07), whereas the farmland sites were significantly less even (E = 0.13 to 0.19; p(t) < 0.001). All the farmland sites were dominated by A. caliginosa.

Table 2. Species composition and frequencies (%) of Lumbricidae recorded at the five study sites








Aporrectodea caliginosa







Allolobophora chlorotica







Aporrectodea rosea







Lumbricus terrestris







Lumbricus castaneus







Dendrobaena octaedra







Abbreviations, as in Table 1

In the park the total weight of the earthworm casts was in both years significantly higher than at all four farmland sites (P(U) < 0.001) (Fig. 2). The total annual cast weight in the park was in both years > 1 kg m-2a-1. The four farmland sites showed the total cast weights between 1 and 486 g · m-2a-1. No clear dependencies of cast weight on C/N ratio, soil humidity, or soil compactness appeared. However, a strong correlation was found between cast weight and soil pH although this result might be a chance product due to the small number of data points.

Fig. 2. Effect of basic soil parameters on mean earthworm cast weight. Error bars denote one standard error of weight. Correlation in? A: R2 = 0.92, p < 0.01

The carbon and nitrogen contents of the casts were significantly higher than those of the surrounding soil (Fig. 3). In case of carbon, this accumulation was more pronounced in the park soil with its also higher C soil content (Fig. 3 A). On the contrary, the farmland sites accumulated relatively more N with respect to the soil N content (Fig. 3 B). The differences in the total cast weight across the five sites caused also significantly higher total amounts of C and N deposition in the park, in comparison to the field sites. The variability within the park plots was higher than within each of the four farmland sites, although this difference is significant only for carbon (p < 0.05). This higher variability was most pronounced when referring to the C/N ratios (Fig. 3 C). The C/N ratios of casts of the four farmland sites were about 60 to 75% lower than those of the surrounding soil. In the park the cast C/N ratio was only 38% lower.

At the farmland sites the pH values of the casts were similar to those of the surrounding soil. Instead in the park the pH values of the casts were on average about 0.4 degrees above the soil pH (Fig. 4).

Fig. 3. Quotients of carbon (A) and nitrogen (B) contents of casts and surrounding soil depending on soil C and N (measured in %). In C the ratio of C / N (casts) through C / N (soil) is plotted against C / N (Soil). P: Park. Errors denote one standard error

Fig. 4. Cast pH in relation to the soil pH. P: Park; T: typical farmland; Errors denote one standard error

Similar mean densities of macropores were found in the park and at the four farmland sites. However, as in case of the cast weights, the variance within the park plots was significantly higher than within the farmland plots (p(t) < 0.01). Instead no differences were found across the farmland sites and there was no correlation detectable between the site variance and soil density. In both habitat types earthworm macropores accounted for about 2 to 9% of the total soil volume with a mean of 5.7 ± 2% ( ± s).


Naturally soils in temperate grasslands or forests have about four to eight earthworm species [9,12]. Polish natural deciduous forests contain about 6 to 15 species [20]. Therefore with respect to species richness, the present results do not deviate from those expected although the park site appears to be at the lower boundary of species richness. Agricultural landscapes instead are depauperate and contain only one to six species. In most farmlands especially nightcrawlers (Lumbricus terrestris) and epigeic species like L. rubellus or Eisenia foetida are missing [12]. The present results are in line with this general pattern.

The higher species richness in the park caused also higher casts weights. However, macropore densities did not differ across park and farmland sites. These seemingly contradictory results are explained by the changes in the community structure. In the farmlands the endogeic Aporrectodea caliginosa dominated with a relative abundance of more than 90%. Instead the anecic and epigeic species that count for a large part of the cast production, but to a lesser degree for macropore formation, were largely missing.

The present results do not unequivocally point to soil compactness as a major factor that influences earthworm species richness. Compactness was neither correlated with species richness, nor with cast weight and macropore density. Of course compactness influences other soil properties like acidity or relative humidity. However, the four farmland sites had similar species numbers irrespective of differences in humidity, pH and organic element contents. Hence earthworms tolerate moderate and even high soil compactness up to 1.7 Mg · m-2. Only in the even higher compacted soil of the ZZ site (1.79 Mg · m-2), were diversity and activity reduced.

The present results favour the interpretation that pH might be of major importance in structuring earthworm communities. It has long been known that low pH values are unfavourable for earthworms and cause a decrease in species numbers [9,11,12,18]. In acid soils only one to six acid tolerant species like Allolobophora clorotica or Dendrobaena octaedra occur [6]. However, the negative correlation between cast weight and pH reported here shows that too high soil pH values also reduce earthworm activity. This result is in line with the findings of Bouché [5] who reported that the vast majority of earthworms favour soils below pH of 7.4.

An unexpected result was the higher cast weight variability of the park site with respect to the farmlands. It seems that the park soil exhibits a clearly small scale heterogeneity not visible from the plant composition. However, this higher variability was seen for pH, C- and N- contents, the C/N ratio (Fig. 3), and soil humidity and compactness.


  1. Farmlands show reduced earthworm activities as measured by cast production and macropore density, as compared with a more natural park site.

  2. Cast production at the farmland sites appeared to be independent of soil compactness and relative humidity.

  3. Soil compactness above 1.7 Mg·m-2 hampered earthworm activity.

  4. Soil pH was inversely correlated with the total cast weight. The two most alkalic sites showed significantly lower total cast weights per m2 than the park site.


Miss Hazel Pearson kindly improved our English. This work was in part supported by a grant from the Committee for Scientific Research to WU (KBN, 3 F04F 034 22).


  1. Anderson J.M., 1975. The enigma of soil animal species diversity. [In:] Progress in Soil Zoology. Ed. J. Vanek. Akademia Press Prague, 51-58.

  2. Andrén O., Balandreau J., 1999. Biodiversity and soil functioning - from black box to can of worms? Appl. Soil Ecol. 13, 105-108.

  3. Bakken L.R., Børrsen T., Njøs A., 1987. Effect of compaction by tractor traffic on soil structure, denitrification, and yield of wheat. J. Soil Sci. 38, 541-552.

  4. Boström U., 1996. The effect of soil compaction on earthworms (Lumbricidae) in a heavy clay soil. Swedish J. Agric. Res. 16, 137-141.

  5. Bouché M.B., 1972. Lombriciens de France. Écologie et systématique [Lumbricidae in France. Ecology and systematics] INRA Publ. Paris.

  6. Bouché M.B., 1983. The establishment of earthworm communities. [In:] Earthworm Ecology from Darwin to Vermiculture. Ed. J.E. Satchell. Chapman & Hall London, 431-447.

  7. Darwin C., 1881. The formation of vegetable mould, through the action of earthworms, with observations in their habits. John Murray, London.

  8. Dziadowiec H., Czarnecki A., Jonczak J., 1999. Raport za rok hydrologiczny 1999 [Hydrological year report of 1999]. ZMŒP, 20-40 [in Polish].

  9. Edwards C.A., Bohlen P.J., 1996. Biology and ecology of earthworms. Chapman and Hall London.

  10. Haukka J., 1988. Effect of various cultivation earthworm biomasses and communities on different soil types. Ann. Agric. Fennici 27, 263-269.

  11. James S., 2000. Earthworms (Annelida: Oligochaeta) of the Columbia River basin assessment area. Gen. Tech. Rep. PNW-GTR-491. Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, 1-13.

  12. Lee K.E., 1985. Earthworms, their ecology and relationships with soils and land use. Academic Press New York.

  13. Lee K.E., Smetten K.P.J., 1995. Identification and manipulation of soil biopores for the management of subsoil problems. [In:] Subsoil management techniques. Eds. N.S Jayawardane & B.A. Stewart. Lewis Publishers London, 211-244.

  14. Ligthart T.N., Peek W.C., Taber E.J., 1993. A method for the three-dimensional mapping of earthworm burrows systems. Geoderma 57, 129-141.

  15. Mckenzie B.M., Dexter A.R., 1983. Size and orientation on burrows made by the earthworms Aporrectodea rosea and A. caliginosa. Geoderma 56, 233-241.

  16. Norgrove L., Hauser S., Weise S.F., 1998. Effects of crop density and species upon surface casting by earthworms and implications for nutrient cycling in tropical intercropping system. Pedobiologia 42, 267-277.

  17. Olszewski T. D., 2004. A unified mathematical framework for the measurement of richness and evenness within and among multiple communities. Oikos 104, 377-387.

  18. Satchell J.E., 1983. Earthworm Ecology from Darwin to Vermiculture. Chapman and Hall London.

  19. Statsoft Inc., 1997. Statistica for Windows (Computer program manuals). Tulsa OK.

  20. Ro¿en A., 1982. The annual cycle in populations of earthworm (Lumbricidae, Oligochaeta) in three types of oak-hornbeam of the Niepolomicka Forest. I. Species composition, dominance, frequency and associations. Pedobiologia 23, 199-208.

  21. Ulrich W., 2004. Soil-living parasitic Hymenoptera: comparison between a forest and an open landscape habitat. Pedobiologia 48, 59-69.

  22. Werner M.R., 1990. Earthworm ecology and sustaining agriculture. Components 1, Vol. 4, retrieved from http://www.sarep.ucdavis.edu/worms/werner.htm.

Werner Ulrich
Department of Animal Ecology,
Nicolaus Copernicus University in Torun, Poland
Gagarina 9, 87-100 Torun, Poland
email: ulrichw@uni.torun.pl

Adam Czarnecki
Laboratory for Landscape Planning and Ecology,
Nicolaus Copernicus University in Torun, Poland
Gagarina 9, 87-100 Torun, Poland
email: czarnecki@boil.uni.torun.pl

Iwona Paprzycka
Laboratory for Landscape Planning and Ecology,
Nicolaus Copernicus University in Torun, Poland
Gagarina 9, 87-100 Torun, Poland
email: ipaprzycka@interia.pl

Responses to this article, comments are invited and should be submitted within three months of the publication of the article. If accepted for publication, they will be published in the chapter headed 'Discussions' and hyperlinked to the article.