BIODIVERSITY PROJECTS

Earthworm Populations at the Wisconsin Integrated Cropping Systems Trial

Jon Simonsen [1], Martha Rosemeyer [2], and Joshua Posner [3]

Introduction

Many farmers, organic gardeners, and researchers have recognized earthworms as important organisms contributing to “healthy” soils (Romig et al. 1995). Earthworms are frequently associated with their ability to mix the soil (Cook and Linden 1996, Marinissen and Hillenaar 1997), increase rates of water infiltration (Bouche and Al-Addan 1997, Linden et al. 1991, Zachmann et al. 1987, Trojan and Linden 1994, 1998), improve soil aeration (Edwards and Lofty 1982), and build soil structure (Ketterings et al. 1997). They also have important roles in nutrient cycling (Blair et al. 1997, Subler et al. 1997, Bouche et al. 1997, Parmelee et al. 1998) and increasing the biological activity of the soil (Brown 1995, Doube and Brown 1998). Less positive functions of earthworms in agricultural soils include increased rates of nitrate and atrazine leaching (Tan et al. 1998, Sigua et al. 1995) and loss of residue cover (Gallagher and Wollenhaupt 1997)

In agricultural soils the earthworms are predominantly of European origin and are broadly classified into three functional groups (epigeic, endogeic, and anecic) according to their morphological attributes and ecological role in the soil (Bouche 1977). Epigeics are small pigmented earthworms that rarely burrow into the soil and live in the litter layer feeding on decaying organic matter. These species are common in forested areas but are rarely found in agricultural fields because of inadequate residue cover and surface moisture. The endogeic species are the most abundant earthworm in agricultural fields. Rarely found at the surface of the soil, these unpigmented earthworms are larger than the epigeics and continuously burrow horizontally through the soil searching for areas high in organic matter content. There is one species of anecic earthworm in Wisconsin, the well known nightcrawler (Lumbricus terrestris). This large pigmented species creates one or two permanent vertical burrows (up to eight feet deep) and feeds nocturnally by pulling vegetation from the soil surface into its burrow.

Research examining earthworms in agricultural soils has primarily focused on their response to a single input such as tillage or pesticide application but few studies investigate populations under field conditions that closely resemble those found on a farmer’s field. Current research priorities include the need to study the effects of conventional and alternative agricultural systems on earthworm populations (Hendrix 1998). The experimental design of the WICST project provides an ideal setting to study the response of earthworm populations to field-scale, systems level management on both conventional and low chemical input cropping systems.

An initial screening took place at the beginning of the trial in 1990 and 1991 (Brown and Posner 1991) and work resumed in 1999 after the cropping systems had gone through at least two full rotations.

Materials and Methods

Soil cores were used to sample the endogeic earthworm populations. After extraction these cores were sorted through a ¼ inch sieve to ensure that all of the soil was thoroughly searched. The total number of earthworms was recorded and biomass was estimated by weighing the preserved specimens in the lab [4]. In April 1999 two soil cores were extracted from all plots [5] at Arlington and Lakeland using a cylindrical coring device that measured 89.03 cm² X 25 cm (0.0022 m³). This was similar to the sampling procedure used by Brown and Posner nine years earlier. A number of pilot studies were also conducted during this period to determine the subsampling effort required and to see if earthworm numbers could be correlated with the field drainage (elevation) or proximity to the grass buffer edge between plots.

Soil sampling was repeated in June 1999 using a flat spade marked at 25 cm to excavate larger soil cores approximately 625 cm2 X 25 cm (0.0156 m³). Again, two samples were extracted in each plot and sorted through a ¼ inch sieve at both Arlington and Lakeland. To determine if earthworm number could be correlated with the groundcover, vegetation, or incorporated crop residue, these parameters were visually estimated for each soil sample. To determine if earthworm numbers were correlated with soil organic matter content[6] or soil moisture[7] subsamples of the cores were taken from all plots at Arlington but only ½ of the plots (2 blocks) at Lakeland.

A third sampling effort followed in July 1999 to quantify the deeper burrowing Lumbricus terrestris (nightcrawler) which, because of their deep permanent burrows, are often missed by the soil cores. Nightcrawler activity was recorded by counting the number of active middens [8] in a 1.0 m² quadrat. Twenty quadrats were taken in each plot at both sites.

Results and Discussion[9]

Earthworm Species Identification:
Earthworms were identified from the soil cores in terms of functional group on the basis of presence or absence of pigmentation. The core sampling rarely captured nightcrawler adults. Juvenile anecics were captured but these earthworms are polyhumic endogeics that feed on organic matter in the upper soil profile until they are large enough to create the vertical macropores (Sam James pers. comm.). Consequently in the data analysis the numbers from the soil cores are used to represent the number of the endogeics and the midden data are used to measure the number of anecic earthworms.

In 1991 Brown and Posner found that the predominant species were Aporrectodea tuberculata and turgida (endogeic) and Lumbricus terrestris (anecic). The earthworms collected in 1999 from the spring (April) and summer (June) soil cores were immediately preserved in 100% EtOH. However this did not preserve the diagnostic characteristics needed for identification. In order to start a reference collection adults were collected from one day of sampling at Arlington and Goose Pond Sanctuary in late June. These specimens were preserved with a 10% formalin solution for 24 hours followed by 70% EtOH (Schwert 1990). Four species were identified [10]: Lumbricus terrestris (Linnaeus), the anecic nightcrawler, and Aporrectodea tuberculata (Eisen), an endogeic, were identified from the WICST plots. The endogeics, A. tuberculata, Octolasion tyrtaeum (Savigny), and Eiseni rosea (Savigny) were found at Goose Pond Sanctuary.

Pilot Studies:
Graphical analysis of sample size for the spring soil core data show that much of the variation in estimating the mean is removed when eight or more soil cores were extracted from a plot. Application of the sample size formula n=(( z * d ) / d)², where (z) = the µ/2 upper quantile of the normal distribution, (d) = standard deviation of the measurement [11], and (d) = the error of margin from the true mean that is tolerated [12] (Yandell 1997), for the data indicate that an average of 21 subsamples are required per plot. Such a large sample size was not feasible and prompted the second sampling of the plots using a larger soil excavation and more hired help to maximize the sampling effort. A more detailed statistical analysis examining the effect of core size and subsampling effort will be done in 2000. The graphical analysis of sample effort for the midden sampling indicated that most variation of the population mean estimate was reduced when 15 or more quadrats were taken per plot. Using the same sample size formula for the data indicated that an average of 11 quadrats were required per plot. Somewhat surprisingly, there were no significant differences between the number of earthworms taken from the grassy edge vs. the center of a plot or the topographically high vs. low areas. Additionally no relationships were observed between the number of earthworms and the amount of incorporated crop residue, percent bare ground, percent litter, percent live vegetation, percent crop, percent weed, soil organic matter content, or percent soil moisture.

Total Earthworm Numbers:
Population levels were typical of agricultural systems (Edwards and Bohlen 1996) and the endogeics greatly outnumbered the anecics (Figures 1a & 1b) [13]. In 1991 numbers were higher on the better drained Arlington soils (279/m²) than the somewhat poorly drained Lakeland soils (197/ m²). In 1999 both Arlington and Lakeland had similar endogeic populations that ranged from 34 to 374 /m² (Figures 2a & 3a). Midden density ranged from 0 to 27 /m² (Figures 4a & 4b) but Lakeland had 56% less than Arlington (Figure 1b). The reduced nightcrawler activity at Lakeland is a consequence of the shallower soils and a heavy rain event that ponded the plots prior to the summer soil core (June) and midden sampling (July). When the fields began to dry many decomposed earthworms were observed at the surface of the soil.

As expected, earthworm numbers decreased in the summer (June) when compared to the spring (April) sampling at both sites. The total population at Arlington decreased 17% but at Lakeland populations decreased significantly by 52%. As soil temperatures rise and the soil moisture levels decrease endogeic earthworms burrow deeper into the soil profile, and enter into a state of quiescence during which they curl themselves into a tight ball to reduce water loss, and reduce their metabolic rates until conditions become favorable (Edwards and Bohlen 1996). During the second soil core sampling both aestivating and active individuals were found in the soil cores at both sites. This indicates that some endogeic earthworms had already burrowed below 25 cm, entered into dormancy and were missed. Mechanical weed control in CS3 and CS5 may also be a factor reducing the populations. The greater decrease of earthworm numbers at Lakeland is a result of the heavy rain event that reduced the populations in all cropping systems.

Cash-Grain vs. Forage-Based Cropping Systems:
By the fall of 1991 (October) differences between treatments were beginning to become obvious with the forage systems having higher numbers of endogeic earthworms than the cash grain rotations. In 1999 we still expected to find higher earthworm populations in the forage-based systems because of the use of manure (Edwards and Lofty 1997) and the reduced tillage during the alfalfa phases of these rotations (Kladivko et al. 1997, Jordan et al. 1997, Clapperton et al. 1997) (Figures 2 & 3). During both soil sampling periods at Arlington the cash grain systems had 59% fewer earthworms than the forage-based systems. The cash-grain systems averaged 133.1 and 110.8 earthworms /m² during the spring core and summer excavation sampling respectively and the forage-based systems averaged 327.9 and 273.1 earthworms /m² during the same periods. Similar results were obtained at Lakeland which had 56% less earthworms in the cash grain systems during the spring sampling (186.2 and 422.9 earthworms /m² respectively) [14]. Nightcrawler activity at Arlington was also lower in the cash grain rotations where there were 28% less middens than in the two forage-based systems (CS4 and CS5) [15].

Cash-Grain Cropping System Comparisons:
We expected the no till cropping system to have the highest number of earthworms because of minimal soil disturbance and CS1 to have the lowest because of yearly tillage and intensive chemical use. We anticipated that CS3 would fall in the middle because the intensive mechanical weed control will reduce populations but the wheat/ red clover phase may promote earthworm numbers (Edwards and Bohlen 1996).

In 1991 more endogeic earthworms were found in the lower chemical input system (CS3) than the higher chemical input continuous corn rotation (CS1). In 1999 during the spring sampling earthworm populations in CS2 and CS3 were similar at both sites (Figure 2a). CS1 had the significantly lowest number of earthworms at Arlington but at Lakeland the continuous corn system was similar to the other grain systems (Figure 2a). During the summer sampling at Arlington CS2 had the highest number of earthworms /m² followed by CS3 and CS1 all of which were significantly different (Figure 3a). At Lakeland the same pattern was evident but there were no significant differences (Figure 3a) and it is likely that cropping system effects were masked by the large rain event. As noted earlier earthworm numbers decreased during the second sampling but not all systems decreased in the same way (Figures 2a & 3a). The mechanical weed control used in CS3 during the corn phase of the rotations can directly influence earthworm populations (Werner and Dindal 1989). Alternatively, in the no till system (CS2), where there is minimal soil disturbance, the endogeics are still present later in the season. This pattern was observed at Arlington where numbers decreased by 74 earthworms /m² in CS3 but did not change much in the continuous corn (-8.2 earthworms /m²) or no till systems (+15.2 earthworms /m²). The same pattern was not observed at Lakeland primarily as a result of the large rain event and all three cropping systems had similar declines in earthworms /m² (-139.7 (CS1), -93.2 (CS2), -104.7 (CS3) earthworms/ m²).

The response of active middens to cash-grain cropping systems were very similar at both sites even though Lakeland had lower midden numbers (Figure 4a). The no-till cropping system, CS2 had significantly higher numbers compared to CS1 and CS3 which were similar to each other [16].

Forage-based Cropping System Comparisons:
In the forage-based cropping systems we expected the rotationally grazed pasture to have the highest earthworm population because there is no tillage, a well developed sod (Springett and Gray 1997), and addition of manure and urine by heifers throughout the growing season. Differences between CS4 and CS5 were more difficult to predict. CS5 promotes earthworm populations because it has low chemical inputs and a less intensive cutting schedule but reduces populations because it receives fall tillage and intensive mechanical weed control once every three years of the rotation. CS4 may reduce populations by using pesticides and an intensive cutting schedule that may lead to soil compaction but would promote earthworm populations because it gets tilled less intensively one out of every four years of the rotation. We hypothesized that CS5 would have lower earthworm populations than CS4 because of the intensive mechanical weed control every three years.

In 1991 there were no differences observed in the forage systems. In 1999 the rotationally grazed pasture did usually have higher populations than the other two systems but the difference was only significant during the spring soil core sampling at Lakeland where high numbers of earthworms were observed (674 /m²) (Figures 2b & 3b)[17]. There were no significant differences found when comparing CS4 and CS5 for any of the soil core sampling dates or the number of middens recorded in these systems.

When comparing endogeic earthworm populations in the spring vs. summer soil sampling at Arlington CS5, the low chemical input system with mechanical weed control, had the greatest decrease (-132 /m2) in the number of earthworms compared to CS4 (-28 /m²) and CS6 (-4 /m²) (Figures 2b & 3b).

Summary

These preliminary results suggest that primary tillage, mechanical weed control, and manure application have a large influence on the earthworm populations. Future analyses will focus on the response of the earthworm functional groups within the context of cropping system phases to further tease apart their population dynamics in these cropping systems.

Bibliography

lair, J.M., R.W. Parmelee, M.F. Allen, D.A. McCartney, and B.R. Stinner. 1997. Changes in soil N pools in response to earthworm population manipulations in agroecosystems with different N sources. Soil Biol Biochem 29(3/4): 361-367.

Bouche, M.B. 1977. Strategies lombriciennes. In Soil organisms as components of ecosystems: Proceedings of the 6th Int. colloquium on soil zoology, U. Lohm and T. Persson (eds.). Swedish Natural Science Research Council Ecological Bulletin No. 25, Stockholm, pp. 122-133.

Bouche, M.B. and F. Al-Addan. 1997. Earthworms, water infiltration, and soil stability: some new assessments. Soil Biol Biochem 29(3/4): 441-452.

Bouche, M.B., F. Al-Addan, J. Cortez, R. Hammed, J-C Heidet, G. Ferriere, D. Mazaud, and M. Samih. 1997. Role of earthworms in the N cycle: a falsifiable assessment. Soil Biol Biochem 29(3/4): 375-380.

Brown , G.C. and J.L. Posner. 1991. Earthworm ecology in agriculture: 1990 and 1991 data from the Wisconsin Integrated Cropping Systems Trial. The Wisconsin Integrated Cropping Systems Trial. First Report.

Brown, G.C. 1995. How do earthworms affect microfloral and faunal community diversity? 1995. Plant and Soil 170: 209-231.

Clapperton, M.J., J.J. Miller, F.J. Larney, and C.W. Lindwall. 1997. Earthworm populations as affected by long-term tillage practices in southern Alberta, Canada. Soil Biol Biochem 29(3/4) 631-633.

Cook, S.M.F. and D.R. Linden. 1996. Effect of food type and placement on earthworm (Aporrectodea tuberculata) burrowing and soil turnover. Biol Fertil. Soils 21:201-206.

Doube, B.M., and G.G. Brown. 1998. Life in a complex community: functional interactions between earthworms, organic matter, microorganisms, and plants. In Earthworm Ecology. ed. Clive Edwards. 1998. St Lucie Press, pp 179-211.

Edwards, C.A. and J.R. Lofty. 1982. Nitrogenous fertilizers and earthworm populations in agricultural soils. Soil Biol Biochem 14:515-521.

Edwards, C.A. and J.R. Lofty. 1982. The effects of direct drilling and minimal cultivation on earthworm populations. J. Appl Ecol 17: 533-543. In Biology and Ecology of Earthworms. eds. Edwards, C.A. and P.J. Bohlen. 1996. Chapman and Hall: pp 207-210.

Edwards, C.A. and P.J. Bohlen. 1996. Biology and Ecology of Earthworms. Chapman and Hall: pp 60-62.

Hendrix, P.F. 1998. Earthworms in agroecosystems: a summary of current research. In Earthworm Ecology. ed. Clive Edwards. 1998. St Lucie Press, pp 259-267.

Jordan, D., J.A. Stecker, V.N. Cacnio-Hubberd, F. Li, C.J. Gantzer, and J.R. Brown. 1997. Earthworm activity in no-tillage and conventional tillage systems in Missouri soils: a preliminary study. Soil Biol Biochem 29(3/4): 489-491.

Kladivko, E.J., N.M. Akhouri, and G. Weesies. 1997. Earthworm populations and species distributions under no-till and conventional tillage in Indiana and Illinois. Soil Biol Biochem 29(3/4): 613-615.

Ketterings, Q.M., J.M. Blair, and J.C.Y. Marinissen. 1997. Effects of earthworms on soil aggregrate stability and carbon and nitrogen storage in a legume cover crop agroecosystem. Soil Biol Biochem 29(3/4): 401-408.

Linden, D.R., M.D. Trojan, F. Zheng, and R.R. Allmaras. 1991. Modeling preferential flow in earthworm burrows. In: Gish, T.J. and A. Shiromohammadi. eds. Preferential Flow: Proc. of the National Symposium. December 16-17, 1991, Chicago, IL. American Society of Agricultural Eng. St. Joseph, MI USA. pp. 268-277.

Marinissen, J.C.Y. and S.I. Hillenaar. 1997. Earthworm-induced distribution of organic matter in macro-aggregates from differently managed arable fields. Soil Biol and Biochem 29(3/4): 391-395.

Parmelee, R.W., P.J. Bohlen, and J.M. Blair. 1998. Earthworms and nutrient cycling processes: integrating across the ecological hierarchy. In Earthworm Ecology. ed. Clive Edwards. 1998. St Lucie Press, pp 123-143.

Romig, D.E., M.J. Garlynd, R.F. Harris, and K. McSeeeney. 1995. How farmers assess soil health and quality. Journal of Soil and Water Conservation 50(3): 229-236.

Schwert, D.P. 1990. Oligochaeta: Lumbricidae. In Soil Biology Guide ed D.L. Dindal. Wiley. pp341-356.

Sigua, G.C., A.R. Isensee, A.M. Sadeghi, and G.J. Im. 1995. Distribution and transport of atrazine as influenced by surface cultivation, earthworm population, and rainfall pattern. Chemosphere 31(10): 4237-4242.

Springett, J. and R. Gray. 1997. The interaction between plant roots and earthworm burrows in pasture. Soil Biol Biochem 29(3/4) 621-625.

Subler, S., and A.S. Kirsch. 1998. Spring dynamics of soil carbon, nitrogen, and microbial activity in earthworm middens in a no-till cornfield. Biol. and Fert. of Soils 26(3): 243-249.

Subler, S., C.M. Baranski, and C.A. Edwards. 1997. Earthworm additions increased short-term nitrogen availability and leaching in two grain-crop agroecosystems. Soil Biol Biochem 29(3/4): 413-421.

Tan, C.S., C.F. Drury, M. Soultani, I.J. Van-Wesenbeeck, H.Y.F. Ng, J.D. Gaynor. 1998. Effect of controlled drainage and tillage on soil structure and tile drainage nitrate loss at the field scale. Water Science and Technology 38(4-5): 103-110.

Trojan, M.D. and D.R. Linden. 1994. Tillage, residue, and rainfall effects on movement of an organic tracer in earthworm-affected soils. Soil Science Soc of Am Journal 58(5): 1489-1494.

Trojan, M.D. and D.R. Linden. 1998. Macroporosity and hydraulic properties of earthworm-affected soils as influenced by tillage and residue management. Soil Science Soc of Am Journal 62(6): 1687-1692.

Werner, M.R. and D.L. Dindal. 1989. Earthworm community dynamics in conventional and low-input agroecosystems. Rev Ecol Biol Sol 26(4): 427-437.

Yandell, B.S. 1997. Practical Data Analysis for Designed Experiments. Chapman and Hall. pp 82-85.

Zachmann, J.E., D.R. Linden, and C.E. Clapp. 1987. Macroporous infiltration and redistribution as affected by earthworms, tillage, and residue. Soil Science Soc of Am Journal 51(6): 1580-1586.

[1] Graduate Student—Dept. of Agronomy, University of WI—Madison. jmsimons@facstaff.wisc.edu
[2] Visiting Assistant Professor—Dept. of Agronomy, University of WI—Madison. rosemeym@evergreen.edu
[3] Professor—Dept. of Agronomy, University of WI—Madison. jlposner@facstaff.wisc.edu
[4] Soil was not dissected from the digestive system.
[5] WICST is designed as a randomized complete block experiment, one block was sampled each day.
[6] Loss on ignition procedure – Soil and Forage Analysis Lab, UW Madison.
[7] Gravimetric method
[8] Middens are the small mounds of litter and casting material created by feeding activity at the soil surface above the burrows.
[9] All significant results are reported at µ£0.05.
[10] Dr Sam James assisted with the identifications and taxonomy. Dept of Biology, Maharishi International Univ., Fairfield, Iowa 52556
[11] Estimated from the data.
[12] An acceptable difference (d) was set at 1 worm /soil core and (z) was set at 0.025.
[13] Biomass data for nightcrawlers will be collected during the 2000 field season.
[14] No comparisons could be made during the summer sampling at Lakeland because of the transitional cropping systems (CS4 and CS5).
[15] Nightcrawler activity could not be estimated in the pasture because of the difficulty in delineating the middens with the dense vegetative cover and trampling by cattle. No comparisons could be made at Lakeland because neither the pasture or the two transitional cropping systems could be included in the analysis.
[16] Middens were not counted in the spring so no seasonal patterns could be observed.
[17] CS4 and CS5 at Lakeland are transitional cropping systems and not included in the summer analysis.

Home | Who we are | Research | Publications | Links | Site map | Contact us

University of Wisconsin-Madison
College of Agricultural and Life Sciences
If you have any difficulty accessing this page, please contact our web master.

Site hosted by UW-Madison Center for Integrated Agricultural Systems