Each microcosm was made of commercially available high-density stable polyethylene container (14 cm length, 14 m width and 7 m depth) with 36 pores (1 mm diameter) of lid. Soils sampled in microcosms treated with three levels (12.5, 25 and 50 tons of dry matter ha-1 year-1) of MSS, ISS, LPS, AFPS and PMC for 4 consecutive years (twice annually) were sieved gently through a 2 mm mesh sieve. In a preliminary experiment, ~40% of water holding capacity was optimal for microcosm test. Amount of 300-g fresh soil was hydrated to ~40% of water holding capacity. Hydrated water required to achieve the desired hydration was calculated according to the method of Greene et al. (1988). Ten earthworms were placed into each microcosm. The microcosms were kept in the controlled chamber at 20°C and 60±5% relative humidity under a 16:8 h light:dark cycle. Mortalities were assessed by emptying the test soil onto a tray and sorting the worms from the soil. Earthworms were considered to be dead if their bodies and anterior did not move or respond when they prodded with fine wooden dowels. Live worms were placed back into their original microcosms. The numbers of live and dead worms in each microcosm were recorded every 2 weeks and the dead worms were discarded. A randomized complete block design with three replicates was used. Mortality percentages were transformed to arcsine square root values for analysis of variance. The Bonferroni multiple-comparison method was used to test for significant differences among the treatments (SAS Institute, 2004).

Toxic effects of MSS, ISS, LPS, AFPS and PMC treatments on E. fetida in microcosm tests were evaluated (Table 4). All treatments did not affect any adverse effects on the organisms 2 weeks after treatment. At 4 weeks after treatment, effect of test waste material (F = 3.73; df = 4,44; P = 0.0141) on the mortality was significant but that of treatment level (F = 1.83; df = 2,44; P = 0.1785) was not significant. The material by level interaction was also significant (F = 2.34; df = 8,44; P = 0.0436). At 8 weeks after treatment, effect of test waste material (F = 200.90; df = 4,44; P < 0.0001) and treatment level (F = 5.37; df = 2,44; P = 0.0101) on the the mortality was significant. The material by level interaction was also significant (F = 9.49; df = 8,44; P < 0.0001). After 16 weaks after treatment, effect of test waste material (F = 124.11; df = 4,44; P < 0.0001) and treatment level (F = 9.73; df = 2,44; P = 0.0006) on the mortality was significant. The material by level interaction was also significant (F = 63.42; df = 8,44; P < 0.0001).

Heimbach et al. (1992) demonstrated that there is a good correlation (r = 0.86) between LC50 values of pesticides from an artificial soil test and the number of earthworms collected from a standardized field test. Our present and previous studies indicate that microcosm soil test using earthworms can predict results from a field test for assessing side effects occurred by long-term exposure of soil contaminants. Burrows & Edwards (2002) have been tried to use integrated soil microcosm based upon earthworms to predict effects of pollutants on soil ecosystems.