Pedobiolog1a 36. 109-120 (1992) Gustav Fischer Verlag Jena National Institute of Public Health and Environmental Protection. Bilthovcn, the Netherlands Influence of environmental conditions on the growth and reproduction of the earthworm Eisenia andrei in an artificial soil substrate C. A. M. VAN GESTEL, E. M. DlRVEN-VAN BREEMEN and R. BAERSELMAN With 4 Figures Synopsis: Original scie111iflc paper In reproduction studies with the earthworm Eisenia andrei, carricdr out in an artificial soil substrate, a considerable variation in cocoon production and growth was observed. Cocoon production appeared to be negatively correlated with earthworm growth, and positively correlated with initial worm masses. To detect potential sources for this variation, the effect on earthworm growth and reproduction of three environmental factors was determined. Cocoon production was reduced at high soil pH ( ~ 7 0), and was optimal at 20 •c and a moisture content of 85% in the artificial soil. This moisture content exceeds field capacity. The 111nuence of these environmental factors could. however, not fully explain the variation in cocoon production and growth Key words: Eiscnia w1drei. earthworm. growth, temperature. soil, moisture content, pH. reproduction. 1. Introduction In ecotoxicology survival. growth and reproduction are commonly used as test parameters. Effects on such parameters in individual organisms, however, cannot be directly translated to the level of populations. Physiologists therefore attempt to describe models linking physiological and population processes (CA LOW & SIBLY, 1990). Such models, which are in fact energy budgets, assume that the energy absorbed by an organism is used for metabolism (scope for metabolism) and production (growth and reproduction; scope for growth). CA LOW & SrnL Y ( 1990) describe in a theoretical way the possible relationships be tween scope for m.::tabolism and scope for growth, and conclude that in case metabolism does not affect survival, scope for growth may be a relevant index of population density change. As scope for growth consists of two parameters, i.e. growth and production of offspring, tt is important to understand the possible relation between these two parameters. YAN GESTEL et al. ( 1989), in a newly developed earthworm reproduction toxicity test, observed a considerable variation in cocoon production and the hatching rate of cocoons produced by the earthworm Eisenia andrei incubated in an artificial soil substrate. There seemed to be some relatton between earthworm growth and cocoon production. To further investigate this phenomenon. a series of reproduction tests with the earthworm species £. andrei was carried out. Additionally, the effect of three environmental parameters (temperature, soil pH. soil moisture content) on the reproduction of E. andrei was determined. 2. Materials and methods 2. I. Earthworms The earthworms used for the experiments wer. . adult, with a well developed clitellum and were of the ;pccies £i.1·c•11i11 undrei Bouc11L 1972. All worms were obtained from our own culture; worms were Pcdobiologia 36 ( 1992) 2 I 09 grown on horse dung as a food source at an ambient temperature of.20 ± S •c. The worms used were between 8.5 and 17.S weeks old, and their individual live masses varied between 276 and 449 mg. 2.2. Artificial soil All cocoon production tests were carried out in artificial soil, which was composed of a dry mass of 10% I mm sieved sphagnum peal, 20% kaolin clay. ea. 69.5% sand and ea. 0.5% CaC03. The moisture content of the substrate was adjusted Lo ea. 55% (m/m) by the addition of deminerali7ed water, and the pH (I M KCI) was 6.0 ± 0.5. For the mcubation of cocoons the artificial soil was made up with finely ground ( ~ 0.5 mm) sphagnum peal. This artificial soil had a moisture content of 35% (m/m). 2.3 Reproduction experiments Tests were performed as described by VAN GESTEL et al. ( 1989). One liter glass jars were filled with 400 or 500 g dry mass of artificial soil, and 4 or 10 g finely ground cow dung was placed in a hole in the middle of the soil <1s a food source for the worms. Ten adult earthworms were introduced into each jar. Before the start of the experiments, the earthworms were pre-incubated for one week (phase A) in the same artificial soil with cow dung as a food source in a hole in the middle. At the start of each experiment, worms were sorted from the pre-incubation substrate, washed. dried on filter paper, weighed. and introduced in freshly prepared artificial soil. Incubation periods (phase B) lasted 3 weeks. Table 1 Results of reproduction tests with Eiseniaandrei in artificial soil. (mean of 3 or 4 replicates). lest pre-i ncu bat ion worm pH % cocoons/ % juvcn./ number mass end growth/ worm/ fertile worm/ cocoons/ % (mg) worm/ week cocoons week worm"/ growth/ week week worm/ week I 38 0.9 449 6.3 -2.2 1.64 71.8 2.25 2 0.98 -5.4 435 6.3 2.6 1.28 84.8 2.72 3 1.30 -4.4 441 6.2 2.7 1.42 91.1 3.05 4 0.98 20.0 338 7.0 - 1.5 1.23 58.0 1.21 4 (C)* 1.23 - 1.5 323 6.5 4.2 1.09 5 1.38 15.2 363 6.7 -2.9 1.49 93.1 3.65 5 (Cl 1.49 -2.9 331 6.5 2.1 1.19 6 0.49 37.7 384 7.2 0.93 91.2 1.85 i 2.15 19.5 447 6.3 -2.2 2.03 94.3 6.72 7 (C) 2.03 -2.2 419 6.2 0.7 1.24 91.3 2.94 8 1.15 15.6 329 6.6 3.0 1.35 85.3 2.59 8 (C) 1.35 3.0 358 5.8 4.7 1.40 9 0.57 0.5 276 5.8 6.4 0.68 86.1 1.21 9 (C) 0 68 6.4 328 6.5 11.7 0.32 71.0 0.53 10 327 7.4 6.8 0.30 63.6 0.26 10 (C) 0.30 6.8 393 7.3 4.5 0.27 II 0.80 13.7 376 5.0 1.03 94.6 2.23 12 0.93 21.8 403 5.9 -1.5 1.13 89.9 2.57 13 0.85 - 1.9 375 7.3 3.3 0.77 83.8 1.27 14 0.87 8.1 348 6.9 13.3 0.53 91. I 1.21 15 1.45 -3.5 369 6.4 3.8 1.40 86.8 3.25 16 1.00 22.2 368 6.0 2.0 1.34 94.4 3.06 17 0.90 11.4 357 6.0 4.4 1.40 94.1 3.61 18 0.98 591 -0.1 1.20 95.2 2.95 • Results of phase C (3 weeks post-incubation) are indicated with (C) after the test number. I I 0 Pcdob1olog1a 36 ( 1992) 2 and al the end worm masses were determined again. In some cases an additional three week post-incubation period (phase C) was included after phase B. All experiments had four replicates, each containing 20 worms. A series of tests was carried out under slightly different conditions. The tests have been numbered I 10 18 (table I). In a number of tests, only 4 g cow dung was added at the start or phase B. and after 1.5 or after I and 2 weeks another 4 g was given. The total amount of food given in a three week lest period ra ngcd bet wecn 8 and 12 g per I 0 worms. Cocoon production al the end of all incubation periods was determined as described by van G1:sTEL er al. ( 1989). The cocoons produced during phase B were incubated for 5 weeks lo assess hatchabilily as described by VMI GESTHI. et al. (1988). Hatchability of cocoons produced in phase C was only determined in two h:sls. For the incubation of the cocoons the artificial soil was made up with finely ground peal and I% (m/m) finely ground cow dung was mixed through it as a food source for the hatching juveniles. For 1es1 6 twelve replicates were used 10 determine cocoon production. but cocoons of only two replicates were incubated lo assess hatchability. In tests I 0 and 13 soil pH was somewhat higher than intended: 7 .3- 7.4. For tests 9 and 10 only three replicates were used, and in lest I 0 all the worms for the whole experiment were pre-incubated in one big container with artificial soil. For all the other experiments, worms were pre-incubated in groups of 10 worms/jar. In tests 4 and 5 pre-incubation lasted 2 weeks instead of I week. · In test 3, horse dung was used as a food source instead of cow dung. In experiment 18 rather old ( ~ 20 weeks) and large adult worms (c. 600 mg) were used to test the influence of body mass on reproduction. All incubations were done in a climatic chamber al 20 ± 3 °C. and constant illumination (light intensity c. 400 lux). 2.4 Influence of pH, soil moisture content and temperature To test the innucnce of soil pH, the artificial soil was made up with different amounts of CaC0 Lo 3 obtain nominal soil pH.( I M KCI) values of 4.0, 4.5, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5 and 8.0) To determine the influence of soil moisture content, cocoon production was determ111cd al soil moisture contents of 35, 45. 55. 65, 85 and 130% (m/m), corresponding with· pF values of 2.8. 2.4, 2.0, 1.7, 1.4, and 0.9, respectively. The soil moisture content of 55% (m/m) corresponds with field capacity. To determine the innuence of temperature, tests were incubated al temperatures of 10, 15, 20, 25, and 30 °C. All tests were carried out with four duplicates, each containing 10 adult earthworms. Experiments comprised pre-incubation and incubation phases A and B, as described above, and were performed under the environmental conditions lo be tested. Cocoons produced during phase B were incubated as dcscri bed a bovc. 2.5 Statistical analysis All data were analyzed by regression and correlation analysis using the software package Genslat 5, and applying Student's 1-1es1 3. Results 3.1 Reproduction experiments Table I summarize~ the results of all experiments. Table 1 shows that both cocoon production and growth varied considerably in the dilTcrcnl tests carried out. Cocoon production was rather low ( < 1.0 cocoons/worm/week) in tests 6, 9, 10, 13 and 14. The use of horse dung instead of cow dung (test 3) did not influence growth or reproduction of £. a11drei. With horse dung cocoon production was slightly (not significantly) higher. When the artificial soil for the incubation of cocoons was made up with horse dung. the substrate was completely overgrown with fungi. With cow dung no or only minor fungal growth was observed. As is shown in figure I. a signil"icanl correlation was found between growth and cocoon production. Pedobiologia 36 (1992) 2 I I I 2.4 2.2 r • ·0.646, P<0.01 2 . 1.8 . ... . -. ~ ~., 1.6 .. . ~ 1.4 .. .,.. ... . ··;. . ·; { 1.2 "".. . ... .~ c ~ 0.8 . .•. • . 0.6 phase B • 0.4 phase C 0.2 0 ·5 .3 ·1 3 7 9 11 13 15 "• growlh/worm/Week Figure I. Relationship between cocoon production and growth of Eisenia andrei in artificial soil. Each point represents one replicate. containing 10 earthworms (n = 98; phase B = 3 weeks incubation; phase C = 3 weeks post-incubation). Cocoon production was also significantly correlated with initial worm mass. This relation is shown in figure 2. In this figure the data for the large worms (test 18) are omitted; when these are also included correlation coefficient is lower: r = 0.286 (p < 0.01). The very large worms (591 mg) did not produce as many cocoons (1.2 cocoon/worm/week), and they did not gain mass as would be expected on the basis of the relation between growth and cocoon production described above. .. 2.4 2.2 r • 0.359, P<0.01 2 .. . . 1 8 . . .. ~" 1.6 \, ... E~: 1.4 .. •.. .. .•.. I· .- = ..,. . . ..•. .. i 12 ':•. c~ 0.8 .. . ...· ' .. . . 0.6 0.4 • phase B • phase C 0.2 0 240 280 320 360 400 440 480 worm weight (mg) Figure 2. Rel<1tio11 between cocoon production and initial worm masses of Eise11ia andrei in artificial soil. Each point represents one replicate containing 10 earthworms (n = 94; phase B = 3 weeks incubation; phase C = 3 weeks post-incubation). There was no correlation between the age of the worms and growth during the experiment (n = 20: r = 0 343. n.s.), but cocoon production was slightly correlated with worm age: r = - 0.454, p < 0.05 (n = 20). From the results presented in table I it can be concluded that cocoon production during incubation and post-incubation phases B and C was significantly correlated with that in the corresponding (pre-)incubation period. This relationship, which is shown in figure 3, indicates that cocoon production was always higher during incubation and post-incubation than during the pre-incubation period. As shown in table 1, in many cases worm growth was very high during pre-incubation compared to that during the incubation period. I 12 Pedobiologia 36 ( 1992) 2 2.• / 2.2 r. 0.667, pc0.01 2 . . .. 1.8 • t 1.6 I . • I ~ ~"~' 11 •2 : .•I • ..• ; ..II "~.I · • _..•. ...s I c 8 . ~ 0068 :I • I• ...... ,.r .• • phase B phase C cocoons/Worm/week (p<o·inc. parod) Figure 3. Rcla11on between cocoon production by Eisenia andrei during the incubation and post incubation periods (phase B and C) and the corresponding pre-incubation periods (phase A for phase B, and phase B for phase C). Each point represents one replicate containing 10 earthworms (n = 91)'. · Growth of the worms was negatively correlated with initial worm mass: r = -0.398. p < 0.01 (n = 98). This corresponds with the increased cocoon production with increasing wom1 masses and the decreased cocoon production with increasing growth rates. Cocoon production was not correlated with soil pH: r = -0.442, n.s. (n = 20), although it looks as if cocoon production is somewhat reduced at high pH. This is for example the case in tests 10, 13 and 14, where at pH 6.9-7.4 the cocoon production was low. In test 4 pH was also high, but cocoon production was normal, and in test 9 cocoon production was very low while pH was about normal. The time of the year the experiments were run was also included in the analysis: there was no correlation between cocoon production or growth rate and the time of the year. Neither cocoon production nor growth were correlated with the minimum, maximum or average temrerature in the culture, which nuctuated between 13.5 and 28.0 °C al the time of the tests. As growth and cocoon production were negatively correlated, we looked for a unifying factor according to which total biomass production could be ca.lculated. It was assumed that the only two ways to produce biomass are the production of cocoons and the production of body mass. So, biomass production was calculated by adding the mass gain of the worms (in mg/worm/week) to the mass of cocoons produced. The latter was calculated from the number of cocoons produced per worm per week and the initial worm masses using the equation derived by HARTENSTEIN et al. ( 1979): Mc = 0.0106 · M. + 6.48 (r = 0.97) In this equation M. = worm mass, and Mc =cocoon mass (both in mg). This resulted in biomass production figures varying between 4.8 and 50.4 mg/worm/week, with the lowest production in test 4, and the highest in test 14, respectively. From this it can be concluded, that also biomass gain shows considerable variations. When this calculation was applied on the individual data (per 10 worms) it appeared that also within one test variation was high, e.g. in test 5, biomass gain varied between -0.3 and 9.5 mg/worm/week. So, no improvement was obtained by this correction. The conclusion may be drawn that individual variation in the biomass production of earthworms is considerable. Be<ween 58.1 and 95.2% of the cocoons hatched, the remainder appeared to be infertile. The number of juvenile worms per fertile cocoon varied between 1.3 and 4.3. In most cases, however, between 2.0 and 3.0 juveniles emerged per cocoon. The number of juveniles per fertile cocoon was significantly correlated with the percentage of fertile cocoons: 8 Pedobiologia 36 ( 1992) 2 1 13 r = 0.556. p < 0.01 (n = 73). The number or juveniles per worm per week. which 1s given 1n 1able I, i~ a good measure or total reproduction. From table I it can be concluded that also this parameter showed great variations: between 0.26 and 6.72 juveniles were produced per worm per week. In most studies it fluctuated, however, between 1.2 and 3.7 juveniles/ worm/week. 3.2. Innuence of environmental.factors 3.2.J. Soil pH Table 2 summarizes the results of the test on the 111nuence of soil pH on the reproduction or £. {1/1(/rei. Tahk 2 lnllucncc or ~oil pi I nn lhc rcrro<luclion and grow1h or !:"1.1°('/liU 11111/n•1 1n artificial soil. ,oilpll pre-incubation worm % growth cocoons •1. juvcn. mass growth/ worm fertile worm/ •1. 11111111nal ac1ual • cocoons (mg) worm week cocoons week worm worm week week week .HJ 4.2 1.20 21.7 419 1.4 1.27 81.7 2.40* -1 ' 4.8 1.33 18.6 423 3.0 1.40 80.5 2.99 us 5.0 5.1 23.7 -133 0.1 1.33 80.8 2.99 5.5 5.4 I.JO 20.7 414 1.2 1.35 87.9 2.89 6.0 60 1.50 :w.2 447 - 1.J l.49 82.6 3.38 6.5 6-1 UJ 18.0 -117 1.6 1.31 86.6 2.60• 7.0 68 1.25 21.7 427 0.5 1.18* 90.1* 2.82 7.5 75 l.15 19.5 418 l.J 0.98** 77.3 1.6&•• 8.0 75 1.30 23.7 452 - 0.7 1.06* 77.0 1.s8•• • .it the end of the J week incubation period • •• significantly different from pH 6.0 alp < 0.05 and p < 0.01, respectively From table ~ it appears that the actual pH recorded at the end or incubation did not d1ITcr much rrom nominal pH. Only in case or pH 8.0 difference was 0.5 pH units. The resulls presented in table 2 indicate that cocoon production is optimal at a soil pH tif 5.0-6.0. Cocoon production appeared to be slightly correlated with soil pH: r = -0.699. p < 0.05 (n = 9). Al initial pH ~ 7.0 cocoon production was significantly reduced. During the pre-incubation period worms at all pH levels showed a strong mass gain (I !'I - 24'!/o). Growth during incubation (phase B) was not affected by soil pH, and no correlation between growth and cocoon production was found in this study (n = 36: r = 0.117, n.s.). Cocoon production during incubation was correlated with that during the pre-111cuba1ion period (n = 36: r = 0.535, p < 0.01); growth was correlated with the initial \\Orm masses (n = 36: r = 0.511. p < 0.01). Cocoon hatchability was signilicantly increased at pH 7.0. The number ofj uvenile worms per rertile cocoon amounted 1.89- 3.22, and when compared to pH 6.0 it was signilicantly reduced at pH 7.5 and 8 (p < 0.01). and at pH 4.0, 5.5 and 6.5 (p < 0.05). The number or JUVcnilcs per worm per week was significantly reduced at pH 4.0, 6.5, 7.5 and 8.0. Estimated biomass gain ranged between 8.4 and 28.0 mg/worm/week and did not show an) relationship with soil pH. When results or the reproducllon studies described above are combined with those or . this pH experiment. a total view on the innuence or soil pH on cocoon production can be obtained. These results arc depicted in figure 4. From this figure it is obvious that especially at high pH ( ~ 7.0) cocoon production decreases. I I 4 Pcdob1ologia 36 ( 1992) 2 2.4 2.2 .. 1 8 . ~ 1 6 . . . . ? :· E I 4 . . 0 1 2 .!'" • . ~ c 8 8 08 06 .. O• 02 0 5 6 8 pH Figure 4 Rcl.111011 bet" ccn cocoon production of £ise11io a11drl!i and the pH of the artificial soil. E;1ch po111t r.:rrcscn1s the average value of 3 or 4 replicates. i.e. 30 or 40 worms. 3.2.2. Temperature Table J summari7c:. the resulb or the test performed to study the innuence ortemperature on the reproduction or E. rmclrei. Tabk 3. I nlluence oftcmpcralurc on lh.: n:production and growth of r:il'l'llia a11drC'i in <1rtilic1al soil. tempcralurc pre-incubation worm % cocoons/ % ju\ en .. ('C) mass growth/ worm/ fertile worm cocoons 0/o (mg) worm/ week cocoons week nominal actual worm/ growth/ week wee I. worm/ week 10 9.:i 0.00 16.9 J II 7.0 0.04** 66.7 0.04*• 15 17.0 0.00 17.2 291 I J.6• 0.46** 86.1 0.97** 20 21.0 0.28 15.0 304 7.7 1.35 89.1 3.37 25 15 () 0.50 9.8 194 9.3 1.28 9J.6 2.68** .>O. . JO 0 0.00 9.4 295 13.0 0.01 ... ~1g111ficantl} diffcrcnl from 20 C al p < 0.05 and p < 0.0 I. respectively From table J is appears that a temperature or 20- 25 C is optimal for E. andrei. Since the temperature of 20 °C is prescribed by Anonymous (1984; 1985) for acute toxicity testing with earth\\Orms. it was taken as the control for the statistical analysis or the results of this leSl During the rre-incuballon period, worm masses increased strongly, but mass gain was much higher al the lower 1emperatures. Al 10. 15 and 30 °C no cocoons were produced. During 111eubation (phase B). cocoon production was also significantly reduced at these temperatures Growth was significantly increased at 15 Conly. Hatchabilitv or the cocoons (% rerule cocoons) was not affected by temperature. The number or JU;.enile worms per fertile cocoon ranged between 2.20 and 2.97; it was signilicanll} lower (p < 0 01) at 25 C. and not affected al the other temperatures. The or number offspring was significantly lower al all tempera lures tested compared to 20 'C. Pcdobiologia 36 ( 1992) 2 115 In this stuc\y again a correlation was found between the cocoon production during the incubation and pre-incubation periods: r = 0.820. p < 0.01 (n = 20). Cocoon production was not correlated with growth (n = 20: r = - 0.321, n.s.) nor with the initial mass or the worms (n = 20: r = -0.127, n.s.). Estimated biomass gain ranged between 36.3 and 39.4 mg/worm/week at 15-30 °C, but was only 21.7 mg/worm/week al I O° C. 3.2.3. Soil moisture co11ten1 Table 4 summarizes the results or the test performed to study the mnuence of soil moisture content on the reproduction or £. andrei. As was indicated above the moisture content or 55% used in all reproduction studies corresponded with l1eld capacity. Since this is also the moisture content prescribed in international guidelines for earthworm toxicity testing (Anonymous, 1984; 1985), it was considered to be the control in this study. During the pre-incubation period, growth was significantly higher at the highest moisture level: this could partly be due lo the absorption of water by the worms (EDWARDS & LnFTY. 1977). During incubation growth was equal at all moisture levels. Table 4 lnnuencc of soil moisture content on the reproduction and growth of l-.'isl'11ia u11tlrei in ,1r1ific1al soil. soil moisture pre-incubation worm % cocoons juven./ content (%) mass growth worm/ fertile worm cocoons % (mg) worm/ week cocoons week non11nal ac1u:1I worm/ growth/ week week worm/ week 35 34.2 O.JS 8.5 406 2.8 0_93•• 88.2 1.91 ** 45 44.3 0 85 8.6 403 0.4 I .JS• 85.1 2.39 .. 55 54.8 1.23 8.4 406 0.9 1.62 88.4 3.36 65 63.0 1.48 9.5 404 0.0 1.70 93.1 3.77 85 81.4 1.63 12.6 425 -0.1 1.89*• 98.2• 4.93. . 130 126.7 1.35 17.s• 417 2.5 1.69 90.3 3.97 ••• significantly different from 55% moisture at p < 0.05 and p < 0.01, respectively Cocoon production during incubation was significantly lower al the moisture contents or 35% and 45% compared to 55%. At 85% moisture, cocoon production was significantly higher. Cocoon production was correlated with growth (n = 23: r = - 0.543, p < 0.01), with cocoon production in the pre-conditioning period (n = 23: r = 0.871. p < 0.01), but not with initial worm masses (n = 23: r = 0.213, n.s.). The number of juveniles per fertile cocoon ranged between 1.38 and 3.25, and was not affected by soil moisture content. The number of fertile cocoons was significantly higher ut the 85% moisture level. The number of juveniles per worm per week was significantly lower at mOlsture levels of 35 and 45%, and significantly higher at 85%. There was no relation between the number of juveniles per fertile cocoon and hatchability of the cocoons. Estimated biomass production ranged between 15.9 and 21.0 mg/worm/week for the moisture levels or35-85%. and was 28.8 mg/worm/week at the highest moisture level tested. 4. Discussion 4.1. Cocoon production In the literature only one paper was found on cocoon production by E.fetida 111 artificial soil. From the study of VONK e1 al. ( 1986), who did not supply the I 16 Pcdob1olog1a 36 (I 992) 2 worms with food, cocoon production rates between 0.4 and 0.6 cocoons/worms/week could be calculated. In substrates of sand, soil or peat with horse dung or in compost E.fetida and E. a11drei produced between 1.0 and 3.3 cocoons/week (CLUZEAU & FAYOLLE, 1989; F1sc11ER, 1989; NEUllAllSER "'al., 1984Cl, b; NEUHAUSER & CALLAHAN. 1990; ZACllARIAE & EBERT, 1970). In studies of REINECKE & VENTER (1985 a, b) E.fe1ida produced 2.6-3.3 cocoons/week in a substrate of 100% cow manure. HARTENSTEIN et al. (1979) determined the innuence of populauon density or E.f1!1idn on cocoon production in horse dung: at a density of 40-53 worms/liler I - 1.5 cocoons/worms/week were produced. while at lower densities average cocoon producllon increased to 2-4 cocoons/worm/week. From the literature cited, it can be concluded that also other authors found a variallon in cocoon production of a factor of 2 or more within one test. In our study. varia1ion between different experiments was much greater, possibly due to differences in age and mass of the worms, soil pH and some unexplained factors. In the reproduction studies described in this paper, a significant correlation was found between cocoon production in the incubation period and the pre-incubation period. Except for the pH study. in all cases cocoon production was higher in the 3 week incubation period. This ind1ca1es the need for a pre-incubation period to stimulate reproductive activity of E. amlrl!i .in ani!icial soil. Worms cultured on horse dung apparently need to get used to the artifit:ial soil. and since only young adults were used cocoon production might not yet be optimal. As indicated by HARTENSTEIN et al. (1979) cocoon production by E.fetida in horse dung tends to increase during the !irst 10 weeks, and decreases thereafter. 4.2. Growth In our studies a clear negative correlation between growth and cocoon produt:tion was found. Sut:h a correlation can also be derived from BOSTROM (1988) for Aporrl!ctodea caligi11osa. NEUI IA USER et al. ( 1980) found a relation between growth and population density of E.fetida in cow dung, horse dung or activated sludge. When incubated individually. worm masses increased to levels as high as 2000 mg. When population densities were I 0 and 53 worm1liter max11num worm masses amounted to 80 and 300 mg/worm. respectively. According to NEUHAUSER et al. ( 1980) this difference can be attributed to cocoon production (individually incubated worms do not produce cocoons), competition for food and the 1oxicity of worm excreta. The negative correlation between growth and cocoon production was not reproduced in the study on the effect of soil pH. but it was found in the studies on the effecl or soil moisH1re and temperature. When the prcincubation periods were considered, the relation between growth and cocoon production was less pronounced. As noticed in tables I through 4. worms often grew very fast during this pre-incubation period. This might also be due to the much lower population density in the test jars compared to the cultures. 4.3. Food BosTRCi~1 ( 1988) mentioned that the quality of the food source might innuence cocoon production. In our studies the earthworms were all fed with the same food source: cow dung. Since only two lots of cow dung were used for all studies described here, this factor cannot have innuenced the Jest results to a great extent. NEUHAUSER et al. ( 1980) mentioned that an amount of 4.2 g food (horse manure) was sufficient for the growth of E.(etida; they did not mention. however, for how many worms and for how long a period this quantity of rood should suffice. ABBOTT & PARKER ((1981) reported that 10 g sheep dung was enough to allow maturation within three weeks and cocoon production by !ive £./e1idn. Pcdobiologia 36 ( 1992) 2 I I 7 CLUZEAU & FA VOLLE ( 1989) supplied 50 mg food per g worm per day in their incubations with £. a11drei. Considering an average worm mass of 400 mg and 10 worms per jar, this corresponds with 200 mg/day. or c. 4 g in three weeks. In our studies at least 8 g food was given per 10 worms during the three week incubation periotl. Considering the literature data mentioned, this should have been sufficient. The results of the diffcren t tests performed at different food rates confirm this. since no significant differences were found between tests carried out with 8 or with 12 g of food. 4.4. Hatchability of cocoons In the studies of REINECKE & VENTER (1985b, 1987). VENTER & REINECKE (1987). HARTENSTEIN('/(//. (1979), VAIL (1974) and WATANABE & TSUKAMOTO (1976) between 33 anti 88% of the cocoons of £.fe1ida hatched. giving between 1.6 and 3.6 juvenile worms per fertile cocoon. VAN GESTEL el al. (1988) reported that on average 91 -96% of the cocoons of E. a11drei incubated in artificial soil at 20 °C was fertile, giving 2.6-2.8 juvenile worms/fertile cocoon. From tables I to 4 it can be concluded that in most of our studies more than 80% of tht! cocoons hatched. The number of juveniles per cocoon produced in these studies ..:orrespontls with those found in the literature. A relation between the number of juveniles per cocoon and the % fertile cocoons as found in this study was not described in the litt:rature. Only in the study of TSUKAMOTO and WATANABE (1977), who incubated cocoons at different temperatures, such a relation was observed. 4.5. Temperature HARTENSTEIN (1984) studied the inOuence of temperature on mass loss and mass gain by £.fe1ida and concluded that its ''metabolic zone of thermal compensation" was between 20 and 25 °C. EDWARDS & LOFTY (1977) mentioned an optimum temperature of 16-23 °C for E..fe1ida. KAPLAN e1 al. ( 1980) studied the inOuence of temperature on growth and survival of E.fe1ida in horse dung and activated sludge: growth was optimal at 20- 29 °C; ,ll 33 ~c high mortality occurred. TSUKAMOTO & WATANAllE (1977) found that growth of l fe1ula was signilicantly higher at tempera lures of20 and 25 °C compared to IO and 15 °C. On the other hand they concluded that cocoon hatchability decreased with increasing temperature: as we did not incubate cocoons at different temperatures, this effect could not be reproduced in our study. Cocoons produced at different temperatures, but incuba!ed a1 the same temperalure. showed the same ha1ching raies. From our resul!s it can be concluded that the temperature prescribed by Anonymous ( 1984. 1985) for acute toxicity testing with earthworms is also optimal for reproduction 1ests. 4.6. Soil moisture content In the literature no daia on the influence of soil moisture con lent on £.fe1ida or£. andrei were found. Some authors (HAUKKA, 1987; REINECKE & VENTER, I 985a, 1987) reported opiimal soil moisture levels for these species in pure cow or horse dung and compost. Since no pF values were determined for these subsirates, results cannot be compared with !hose or this study. For other earthworm species optimal soil moisture content seemed to be at pF values be1wcen 2.0 and 3.3. At higher pF values activity decreases, whereas at pF below 2.0 mortality may occur due to saturation of the soil (Lee, 1985). Our results demonslrate that for£. andrei pF values below lield capacity (85%, pF 1.4) are op1imal for reprodµction. Unlike mentioned by LEE (1985) moisture contents below I 18 Pcdobiolog1a 36 ( 1992) 2