Effect of soil on microbial responses to metal contamination

Effect of soil on microbial responses to metal contamination

Environmental Pollution 110 (2000) 115±125 www.elsevier.com/locate/envpol E€ect of soil on microbial responses to metal contamination M. Khan, J. Sc...

275KB Sizes 0 Downloads 34 Views

Environmental Pollution 110 (2000) 115±125

www.elsevier.com/locate/envpol

E€ect of soil on microbial responses to metal contamination M. Khan, J. Scullion* Soil Science Unit, Institute of Biological Sciences, University of Wales, Aberystwyth SY23 3DE, UK Received 17 December 1998; accepted 15 October 1999

``Capsule'': The e€ect of the treatments on microbial and extractable indices was greatest in loam soils. Abstract An experiment was conducted to investigate microbial responses to metal inputs in ®ve soils with varying clay and organic contents; one soil had also a higher pH. These soils were treated with a low metal, sewage sludge control or with this sludge contaminated to achieve Cu=112, Ni=58 and Zn=220 mg kgÿ1 in medium and Cu=182, Ni=98 and Zn=325 mg kgÿ1 in high metal soils. CO2 evolution rates were measured at 1 week and at 4±5-day intervals thereafter until the end of the incubation (7 weeks). Extractable metals (CaCl2 and water), biomass C, metabolic quotient, ergosterol, bacterial±fungal phospholipid fatty acid (PLFA Ð 3 weeks only) ratio and mineral N were measured at 3 and 7 weeks. Metal inputs caused a marked increase in metal availability in the slightly acidic sandy loams, a smaller increase in slightly acidic clays and had little e€ect in the alkaline loam. After an initial increase in CO2 evolution with metal inputs in all soils, the high metal treatment alone caused a signi®cant decrease at later stages, mainly in sandy loams. Although biomass C and metabolic quotient decreased in all soils with higher metal inputs, the e€ect was more pronounced in the sandy loams. Metal inputs increased ergosterol and decreased bacterial±fungal PLFA ratios in most soils. Larger mineral N contents were found in all high metal soils at 3 weeks but, after 7 weeks, metals caused a signi®cant decrease in sandy loams. CaCl2 and water-extractable Cu, Ni and Zn contents were closely correlated with microbial indices in sandy loam but not in clay soils. Overall, the e€ect of treatments on microbial and extractable metal indices was greater in loams. Within a single series, higher organic soils showed less pronounced responses to metal inputs, although this trend was not always consistent. # 2000 Elsevier Science Ltd. All rights reserved. Keywords: Soil types; Available metals; Respiration; Ergosterol; Mineral N

1. Introduction Although metals such as Cu, Ni and Zn are essential nutrients for living organisms (Alloway, 1995), they may be toxic at high concentrations. Soil metal content is determined by the nature of the parent material and by inputs of metals from sources such as sewage sludge, mining and smelting, impurities in agricultural and horticultural materials and fossil fuel combustion. Atmospheric deposition results in di€use contamination, whereas a more concentrated contamination arises where sewage sludge, metal-containing pesticides or fertilisers are used extensively (McGrath et al., 1995). Owing to its relatively high concentrations of metals, sewage sludge is a major potential source of metals in * Corresponding author. Tel.: +44-1970-622303/04; fax: +441970-622307. E-mail address: [email protected] (J. Scullion).

agricultural soils to which it is applied to improve fertility (Smith et al., 1989) and physical conditions (Pagliai et al., 1981). Soil microorganisms play an essential role in the decomposition of soil organic matter. Any reduction in the diversity or abundance of microorganisms may a€ect the cycling of plant nutrients in the soil (Giller et al., 1998). Any disturbance to the soil ecosystem can disrupt microbial activity and hence nutrient cycling. Numerous laboratory and ®eld studies have demonstrated the adverse e€ect of metals on the soil ecosystem. Signi®cant reductions in microbial biomass have been found in metal-contaminated compared to uncontaminated soil (FrostegaÊrd et al., 1993b; Flieûbach et al., 1994; Leita et al., 1995). Soil respiration responses to metal contamination are, however, less consistent. Doelman and Haanstra (1984), BaÊaÊth et al. (1991) and Hattori (1992) found a signi®cant decrease in CO2 evolution in metal-contaminated soil. In contrast, others

0269-7491/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved. PII: S0269-7491(99)00288-2

116

M. Khan, J. Scullion / Environmental Pollution 110 (2000) 115±125

would in¯uence any toxic e€ects on microorganisms. Metals were added in sewage sludge since this is a major pathway for metal input to agricultural soils and there is evidence (Aoyama et al., 1993) that metals are less toxic when applied with organic wastes than where they are applied as mineral salts.

(Bardgett and Saggar, 1994; Flieûbach et al., 1994) reported increased CO2 evolution in metal-polluted soils. These di€ering ®ndings may result from variations in levels of metal contamination, in the source of the contamination (e.g. sewage sludge or mining), in the period of time over which responses were monitored and in characteristics of the receiving soil. Mineral N responses to metal contamination, which are usually positive, have been attributed to higher mineralization rates and the release of N from dead microbial cells (Bogomolov et al., 1996), and to decreased microbial immobilisation (Westermann and Tucker, 1974). Studies have shown that metal contamination causes also a shift within the soil microbial community from sensitive to less sensitive microorganisms (Maliszewska et al., 1985; Giller et al., 1989). Some species may be eliminated whilst others increase in abundance because of reduced competition for substrate (van Beelen and Doelman, 1997). The replacement of sensitive by resistant species may not result in any net e€ect on broad microbial indices such as soil respiration or total biomass. As noted above, contradictory ®ndings in relation to microbial responses to soil contamination may be due, at least in part, to variations in metal availability between di€erent soils. Changes in organic matter and clay content can markedly a€ect the microbial availability of metals in soils (Dar, 1996). Only limited research has been carried out into the in¯uence of soil type on microbial responses to metal inputs (e.g. Hattori, 1991). Most comparisons between soils are reported in reviews of studies where experimental procedures are not always comparable (e.g. Giller et al., 1998). Also, the natural variability of soils, and the e€ects of management on organic matter contents and pH, make it dicult to specify a particular index which represents the e€ect of increased metal concentrations on soil microorganisms (Huysman et al., 1994). The aim of the study reported here was to investigate the combined e€ect of metals (Cu, Ni, Zn), applied at rates close to current UK limits (DoE, 1989), on a wide range of microbial indices in ®ve soils with contrasting properties. The main hypothesis was that soils di€ering in their clay and organic contents would have varying metal availability and that these variations

2. Materials and methods The experiment involved incubating ®ve di€erent soils, amended with three sludges of varying metal content, in ¯asks under controlled conditions (in the dark at 20‹1 C). Respiration measurements started after 1 week and were repeated at 4±5-day intervals to the end of the experiment at 7 weeks. Extractable metals, pH, microbial indices, and mineral N in soils were measured after 3 weeks, in the ®nal stages of a period of marked decline in microbial respiration, and at the end of the experiment (7 weeks) when respiration rates had stabilised. All analytical results are means of three replicates. 2.1. Soils and treatments Two of the soil series used (Ru€ord Ð sandy clay loam; Bromyard Ð clay) were chosen because of their marked variation in clay content but similar, slightly acidic pH (Table 1). For these soil series, adjacent grassland [Ru€ord Grass (RG) and Bromyard Grass (BG)] and arable [Ru€ord Arable (RA) and Bromyard Arable (BA)] sites were sampled to provide soils of similar clay but di€ering organic content. The ®fth soil used in the experiment was mapped as Salop (clay loam) series and under arable cropping (SA), with a clay content somewhat higher than that of the Ru€ord series soils and a slightly alkaline pH. Cation exchange capacity was markedly higher in the two Bromyard soils compared with the others used in the experiment. Samples were taken to a depth of 10 cm and bulked. Plant material, stones and visible soil fauna were picked out by hand and the soils were then sieved at 2 mm prior to use in the experiment. Total Cu and Ni contents (Table 2a) varied little between soils. However, total Zn contents were markedly higher in the Ru€ord grassland compared with

Table 1 Some physicochemical properties of soils used in this experiment Soil code

Soil series

Clay (%)

Silt (%)

Sand (%)

Loss on ignition (%)

pH

CECa (meq kgÿ1)

RA RG SA BA BG

Ru€ord Ru€ord Salop Bromyard Bromyard

22.7 24.5 30.3 53.2 57.4

15.1 16.8 22.8 43.4 41.5

61.9 58.7 40.8 3.4 1.1

3.9 3.9 4.5 8.5 10.6

6.3 6.5 7.4 6.3 6.6

69.3 73.4 99.2 254.8 335.7

a

CEC, cation exchange capacity.

M. Khan, J. Scullion / Environmental Pollution 110 (2000) 115±125

117

Table 2 Total metal concentrations (mg kgÿ1) in (a) unamended soils and (b) sludge-amended soils as estimated from metal contents in added sludge and measured soil total metal contents prior to sludge inputa

this moisture potential. The moisture condition of each soil incubated was adjusted to the gravimetric water content corresponding to this moisture potential.

(a) Soil type

Cu

Ni

Zn

2.2. Soil analyses

RA RG SA BA BG

22.4 21.2 25.7 16.0 24.0

20.2 26.5 28.3 19.2 24.0

72.4 104.6 72.0 50.0 59.7

(b) Treatment

Cu

Low Medium High UK limits (DoE, 1989)b a b c

Ni c

19±28 112 182 135

Zn c

20±29 58 98 75

62±116c 220 325 300

See Table 1 for key to soils. Background total metal concentrations varied between soils. DoE (1989), pH ranges from 6 to 7.

others. As a result, inputs of Zn in added sludge were less for this soil. Total metal contents in sludgeamended soils were calculated from measured soil concentrations and sludge metal loadings. All soils were amended with a standard sewage sludge, containing three levels of metals (Table 2b). The `control' or low metal treatment was a sludge obtained from Aberystwyth sewage works, UK; this source had a very low sludge metal content because it serves a rural area. Other sludge treatments were arti®cially contaminated with salts of Cu, Ni and Zn (copper sulphate, nickel chloride, zinc sulphate) to achieve soil levels slightly below (medium) or above (high) the UK limits at the rate of sludge application (DoE, 1989). Preliminary studies (Khan, 1999) had indicated that these metals had a particularly marked e€ect on microorganisms at such concentrations; these studies also showed that the salt used had little e€ect on microorganisms at the input rates applied. Prior to mixing with soils, contaminated sewage sludge was incubated for at least 1 week after metal treatment to allow for some metalsludge equilibration. Sewage sludge was applied to soils at a rate of 40 g (oven dry) per kg dry soil. This sludge input rate was high on a per hectare basis, depending on assumptions of incorporation depth, but not unrealistic given the uneven distribution of applied sludges on the micro-scale relevant to micro-organisms. Microbial activity has been shown to be at a maximum for soil moisture contents slightly below water holding capacity (Cheng and Coleman, 1989). However, this optimal moisture condition is likely to vary from one soil to another. As the soils used in this experiment di€ered in texture and organic matter, a standard moisture potential of ÿ50 kPa was maintained during incubations. Sub-samples of each soil were saturated with water and then centrifuged (Piper, 1950) to achieve

Loss on ignition (Ball, 1964), cation exchange capacity (CEC) (Chapman, 1965), particle size analysis and pH (water) (MAFF, 1986), `total' (hot, concentrated HNO3 digestion), 0.1 M CaCl2 and water-extractable metals (McGrath and Cegara, 1992) were measured in all soils. CaCl2 and water-extractable metals are considered (Ure, 1995) to provide a better indication of bioavailability than total metal contents. 2.3. Microbial indices For soil respiration, incubation ¯asks were ®tted with Suba Seals for 6 h and CO2 concentration in the headspace measured by gas chromatography (Pye-Unicam Series 104 Chromatograph) (Anderson and Domsch, 1978). The fumigation±extraction method was used for estimating microbial biomass C, with a standard conversion factor for transforming extractable to biomass C (Vance et al., 1987). Soils were fumigated with chloroform for 24 h in a desiccator. Carbon was extracted using 1 M K2SO4 and the ®ltrates were analysed in a Shimadzu Total Organic Carbon Analyser (TOC-5050). Metabolic quotient (Odum, 1985), the ratio of respired to biomass C, was calculated as an indicator of stress within the microbial community. Mineral N in the soil was determined by the steam distillation method of Keeney and Nelson (1982). Ergosterol is found only in fungi and certain green microalgae (Newell et al., 1987) and ergosterol content has been used as an indicator of fungal abundance in soil. Ergosterol was extracted from soil following the methodology of Donnison (1997). Cyclohexane was added to soil samples that were then saponi®ed and sonicated. The upper organic phase was removed and evaporated under nitrogen gas at 40 C and stored at ÿ20 C until the ergosterol content was measured by high pressure liquid chromatography (HPLC). Phospholipid fatty acids (PLFAs) may provide an indication of changes in the composition of microbial populations in polluted soils (e.g. FrostegaÊrd et al., 1993b). Because of limited access to this analytical procedure, PLFAs were measured on only four soils (excluding BG) and at 3 weeks incubation. The extraction procedure used was that of White et al. (1979) and the resulting fatty acid methyl-esters were analysed on a Hewlett Packard 5890II GC equipped with a 5972A mass selective detector (MSD II). Components were identi®ed by chromatographic retention time and mass spectral comparison using a standard qualitative bacterial acid methyl ester mix (Supelco) that ranged from

118

M. Khan, J. Scullion / Environmental Pollution 110 (2000) 115±125

C11 to C20. Combined signals for the fatty acids i15:0, 15:0, i16:0, 16:0, 17:0, i17:0 and 18:1 were used as an indicator of total bacteria PLFA (FrostegaÊrd et al., 1993a). The 18:2 fatty acid was used as an indicator of fungal PLFA (FrostegaÊrd et al., 1993a). From these two measurements bacterial:fungal PLFA ratios were calculated. Full PLFA data are reported elsewhere (Khan, 1999). 2.4. Statistical analysis Soil and metal input treatment means were compared by two-way analysis of variance using the STATGRAPHICS Version 6.0 statistical package (Manugistics, 1992). Relationships between extractable metal concentrations and microbial indices were also assessed by correlation analysis using the same package; only data for 7 weeks were directly comparable for correlation analyses. 3. Results

CaCl2-extractable Cu decreased in most soils, Ni increased markedly in the two loam soils only and Zn concentrations showed no consistent pattern. Extractable Ni and Zn levels were highest in the Ru€ord soils and lowest in the high pH Salop (SA) soil, with the clays (BA, BG) intermediate. Cu concentrations were again highest in the sandy loams but lowest in the clays. As expected, both medium and high input treatments signi®cantly increased extractable levels of added metals after both 3 and 7 weeks incubation. However, for all metals and both extraction procedures there were highly signi®cant ( p<0.001) soil-metal input treatment interactions. These took two forms as illustrated by data for CaCl2 metal concentrations at 7 weeks (Fig. 1). For all Cu extractions, metal inputs increased concentrations markedly only in the sandy loams, particularly that under arable cropping. For Ni and Zn, inputs caused a similar increase in the sandy loams, a less marked increase in the two clays, but had little e€ect on concentrations in the higher pH Salop soil. 3.2. Microbial analyses

3.1. Soil analyses Soil pH and extractable metal data are presented in Tables 3 and 4 . The pH of all soils at 3 and 7 weeks was lower than that prior to treatment, with this reduction being more pronounced in the Ru€ord (RA, RG) soils. Increased metal inputs caused a progressive decline in pH, although these e€ects were small in comparison to soil di€erences. Water-extractable Cu and Zn concentrations decreased between weeks 3 and 7, whereas Ni concentrations increased. Over the same period,

Respiration data for di€erent soils and metal inputs are shown in Fig. 2a and b. Throughout the incubation period, respiration was highest in the two sandy loams (RA, RG) and, in most cases, lowest in the two clays (BA, BG). Responses to metal inputs varied with time. At 7 days, CO2-C evolution was higher in the medium and high input, compared with the low input soils. Between days 11 and 21, di€erences between metal treatments were generally non-signi®cant. After this time, the high input treatment only signi®cantly decreased respiration.

Table 3 E€ects of (a) soil type and (b) metal inputs on pH, CaCl2 and water-extractable metals (mg kgÿ1) after 3 weeks of incubationa (a)

RA

RG

SA

BA

BG

Signi®canceb

pH

5.67d

5.68d

7.22a

6.27b

6.06c

***

Cu±CaCl2 Cu±water

3.23a 1.54a

2.34b 1.41a

1.08c 1.35a

0.68d 0.72c

1.14c 1.01b

*** ***

Ni±CaCl2 Ni±water

12.55a 2.00a

11.05b 1.55b

3.03e 0.52e

6.46d 0.76d

7.31c 1.01c

*** ***

Zn±CaCl2 Zn±water

71.56a 9.01a

54.42b 6.42b

4.35e 1.81d

24.60d 1.78d

38.72c 2.88c

*** ***

(b)

Low

Medium

High

Signi®canceb

pH

6.30a

6.20b

6.04c

***

Cu±CaCl2 Cu±water

0.42c 0.03c

1.69b 1.02b

2.97a 2.31a

*** ***

Ni±CaCl2 Ni±water

1.14c 0.25c

4.76b 0.67b

18.34a 2.57a

*** ***

Zn±CaCl2 Zn±water

6.44c 0.56c

37.19b 3.42b

72.57a 9.16a

*** ***

a b

Mean values within rows which have a common letter sux do not di€er at a 5% level of probability. See Table 1 for key to soils. ***p<0.001 for overall treatment di€erence.

M. Khan, J. Scullion / Environmental Pollution 110 (2000) 115±125

119

Table 4 E€ects of (a) soil type and (b) metal inputs on pH, CaCl2 and water-extractable metals (mg kgÿ1) after 7 weeks of incubationa (a)

RA

RG

SA

BA

BG

Signi®canceb

pH

5.17e

5.39d

7.12a

5.92b

5.76c

***

Cu±CaCl2 Cu±water

3.25a 2.17a

1.81b 1.5lb

0.37c 0.87d

0.40c 0.85d

0.32c 1.15c

*** ***

Ni±CaCl2 Ni±water

16.78a 3.58a

13.20b 3.61a

2.79e 0.70c

6.34d 0.88b

6.64c 0.99b

*** ***

Zn±CaCl2 Zn±water

72.35a 5.35a

66.83b 3.35b

2.44e 0.23d

27.93d 0.38cd

31.09c 0.58c

*** ***

(b)

Low

Medium

High

Signi®canceb

pH

6.00a

5.84b

5.77c

***

Cu±CaCl2 Cu±water

0.39c 0.66c

0.63b 1.22b

2.66a 2.05a

*** ***

Ni±CaCl2 Ni±water

2.18c 0.62c

6.24b 1.13b

19.03a 4.11a

*** ***

Zn±CaCl2 Zn±water

2.44c 0.26c

38.75b 1.24b

79.19a 4.43a

*** ***

a b

Mean values within rows which have a common letter sux do not di€er at a 5% level of probability. See Table 1 for key to soils. ***p<0.001 for overall treatment di€erence.

Fig. 2. Respiration data for soil incubations; e€ect of (a) soil type and (b) metal input rate. Error bars indicate standard errors for overall mean. See Table 1 for key to soils. Fig. 1. Soil-metal input interactions for CaCl2-extractable metals (a) Cu ( p<0.001), (b) Ni ( p<0.001) and (c) Zn ( p<0.001) after 7 weeks. See Table 1 for key to soils.

As with extractable metal data, there were highly signi®cant soil-metal input interactions on most sampling

occasions, but the nature of these interactions varied with time. Results for 7, 21 and 49 days are given as representative of these variations (Fig. 3). At 7 days, metal inputs caused a small increase in most soils, but this increase was more marked for high inputs to the SA

120

M. Khan, J. Scullion / Environmental Pollution 110 (2000) 115±125

soil and medium inputs to the RA soil; high inputs to the RA soil decreased respiration. At 21 days, most soils showed a small increase in respiration with additional metal inputs but the SA soil gave a progressive decrease.

At the end of the incubation period, when respiration rates had broadly stabilised, the three loam soils (RA, RG, SA) showed decreased respiration with high metal inputs whereas the clay soils (BA, BG) showed a slight increase. Over the full incubation, elevated metal inputs decreased cumulative CO2-C evolution by up to 12% in the loam soils but had no e€ect on the clays. Biomass C data for 3 and 7 weeks show a marked reduction for all soils over this period (Tables 5 and 6 ). Grassland soils had higher biomass than corresponding arable soils, and clay soils higher biomass than equivalent sandy loam soils; biomass in the high pH arable soil (SA) was at intermediate levels. Metal inputs consistently reduced biomass, but di€erences between input levels were not always signi®cant. Soil-metal input interaction e€ects were not highly signi®cant. Metabolic quotients (Tables 5 and 6) decreased in most soils between 3 and 7 weeks. Quotients were highest in the sandy loams (RA, RG) and lowest in the clay grassland soil (BG). Only the high metal treatment at 3 weeks signi®cantly increased this index, although higher input treatments tended to have high quotients. There were highly signi®cant soil-metal input interactions on both sampling occasions (Fig. 4). At 3 weeks only the RA soil showed a marked and progressive increase in quotients with higher metal inputs. At 7 weeks, the two clay soils showed a small but progressive increase with higher metal inputs, quotients for RG and SA soils were increased only by medium metal inputs, whereas in the RA soil metal inputs caused a progressive decrease. Ergosterol and bacterial±fungal PLFA ratio (3 weeks only, BG excluded) data are given in Tables 5 and 6. Ergosterol levels di€ered signi®cantly between soils on both sampling occasions, but there was no consistent relationship with soil type or management. Bacterial± fungal PLFA ratios were lower in RA and BA soils than

Fig. 3. Soil-metal input interactions for CO2-C evolution (a) after 1 week ( p<0.001), (b) after 3 weeks ( p<0.001) and (c) after 7 weeks ( p<0.001) of incubation. See Table 1 for key to soils.

Table 5 E€ects of (a) soil and (b) metal input rate on microbial biomass C (mg kgÿ1), ergosterol content (mg kgÿ1), metabolic quotient (qCO2, mg of CO2-C mgÿ1 biomass C hÿ1), bacterial to fungal phospholipid fatty acids (PLFAs) ratio (Baci:Fung ratio) and mineral N content (mg kgÿ1) after 3 weeks incubationa (a)

RA

Biomass C Ergosterol qCO2 Bact:Fung PLFA Mineral N

875.82d 14.33b 5.23a 4.88b 56.90b

(b)

Low

Biomass C Ergosterol qCO2 Bact:Fung PLFA Mineral N

1442.21a 12.51c 3.36b 5.66a 24.59c

a b c

RG 1187.99c 15.90a 4.45b 5.56a 70.16a Medium 1409.46a 14.12b 3.58b 5.81a 32.77b

SA

BA

BG

Signi®canceb

1315.72b 14.89b 3.51c 5.79a 17.08d

1366.30b 15.26ab 3.15c 5.02b 14.18d

2042.86a 11.29c 1.94d NDc 20.98c

*** *** *** ** ***

High

Signi®canceb

1221.42b 16.37a 4.03a 4.47b 49.63a

*** *** ** *** ***

Mean values within rows which have a common letter sux do not di€er at a 5% level of probability. See Table 1 for key to soils. ***p<0.001, **p<0.01 for overall treatment di€erence. Not determined.

M. Khan, J. Scullion / Environmental Pollution 110 (2000) 115±125

121

Table 6 E€ects of (a) soil and (b) metal input rate on microbial biomass C (mg kgÿ1), ergosterol content (mg kgÿ1), metabolic quotient (qCO2, mg of CO2-C mgÿ1 biomass C hÿ1) and mineral N content (mg kgÿ1) after 7 weeks incubationa (a)

RA

Biomass C Ergosterol QCO2 Mineral N (b)

Low

Biomass C Ergosterol qCO2 Mineral N a b

432.27e 12.39a 4.84a 107.34b

766.35a 10.84b 3.15 81.11

RG

SA 497.38d 12.57a 4.71a 117.59a

Medium 697.93b 11.32b 3.43 83.12

827.52b 9.97b 3.01b 57.09d High 603.63c 12.20a 3.26 78.80

BA

BG

Signi®canceb

583.56c 12.54a 2.64c 53.43d

1105.79a 9.79b 1.22d 69.60c

*** *** *** ***

Signi®canceb *** *** NS NS

Mean values within rows which have a common letter sux do not di€er at a 5% level of probability. See Table 1 for key to soils. ***p<0.001, **p<0.01; NS, non-signi®cant for overall treatment di€erence.

Fig. 4. Soil-metal input interactions for qCO2 (metabolic quotient) (mg CO2-C mgÿ1 biomass C hÿ1) (a) after 3 weeks ( p<0.01) and (b) after 7 weeks ( p<0.001) of incubation. See Table 1 for key to soils.

in either RG or SA soils. Metal inputs resulted in a progressive increase in ergosterol content but decreased bacterial±fungal PLFA ratios only at the high metal input rate. Again there were highly signi®cant soil-metal input interactions for ergosterol and bacterial±fungal PLFA ratios, but at 3 weeks only. Whereas most soils showed progressive increases in ergosterol content with higher metal inputs, there was a decrease in the RA soil (Fig. 5a). Medium inputs of metals caused a particularly marked increase in the bacterial±fungal PLFA ratios of the SA and BA soils; high inputs increased these ratios in the SA soil but caused a reduction in RA, RG and BA soils (Fig. 5b). Mineral N (Tables 5 and 6) increased in all soils between weeks 3 and 7, being particularly high in the

Fig. 5. Soil-metal input interactions for (a) ergosterol content ( p<0.001) and (b) bacterial to fungal phospholipid fatty acids (PLFAs) ratio ( p<0.001) after 3 weeks of incubation. See Table 1 for key to soils.

sandy loams. Metal inputs caused a marked and progressive increase in mineral N only at 3 weeks. On both sampling occasions there were highly signi®cant interaction e€ects (Fig. 6). At 3 weeks, only the sandy loams showed a marked positive response to metal inputs. At 7 weeks, metal inputs tended to decrease mineral N in these loams, but the reverse was the case in the other soils. 3.3. Relationships between microbial and extractable metal indices Correlations between microbial indices±mineral N and both CaCl2 and water-extractable metals are presented in Table 7. Correlation coecients were

122

M. Khan, J. Scullion / Environmental Pollution 110 (2000) 115±125

Fig. 6. Soil-metal input interaction for mineral N content (a) after 3 weeks ( p<0.001) and (b) after 7 weeks ( p<0.001) of incubation. See Table 1 for key to soils.

Table 7 Correlation coecients between microbial indices-mineral N and extractable metal concentrations (a) all soils (n=45) (b) sandy loam (n=18) (c) clay (n=18) Ergosterola

CO2-Ca

Mineral Na

(a) CaCl2 extractable Cu ÿ0.4846***b Ni ÿ0.4775*** Zn ÿ0.5072***

0.5829*** 0.6194*** 0.6253***

0.0640 ÿ0.2165 ÿ0.2334

0.2512 0.2624 0.3461*b

Water extractable Cu ÿ0.4679**b Ni ÿ0.4420** Zn ÿ0.4790***

0.5872*** 0.5032*** 0.5944***

ÿ0.0967 ÿ0.1990 ÿ0.0883

0.3039* 0.3227* 0.2947*

(b) CaCl2 extractable Cu ÿ0.5850* Ni ÿ0.6579** Zn ÿ0.7546***

0.8310*** 0.8528*** 0.8048***

ÿ0.6717** ÿ0.7668*** ÿ0.7947***

ÿ0.6679** ÿ0.5935** ÿ0.5480*

Water extractable Cu ÿ0.5324* Ni ÿ0.6547** Zn ÿ0.6303**

0.8221*** 0.8334*** 0.8776***

ÿ0.6848** ÿ0.6948** ÿ0.7071**

ÿ0.5503* ÿ0.6509** ÿ0.6133**

(c) CaCl2 Cu Ni Zn

0.3438 0.2228 0.1869

0.4225 0.2217 0.1250

0.1296 0.4420 0.4984*

ÿ0.1489 0.1738 ÿ0.1586

ÿ0.2239 0.1680 ÿ0.1585

Metal

Biomass Ca

extractable ÿ0.4938 ÿ0.2336 ÿ0.1922

Water extractable Cu 0.0866 Ni ÿ0.1449 Zn 0.1140

0.5947** 0.5382* 0.6433**

a Biomass C=microbial biomass C (mg kgÿ1), ergosterol= ergosterol content (mg kgÿ1), CO2-C=CO2-C evolution in terms of C output (mg gÿ1 hÿ1); mineral N, mineral nitrogen (mg kgÿ1). b *p<0.05, **p<0.01, ***p<0.01.

calculated for all soils, for clay (BA, BG) soils and for loam soils with similar pH (RA, RG). Respiration rates and soil mineral N levels were signi®cantly and negatively correlated with extractable metals within the Ru€ord group of soils. Mineral N was weakly but positively correlated with water-extractable metals across all soils and within the clays; of the three metals added, extractable Zn tended to have the closest correlations. Biomass C showed signi®cant, negative relationships with CaCl2 and water-extractable metals across all soils, for sandy loams (RA, RG), but not for clays (BA, BG). In most cases, biomass C was most closely related to extractable Zn; there was no consistent pattern in favour of CaCl2 or water-extractable metals. Ergosterol contents correlated positively with extractable metals across all soils and within the sandy loams only. There was no consistent pattern in favour of any single metal or extraction procedure with respect to the closeness of these relationships. 4. Discussion The main hypothesis in the present study was that metal availability is controlled by soil physicochemical properties and that, for this reason, similar inputs of metals to contrasting soils may have di€ering e€ects on biological activity. The discussion will concentrate on this aspect of the experimental ®ndings. In a previous study, Hattori (1992) found that metals inhibited CO2 evolution more in a soil with low organic matter and cation exchange capacity (4.4 meq 100 gÿ1) than in a soil of higher organic matter and cation exchange capacity (19.0 meq 100 gÿ1). Similarly, Maliszewska et al. (1985) found that metals were, in general, less toxic to microorganisms in ®ne-textured alluvial than in sandy loams, a ®nding attributed to variations in sorption capacity. Results from the present study broadly support these ®ndings. Both metal availability and microbial responses to metal inputs di€ered markedly between soils. With higher metal inputs, increases in exchangeable (CaCl2) and water-soluble metal concentrations were much more pronounced in soils with lower cation exchange capacities. These results were obtained despite the somewhat lower inputs of sludge Zn to the Ru€ord soils. The high pH of the Salop soil was an additional factor limiting availability of Ni and Zn. Studies (e.g. Je€ery and Uren, 1983) have shown that Zn solubility is more sensitive to pH than that of Cu. There was clear evidence that metal inputs applied at rates within current UK guidelines (DoE, 1989) a€ected the short-term size, composition and activity of microbial populations, particularly in soils with higher metal availability. In relation to microbial respiration, it has been argued that high metal concentrations decrease CO2 evolution

M. Khan, J. Scullion / Environmental Pollution 110 (2000) 115±125

(Doelman and Haanstra, 1984) whereas moderate contamination leads to higher respiration rates (Bardgett and Saggar, 1994; Flieûbach et al., 1994; Leita et al., 1995). Since most of the above studies report only total rather than available metal contents, it is not possible to make any direct comparison between responses obtained in these investigations and those reported here. The direction of respiration response varied with time in the present study, being positive shortly after metal inputs but negative some weeks later. Since metal extractions were not carried out at week 1, this change cannot be related to varying metal availability, although responses were in general more pronounced in soils where metal inputs caused a greater increase in availability. Changes in microbial population structure may also be a factor in respiration responses to metals (Flieûbach et al., 1994). Between weeks 3 and 7, there was an apparent increase in the proportion of the microbial biomass as fungi, particularly in high metal soils, as indicated by the much smaller reduction in ergosterol compared with biomass C content over this period. This pattern may have represented a continuation of trends set in the period prior to 3 weeks. If this were so, early respiration data would have been determined largely by bacterial responses to metal inputs whilst later responses increasingly re¯ected those of soil fungi. In contrast to respiration, microbial biomass C has been found to be consistently lower in higher metal soils (Bardgett and Saggar, 1994; Leita et al., 1995). Data reported here con®rm this trend. There was a slight tendency towards more pronounced reductions in biomass C with metal inputs to loams with less organic matter than in clays with more organic matter. However, the in¯uence of soil type on biomass response was not large. An increase in metabolic quotient (qCO2) with metal pollution has been used as an indicator of microbial stress and interpreted as microorganisms using more energy for maintenance rather than biomass synthesis (Bardgett and Saggar, 1994; Flieûbach et al., 1994). Results obtained here could be explained in part by variations in available metal concentrations in di€erent soils, but there was little change in metabolic quotient between weeks 3 and 7 despite an increase in metal availability in many soils. At 3 weeks, qCO2 responses are in close accordance with those of Brookes and McGrath (1984), Bardgett and Saggar (1994) and Flieûbach et al. (1994). However at 7 weeks, there was no clear trend, indicating some of the reported limitations of this index of microbial status (Insam et al., 1996) where di€erent soils are compared. The absence of any clear response to metal inputs at 7 weeks may be related to the increasing proportion of fungi in the microbial population. Huysman et al. (1994) found that fungi are more resistant to copper inputs than bacteria. In the present experiment there was a consistent increase in ergosterol

123

content in the high metal soils after 3 weeks, although this di€erence remained at 7 weeks in only three of the ®ve soils. This increase in fungal biomass was associated with a decrease in bacterial:fungal PLFA after 3 weeks, due to a combination of lower bacterial and higher fungal PLFA (Khan, 1999). Other authors have found an increase in fungal counts (Ohya et al., 1985; Hattori, 1992) and fungal phospholipid fatty acids (FrostegaÊrd et al., 1993b) in metal-contaminated soils. It is likely that increases in fungal biomass are due to reduced bacterial competition for substrate. Studies on N mineralization in metal-contaminated soils, as reviewed by BaÊaÊth (1989), have produced variable ®ndings. Similar inconsistencies between soils and sampling times were obtained here. Only the sandy loams produced a marked response to metal inputs, positive at 3 but negative at 7 weeks. This response pattern was similar to that for respiration. It may be explained largely on the basis of metal e€ects on sludge N mineralization at 3 weeks and, perhaps, both sludge and biomass N mineralization at later stages. Bogomolov et al. (1996) considered that where increased mineral N occurred in metal-contaminated soils, it was released from dead microbial cells. Assuming microbial C:N ratios in the normal range, all of the increase in mineral N observed in the present study between weeks 3 and 7 could be attributed to the decline in microbial biomass over this period. As discussed above, similar metal inputs lead to varying microbial responses in di€erent soils and these variations have been attributed, at least in part, to variations in extractable metal concentrations. However, correlation analyses show that relationships between soil metal concentrations and various microbial indices are not consistent, either between individual indices or between di€erent soil types. Indeed, few signi®cant correlations were obtained for clay-rich soils. These ®ndings suggest that reliable predictions of microbial status, based on extractable metal concentrations, can be obtained only where predictions are restricted to soils having similar properties. 5. Conclusions It can be concluded that variations in soil properties have a major e€ect on the proportion of the total metals present in sludge to which microorganisms are exposed and that these variations a€ect microbial response, at least in the short term. Metal inputs in sludges within current UK limits a€ected microbial populations and their activity, ®ndings consistent with those of Flieûbach et al. (1994). Although soil pH in¯uences the availability of metals and their toxicity, the role of organic matter and clay content should also be considered, as in the Netherlands (Berg et al., 1993),

124

M. Khan, J. Scullion / Environmental Pollution 110 (2000) 115±125

when devising guidelines for soil metal pollution. Assessments of metal e€ects on microbial populations require a range of parameters to be measured, particularly where comparisons involve soils with contrasting properties. Acknowledgements The ®rst author acknowledges the Ministry of Education, Government of Pakistan for ®nancial support for this project. Thanks are also due to Dr. Gwyn Grith, Institute of Grassland and Environmental Research, Aberystwyth, UK, for his assistance in ergosterol and PLFA determinations.

References Alloway, B.J., 1995. Soil processes and the behaviour of metals. In: Alloway, B.J. (Ed.), Heavy Metals in Soils. Blackie Academic and Professional, Glasgow, pp. 3±10. Anderson, J.P.E., Domsch, K.E., 1978. A physiological method for the quantitative measurement of microbial biomass in soils. Soil Biology and Biochemistry 10, 215±221. Aoyama, M., Itaya, S., Otowa, M., 1993. E€ects of copper on the decomposition of plant residues, microbial biomass and B-glucosidase activity in soils. Soil Science and Plant Nutrition 39, 557±566. BaÊaÊth, E., 1989. E€ects of heavy metals in soil microbial processes and populations (a review). Water, Air and Soil Pollution 47, 335±379. BaÊaÊth, E., Arnebrandt, K., Nordgren, A., 1991. Microbial biomass and ATP in smelter-polluted forest humus. Bulletin of Environmental Contaminant and Toxicology 47, 278±282. Ball, D.F., 1964. Loss on ignition as an estimate of organic matter and organic carbon in non-calcareous soils. Journal of Soil Science 15, 84±92. Bardgett, R.D., Saggar, S., 1994. E€ect of heavy metal contamination on the short-term decomposition of labelled (14C) in a pasture soil. Soil Biology and Biochemistry 26, 727±733. Berg, R., van de Dennem, C.A.J., Roels, J.M., 1993. Risk assessment of contaminated soil: proposals for adjusted toxicologically based Dutch soil clean-up criteria. In: Arendt, F., Annokkee, G.J., Bosman, R., vd Bank, W.J. (Eds.), Contaminated Soil. Kluwer Academic Publishers, Dordrecht, pp. 349±361. Bogomolov, D.M., Chen, S.K., Parmelee, R.W., Subler, S., Edwards, C.A., 1996. An ecosystem approach to soil toxicity testing: a study of copper contamination in laboratory soil microcosms. Applied Soil Ecology 4, 95±105. Brookes, P.C., McGrath, S.P., 1984. E€ects of metal toxicity on the size of the soil microbial biomass. Journal of Soil Science 35, 341±346. Chapman, H.D., 1965. Cation-exchange capacity. In: Black, C.A., Evans, D.D., White, J.L., Ensminger, L.E., Clark, F.E. (Eds.), Methods of Soil Analysis, Chemical and Microbiological Properties, Part 2. American Society of Agronomy, Madison, WI, pp. 902±904. Cheng, W., Coleman, D.C., 1989. A simple method for measuring CO2 in a continuous air-¯ow system. Modi®cation of the substrateinduced respiration technique. Soil Biology and Biochemistry 21, 385±388. Dar, G.H., 1996. E€ects of cadmium and sewage-sludge on soil microbial biomass and enzyme activities. Bioresource Technology 56, 141±145. DoE, 1989. Code of Practice for the Agricultural Use of Sewage Sludge. HMSO, London.

Doelman, P., Haanstra, L., 1984. Long-term and short-term e€ects of cadmium, chromium and zinc on soil microbial respiration in relation to abiotic soil factors. Plant and Soil 79, 317±327. Donnison, L., 1997. Mycology of haymeadows under management change. PhD thesis, University of Wales Aberystwyth, UK. Flieûbach, A., Martens, R., Reber, H.H., 1994. Soil microbial biomass and microbial activity in soil treated with heavy metal contaminated sewage sludge. Soil Biology and Biochemistry 26, 1201±1205. FrostegaÊrd, A., BaÊaÊth, E., Tunlid, A., 1993a. Shifts in the structure of soil microbial communities in limed forests as revealed by phospholipid fatty acid analysis. Soil Biology and Biochemistry 25, 723± 730. FrostegaÊrd, A., Tunlid, A., BaÊaÊth, A., 1993b. Phospholipid fatty acid composition, biomass, and activity of microbial communities from two soil types experimentally exposed to di€erent heavy metals. Applied and Environmental Microbiology 59, 3605±3617. Giller, K.E., Witter, E., McGrath, S.P., 1998. Toxicity of heavy metals to microorganisms and microbial processes in agricultural soils: a review. Soil Biology and Biochemistry 30, 1389±1414. Giller, K.E., McGrath, S.P., Hirsch, P.R., 1989. Absence of nitrogen ®xation in clover grown on soil subject to long-term contamination with heavy metals is due to survival of only ine€ective Rhizobium. Soil Biology and Biochemistry 21, 841±848. Hattori, H., 1991. In¯uence of cadmium on decomposition of glucose and cellulose in soil. Soil Science and Plant Nutrition 37, 39±45. Hattori, H., 1992. In¯uence of heavy metals on soil microbial activities. Soil Science and Plant Nutrition 38, 93±100. Huysman, F., Verstraete, W., Brookes, P.C., 1994. E€ect of manuring practices and increased copper concentrations on soil microbial populations. Soil Biology and Biochemistry 26, 103±110. Insam, H., Hutchinson, T.C., Reber, H.H., 1996. E€ects of heavy metals stress on the metabolic quotient of the soil micro¯ora. Soil Biology and Biochemistry 28, 691±694. Je€ery, J.J., Uren, U.C., 1983. Copper and zinc species in the soil solution and the e€ect of soil pH. Australian Journal of Soil Science 21, 479±488. Keeney, D.R., Nelson, D.W., 1982. Nitrogen Ð inorganic forms. In: Page, A.L., Miller, R.H., Keeney, D.R. (Eds.), Methods of Soil Analysis Ð Chemical Microbiological Properties, 2nd Ed., Agronomy Vol. 9, Part 2. Madison, WI, pp. 643±638. Khan, M., 1999. Microbial responses to metals in sludge treated soils. PhD thesis, University of Wales Aberystwyth, UK. Leita, L., De Noblil, M., Muhlbachova, G., Mondini, C., Marchiol, L., Zerbi, G., 1995. Bioavailability and e€ects of heavy metals on soil microbial biomass survival during laboratory incubation. Biology and Fertility of Soils 19, 103±108. MAFF, 1986. The Analysis of Agricultural Materials (Ministry of Agriculture, Fisheries and Food. Technical Bulletin 27). HMSO, London. Maliszewska, W., Dec, S., Wierzbicka, H., Wozniakowska, A., 1985. The in¯uence of various heavy metal compounds on the development and activity of soil micro-organisms. Environmental Pollution 37, 195±215. Manugistics, 1992. Statgraphics, Statistical Graphics System: Version 6.0: Example Manual. Statistical Graphics Corporation, Rockville, MD. McGrath, S.P., Cegarra, J., 1992. Chemical extractability of heavy metals during and after long-term applications of sewage sludge to soil. Journal of Soil Science 43, 313±321. McGrath, S.P., Chaudry, A.M., Giller, K.E., 1995. Long-term e€ects of metals in sewage sludge on soils micro-organisms and plants. Journal of Industrial Microbiology 14, 94±104. Newell, S.Y., Miller, J.D., Fallen, R.D., 1987. Ergosterol content of salt marsh fungi: e€ect of growth conditions and mycellial age. Mycologia 79, 688±695. Odum, B.P., 1985. Trends expressed in stressed ecosystems. Bioscience 35, 419±422.

M. Khan, J. Scullion / Environmental Pollution 110 (2000) 115±125 Ohya, H., Komai, Y., Yamaguchi, M., 1985. Zinc e€ects on soil micro¯ora and glucose metabolites in soil amended with 14Cglucose. Biology and Fertility of Soils 1, 117±122. Pagliai, M., Guildi, G., La Marca, M., Giachetti, M., Lucamente, M., 1981. E€ects of sewage sludges and composts on soil porosity and aggregation. Journal of Environmental Quality 10, 556±561. Piper, C.J., 1950. Soil and Plant Analysis. University of Adelaide, Adelaide. Smith, S.R., Hall, J.E., Hadley, P.H., 1989. Composting sewage sludge wastes in relation to their suitability for use as fertiliser materials for vegetable crop production. International Symposium of Compost Recycling of Wastes, Athens, 4±9 October. Ure, A.M., 1995. Methods of analysis for heavy metals in soil. In: Alloway, B.J. (Ed.), Heavy Metals in Soils. Blackie Academic and Professional, London, pp. 58±162.

125

van Beelen, P., Doelman, P., 1997. Signi®cance and application of microbial toxicity tests in assessing ecotoxicological risks of contaminants in soil and sedimentsa. Chemosphere 34, 455± 499. Vance, E.D., Brookes, P.C., Jenkinson, D.S., 1987. An extraction method for measuring microbial biomass C. Soil Biology and Biochemistry 19, 703±704. Westerman, R.L., Tucker, T.C., 1974. E€ect of salts and salts plus nitrogen-15-labelled ammonium chloride on mineralization of soil nitrogen, nitri®cation, and immobilization. Soil Science Society of America, Proceedings 38, 602±605. White, O.C., Davis, W.M., Nickels, J.S., King, J.C., Bobbie, R.J., 1979. Determination of the sedimentary microbial biomass by extraction of lipid phosphate. Oecologia 40, 51±62.