FEMS Microbiology Ecology 52 (2005) 139–144 www.fems-microbiology.org
The effects of soil horizons and faunal excrement on bacterial distribution in an upland grassland soil Patricia M.C. Bruneau a, Donald A. Davidson b, Ian C. Grieve Iain M. Young c, Naoise Nunan d,1
b,*
,
a Scottish Natural Heritage, 2 Anderson Place, Edinburgh, EH6 5NP, Scotland, UK School of Biological and Environmental Sciences, University of Stirling, Stirling FK9 4LA, Scotland, UK c SIMBIOS Centre, University of Abertay Dundee, Bell Street, Dundee DD1 1HG, Scotland, UK Biomathematics and Statistics Scotland, Scottish Crop Research Institute, Invergowrie, Dundee, DD2 5DA, Scotland, UK b
d
Received 24 June 2004; received in revised form 27 October 2004; accepted 29 October 2004 First published online 11 November 2004
Abstract The density and spatial location of bacteria were investigated within different horizons of an upland grassland soil before and after a liming treatment to increase the numbers of large soil fauna. Bacterial cells were located by image analysis of stained thin sections and densities calculated from these data. Excrement from macro- and meso-fauna was identified using micromorphology and the densities of bacteria on specific areas of excrement measured by image analysis. There were significant differences among horizons in the density of bacterial cells, with the minimum density found in the horizon with least evidence of earthworm activity, but no difference in density between the organic H and organo-mineral Ah horizons. Soil improvement by liming significantly increased bacterial densities in all three horizons, with the greatest increase found in the horizon with the smallest density before liming. There were no differences in bacterial density between areas dominated by excrement from earthworms and excrement from enchytraeids, although densities in both areas were significantly increased by liming. Variability in bacterial density at spatial scales of less than 1 mm was linked to the occurrence of excrement. Bacterial densities within areas of both types of excrement were significantly greater than those in the surrounding soil. However, the frequency distribution of the ratios of density in excrement to that in the soil was bimodal, with a majority of occurrences having a ratio near 1 and only some 20–30% having a much larger ratio. These variations can probably be explained by variations in the age of the excrement and its suitability as a substrate. 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. Keywords: Soil micromorphology; Soil structure; Biological thin sections; Bacteria; Faunal excrement; Temperate grassland
1. Introduction The nature of the habitat is of crucial importance in influencing biological activity within soils [1]. In soil ecology, the concept of functional domains has been used to *
Corresponding author: Fax: +44 1786 467846. E-mail address:
[email protected] (I.C. Grieve). 1 Present address: CNRS, Bioge´ochemie des Milieux continentaux, Baˆtiment EGER, aile B, INRA INA-PG, 78850 Thiverval Grignon, France.
explain the spatial relationships between soil structure and faunal communities [2,3] and a similar approach has been successfully applied to biogenic structures created by the ‘‘engineering’’ activities of meso- and macro-fauna [4,5]. However, most of the research on soil–faunal interactions has been concerned with mineral or cultivated soils and much less is known about the spatial links between soil structure and fauna in more organic upland soils. In such soils, fauna have been shown to play important roles in soil function, most
0168-6496/$22.00 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.femsec.2004.10.010
140
P.M.C. Bruneau et al. / FEMS Microbiology Ecology 52 (2005) 139–144
notably in the formation of horizons and structure [6] and in carbon cycling and organic matter decomposition [7]. Soil micromorphology offers a unique methodology for studying the nature and spatial distribution of soil components seen in thin sections [8]. Recent developments in image analysis of soil thin sections [9] offer exciting possibilities for relating the spatial distribution of soil microbial populations to specific micromorphological features. Sampling and analysis of bulk soils have demonstrated variability in microbial activity at scales of a few mm in the rhizosphere [10] to several meters within a pasture [11]. Micromorphological techniques provide ways of analysing the spatial distribution of microbial clusters which may be hotspots of microbial activity and which are likely to be associated with aggregates, particulate organic matter and the rhizosphere [12]. In one recent study [13], the distribution of bacteria was examined at a range of scales and bacteria were found to occur in preferentially colonised patches, close to pores in the subsoil, but more randomly in the topsoil. However, it is not known whether such spatial patterns also occur within upland soils which have not been under recent cultivation. This paper uses micromorphological analysis at two scales to quantify for the first time the spatial association between faunal excrement and bacteria in upland soils. The effects of faunal activity in such soils are most clearly expressed in the dominant occurrence of excremental features, which varies according to soil horizon and in a variety of structures ranging from distinctive vermiform features of earthworm origin to amorphous material resulting from aged or fused excrement. Faunal excrement has been shown to constitute up to 80% of the overall soil volume in the organic horizons and as much as 50% in mineral horizons in upland soils [14]. Excrement of larger soil fauna provides a habitat for other fauna and microbes and it is hypothesised here that the spatial distribution of bacterial colonies will be associated with excremental features. To test this hypothesis, the spatial distribution of bacteria was studied by image analysis of thin sections made from soils from an upland grassland site. Soils to which a surface application of lime had been made were compared with unlimed control soils. The aim of the liming treatment was to stimulate the activity of soil fauna, particularly earthworms, and thus provide more substrate material for detrivores. Data on bacterial distributions derived from soil micromorphology and image analysis were used to test differences in bacterial density between soil horizons and between different types of excrement in the limed and control soils.
2. Experimental site and methods Soil samples were obtained from an upland grassland at Sourhope Research Station, approximately 25 km
south of Kelso in south-eastern Scotland (5528.5 0 N, 214 0 W). The altitude of the site is approximately 300 m, mean annual precipitation is 952 mm [15] and mean annual soil temperature at 5 cm depth is approximately 8 C. Vegetation is dominated by Agrostis capillaris (L.) with Festuca ovina (L.), F. rubra (L.) and Anthoxanthum odoratum (L.). Sheep and cattle were excluded from the experimental site and plots were mowed and cuttings removed at 3-week intervals during the growing season. The soil parent material is glacial till derived from andesitic lavas of Devonian age. Soils are brown forest soils (Cambisols) belonging to the Sourhope series [16]. The upper part of the soil profile typically comprises thin L and F horizons (total thickness 1–4 cm), a 3–8 cm thick H horizon and an Ah horizon with total thickness of between 10 and 20 cm. Some profiles exhibited a thin (1–1.5 cm) dark reddish grey to dark grey (5YR 4/2 to 10YR 4/1) horizon at the base of the H horizon. Microscopic study of thin sections from this horizon revealed that it was organic, but contained a high concentration of phytolith fragments and this horizon was designated Hphy [14]. This paper examines differences in bacterial distributions between the control plots and plots which received a surface dressing of calcium carbonate (39% Ca) at 0.6 kg m2 in the spring of 1999, 2000 and 2001. The annual application of calcium carbonate increased the mean pH of the organic (H) horizon from 3.4 to 4.1 between 1999 and 2001 [14]. Two years after the first liming treatment was applied, the limed plots had significantly larger abundance and biomass of earthworms than the unlimed control plots [17]. When each species was analysed separately, there were significantly more individuals of Dendrobaena rubidus and Aporrectodea rosea in the limed plots. One year after the first application of lime, there was a significant reduction in the total number of enchytraeid worms, but no reduction in total biomass. Cognettia was replaced by Fridericia as the dominant genus [18]. These changes in fauna following liming were also reflected in the excrement visible on thin sections and the liming treatment increased the amount of earthworm excrement [14]. Earthworm excrement was mainly found in the upper, organic horizons and that of enchytraeids was dominant in the organo-mineral horizon (Ah). Despite significant heterogeneity, soil improvement treatments also influenced the structure and functioning of the bacterial community, with liming having the greatest impact. Autotrophic ammonia oxidation was significantly increased by the combined action of liming and a sewage sludge application [19]. Undisturbed soil samples were taken from the organic and upper Ah horizon using cores (6 cm diameter by 5 cm length). Two replicates samples were obtained from each plot in June 1999, November 1999 and September 2001. Bacteria were fixed as soon as possible after sampling with an aqueous solution of glutaraldehyde and
P.M.C. Bruneau et al. / FEMS Microbiology Ecology 52 (2005) 139–144
subsequently stained with an aqueous solution of Calcofluor white MR [9]. Thin sections were prepared using standard water/acetone exchange in the liquid phase and impregnated using Crystic XB52n9 resin. Vertically oriented, diamond-polished thin sections, 30 lm thick, were produced from the impregnated blocks. Fig. 1 summarises the methodology by which faunal excrement and bacterial locations and density statistics were determined. At the macroscale each thin section was divided into homogeneous areas which corresponded to the principal soil horizons. It has already been shown that there are two distinctive types of faecal material in this upland grassland soil [6]. Small pedofeatures (60–100 lm), spherical to ellipsoidal in shape, loose to moderately compacted, mainly organic and with no internal fragments larger than 20 lm were derived primarily from enchytraeids. Larger excremental pedofeatures, up to several millimetres in size, often with mammillated shapes and including mineral and plant fragments and with diagnostic vermiform features, were derived from earthworms [8]. These morphological features were identified using AnalySIS v3.0 (Soft-Imaging System GmbH, Munster, Germany) with an analysis procedure specifically designed for this project [20]. To identify bacterial clusters, thin sections were examined using a Zeiss Axioplan 2 microscope fitted for epifluorescence (100W OSRAM mercury UV lamp, HBO 103 W/2) at a magnification of ·630. The slides were mounted on a motorised stage and tessellated images of 25 (5 · 5) contiguous fields of view, representing an area of 0.282 mm2, were acquired within each defined horizon. Bacterial locations were determined using
141
a Zeiss KS300 Imaging System 3.0 [9]. The x and y coordinates of each individual bacterial cell and its shape and size were determined. Total bacterial counts in each ·630 image were then converted to bacterial densities (mm2). These densities can be related to soil volumes on the assumption that the thickness of the analysed soils in which bacteria could be seen was 2 lm (±1 lm from the plane of observation). A measured density of 1000 cells mm2 is thus equivalent to a density of 5 · 108 cells cm3. In order to compare the locations of bacteria and faunal excrement, bacterial locations were converted to a new reference system at a ·50 scale of magnification and overlain on the image of excremental features. One consequence of this scale change was that bacterial locations merged where distances apart were less than the new pixel size, although bacterial colonies in soils consist of very small numbers of cells [21]. Each ·50 image was segmented into areas of similar excremental features and bacterial density (mm2) was calculated for each defined area of excrement. These densities could not be compared directly with densities calculated at the ·630 magnification scale due to the effects of the change of scale on bacterial clusters. Densities were therefore calculated for each whole image at the ·50 magnification scale; the two density measures were very strongly correlated (r = 0.953; P < 0.001). Total void space and void space per area of interest were measured by applying a manual threshold to the ·50 image. A threshold value was chosen for each image to separate void space from dark material. Bacterial densities on each type of excrement were then corrected for void space.
Fig. 1. Identification of bacteria and excremental features in soil thin sections. (a) Single image at magnification ·50, identification of areas with similar pedofeatures; (b) tessellated images (5 · 5) at magnification ·630, identification of bacteria; (c) Details showing a bacteria cluster.
142
P.M.C. Bruneau et al. / FEMS Microbiology Ecology 52 (2005) 139–144
3. Results and discussion 3.1. The effect of horizons and a liming treatment on bacterial density Table 1 shows bacterial densities in each horizon for the control and limed soils and striking contrasts are suggested. Mean densities ranged from under 200 to over 1300 cells mm2, equivalent to 108 and 6.5 · 108 cells cm3, respectively. These densities are similar to densities measured by other methods for upland soils. Densities of bacterial colony forming units (cfu) of 4 · 1012 cfu m2 in the uppermost 5 cm of improved grassland soils have been reported for a range of sites in the UK uplands [22], equivalent to just less than 108 cfu cm3. Frequency distributions of the raw bacterial densities showed slight positive skewness and the data were normally distributed following a square root transformation. Statistical analyses were carried out on the transformed dataset and all tables report back-transformed means and 95% confidence limits. Two-way analysis of variance showed that both the treatment effect (P < 0.001) and the horizon effect (P < 0.05) were statistically significant, but the interaction between treatment and horizon was not significant. Mean bacterial density was lowest for the Hphy horizon of the unlimed soil, probably reflecting the limited current faunal activity in this horizon. The phytoliths which were a prominent feature of the micromorphology of the Hphy horizon were highly fragmented and this horizon contained a large amount of older undifferentiated excrement, the result of substantial past bioturbation [14]. Microbial activity and diversity are smaller in deeper soil organic horizons that contain older organic matter [23] as a result of the rapid decrease in microbial activity and biomass with age of the substrate [24]. The only micromorphological evidence for earthworm activity in the Hphy horizon at Sourhope was the existence of vertical channels, suggesting that this horizon may be primarily a zone of transit for earthworms between the upper and lower horizons, as opposed to their main habitats.
Table 1 Means and 95% confidence limits of density of bacterial cells (backtransformations of square root-transformed data) in three horizons of the Sourhope soil Horizon
H Hphy Ah
Control
Limed
Mean density (mm2)
95% CL
Mean density (mm2)
95% CL
944 183 804
654–1287 42–423 559–1094
1275 1085 1316
1004–1579 728–1516 1016–1654
Bacterial densities in the H and Ah horizons were not significantly different, and the structure of both these horizons contained distinctive faunal excrement. Earthworm excrement was most clearly expressed in the H horizon and excrement of enchytraeids was the dominant form observed in the Ah horizon [14]. It would thus appear that both organisms provide substrates to sustain bacterial activity within these horizons. Bacterial densities were greater in all three horizons of the limed soil and this treatment effect was highly significant (P < 0.001). This is consistent with previous studies which have reported increased bacterial activity and diversity, community respiration and dissolved organic carbon leaching following liming of acidic forest soils [25]. In a range of upland grassland soils in the UK [22], liming increased bacterial numbers, bacterial activity and carbon utilisation, although the almost 4fold differences in bacterial numbers between improved and unimproved grassland soils in that study were much larger than the differences in bacterial densities with liming found here. 3.2. Associations between bacteria and faunal excrement Table 2 shows the mean densities of bacteria on areas of excrement from: (a) enchytraeids and (b) earthworms, again as back-transformations of statistics calculated on square root-transformed data. Densities were calculated on the basis of the areas of mineral and organic matter excluding void space. Liming effects were highly significant (P < 0.001) for bacterial density on both types of excrement, with significantly greater densities in the limed soils. Horizon effects were also significant (P < 0.01 for enchytraeid excrement and P < 0.05 for earthworm excrement). The smallest densities were again found on excrement in the Hphy horizon, and there was no significant difference between the H and Ah horizons or earthworm and enchytraeid excrement. Bacterial densities on each type of excrement were also compared with densities in the soil. For every sampling area, the density of cells on the excrement and within the whole sampling area was calculated. Differences between the densities were normally distributed and, for both excrement types, a paired t-test showed significantly greater densities (P < 0.01) on the areas of excrement than in the whole sampling area. The mean difference for earthworm excrement was 207 (standard error = 66) cells mm2 and that for enchytraeid excrement was 141 (standard error = 47) cells mm2. Fig. 2 shows frequency distributions of the ratio of density on excrement to density in soil for: (a) enchytraeid and (b) earthworm excrement. The distributions were bimodal in both cases, with most excrement densities similar to soil densities (ratios of near 1) but between 20% and 30% of excrement densities substantially greater than in the soil. These greater densities
P.M.C. Bruneau et al. / FEMS Microbiology Ecology 52 (2005) 139–144 Table 2 Means and 95% confidence limits of density of bacterial cells on: (a) enchytraeid and (b) earthworm excrement (back-transformations of square root-transformed data) in three horizons of the Sourhope soil Horizon
Control
Limed
Mean density (mm2)
95% CL
Mean density (mm2)
95% CL
(a) Enchytraeid excrement H 778 Hphy 91 Ah 710
533–1070 43–157 500–966
1481 805 1373
1175–1822 486–1206 1046–1744
(b) Earthworm excrement H 951 Hphy 212 Ah 656
591–1397 67–438 385–997
1306 914 1481
1016–1633 585–1318 1105–1914
143
H and Hphy horizons, formed by coalescence and ageing of excrement [14]. The existence of high densities of bacteria linked to the spatial distribution of faunal excrement in the structure of these soils provides further support for the spatial association of hotspots of microbial activity with labile substrates [12]. The analysis of stained soil thin sections has also revealed substantial variations in microbial abundance at a much more detailed spatial scale than previously seen [10,11]. The application of similar analytical methods to cultivated soils revealed a spatial association between bacteria and pores [13]; this study demonstrates a further link with faunal excrement in an uncultivated soil.
4. Conclusions (a) Enchytraeid
70 60
Frequency
50 40 30 20 10 0 0
1
2
3
4
5
Ratio (b) Earthworms
Frequency
30
20
10
0 0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Ratio Fig. 2. Frequency distributions of the ratio of bacterial density on excrement to that in whole soils for: (a) enchytraeid and (b) earthworm excrement.
may be on areas of relatively fresh excrement as suggested by the distinct clusters of bacteria within excremental features in Fig. 1(c). Microbial biomass and activity on earthworm excrement decrease rapidly with substrate age [24]. Large amounts of undifferentiated excrement occur in these soils, particularly in the lower
In this upland grassland soil, the upper horizons are dominated by organic matter consisting largely of excrement. Although there are many difficulties in associating excremental features with particular organisms, earthworms and enchytraeids seem to have been the main sources. Densities of bacteria, determined by counting under high magnification on stained thin sections of soil, were similar to those determined by other methods. Bacterial densities were similar in the organic H horizon and the uppermost organo-mineral Ah horizon in the horizon, dominated, respectively, by earthworm and enchytraeid excrement. Bacterial densities were significantly less in an intervening horizon rich in phytolith fragments and which had the smallest concentration of earthworm excrement, presumably a consequence of the unpalatable phytolith fragments. The effect of liming was to increase densities significantly in all three upper horizons, with the most marked change in the phytolith-rich horizon. The changes in the earthworm and enchytraeid populations stimulated by liming were followed by increased amounts of faunal excrement, an attractive substrate for bacteria [14]. There was no significant difference in density between sites on earthworm and enchytraeid excrement, although mean densities on both substrates were significantly increased by liming and were significantly greater than densities in the whole soil. The major implication of the data presented here, however, is the importance of fresh faunal excrement in determining the distribution of high concentrations of bacteria at spatial scales of less than 1 mm. These data suggest than spatial variability in microbial function exists at the sub-mm level within the spatial structure of upland soils.
Acknowledgements This work was funded by a NERC Grant (GST/02/ 2127) under the Soil Biodiversity and Ecosystem
144
P.M.C. Bruneau et al. / FEMS Microbiology Ecology 52 (2005) 139–144
Function thematic programme. We thank David Crabb, Kirsty Harris and Dr. Linda Deeks (Soil-Plant Dynamics Unit, Scottish Crop Research Institute) for preparation and analysis of the thin sections for biological analysis and George MacLeod (Stirling University) for preparation of thin sections for micromorphological analysis.
References [1] Young, I.M. and Crawford, J.W. (2004) Interactions and selforganisation in the soil–microbe complex. Science 304, 1634–1637. [2] Lavelle, P. (2000) Ecological challenges for soil science. Soil Science 165, 73–86. [3] Lavelle, P. (2002) Functional domains in soils. Ecological Research 17, 441–450. [4] Topoliantz, S., Ponge, J.-F. and Viaux, P. (2000) Earthworm and enchytraeid activity under different arable farming systems, as exemplified by biogenic structures. Plant and Soil 225, 39–51. [5] Lavelle, P., Bignell, D., Lepage, M., Wolters, V., Roger, P., Ineson, P., Heal, O.W. and Dhillion, S. (1997) Soil function in a changing world: the role of invertebrate ecosystem engineers. European Journal of Soil Biology 33, 159–193. [6] Davidson, D.A., Bruneau, P.M.C., Grieve, I.C. and Young, I.M. (2002) Impacts of fauna on an upland grassland soil as determined by micromorphological analysis. Applied Soil Ecology 20, 133–143. [7] Cole, L., Bardgett, R.D. and Ineson, P. (2000) Enchytraeid worms (Oligochaeta) enhance mineralization of carbon in upland soils. European Journal of Soil Science 51, 185–192. [8] Fitzpatrick, E.A. (1993) Soil Microscopy and Micromorphology 304 pp. Wiley, Chichester. [9] Nunan, N., Ritz, K., Crabb, D., Harris, K., Wu, Crawford, J.W. and Young, I.M. (2001) Quantification of the in situ distribution of soil bacteria by large-scale imaging of thin sections of undisturbed soil. FEMS Microbiology Ecology 36, 67–77. [10] Gaillard, V., Chenu, C., Recous, S. and Richard, G. (1999) Carbon, nitrogen and microbial gradients unduced by plant residues decomposing in soil. European Journal of Soil Science 50, 567–578. [11] Ritz, K., McNicol, J.W., Nunan, N., Grayston, S., Millard, P., Atkinson, D., Gollotte, A., Habeshaw, D., Boag, B., Clegg, C.D., Griffiths, B.S., Wheatley, R.E., Glover, L.A., McCaig, A.E. and Prosser, J.I. (2004) Spatial structure in soil chemical and microbiological properties in an upland grassland. FEMS Microbiology Ecology 49, 191–205. [12] Nannipieri, P., Ascher, J., Ceccherini, M.T., Landi, L., Pietramellara, G. and Renella, G. (2003) Microbial diversity and soil functions. European Journal of Soil Science 54, 655–670.
[13] Nunan, N., Wu, K., Young, I.M., Crawford, J.W. and Ritz, K. (2003) Spatial distribution of bacterial communities and their relationships with the micro-architecture of soil. FEMS Microbiology Ecology 44, 203–215. [14] Davidson, D.A., Bruneau, P.M.C., Grieve, I.C. and Wilson, C.A. (2004) Micromorphological assessment of the effects of liming on faunal excrement in an upland grassland soil. Applied Soil Ecology 26, 169–177. [15] Burt-Smith, G (2003) Sourhope Field Experiment Report IV (1999–2002). Available from: . [16] Kenny, C (1998) Soil Biodiversity NERC Thematic Programme Baseline Data. Available from: . [17] Bishop, H.O. (2003) Effects of earthworms on the distribution and chemical characteristics of soil organic matter. PhD thesis, University of Stirling, p. 221. [18] Black, H.I.J., Piertney, S.B., MacDonald, C., Standen, V., Bull, I.D., Evershed, R.P. (In press) Does species diversity affect rhizosphere carbon assimilation by enchytraeids? A case study from a Scottish pasture. European Journal of Soil Science. [19] Gray, N.D., Hastings, R.C., Sheppard, S.K., Loughnane, P., Lloyd, D., McCarthy, A.J. and Head, I.M. (2003) Effects of soil improvement treatments on bacterial community structure and soil processes in an upland grassland soil. FEMS Microbiology Ecology 46, 11–22. [20] Bruneau, P.M.C., Davidson, D.A. and Grieve, I.C. (2004) An evaluation of image analysis for measuring changes in void space and excremental features on soil thin sections in an upland grassland soil. Geoderma 120, 165–175. [21] Grundmann, G.L. (2004) Spatial scales of soil bacterial diversity – the size of a clone. FEMS Microbiology Ecology 48, 119–127. [22] Grayston, S.J., Campbell, C.D., Bardgett, R.D., Mawdsley, J.L., Clegg, C.D., Ritz, K., Griffiths, B.S., Rodwell, J.S., Edwards, S.J., Davies, W.J., Elston, D.J. and Millard, P. (2004) Assessing shifts in microbial community structure across a range of grasslands of differing management intensity using CLPP, PLFA and community DNA techniques. Applied Soil Ecology 25, 63–84. [23] Agnelli, A., Ascher, J., Corti, G., Ceccherini, M.T., Nannipieri, P. and Pietraellara, G. (2004) Distribution of microbial communities in a forest soil profile investigated by microbial biomass, soil respiration and DGGE of total and extracellular DNA. Soil Biology and Biochemistry 36, 859–868. [24] Tiunov, A.V. and Scheu, S. (2000) Microbial biomass, biovolume and respiration in Lumbricus terrestris L. cast material of different age. Soil Biology and Biochemistry 32, 265–275. [25] Ba¨ckman, J.S.K., Hermansson, A., Tebbe, C.S. and Lindgren, PE. (2003) Liming induces growth of a diverse flora of ammoniaoxidising bacteria in acid spruce forest soil as determined by SSCP and DGCE. Soil Biology and Biochemistry 35, 1337–1347.
