Chapter 4R

Methods used for Research Chapter 4 Dr. Ellen van Mulligen IV Methods Sometimes, surely, truth is closer to imagination - ............ - then to fact? To be accurate is not to be right.*1 The methods used in this research are presented in this Chapter. In “Parametrisation of the indicators” measurements are linked to indicators and in “General Research Methods”, sampling scheme is presented. In “Field and Laboratory Methods', original procedures are described. This section also considers some issues concerning the reliability of the data obtained. Finally methods of data processing are described in “Calculations”, .1 Parameterisation of the indicators In table IV.1 all candidate indicators considered for this study are represented. To decide which indicator was preferable all cadidate indicators were given scores (-2 upt to 2, except when mentioned different) for the following evaluation criteria: theoretical preferability (0-2); representativenes (0-2); ability to discriminate; measure errors (0-2); disturbance research areas; time consuming method; direct or indirect measure; need of a specialistic method; availability of a standard method and; simplicity of the possible methods to access the indicator. Eventually the the indicators with a total evaluation equal or higher than 7 were selected. Precipitation, (air) temperature and insulation were selected to indicate climatic differences between the sites and to describe the temporal changes that occurred. Precipitation is also a major factor influencing biological activity. Temperature (insulation) is important with respect the effect on evapo-transpiration of the system and hence water availability. Calculated solar radiation as measure for insulation (see section IV.4) and soil water content were excepted as indicators for climatic at a small scale. For “soil degradation” the physical indicators were, which is in accordance with the rather physical definition of soil degradation (see section I). Many forms of organic matter are present in and on the soil. The parameters measured were, the litter cover (see section 2.1 in this chapter), and the carbon content of soil samples. In the soil thin sections, recognisable organic matter was studied. Soil aggregation was thought to be of great importance in this study. As discussed in chapter II macro and micro aggregation seems to be resulting from different processes and are expected to shown different life cycles (see section II.3). Consequently indicators for both were included. The water-stable soil microaggregation was determined for soil material  106 m by analyses with a microscan.(reference ). Macroaggregates ( 250 m) present in the soil were determined by gently sieving samples at field soil moisture content. In soil thin sections both the micro- and macro-aggregation were studied qualitatively and quantitatively (Ringrose-Voase, 1991). The aggregate stability was assessed using the single drop test. (Low, 1954). This was done for macro-aggregates at two moisture contents.. Methods As discussed above, the ability of the soil to absorb and retain water is one of the most important properties in dry areas.. This was estimated from the maximum infiltration rate, for 70 mm water supplied at a maximum intensity of ca. 200 mm/h (see section IV.2.1). The water storage capacity of the soil was not determined. The soil moisture content can not be regarded as an indicator of the water storage capacity of the soil because it only indicates the amount of water present at one moment and in a certain part of the soil, while many biota are able to collect water from relative moist, deeper parts of the soil or sub-soil. It was not expected that the storage capacity of the soils studied here would be less than the amount of precipitation. Table IV.1 List of candidate indicators with their scores on several for the present study relevant criteria in order to establisch the preferability of using them as indicator in the present study. Evaluation criteria: theoretical preferable representativenes ability to discriminate 1 measure errors disturbance research areas time consuming method direct or indirect measure specialistic method standard method Simplicity total evaluation Candidate indicators Climate & environment Air temperature Soil temperature Precipitation Insulation at soil surface Soil degradation Chemicals/nutrients Carbon content Soil moisture content Water holding capacity soil Infiltration capacity Soil aggregation Waterstable aggregation Biological activity LAI Soil biota Roots Biomass Soil meso fauna Soil micro organisms Fungi Bacteria Streptomycetes 2 1 2 0 -1 -1 -1 2 2 -1 2 -2 2 -1 2 0 2 12 0 5 2 12 0 9 2 2 2 2 2 2 2 1 -1 -1 1 0 0 0 0 0 0 1 0 0 1 -1 0 -1 -1 1 -1 -1 1 2 -1 0 0 2 1 -2 2 2 2 -1 -1 0 1 2 2 0 1 1 2 0 2 2 13 2 10 0 7 2 7 2 8 0 8 2 1 2 2 1 1 2 1 2 1 0 2 2 2 2 -1 -1 -2 -2 -2 0 2 2 1 0 -2 -2 0 -2 -2 1 -1 1 1 0 -2 2 -2 0 -1 5 -2 2 -2 -2 2 2 2 1 1 1 0 0 1 1 -1 1 -1 1 -1 2 2 2 -1 -1 -1 2 2 -1 1 2 1 1 1 1 1 2 -1 2 0 2 1 1 1 0 1 2 2 0 1 2 2 2 2 2 1 1 2 2 2 2 2 2 2 1 1 2 2 1 2 2 2 -2 -2 -2 -2 0 0 0 8 8 6 1 this ability to discriminate should be regarded in the present study: effectivity in relation to spatial dimension of this research; effectivity in relation to temporal dimension of this research. 2 Methods Because “biological activity” was thougt to be of great importance, but it was unknown what specific aspect of this activity was most influential it was decided to measure as much biological indicators as possible. Only when the indicators score on costs, time consumption, disturbance of the reseach site, specialisme involved and consequently the risk of errors was high the indicator was rejected. In most instances an, in the regard as mentioned above, acceptable measure was found for the candidate indicator. So despite of the low evaluation the “biological activity” was assessed in terms of the measurements that describe severall parameters of the vegetation (composition and cover). Common methods for studying soil meso-fauna in semi-arid conditions (Pantis et al., 1988; Santos & Whitford, 1981; Whitford & Parker, 1989), would have either involved much time consuming fieldwork and analysis, and/or specialised laboratory equipment. The abundance of roots in the soil, signs of meso-faunal activity fauna (e.g. channels and excrement) were studied in soil thin sections(Pawluk, 1985, see section IV.3). Furthermore, the study of phenomena related to the activity of the meso fauna, also results in knowledge of the effect of the meso-fauna in relation to aggregation of soil material (excrements) and soil porosity (channels and chambers). The bacteria and fungi were accessed by determining numbers of viable bacteria (Olsen & Bakken, 1987), length of viable hyphae (Ingham & Klein, 1984a and b) and the bacterial, fungal and microbial biomass (Lodge & Ingham, 1991). These methods to access bacteria and fungi were chosen after consultation with a soil ecologist (C. Heijnen personal communication). Reasons for the preference for the counting of viable organisms in favour of the fumigation methods, were the simplicity of the method and the limited requirements for specialised microbiological research materials. This last was important as the research could be most conveniently undertaken in a geomorphological specialised laboratory in Valencia, Spain. Furthermore, only viable organisms are considered by this method and it permits the calculation of biomass because for both the bacteria and fungi, size classes were selected with known volumes (see section IV.3 and 4). The volume of the organisms is likely to be related to the influence that they have in binding soil particles. Microbial biomass is the most commonly used measure for the presence of microbial organisms in soil research. Because soil bacteria and fungi are able to react quickly to changes in their environment, (see chapter II) air temperature and precipitation values were analysed for the seven days antecedent to sampling. In addition to precipitation, the organic matter and soil moisture content of the soil and the volume of pores with sizes similar to those of the organisms (see chapter II), were also considered important for soil bacteria and fungi, and consequently determined. Table IV.2 .2 List of abbreviations, as used in the present study, with explanation General research methods - field sampling The research area was visited four times, in July 1992, March and September 1993, and in March 1994. From the information presented in Chapter II , it may be expected that the soil properties that are of concern will be sensitive to changes in the soil micro-environment. It is essential, therefore, to minimise both the time between soil sampling and sample processing in the laboratory and the disturbance of the micro-environment One consequence of this was that field visits had to be of limited duration, with the result that the number of measurements which could be performed in the field was also limited. 3 Methods To be able to measure the in situ living populations of bacteria and fungi, the microscopic sections were prepared in a laboratory in Valencia during the field visit periods; this took place within a few days of sampling. Other samples, including mixed bulk samples, pF cores and samples for soil thin sections (see below), were taken at the field locations and transported for analysis to the laboratory in Amsterdam. Samples were transported in ways that minimised its effect on the analyses that were to be performed . Figure IV.3 Timing of laboratory analyses with respect to the date of sampling in weeks. - sampling procedures At each location, a representative 25 m2 square plot was selected mid-way along a representative slope. This plot was orientated parallel to the contour lines. The vegetation and other surface cover characteristics of the plots were mapped as described below. All measurements and sampling during field visits took place within, or in the direct vicinity, of this plot (see Annexe A). To be able to study the micro-gradient (see chapter III), the extremes of the vegetation mosaic were sampled, e.g. under the vegetation and on bare spots (figure IV.1 a & b). An area was regarded as being bare when vegetation (with the exception of lichens) was completely absent or very scarce (cover  15%). At all locations one or two most representative plants were selected from which to sample beneath. For the Northeast slope at Benidorm (BN *2) this was Rosmarinus officinalis, for the Southwest slope (BS) Stipa tenacissima, for Callosa (CA) Ulex parviflorus, and for Cocoll (CO) Cystus albidus and Rosmarinus officinalis. Mixed samples were collected at eight locations in each plot. Care was taken to sample at places that appeared undisturbed. Because of the shallowness of the soil, the need to minimise disturbance, and the minimum sample volume needed (see section 2.2 of this chapter), it was decided to combine material from two sample points into one mixed sample. This resulted in two mixed samples being collected for vegetated and two for bare surfaces per location. For the pF cores, attempts were made to take samples at all eight sampling points. This was almost never possible because of the shallowness of the soil. The soil thin sections were taken for each cover type at one location, giving two samples per location (see section 2.2 of this chapter). - laboratory analyses As mentioned in section II, micro-organism populations, such as bacteria and fungi, are known to respond rapidly to changes in the micro-environment, example with respect to temperature and moisture. This could produce several undesirable effects. It is obvious for example that while dead micro-organisms can not be measured by techniques for measuring living (or viable) bacteria and fungi .and that these also no longer interact with their environment, there are micro-organisms that might prefer the new conditions and increase to outnumber the original population. This could lead to a complete change of populations and consequently of the characteristics of the soil sample. a Figure IV.1 4 b Examples of bare (a) for ...... in.......and vegetated (b) surface at ...... for....... Methods Figure IV.2 a b c Comparison of the results of cooled and uncooled samples. Samples were taken at Southwest slope at Benidorm and Cocoll. To check for such effects, replicate samples were taken at a few sampling points in July 1992 and at all sampling points at locations BS and CO during the second sampling series in March 1993. One subset of this series was stored and transported cooled, between 2 and 6 ºC; the other subset was transported and stored at room/transport facility temperatures. Figure IV.2 shows that both series differ. This is confirmed by the results of the t-test for dependent samples for the fungi and the Colony Forming Units, in which the cooled and noncooled sample series showed significant difference at a level of p  0.05. The viable bacteria did not show a significant difference. All laboratory analysis were performed as soon as possible after the field visit to reduce the effect of storage. The other laboratory analyses were performed in the order presented in figure IV.3. Because so many variables were studied once every 6 months, the number of samples which could be processed was limited to approx. 30. The standard sampling programme (see above) consisted of treating 16 samples. The other 14 samples were taken for additional studies, such as the effect of the sampling depth or the effect of the storage temperature on the samples as mentioned above. .3 Field Methods .3.1 Field Observations and Measurements - measurement of environmental conditions The temperature (air as well as soil temperature) and the precipitation were recorded for the ERMES project at the same locations (see section III.1). Due to practical and financial problems the installation of the measuring devices was delayed until the end of 1992. Climatological conditions, in this case average air temperature and total precipitation were calculated for the 7 days antecedent to sampling. - vegetation and surface cover During each field visits, the larger elements of the vegetation (15 cm high, third column of table IV.2) were mapped (see Annexe B). These are the vegetated areas. Areas with less than 15% cover of low vegetation (15 cm) were considered as bare areas. The cover percentage of the larger vegetation elements was calculated from the maps. A screen was placed over the map (19 * 19 lines) and the vegetation types (third column of table IV.2) below the intersecting lines were recorded and summed per vegetation type. The character of the surface or the type of cover of the surface of bare areas was recorded in the field using a quadrate (0.25 m2). A quadrate is a square divided in smaller squares by a grid, here 5 * 5 lines). The type of surface characteristics or cover, mentioned in the first two columns (except: 'vegetation') of table IV.2 were used. The surface characteristics or cover on the intersections were added up. According to personal observations four replications at randomly chosen points in the bare areas of each plot was sufficient. Table IV.3 Classes of surface cover 5 Methods Figure IV.5 Infiltrometer with trickle system as used for this research - infiltration capacity Initially, infiltration measurements collected for the ERMES Project (Calvo et al., 1998), made at the the same general locations (see part III.1) seemed useful. However after the first field visit it appeared essential to obtain new data for the exact sites were the other measurements were being made. Infiltration was measured with an infiltrometer using a trickle system (figure IV.5), similar to the portable rainfall simulator of Imeson (1977). The trickle system (ca. 710 cm2) was supported by legs which could be extended to a maximum height of 60 cm. To randomise drop impacts, a mesh was placed 10 cm below the trickle system. Water was stored in a reservoir on top of the infiltrometer and was applied with a constant head. This measurement was preferred because the more standard method of a ring with a layer of water with a constant head on the soil surface held severall negative aspects with regard to the present research. Firstly a ring has to be inserted in the soil, damaging the soil surface. Secondly this ring has to be water tight connected to the soil to prevent sepage of water which is very dificult in dry conditions at the sites of this study. Thirdly because of high slope angles the layer of water on the soil surface would vary considerably in height. Fourthly the water layer which has to be applied to the soil surface was thought to be undisirable in relation to the natural relative dry circumstances in the research area. To prevent the damaging of the soil surface by the impact of drops the height above the soil surface of the mesh was kept as small as possible because in this way measured infiltration rate can be regarded as mainly determined by the characteristics of the soil below the soil surface, like porevolume, connectivity of pores, antecedent soil moisture content and stability of the soil structure. Seventy mm of of water was applied. The intensity of the application of water, with a maximum of 200 mm/h, was regulated with a valve between the container and the infiltrometer. The amount applied water was recorded by noting down the water level in the container at constant intervals of time after starting the measurement. The intensity of the flow was regulated so as to prevent the accumulation of water on the soil surface. It was assumed that this was the maximum possible rate at which water could infiltrate into the soil. The water used for this measurement was natural spring water from Callosa d'Ensarria and it contained: 220 mmol/ltr SO 4 , 419 mmol/ltr Cl, 4.0 mmol/ltr NH4 , 37.8 mmol/ltr NO 3 , 0.3 mmol/ltr PO 4 , 3.8 mg/ltr DOC. The high value of Dissolved Organic Carbon (DOC) points at an organic dilution of the water. Figure IV.6 a b Results for both repetitions of (a) bacteria and (b) fungi for samples taken between 0 - 5, 5 - 10 and 10 20 cm depth (represented at 2.5, 7.5 and 15 cm respectively) . The line runs through the averages. .3.2 Soil sampling In July 1992, at the BN and BS locations all samples were taken at three depths (0 - 5, 5 - 10 and 10 - 20 cm). At the other locations sampling was at a depth range of 5 - 10 cm The maximum activity of fungi is found 6 Methods near the surface, while that of bacteria is found at greater depth (see figure IV.6). For fungi the statement by Garcia-Alvarez & Ibañez (1994): "A clear vertical stratification can be seen in biological activity in the first 10 cm of the upper horizon it shows more intensity in the uppermost layer ( 0- 4 cm depth)" was found to be true. They continue: "That fact is probably related to the alternating favourable and unfavourable seasons, as far as temperature and water availability are concerned. With the increase in aridity, the microbial activity tends to be generally concentrated near the surface.........". This argument was proved to be not true for bacteria at this location in July 1992. For the other field visits it was decided to sample at a depth of 2 to 7 cm. The soil surface was not sampled, on the one hand to avoid additional processes which affect this part of the soil, on the other hand to be sure of the presence of fungi and bacteria in the samples. As already argued in part 3.1 for the infiltration capacity the soil surface characteristics are diregarded in the present study and the characteristics of the soil at a depth between 2 and 7 cm depth are regarded as representative for the soil below the soil surface. - mixed soil samples For the laboratory analyses, between 0.5 1 litre of mixed sample was needed. Stones up to a diameter of 4 cm were included. Bigger stones were regarded as hard rock. The samples were collected with a large strong spoon-like device. This was first inserted into the soil, to make sure that it was populated with organisms present in the soil being sampled. The sample was stored in a plastic box which was cleaned with soap and water and with alcohol, to prevent contamination with any micro-organisms, that might be present in the box. The risk of contamination by organisms present in the boxes before sampling was further reduced by cooling the boxes, to keep organisms alive but static. The boxes were hermetically sealed, to prevent contamination and desiccation. - pF cores The standard procedure for taking pF cores (100 cm3) was applied (Stakman, 1974). Because of the stoniness and shallowness of the soil, on many occasions sampling did not work. Much force was needed to press the rings into the soil. The soil in the cores sometimes showed cracks or other signs of disturbance. Some cores proved to include stones when emptied after pF measurements. The results of these measurements should, therefore, be regarded as indicative and as a relative measure for comparing the soil at different locations and times. - undisturbed samples for thin sections The samples for the soil thin sections were when possible taken according to Brewer (1964) and Bullock et al. (1985). Firstly a small pit was dug. The sampling was performed top down and included stones. The samples were taken in Kubiëna size sampling boxes (8 x 5 x 4 cm) or in Petrographic size boxes (5 x 2 x 1.5 cm). In July and September, due to dry conditions it was not possible to follow the standard procedure. The standard procedure involves cutting the soil so as to let the sides of the box enter into it. Under dry conditions this resulted in the fracturing of the soil into aggregates and in soil spilling out of the box. To solve this problem the soil was impregnated by a fast hardening polyester resin (Araldite BY158 and hardener HY2996, 7 Methods proportion 100:28) in situ. After one night of hardening the soil sample was dug out and moulded to meet normal box sizes as much as possible. These samples eventually also included hardrock parts. Soil was sampled vertically and notes were made of the orientation in order to saw the samples in the right direction. The soil thin sections were produced according to Jongerius & Heintzberger (1975). The terminology as proposed in Bullock et al. (1985) was used for the micromorphological descriptions. .4 Laboratory methods To be able to link the results of the measurements in the laboratory to the behaviour of the soil in situ, care was taken to disturb the samples as little as possible. For this reason laboratory analyses involving as little manipulation of the soil samples as possible were preferred. - soil aggregation The amount of water stable micro-aggregates (MIA) with a diameter between 1 and 106 m was measured as follows. Firstly, the sample was gently sieved at field moisture content through a 106 m. sieve . Next, using a Microscan particle size analyser (Brey & Cammeraat, 1992; Cammeraat et all., in press) size distributions were obtained for samples that were either allowed to disperse naturally in distilled water (water dispersed) or which were completely dispersed ultrasonically (1 minute at an output level of 50% of 50 Watt) Edwards & Bremner, 1964; North, 1976; Imeson & Vis, 1984) The differences in size distributions were interpreted as the water stable micro-aggregate size distribution.. MIA was expressed as mass percentage of the total oven dry material at 105 ºC, material (including stones). The stable macro-aggregation (MAA) (i.e. aggregates with a diameter of more than 250 m - Tisdall & Oades, 1982-), for all samples except those of July 1992, was obtained by gently manually sieving the samples for 2 minutes at field moisture content over a series of sieves of: 250, 1000, 2000, 4000, 4800 or 5000 m. The amount of primary material, e.g. sand and stones, was not recorded for March 1993. The amount of this material was only roughly indicated by visual estimation in percentage classes of 5% for each fraction for September 1993 and March 1994. MAA was expressed as mass percentage of the total oven dry (105 ºC) material. Both micro an macro-aggregation were also studied in soil thin sections. (See section " Other soil biota activities and other features studied in soil thin sections" below. 8 Methods - aggregate stability The single drop test (Low, 1954; Imeson & Vis, 1984) was performed on aggregates between 4 and 5 mm in diameter for March 1993 and between 4 and 4.8 mm for July 1992, September 1993 and March 1994. The number of drops needed to break an aggregate into fragments  2.8 mm were counted. When aggregates broke into fragments larger than 2.8 m, care was taken to continue counting until the largest aggregate fragment had broken down. A maximum of 200 drops was counted, above which the aggregate was considered stable. The test was performed on aggregates at a standard moisture content corresponding to pF = 1 (ASm), but also air dry (a moisture content corresponding with pF varying between 5.75 and 6.1 as a consequence of varying room temperatures, ASd). This moisture content was thought to resemble dry field conditions. The drop test could not always be done because an insufficient number of aggregates of the right size were present. When fewer aggregates were available, the test at pF  6 was preferred. The aggregate stability tests were only performed on samples collected during the first visit. The results of the drop tests is expressed as the average number of drops needed to destroy an aggregate. The standard deviations for these averages are very high. This is caused by a very high variation in the numbers of drops needed to break down the aggregates. As can be seen in figure IV.7. a and b, these numbers varied widely for all samples. A significant detail is that curves get steeper at a low average drop number. Therefore the average drop number is used in this study as measure for the stability. Figure IV.7 a b Examples of aggregate stability curves for high, intermediate and low average numbers of d rops for moist (a) and for dry conditions (b). The x axis number of drops needed to destroy the aggregates. The y axis the tested aggregates as percentage of the total number of aggregates. - pF Water retention curves were constructed using the standard porous medium method (Stakman et al., 1969; Stakman, 1974). For suction values up to -100 cm water (pF = 2) the cores were used on the 'sandbox'. At suction values of - 1000 cm (pF = 3) and - 16000 cm (pF = 4.2) of water, small natural clods were placed in a pressure box. At suction values of - 1000000 (pF = 6) of water, small natural clods were exposed to pressure in the laboratory. To obtain soil moisture content values between the measured pFnumbers a non-linear iteration curve-fitting program using the simplex method (Freyer, 1989). It resulted in a S-shaped (Van Genuchten model) curve running from the minimum to the maximum measured pF. - carbon content Measurements of carbon content (CC) were made according to Allison (1935). This method makes use of the oxidation reaction between the carbon and the applied chromic acid. The CC is expressed as mass percentage of the total dry material. 9 Methods - moisture content Volumetric moisture content (SM) of the samples (including stones) was determined by a weighting, drying (ovendry at 105 ºC), weighting cycle. SM is expressed as volume per unit volume soil (cm3/cm3). - number of bacteria Direct counts of numbers of viable soil bacteria (VB) were performed by direct visual enumeration of fluorescent stained bacteria in soil smears according to Anderson & Slinger (1975). The soil smears were derived by shaking 10 g soil (at field moisture content) in 95 ml sterile water in a Warning blender at the maximum speed for 90 sec. to facilitate dispersion of the organisms. From this suspension, two 10 ml aliquots were taken from a depth of ca 1 cm and spread evenly over 1 cm2 area of degreased and sterile microscopic slide. The smears were allowed to dry at room temperature prior to gentle heat fixing over a gas flame. The differential fluorescent stain (DFS) solution was permitted 60 min. contact with the soil smears at room temperature after which they were rinsed with abundant 50 % ethanol. Thorough air-drying at room temperature preceded mounting under U.V. inert mountant Eukitt. The DFS stain was prepared by adding 173.9 mg europium (III) thenoyltrifluoroacetate (europium chelate Eu(TTA)3) and 2.4 mg fluorescent brightener (photine) to 100 ml 50% ethanol (diluted with Millipore). The cloudiness, crystals and other possible dilutions were removed by filtration over a 0.22 mm filter. A Leica Orthoplan fluorescence microscope equipped with an HBO 200 W mercury lamp was used. The instrument was fitted with filters and reflectors to isolate the 300 - 400 nm wavelengths directly onto the specimens without attenuating the returning fluorescent emissions. This microscope was used with an 100 x water immersed objective and 10 x eyepieces, of which one was fitted with an micrometer and the other with a grid (25 squares). Table IV.4 Classes for counting bacteria Table IV.5 Classes for counting hyphae Counting took place with use of a GWBasic counting program developed by Bloem et al. (1992). This program from the Institute for Soil Fertility Research (DLO), Haren, calculates species abundances and statistics. The possibility of using a correction factor was used to recalculate the numbers for dry soil. For each soil smear an area of 40,164 m2 was counted. Bacteria were counted in four size classes as mentioned in table IV.3. These classes were developed following the 5 classes of Lundgren (1982), modified to fit the populations studied here and in response on practical problems, such as instrument availability. Up to 300 soil smears were processed in this way. VB are expressed as the logarithm of the numbers bacteria per gram dry soil. Plate counts of bacteria (Colony Forming Units) were performed as a check on the direct counts, according to the standard treatment for soil bacteria as mentioned in Heijnen & van Veen (1991). This involved the suspension of 10 gr. soil in 95 ml of a 0.1% sodium pyrophosphate solution and 10 gr gravel (2 - 4 mm) to improve dispersal of bacterial cells. This suspension was shaken for 10 min. at 200 rpm at 28 ºC. Serial 2-fold dilutions were plated on a 0.3 % Tryphicase Soy broth agar (TTS-agar, Heijnen & van Veen, 1991). The pH of this TTS-agar is 7.3 ± 0.2 which is close to the pH of the soils used in this research (pH Cl2Ca 7.39 - 7.74, 10 Methods Soriano-Soto, 1995). The numbers of colonies produced on the plates were counted after incubation for 7 days at a temperature of 28 ºC. The positive correlations coefficient calculated with use of the Pearson Product Moment Correlation between the numbers of viable bacterial cells and the number of Colony Forming Units (CFU) are significant (see figure IV.8). The differences between CFU and VB could be caused by two circumstances. Firstly, the fact that the plate counts took place relatively late in Amsterdam (see figure IV.3). According to the concepts mentioned in part IV.1 bacteria could have died in the period after sampling, or even the whole populations could have changed. Secondly, TTS-agar differs fundamentally from food material available in the soil. Other conditions, such as temperature, moisture, but also the available space, during maturation of the colonies are different from natural conditions. These conditions could have been more suitable for some genera or species of bacteria. Therefore in this study I chose for the counts of viable bacterial cells. - number of fungi Figure IV.8 Relationship between number of Colony Forming Units and Viable Bacteria (bot h in LOG no/g dry soil) correlation coefficient r = 0.39 at p = 0.000. Direct counts of viable hyphae (VFH) were performed by direct visual enumeration of fluorescent stained hyphae in agar films (Anderson & Slinger, 1975, Söderström, 1977). Nine ml of 2 % water agar suspension in de-ionised water was put asceptically in test tubes and autoclaved at 121 ºC for 20 min. The same suspension was used as for the soil smears for counting viable bacteria. From the suspension, 1 ml aliquots were taken from a depth of ca 1 cm, and suspended in the agar. This was directly thoroughly mixed for 60 seconds with a wristaction shaker. Two aliquots were taken from 1 cm depth, using a sterile pipette, and rapidly introduced in sample chambers with a volume of 1 ml (h x w x l : 1 x 20 x 50 mm). The agar was gently spread over the whole chamber. After gelling, the soil-agar film was removed from the chamber and floated free in the photine solution for 20 min. at room temperature. It was then rinsed by free floating in 50 % ethanol and fixated for 30 sec in 96 % ethanol at room temperature. The soil-agar film was fitted gently on a microscopic slide and dried thoroughly at room temperature. The soil-agar film was mounted under U.V. inert mountant Eukitt. The fluorescent brightener (photine) stain was prepared by adding 2.4 mg to 100 ml 50% ethanol (diluted with Millipore). The solution was stored in a cool and dark place. For the visual enumeration of the viable hyphae the same Leica Orthoplan fluorescence microscope with the same equipment, was used as for the count of viable bacteria. This microscope was used with an 25 x water immersed objective and 10 x eyepieces, one of which was equiped with an micrometer. Counting took place with the same counting program mentioned above for the viable bacteria, producing similar results. For each soil-agar film an area up to 2,904,402.4 m2 was counted, depending on the statistics. Hyphae were counted per unit of length (1.39 m) according to the five diameter classes mentioned in table IV.4. These classes were developed to fit the populations studied here and also in response to possibilities offered by the available instruments and time. Up to 300 soil-agar films were processed in this way. VFH were per diameter class eventually expressed as meter hyphal length per gram dry soil. 11 Methods - other indicators studied in soil thin sections Two sizes of soil thin sections were derived from undisturbed soil samples (section IV.2.2), (Kubiëna size thin sections) measuring 8 x 4 cm, with a soil volume of 9.6 mm3 and in Petrographic size boxes (5 x 2 x 1.5 cm). I These were examined macroscopically and microscopically (Bolt & Mücher, 1996). By the study of undisturbed soil samples an insight in the relationships between several soil characteristics was obtained. The characteristics mentioned in table IV.5 got special attention in the microscopic description. Visual quantification in 6 percentage classes by volume of the soil thin section (very dominant 70%, dominant 5069%, common 30-49%, frequent 15-29%, few 5-14% and very few 5%, Bullock et al., 1985) was made at an magnification of 32 times. The classes were expressed in a minimum and a maximum. This was done for each macroscopically distinguished layer or area with similar visual characteristics. However, the results were combined to produce results per location per vegetated and bare area. Therefore the resulting numbers were the minima and maxima found in several relevant soil thin sections and sometimes several parts or areas within these thin sections. Up to 35 soil thin sections were examined in this way. Many features which were eventually used were obtained by addition of percentages of volumes of several characteristics per soil thin section. Most of them can be deduced from the names. It should be stressed that: - 'aggregates' include excrements; - 'pores' include channels; - 'Signs of soil meso fauna activity' consist of channels and excrements. Table IV.6 .5 Micromorphological characteristics in soil thin sections Calculations - pores To calculate the pore diameter the following equation after Koorevaar et al. (1983) was used: zc = 2 cos gr (1) where zc = height of capillary rise (m),  = surface tension (N m-1),  = wetting angle (rad) ,  = density of liquid phase (kg m-3), g = magnitude of gravitational field strength (N kg-1) and r = radius of the capillary pore (m). When ideal circumstances are assumed as: pure water moving through ideal pores with a surface which behaves like glass at complete wetting ( = 0), and also  for water = 0.07 N m-1, for  for water = 1000 kg m-3 and for g = 10 N kg-1 the following relation between the height of the capillary rise (which is inversely proportional to negative suction values) and the radius of the capillary is valid: 12 r = 2 cos gzc Methods (2) This equation leads to the relation between pF and radius of the capillary pore shown in figure IV.9. As described in section IV.3 soil moisture content corresponding to many pF-values were measured and interpolated by way of a Non-linear Iteration Curve-fitting Program using the Simplex Method (Freyer, 1989). By substituting the pF values by the radiuses of capillary pores calculated as described above, the volume of all pores with a diameter less or equal to these radiuses is consequently found. However, as the pF values Figure IV.9 Relation pF and radius of the capillary calculated in the way described in section IV.3 did not match the desired values, the nearest match was used, as can be seen in the fourth column in table IV.6. For these pF values the corresponding values of zc and pore diameter derived with equation (1), are given in the fifth and sixth column of table IV.6. Table IV.7 Pore radius with corresponding z c and pF - microbial biomass To calculate the bacterial (BBM) and fungal (FBM) biomass firstly their volume is needed and secondly their specific gravity should be known. For calculation of the volume of the viable bacteria the volumes corresponding with the size classes as developed by Lundgren (1982) were used. Lundgren (1982) distinguished 5 classes of slightly different sizes. But according to Lundgren (1984) "most (98%) of the bacterial cells fell into the two smallest classes, i.e. classes I and II of Lundgren (1982)". Consequently it was assumed that this was also true for the populations studied here. For the big spheres and rods ( 0.71 m) it was decided, as a consequence of the characteristics of the population encountered in this research, to utilise the volumes for the biggest classes of spheres and rods of Lundgren (1982). For fungal hyphae the length and diameter was known, except for the diameter classes  1.39 m and  4.17 m and for black hyphae. As a consequence of the rather high percentage of very thin hyphae in the class  1.39 m it was decided to assume 1 m as the average diameter of the hyphae in this class. Because hyphae, belonging in the class of  4.17 m, varied in width up to 5 m, it was decided to assume 4.5 m as the average diameter for this class. The few black hyphae present were of a very common size, therefore it was decided to assume 2.78 m as their average width. The specific gravity of fungi was mentioned by Alexander (1977) as 1.2. The specific gravity of bacteria was not mentioned but was assumed to be 1(g/cm3), for the cell mass consist for 70 up to 86% (Schlegel, 1986) of water. The weight of the remaining 14 - 30 mass percent was left out of consideration. By adding up the bacterial and fungal biomass an indication of microbial biomass (MBM) was calculated. Following common practice MBM as well as BBM and FBM were expressed in microgram per gram dry soil (g/g dry soil). 13 Methods Figure IV.10 Relationship between solar radiation (%) and soil temperature (C°) for vegetated (cube) and bare (spheres) sample points. The x axis represents (direct) radiation as a percentage of the direct radiation at the same location on a flat surface at the summer solstice. The y axis represents the maximum soil temperature in °C. - solar radiation Table IV.8 Shading factor per plant species. The solar radiation was calculated by applying the SILVI-STAR Program (Koop & Bijlsma, 1993). This program calculates daily direct and diffuse radiance at ground level by simulated fisheye photography from the soil surface upwards. Digital vegetation maps (5 x 5 m) of all sites together with information about the shading of the direct environment, such as mountains and opposite slopes, constituted the input for the program. The shading factor, as presented in table VI.7, was estimated for each plant species by regarding the actual cover with the total extention of the plants. The calculations were performed for all sites for the equinoxes and solstices. The SILVI STAR Program calculates direct and diffuse radiance for clear sky days and completely overcast days in MJ/m2. By reason of comparability this was recalculated into percentage of the maximum of direct radiance on a flat and bare surface on the same location. These maxima of direct radiation were not similar for the location. The lowest value occurred for CA which accounted for 95% of the highest maximum found for BS. To check the validity of the calculated radiation it was compared with the soil temperature. The soil temperature was recorded by Calvo Cases between July 1993 and March 1994 (see figure IV.10). It was measured continuously, by means of 8 sensors at each location at a depth of 2 to 3 cm in the soil. The result of this comparison of the highest direct radiation and the maximum soil temperature at solstices and equinoxes is shown in figure IV.9. The corresponding significant (at p  0.05) correlation is 0.58 (0.58 for bare and 0.65 for vegetated parts). This correlation is high when one realizes how the characteristics of the material vary. - tests of relationships Hypothesised relationships were tested with the Pearson Product Moment test or by visual judgement. .6 Observations The following analyses were duplicated in one sample: micro and macro-aggregation, carbon content, number of bacteria, viable as well as CFU, and number of fungi. All of these analyses were reasonably reproducible. The biggest differences occur in the aggregation of material  106 m and the number of viable bacterial cells per gram dry soil. The highest variation in percentage of material  106 m aggregated is 10 % between two repetitions. The number of viable bacterial cells sometimes varied more than the standard error of the individual analyses. 14 Methods Some soil smears for the enumeration of viable bacteria were not sufficiently rinsed with 50% ethanol. As a result of this the bacteria could not be counted: the slides had become fluorescent red all over. In this way 16 (of a total of 280) soil smears were excluded and others were only partially countable. The series of soil smears of July 1992 were enumerated in January 1994. It is suspected this has led to a underestimation of the numbers of cells because of a reduction of fluorescent emission. Some soil agar films for the enumeration of viable fungal hyphae were too thick. The cause was either the insufficient shrinkage of the film by drying or a very thick layer of mountant. This caused the hyphae in some slides to be uncountable, because it was not possible to focus the hyphae in the agar films under the microscope. The soil agar films were meant to dry at room temperature. Obviously room temperature, and consequently the air moisture, is in Spain quite different compared to Holland and the drying process was also different. Because of the high acceleration of the drying process the agar films tended to rumple and even to tear. This led not only to increased difficulty with mounting the films but also to a reduced surface area. This reduction of surface area was quantified by comparing the surface of the chamber with the actual surface area of the mounted agar films and expressed as a fraction. This reduction of surface area was allowed for during the enumeration by adding the calculated fraction as a correction factor for the calculations performed by the GWBasic program. The series of pF-cores of July 1992 did not enter into a equilibrium at - 2.5 cm suction of water, which is normally the first step for determination of the water retention curves. Because this was thought to be the effect of a hydrophobic reaction, it was decided to wet the soils completely (0 cm suction). It did not occur in other series. *1 Shirley Hazzard, The Evening of the Holiday, 1965 * 2 For abbreviation see table IV.1 15