A Standardized Method for Estimating the Functional Diversity of Soil Bacterial Community by Biolog® EcoPlatesTM Assay—The Case Study of a Sustainable Olive Orchard

Biolog® EcoPlatesTM (Biolog Inc., Hayward, CA, USA) were developed to analyse the functional diversity of bacterial communities by means of measuring their ability to oxidize carbon substrates. This technique has been successfully adopted for studying bacterial soil communities from different soil environments, polluted soils and soils subjected to various agronomic treatments. Unfortunately, Biolog® EcoPlatesTM assay, especially working on soil, can be difficult to reproduce and hard to standardize due to the lack of detailed procedures and protocols. The main problems of this technique mainly regard soil preparation, bacterial inoculum densities and a correct definition of blank during the calculation of the diversity indices. On the basis of our previous research on agricultural soils, we here propose a standardized and accurate step-by-step method for estimating the functional diversity of a soil bacterial community by Biolog® EcoPlatesTM assay. A case study of soils sampled in a Mediterranean olive orchard managed accordingly to sustainable/conservation practices was reported for justifying the standardized method here used. The results of this methodological paper could be important for correctly evaluating and comparing the microbiological fertility of soils managed by sustainable/conservation or conventional/non-conservation systems.


Introduction
Microorganisms are present in all ecosystems and due to their rapid responses to physical and chemical changes, they can be used as bioindicators of environmental quality. The Community-Level Physiological Profiling (CLPP) is a rapid and relatively inexpensive technique to relate microbial functional diversity over space and time to changes in the environment [1][2][3][4].
Biolog ® EcoPlates™ (Biolog Inc., Hayward, CA, USA) were developed to analyse the functional diversity of bacterial communities by means of measuring their ability to oxidize carbon substrates. An EcoPlate is a 96-well microplate that contains 31 common carbon sources from altogether five compound groups-that is, carbohydrates, carboxylic and ketonic acids, amines and amides, amino acids and polymers-plus a blank well as a control, all these replicated thrice to control variation in inoculum densities. Each EcoPlate is filled with a dilution of one soil suspension, thus representing one soil sample. The utilization rates of carbon compounds in the wells are quantified spectrophotometrically by following the reduction of water-soluble colourless triphenyl tetrazolium chloride to purple triphenyl formazan. For the measurements of optical density (OD), two filters are used: (a) 590 nm (absorbance peak of tetrazolium) to evaluate colour development plus turbidity values and (b) 750 nm to measure turbidity values only. The turbidity of dilutions is due to clay and humic particles in soil colloidal suspension.
Every bacterial community has a characteristic reaction pattern with different OD values for different carbon compounds, called a 'metabolic fingerprint' [2,5]. While the inoculated bacterial density significantly affects the rate of colour development in the wells, on the other side the choice of a wrong inoculum can compromise the results of this techniques [5,6]. For this reason, it should be necessary to accurately choose the inoculum densities for different soil samples using a plate count culture-based method. Moreover, a correct definition of blank is essential for calculating the related diversity indices. Biolog ® EcoPlates™ have been successfully adopted in our laboratories for studying bacterial soil communities from different soil environments, polluted soils and soils subjected to various agronomic treatments [6][7][8][9]. Unfortunately, the studies regarding soil as a matrix for bacterial communities often include a Biolog ® analysis but the reagents and apparatus for soil preparation, dilution and incubation time used, standard deviations and number of replicates and precise formulas for calculations are often not reported, so it becomes difficult for other researchers to understand how to correctly use this method.
On this basis and taking into account our previous research on soils, we here propose for the first time a standardized and accurate step-by-step method for estimating the functional diversity of a soil bacterial community by Biolog ® EcoPlates TM assay. We tried to adopt a clear terminology and parameters, explaining them in detail, in order to facilitate the calculation of the most relevant and used Biolog ® -related indices of microbial functional diversity. Even with its limitations [10], the Biolog ® EcoPlates TM method remains a quick and relatively simple and inexpensive technique for comparing microbial communities. The standardization of this technique could be essential for indicating differences in CLPPs between samples within a single experiment, for comparing different experiments, soil types and management systems or simply soils sampled at different times. The aim of this report is not describing the limitations of the method itself but proposing a detailed methodological procedure, easy to be followed and reproduced and based on previous data and experiments. A case study of soils of a Mediterranean olive orchard was reported for comparing the methods and justify the methodological protocol here proposed.
For making and plating, 3 g of TSA powder were added to 1 L of distilled water in a 2-L glass bottle. This solution was sterilized (autoclaved) at 121 • C for 20 min, cooled at 50 • C and then 100 µg cycloheximide mL −1 were added mixing thoroughly. Finally, 20 mL of the solution were poured in each Petri dish.
For plate counting, in order to obtain a 0.18% (w/v) Na 4 P 2 O 7 • 10 H 2 O final concentration, 4.5 mL of 1.8% sodium pyrophosphate and 40.5 mL of 25% Ringer solution were added to 5 g (dry weight equivalent) of fresh soil. The suspension was sonicated for 2 min and soil particles were allowed to settle at 4 • C for 15 min. Then, ten-fold serial dilutions of the supernatant up to 10 −7 in sterile Ringer solution were done, spreading a 100 µL-aliquot of each dilution onto a TSA plate (3-5 replicates for each dilution) and incubating at 28 • C for 72 h. For each sample the suited dilution to enumerate colony forming units (CFUs) for g of dried weight soil was chosen. Then, for microplate incubations, the dilution leading to~10 4 CFUs mL −1 solution was selected for microplate incubation.

Microplate Incubation
The following materials are used: multichannel pipet and sterile tips, sterile plastic multichannel reservoir, Biolog ® EcoPlates™ (Biolog Inc., Hayward, CA, USA). The apparatus includes laminar flow hood, incubator, agitator and Biolog ® Microplate Reader TM equipped with 750-nm and 590-nm filters.
For preparing the microplate, 10 mL of the dilution that was chosen in plate counting was poured into a sterile reservoir of an 8-channel pipet (be careful there are no bubbles in the dilution) and 120 µL of the dilution were inoculated into each well of a microplate. Not the same dilutions of the counting assay were used but new fresh dilutions were prepared. Then, the microplate was placed in its bag to avoid desiccation and incubated at 25 • C in dark, continuously shaking at 50 rpm on an agitator with tilting platform to obtain a uniform distribution of triphenyl formazan. Finally, spectrophotometric readings at both 590 nm (OD 590 ) and 750 nm (OD 750 ) were taken at time 0 and at 12-h increments (±12 Xh ) up to a 144-h incubation.
In order to select the optimal incubation time for microplate analyses (as explained later in the Results and Discussion Section), it is recommended to follow this pattern for each incubation time (as an example, measuring times at 0 h and X h are given here): • calculate a colour value for each substrate well i and the blank (water) well b for each incubation time by subtracting the OD 750 value from the OD 590 value: • subtract the blank well OD reading from the OD value of each substrate well to obtain a blank-corrected value (i bc ) for each well: For a more detailed analysis, the carbon substrates can eventually be divided into eight classes of compounds (polysaccharides and complex compounds, cellulose, hemicellulose, chitin, phosphorylated compounds, organic acids, amino acids and biogenic amines) and the AWCD and diversity indices calculated for each group separately (see examples in References [4,5] and in Figure 1). were calculated subtracting from the raw data (OD 590 -OD 750 ) the blank well optical density (OD) reading from the OD value of each substrate well (i bc ) for each well and successively subtracting the blank-corrected OD reading at time 0 from subsequent blank-corrected readings at ±12 Xh (c i ) (in purple). The same methodology was carried out without subtracting the blank well OD reading (in green), without subtracting the blank-corrected OD reading at time 0 (in blue) or without subtracting both (raw data in red).

Experimental Field and Soil Sampling
The trial was done in a 1-ha olive orchard (Olea europaea L., cv. 'Maiatica'; 70-year-old plants with a distance of 8 × 8 m; NE orientation) located in Ferrandina (Southern Italy, Basilicata region; N 40 • 29 ; E 16 • 28 ). The area has a semi-arid climate, with an annual rainfall of 560 mm (mean 1995-2018), concentrated mostly in the winter and a mean annual temperature of 16.3 • C (mean 1995-2018). The soil is a sandy loam, a Haplic Calcisol, according to the World Reference Base for Soil Resources, with a mean bulk density of 1.30 g cm −3 and sediment as parental material.
The orchard was managed accordingly to sustainable/conservation agricultural practices, namely: In October 2018, soil sampling was done in the undisturbed inter-row area. Soil sub-samples were picked in 10 points over a 1-m radius area around each olive tree from the topsoil layer (0-10 cm) for bacterial communities' analysis by Biolog ® assay. The 10 sub-samples were pooled on site to constitute a composite soil sample of about 1 kg. Five composite samples (n = 5), each composed of 10 different sub-samples, were prepared, in order minimize spatial variability. After removing visible crop residues, roots and pebbles with sterile tweezers and slightly mixing and homogenizing with a sterile spatula, the soil composite samples were immediately stored in sterilized plastic bags at 4 • C and subsequently analysed within 5 days using the procedure described in the previous paragraph.

Influence of Bacterial Inoculum Densities
For the best discrimination of bacterial communities, it is of fundamental importance to choose the shortest incubation time at which AWCD reaches a peak value before the following constant phase. Figure 2A shows the AWCD values measured every 12 h during a 144-h incubation from different soil dilutions. In our case the dilutions examined were 10 −5 , corresponding to~10 4 colony forming units (CFUs) mL −1 measured by plate counting, 10 −4 and 10 −6 . It appears that the dilution with 10 −6 dilution contained a relatively low number of bacteria (~10 3 CFUs mL −1 ) causing an increasing AWCD trend over time, without reaching a constant phase before 144 h and high variability among replicates, as showed by the high values of standard deviation (Figure 2A). On the other side, the 10 −4 dilution likely had too many bacteria (10 5 CFUs mL −1 ) and AWCD reached a peak after only 48 h, followed by a slight decline likely due to substrates or triphenyl formazan degradation, without a clear constant phase (Figure 2A). Instead, the dilution of 10 −5 (~10 4 CFUs mL −1 ) reached a constant phase at 96 h and so it was chosen as the time for measuring AWCD values (Figure 2A). Similar trends were observed both for H' (Figure 2B) and E values ( Figure 2C), that had a constant phase starting from 96 h in the dilution of 10 −5 (~10 4 CFUs mL −1 ), with low values of standard deviation from 96 to 120 h of incubation ( Figure 2B,C). The common pattern of the graphs of Figure 2 likely depends on the fact that the calculation of AWCD, H' and E are all based on the values of c i , as previously explained. Comparing our results with those of other studies [10][11][12][13][14][15][16] using our protocol, we managed to obtain more stable growth curves for AWCD, particularly in the constant phase and relatively low values of SD (Figure 2A). Another important remark is that the inoculum could be relativized simply resuspending soil mass, in order to not to artificially homogenize differences that are intended to be detected among the different soil samples of the same experiment [17][18][19][20]. In order to avoid this possible artefact, in our case, the same dilution (leading to~10 4 CFUs mL −1 ) was chosen for all the samples analysed in the same experiment.

Correct Calculation of Blank
The AWCD values measured on the 10 −5 dilutions at 96 h were calculated subtracting from the raw data (OD 590 -OD 750 ) the blank well OD reading from the OD value of each substrate well (i bc ) for each well and successively subtracting the blank-corrected OD reading at time 0 from subsequent blank-corrected readings at ±12 Xh (c i ) for both 0-h and 96-h incubation times. The same methodology was carried out (a) without subtracting the blank well OD reading, (b) without subtracting the blank-corrected OD reading at time 0 or (c) without subtracting both. The results, depicted in Figure 1, were quite different on the basis of the type of calculation done, particularly for some classes of carbon substrates (e.g., phosphorylated compounds), showing that it is necessary to apply the sequential subtractions and that non-specific OD values must be taken into account for obtaining reliable AWCD data.

Influence of Soil Preparation
On the same soil samples (10 −5 dilution, reading at 96 h, subtraction of both blank well and blank-corrected OD readings) and on the basis of previous experiments [7,9], eight concentrations of Na 4 P 2 O 7 • 10 H 2 O ranging from 0.0% to 3.0% (w/v) (0, 0.6, 1.0, 1.4, 1.8, 2.2, 2.6 and 3.0%) were used for finding the optimal concentration to be used. Due to its high number of electric charges once in solution, it is known that this compound has the ability to disperse charged soil colloids. This is particularly important for soil microbiological assays, as bacteria are not only present on the surface of soil aggregates but also inside the aggregates, that is, on the micropore and macropore walls and in their internal spaces [6]. Figure 3 shows that with more than 1.8% (w/v) Na 4 P 2 O 7 soils did not show an increase in AWCD, H' and E, probably because the maximum possible number of bacteria was extracted from soil. Of course, this value could vary for different soil types but we can suggest that a concentration of around 2.0% (w/v) is necessary for not underestimating AWCD values. The use of the appropriate concentration of Na 4 P 2 O 7 allows to overcome the technical problem in extracting the highest number of microorganisms from soil samples, that was one of the main remarks of this technique highlighted by Preston-Mafham et al. [10].

Conclusions
A detailed and step-by-step description in the materials and methods sections of the scientific articles dealing with Biolog ® is often lacking, making this method hardly reproducible. For this reason, we hope that the description of this method can help to solve the main issues related to the Biolog ® EcoPlates™ assay: (1) soil preparation, dispersion and dilution (importance of using composite samples and Na 4 P 2 O 7 ); (2) the optimal bacterial density of the inoculum (~10 4 CFUs mL −1 ); (3) a correct definition for blank (c iXh = i bcXh − i bc0h ); and (4) the use of unequivocal terminology and parameters in the calculation of diversity indices, as reported in the formulas. The previously-cited problems make this technique apparently difficult to reproduce and hard to standardize. This notwithstanding, we tried to demonstrate that, if rigorously executed, Biolog ® EcoPlates™ assay can be a relatively simple and powerful technique, extremely useful for evaluating soil bacterial functional and metabolic diversity. The results of this methodological paper could be important for correctly evaluating and comparing the microbiological fertility of soils managed by sustainable/conservation or conventional/non-conservation systems.