Vermiculite’s strong buffer capacity renders it unsuitable for studies of acidity on soybean (Glycine max L.) nodulation and growth

Background Vermiculite is the most common soil-free growing substrate used for plants in horticultural and scientific studies due to its high water holding capacity. However, some studies are not suitable to be conducted in it. The described experiments aimed to test the suitability of vermiculite to study the effect of acidity on nodulation and growth of soybean (Glycine max L.). Methods Two different nutrient solutions (Broughton & Dilworth, and modified Herridge nutrient solutions) with or without MES buffer addition were used to irrigate soybean grown on vermiculite growth substrates. The pH of nutrient solutions was adjusted to either pH 4.0 or 7.0 prior its use. The nodulation and vegetative growth of soybean plants were assessed at 3 and 4 weeks after inoculation. Results The unsuitability of presumably inert vermiculite as a physical plant growth substrate for studying the effects of acidity on soybean nodulation and plant growth was illustrated. Nodulation and growth of soybean grown in vermiculite were not affected by irrigation with pH-adjusted nutrient solution either at pH 4.0 or 7.0. This was reasonably caused by the ability of vermiculite to neutralise (buffer) the pH of the supplied nutrient solution (pH 2.0 – 7.0). Conclusions Due to its buffering capacity, vermiculite cannot be used as growth support to study the effect of acidity on nodulation and plant growth.


Background
Vermiculite is the mineralogical name given to hydrated laminar magnesium-aluminium-iron silicate ((Mg, Fe, Al) 3 ((Al, Si) 4 O 10 )(OH) 2 .4H 2 O). It is an expandable 2:1 mineral and often forms from alteration of mica [1]. It is widely available, easily handled, odourless, and low-cost material [2]. Vermiculite is the most common physical growth substrate used for plants in horticultural and scientific studies due to its high water holding, inert chemical nature, moderate level of aeration, absence of substrate for microbial growth and effective cation-exchange capacities compared to sand or gravel to promote better plant growth.
However, some studies are not suitable to be conducted in this substrate. For example, vermiculite was used as an 'N-free' growth medium for the study of associative N 2 -fixation [3], but later research showed that significant quantities of mineral N were released from vermiculite when it was incubated under warm, moist conditions [4]. Marx and Zak [5] found that vermiculite was not suitable for studying the effect of pH on mycorrhizal formation of slash pine (Pinus elliottii), as this substrate neutralised the pH of Melin-Norkrans nutrient solutions added to this medium.
Acid soil stress factors have been reported to affect stages of the nodulation process, nitrogen fixation, and plant growth [6][7][8][9]. Several types of growth substrates such as liquid (hydroponic) solution, vermiculite, quarzt sand culture, or direct testing of acid soils were commonly used in acid stress studies. The choice of an appropriate growth substrate is very important, because some have the ability to buffer acid treatment. Here we evaluated the properties of vermiculite as plant growth substrate to study its utility for the study of nutrient solution pH on nodulation of soybean.

Nutrient solution preparation
Two nutrient solutions lacking of nitrogen, Broughton and Dilworth [10] and modified Herridge nutrient solution [11] were used to test the effect of acidity stress on soybean grown on vermiculite medium. The pH of nutrient solutions was adjusted to pH 4.0 or 7.0 using 5 N HCl or 5 N NaOH, and each of them was either with or without MES buffer addition. The composition of each nutrient solution is presented in Table 1. In general, Herridge nutrient solution contained higher concentration of each element than B&D nutrient solution. This higher concentration of Herridge medium was expected to have more buffering capacity than B&D plant growth medium.

Soybean planting and Bradyrhizobium japonicum inoculation
Soybean seeds of cv. Bragg were surface-sterilised using 6% of H 2 O 2 in 70% ethanol for 5 min and rinsed seven times with sterile water. Axenic seeds were planted in sterilised 4 L black plastic pots containing autoclaved vermiculite growth substrate and irrigated with B&D or Herridge nutrient solution at pH 7.0. Plants were grown in a controlled environment glasshouse at 28°/24°C day/night temperatures and 16 h day length for 5 weeks. At day 5, pH treatments commenced by adding B&D or Herridge nutrient solution with or without MES buffer at its corresponding pH to soybean seedlings twice daily. The water availability of the system was kept at field capacity (run-off). At day 7, soybean seedlings were inoculated with a 4-day-old YMB culture of Bradyrhizobium japonicum strain CB1809 (10 7 cell per mL, 5 mL per seedling). Each treatment was replicated three times with four soybean plant placed in each pot. The pH of growth medium was measured in the supernatant suspension of a 1:2.5 (v/v; vermiculite: distilled H 2 O mixture) at day 0, 1, 2 and 3, and week 1, 3 and 4 after inoculation. Uninoculation control was provided to check if there was any contamination. This control was irrigated with B&D or Herridge nutrient solution at pH 7.0 with or without MES addition. Plants were harvested at 3 rd and 4 th weeks after inoculation for the assessments of nodule number, nodule dry weight, shoot dry weight, and root dry weight. For dry weight measurement, the tissue was placed in an oven at 65°C for 5 d prior to weighing.

Experimental design and data analysis
The experiment was conducted as a completely randomised design with three replicates. Duncan's multiple range tests at 1% and 5% probability levels were used for mean separation.

Bradyrhizobium japonicum inoculation
Bradyrhizobium japonicum inoculation worked well in this experiment ( Figure 1) as shown by green leaves and good growth of soybean Bragg (a) and the formation of good and effective nodulation (b) as indicated by pink pigmentation of nodules from expression of rhizobiaassociated leghaemoglobin. Contamination was also successfully avoided during the time of this experiment as shown in Figures 1c and d. No nodulation was found in this treatment suggesting that there was no contamination prior and after acid treatment commenced.

Treatment effects on nodulation and growth of soybean
Nodule number was not affected by all treatments tested (types of nutrient solution, pH of nutrient solutions, or MES buffer addition; Figure 2a  significant to other treatments (Herridge medium pH 7.0, B&D medium pH 4.0 and 7.0). Interestingly, pH level of nutrient solution did not affect nodule number and dry weight at both times of analysis. We also found that there was no increase in nodule number and nodule dry weight from week 3 to week 4 after inoculation. This shows the process of autoregulation of nodulation (AON) [12,13] to maintain the balance of nodule formation. Shoot dry weight at both harvest times was not affected by any treatment tested (Figure 2e,f ). Similarly, pH level did not affect root dry weight of soybean at both harvest times (Figure 2g,h). Other treatments (type of nutrient solution and MES addition) did not affect root dry weight at week 4. However, at week 3, soybean watered with B & D nutrient solution (pH 4.0 + MES) had significantly higher root dry weight than soybean watered with pH 4.0 of Herridge medium, but not to other treatment combinations.

Effects of MES buffer addition on soybean nodulation and growth
The effects of MES buffer addition on nodulation and growth were measured by comparing the average values of all treatment combinations under MES addition with the average values of all treatment combinations under no-MES addition. We found that addition of MES did not significantly affect nodulation and growth of soybean at both 3 and 4 weeks after inoculation (data not shown). Bugbee and Salisbury [14] also found that MES was biologically inert and does not interact significantly with other ions in solution. They used MES buffer at concentration 1 and 10 mM to test the growth of beans (Phaseolus vulgaris L.), maize (Zea mays L.), lettuce (Lactuca sativa L.), tomatoes (Lycopersicon esculentum Mill.), and wheat (Triticum aestivum L.). They found that the relative growth rates and plant dry weight among controls and MES treatments were nearly identical for each species during the trial period (3 and 4 weeks). In this experiment we used higher MES concentration (20 mM) to increase the buffering capacity of nutrient solution against high buffering ability of vermiculite. Here we showed that 20 mM MES buffer did not affect nodulation (nodule number and dry weight, both per plant) and growth (shoot and root dry weight) of soybean.

Effects of nutrient solution types on soybean nodulation and growth
The effects of nutrient solution types on nodulation and growth were measured by comparing the average values of   It shows that at week 3 after inoculation, soybean watered with Herridge medium had significantly higher nodule dry weight than soybean watered with B & D medium (Figure 3a). However, there was no effect on nodule number, shoot dry weight and root dry weight per plant. At week 4, watering with Herridge nutrient solution increased not only nodule dry weight but also shoot dry weight of soybean (Figure 3b). The better nodule development (weeks 3 and 4) and shoot growth (week 4) of plants watered with Herridge medium were caused by the higher, but clearly non-toxic concentration of most elements in this substrate. As shown in Table 1, Herridge nutrient solution contains higher concentration of each element than B&D nutrient solution. The different forms of the elements in Herridge nutrient solution might also determine a different bioavailability of the elements to the plants.

Effects of nutrient solution pH on soybean nodulation and growth
The effects of nutrient solution pH on nodulation and growth were measured by comparing the average values of all treatment combinations under pH 4.0 with the average values of all treatment combinations under pH 7.0. We found that there was no significant effect of nutrient solution pH on nodulation and growth of soybean at both 3 and 4 weeks after inoculation (data not shown). This is interesting because we expected that nodulation and growth of soybean watered with nutrient solution of pH 4.0 were significantly less than soybean watered with nutrient solution of pH 7.0. Previous experiments also showed that acidity had negative effect on nodulation and growth of soybean [6][7][8]. To find out the cause of these unexpected results, we analysed the vermiculite properties especially on its effects on pH of the substrate medium.

Substrate properties
To check the pH values of growth substrate, samples were collected from each pot at days 0, 1, 2 and 3, and 1, 3 and 4 weeks after inoculation (see Figure 4). Here we measured the pH of the bulk vermiculite medium and assume that it corresponds to the pH of the solution around the roots as we did not observe alterations of nodulation and growth using nutrient solutions adjusted to pH 4.0. It is shown here that the real pH of the medium was not similar to the pH of the input nutrient solution.
Vermiculite increased each of the tested nutrient solution pH. In general, the pH increase of vermiculite substrate watered with Herridge nutrient solution was lower than vermiculite substrate watered with B&D nutrient solution. This result shows that Herridge medium has higher buffering capacity as a result of its higher concentration of most elements in this nutrient solution ( Table 1). The actual pH values of vermiculite substrate watered with pH 4.0 of Herridge and B&D nutrient solutions were around 7.0 and 7.5, respectively. Moreover, the actual pH of vermiculite watered with pH 7.0 of nutrient solutions was around 7.5 and 8.0. We also found that addition of MES buffer into the nutrient solutions was not able to maintain the initial pH of nutrient solutions. It only slightly lowered the pH increase of substrate watered with pH 4.0 nutrient solutions but not to substrate watered with pH 7.0 nutrient solution ( Figure 4).

Vermiculite substrate axenically neutralised nutrient solution pH
To check the 'actual' pH of vermiculite after the addition of nutrient solutions, a simple experiment was conducted. Twenty-five mL of nutrient solution with different pH values was added to 10 mL (2.5 g) of vermiculite. The axenic mixture was shaken (in a Falcon tube; 120 rpm; 28°C) and the pH value of the suspension was measured after 1 min to 96 hours of shaking ( Figure 5a). As shown in this figure, vermiculite easily neutralised the pH of both nutrient solutions whether MES buffer was added to the nutrient solution or not. Nutrient solution with MES buffer addition had slightly better buffering capacity than nutrient without MES, but the final pH values were neutral, even alkali in both nutrient solutions. We also found that vermiculite was able to neutralise very acid nutrient solutions (2.0 and 3.0) to around pH 5 and 6 after 4 and 24 hours of shaking, respectively ( Figure 5b). From this experiment it is clearly shown that vermiculite has very strong buffering capacity. Only as little as 2.5 g of vermiculite was able to neutralise 25 mL of nutrient solutions with or without MES buffer addition. Buffering ability of vermiculite was also shown by Duman and Tunç [15] when water of different pH (2.20, 3.0, 5.95, and 10.58) was added to vermiculite substrate.