Hydrological Behavior of Peat-and Coir-Based Substrates and Their Effect on Begonia Growth

The physical–hydraulic properties of eight substrate mixtures based on sphagnum peat and coir were determined and their effect on the growth of Begonia xelatior was studied. The particle size distribution, water retention curve, saturated hydraulic conductivity, and pore size distribution of the substrates were determined. All substrates exhibited high total porosity, satisfactory water retention capacity, and high saturated hydraulic conductivity. Increasing the percentage of perlite in the mixtures contributed to the reduction of water retention capacity and the increase of large pores. Unsaturated hydraulic conductivity estimated by the Mualem–van Genuchten model showed a sharp decrease within a range of water pressure heads (0 to −50 cm) observed between two successive irrigations. To assess aeration and water retention capacity, total porosity; airspace; and easily, and nonavailable water, as well as the bulk density of the substrates, were determined and concomitantly compared with the “ideal substrates” determined by De Boodt and Verdonck. The comparative results showed that substrate porosity alone is not efficient to create ideal plant growth conditions, but dynamic parameters are also needed. Plants grown in a substrate classified as “nonideal” showed significantly greater growth than the plants grown in most of the other substrates studied.


Introduction
Soilless growing substrates used in greenhouse, container, and green-roof production systems consist of either organic materials or mixtures of organic and inorganic materials.Suitable attributes of the materials used in these substrates for horticultural use are low cost and physical properties supporting adequate aeration and water retention for optimal plant growth.
Sphagnum peat is a widely used substrate for ornamental plant production [1][2][3][4].Altogether, economics, availability of new materials, and peat environmental awareness led to the interest in new substrates [5].Research has been heavily focused on the use of industrial, municipal, or agricultural wastes (e.g., [6,7]), but some of these either are not produced in sufficient quantities and/or may contain undesirable materials or properties (chemical and physical).The most promising alternative media are coir, bark, wood fibers, and composts [8].Also, there are other products that are considered alternative substrates to peat, such as different mixtures based on biochar or by-products of the agri-food industry [9].
Water 2018, 10, 722 2 of 15 Coconut coir is an organic by-product derived from the mesocarp tissue or husk fiber treatment of the coconut (Cocos nucifera) fruit.The long fibers are used for industrial purposes, and the remaining materials, consisting of short fibers and dust, undergo aerobic composting.After composting, the stable material is dehydrated and compressed into a compact form (bricks) for easy transportation.With the addition of water, coir expands to 5 to 9 times its compressed volume.
Substrates such as peat and coir offer many advantages, such as low bulk density, being pathogen-and weed seed-free, high water retention capacity, and easy root penetration.The primary disadvantages of peat and coir are poor aeration and low pressure heads due to a low percentage of large particles.This creates a problem in the water-air balance and gas exchange under different watering regimes [1,3,14,24].To mitigate this disadvantage in pure substrates, various materials characterized by large particle sizes are added, with perlite being the most popular one used [2,14,20,25].The addition of coarse perlite to peat improves aeration [2,14,20,[25][26][27][28].
However, the addition and mixture of different materials raises questions about the ratios of the substrate's mixture components to create a high-quality substrate for optimum plant growth.
Standards for substrates were developed by De Boodt and Verdonck [24], determining appropriate physical values, such as total porosity, airspace, and water capacity for optimum plant growth in containers and in bed cultivation.De Boodt and Verdonck [24] recommend a limit of water availability in a water pressure range of −10 to −100 cm.This limit was later confirmed by other researchers [29,30].These standards have been implemented and used commercially at a large scale [31][32][33] and still form the basis for manufacturing growth substrates [34].However, less attention was paid to dynamic parameters and processes such as water use and gas diffusivity [34].In this direction, we try to point out the necessity of using, also, dynamic parameters in substrate evaluation.
This paper presents a study of the physical-hydraulic characteristics of eight substrate mixes based on peat and coir and assessment of their effect on the growth of Begonia xelatior.The main objective was to determine the hydraulic characteristics of these substrates, widely used in floriculture, and the secondary, to evaluate them with the properties identified by De Boodt and Verdonck [24] in making the "ideal substrates".Also, it is investigated whether it is sufficient to assess the suitability of the substrates based only on their static hydraulic characteristics.

Plant Materials and Cultivation Treatments
Uniform begonia plants (Begonia xelatior, "The President") supplied by a local nursery were individually planted into 4938 cm 3 pots (height 19.9 cm, base diameter 16.4 cm, and top diameter 19.3 cm) with useful substrate volume 3947 cm 3 (height 16.3 cm, base diameter 16.4 cm and top diameter 18.7 cm).The required amount of each substrate was calculated by multiplying the volume of the substrate and its bulk density, taking into account the initial moisture of the substrate.All pots were packed in the same way as to receive the same substrate volume.Plants were placed in a glass greenhouse at the Agricultural University of Athens (Athens, Greece) and were arranged in a randomized complete block design with 10 replications.Throughout the experiment, the daytime temperature ranged between 20-35 • C and at night-time was approximately 12 • C; the humidity during the daytime ranged between 30%-60% and at night-time was 90%.
The irrigation was initiated manually through an automated irrigation system which consists of an eight-station controller (GreenKeeper 212, TORO, Bloomington, MN, USA), with eight valves and eight corresponding pressure regulators and PE pipelines to connect the water supply through the controller to the two drippers positioned opposite each other at the surface of each pot at a water flow rate of 2 L•h −1 .Tensiometer readings were taken each morning throughout the study.Each pot was irrigated with 350 cm 3 of water when the mean tensiometer pressure head reached −50 cm, because it sets the lower limit of easily available water as defined by De Boodt and Verdonck [24].The amount of irrigation water was enough to barely leach from the bottom of the pot.The values of water tension were measured using the tensiometers as presented by [16], positioned randomly in three pots for each substrate.
Therefore, all plants received the same cultivation procedures (i.e., applications of fertilizer, fungicide, etc.).A soluble 20N/8.7P/16.6Kfertilizer at 136 mg•L −1 N, fortified with micronutrients (Rosso, ALFA Agricultural Supplies S.A., Athens, Greece), was used at each irrigation event for all plants throughout the duration of the experiment.

Physical-Hydraulic Properties of Growing Substrates
The following physical-hydraulic properties of the substrates studied were determined:

•
Water retention curve measurements were performed on a tension plate apparatus in a Haines-type assembly [35], with an air-entry value of −190 cm of a water column (three samples of each substrate were used).Substrate-sample size approximated the pot diameter used in the experiment; that is, about 3 cm in height and 18.4 cm in diameter.

•
The water content at 15,000 cm pressure head was determined using a high-range membrane method [36].

•
The saturated hydraulic conductivity, K s , was determined using the constant-head method [37].

•
Pore size distribution was obtained from the retention curves by plotting their slopes as functions of the pressure potential [38].

•
The bulk density of each substrate sample was determined for the volume obtained immediately after the last measurement of the water retention curve and drying in an oven for 48 h at 105 • C.

•
Screen analysis was performed to determine the particle distribution of the substrates.Three air-dried samples, 100 g mass each, from each substrate used for the plant experiment were placed in the top sieve of a column of sieves of 0.068, 0.125, 0.25, 0.50, 1.00, 2.00, 4.00, and 8.00 mm screen mesh sizes according to decreasing aperture and rested on a sieve shaker for 5 min at 30 shakes per minute.

•
Oxygen concentrations of substrates were determined along the depth of the pots using a portable gas analyzer (multi-gas analyzer LMSx, Eijkelkamp, Giesbeek, The Netherlands).Air samples along the depth of the pots were extracted via the perforated chambers into acrylic tubes as presented by Londra et al. [14].The oxygen-measuring devices were positioned randomly in three pots for each substrate, and O 2 concentrations were measured at 5, 10, and 15 cm height from the base of the pot.The measurements were made before each irrigation event.

Calculation of the Hydraulic Properties by the Closed-Form Hydraulic Model
The RETC (RETention Curve) program [39], was used to calculate the fitting hydraulic parameters of the popular Mualem-van Genuchten [40,41] model on the experimental water retention data.
Van Genuchten [41] described the water retention curve as where θ is the soil water content (cm 3 •cm −3 ); the subscripts s and r indicate saturated and residual values of water content, respectively; H is the water pressure head (cm); and α (cm −1 ), m (−), and n (−) are curve-fitting parameters: m = 1 − 1/n and 0 < m < 1.
Combining Equation ( 1) with the model developed by Mualem [40], the hydraulic conductivity in relation to water content, K(θ), or pressure head, K(H), can be calculated as The model fitting parameters described above were evaluated by the RETC program using the measured water retention and saturated hydraulic conductivity data.In the parameter optimization process to fit the water retention function, the unknown parameters of the Mualem-van Genuchten model were α, n, and θ r .

Plant Growth
Plants were harvested at the end of the cultivation period (after 16 weeks) and divided at soil level into shoot and root.The harvested roots were carefully washed clean from the substrate.Fresh weight of shoot and root samples were determined.Subsequently, samples were dried at 70 • C for 48 h and their dry weights determined.The percent growth increase was calculated as the ratio of the difference between the total dry weight at the end (n = 10) and start of the experiment (n = 10) by the total dry weight at the end of the experiment.As all plants at the beginning of the experiment were uniform in size, the calculated dry weights were considered the same for all treatments.
Leaf samples for nutrient analysis were collected at the end of the experiment and were dried at 70 • C for 48 h.Dried leaf samples were ground in a grinding mill to pass a 0.84 mm (20-mesh) screen.The ground samples were subjected to a wet digestion procedure (HNO 3 and 30% H 2 O 2 ) for nutrient analysis [42].N concentration was determined by Kjeldahl digestion (Velp Scientifica, UDK 132, Via Stazione, Italy).K and Na concentrations were determined by flame photometer (Corning 410 model, Sherwood Scientific, Cambridge, UK), and all the other nutrients were measured using the atomic absorption spectrophotometer (Spectr AA-20 model, Varian, Sydney, Australia).

Experimental Design and Statistical Analysis
The experiment was a randomized complete block design with eight treatments (substrates) and ten replications (ten plants per substrate).The analysis of variance of the experimental data was performed using JMP (SAS Inst., Cary, NC, USA) statistical software, and treatment means were compared using Tukey-Kramer's test at a probability level p = 0.05.

Physical-Hydraulic Properties of Growing Substrates
To assess the water retention capacity of the substrates studied, a comparative presentation of water retention curves is given in Figure 1a-c.
The addition of perlite into peat and coir decreases the total porosity (water content at saturation) and water retention capacity of the peat-perlite (Figure 1a) and coir-perlite (Figure 1b) mixtures compared to the pure substrates.The decrease was higher in mixtures with higher percentages of perlite.In the case of the two commercial substrates, the comparison showed that Ps 60 :Pb 30 :P 10 has a greater total porosity and water retention capacity than A 70 :Ps 24 :P 6 (Figure 1c).The knowledge gained from the determination of the substrates' water retention curves plays a prime role in proper irrigation management.
In Table 3, the Mualem-van Genuchten model fitting parameters α, n, and θ r , as well as the experimental values θ s for all substrates examined, are given.As shown in Figure 1, it is apparent that there is a very good agreement of the results between the experimental and predicted values of θ(H), indicating that the corresponding soil hydraulic parameters listed in Table 3 provide an adequate description of the θ(H) relationship with a coefficient of determination R 2 ≥ 0.993 for all substrates.
Differentiating the water retention curve (dθ/dH at each pressure head H) of all substrates studied (Figure 1a-c), the pore size distribution can be revealed [38,43] and is presented comparatively in Figure 2 as a relationship between dθ/dH and equivalent pore diameter.As shown, generally, A 70 :Ps 24 :P 6 and coir-perlite mixtures (C, C 75 :P 25 , C 50 :P 50 ) have larger macropores (equivalent pore diameter >75 µm), followed by Ps 60 :Pb 30 :P 10 and then peat-perlite mixtures.With regard to the peat-perlite mixtures (Ps, Ps 75 :P 25 , Ps 50 :P 50 ), the addition of perlite increased the size of large pores, especially those associated with the water tension of −5 cm.Pore size increased as the percentage of perlite in the mixture increased.The same was observed in the case of coir-perlite mixtures.On the other hand, Ps 60 :Pb 30 :P 10 and peat-perlite mixtures have larger mesopores (equivalent pore diameter 30-75 µm) than A 70 :Ps 24 :P 6 and coir-perlite mixtures.Finally, all substrates analyzed have the same micropores (equivalent pore diameter 5-30 µm).

Physical-Hydraulic Properties of Growing Substrates
To assess the water retention capacity of the substrates studied, a comparative presentation of water retention curves is given in Figure 1a-c The addition of perlite into peat and coir decreases the total porosity (water content at saturation) and water retention capacity of the peat-perlite (Figure 1a) and coir-perlite (Figure 1b) mixtures compared to the pure substrates.The decrease was higher in mixtures with higher percentages of perlite.In the case of the two commercial substrates, the comparison showed that Ps60:Pb30:P10 has a greater total porosity and water retention capacity than A70:Ps24:P6 (Figure 1c).The knowledge gained from the determination of the substrates' water retention curves plays a prime role in proper irrigation management.
In Table 3, the Mualem-van Genuchten model fitting parameters α, n, and θr, as well as the experimental values θs for all substrates examined, are given.As shown in Figure 1, it is apparent that there is a very good agreement of the results between the experimental and predicted values of θ(H), indicating that the corresponding soil hydraulic parameters listed in Table 3 provide an adequate description of the θ(H) relationship with a coefficient of determination R 2 ≥ 0.993 for all substrates.regard to the peat-perlite mixtures (Ps, Ps75:P25, Ps50:P50), the addition of perlite increased the size of large pores, especially those associated with the water tension of −5 cm.Pore size increased as the percentage of perlite in the mixture increased.The same was observed in the case of coir-perlite mixtures.On the other hand, Ps60:Pb30:P10 and peat-perlite mixtures have larger mesopores (equivalent pore diameter 30-75 μm) than A70:Ps24:P6 and coir-perlite mixtures.Finally, all substrates analyzed have the same micropores (equivalent pore diameter 5-30 μm).With regard to particle size distribution (Figure 3), in the case of the peat-perlite mixtures (Ps, Ps75:P25, Ps50:P50), the addition of perlite decreased the percentage of particle sizes greater than 4 mm and less than 1 mm and increased the percentage of particle sizes between 4 and 1 mm.The same was observed in the case of the coir-perlite mixtures, with the exception of particles greater than 4 mm, where the percentage of these particles increased slightly with the addition of perlite in pure coir.Comparison between peat-perlite and coir-perlite mixtures showed that the latter was characterized by a lower percentage of particle sizes greater than 4 mm and a higher percentage of particle sizes less than 1 mm.In the case of prefortified substrates, Ps60:Pb30:P10 was characterized by a lower percentage of particle sizes greater than 1 mm and a higher percentage of particle sizes less than 1 mm compared to A70:Ps24:P6.It is worth noting that Ps60:Pb30:P10 was characterized by the highest percentage of particle sizes less than 1 mm (≈77%) compared to all the other substrates studied.The abovementioned variation of the substrate particle sizes led to differences in various physical properties of the substrates.Large differences in particle size can result in the migration of fine particles into the spaces between larger fragments, resulting in reduced air-filled porosity, increased volumetric water content at specified pressures, and increased bulk density [28].With regard to particle size distribution (Figure 3), in the case of the peat-perlite mixtures (Ps, Ps 75 :P 25 , Ps 50 :P 50 ), the addition of perlite decreased the percentage of particle sizes greater than 4 mm and less than 1 mm and increased the percentage of particle sizes between 4 and 1 mm.The same was observed in the case of the coir-perlite mixtures, with the exception of particles greater than 4 mm, where the percentage of these particles increased slightly with the addition of perlite in pure coir.Comparison between peat-perlite and coir-perlite mixtures showed that the latter was characterized by a lower percentage of particle sizes greater than 4 mm and a higher percentage of particle sizes less than 1 mm.In the case of prefortified substrates, Ps 60 :Pb 30 :P 10 was characterized by a lower percentage of particle sizes greater than 1 mm and a higher percentage of particle sizes less than 1 mm compared to A 70 :Ps 24 :P 6 .It is worth noting that Ps 60 :Pb 30 :P 10 was characterized by the highest percentage of particle sizes less than 1 mm (≈77%) compared to all the other substrates studied.The abovementioned variation of the substrate particle sizes led to differences in various physical properties of the substrates.Large differences in particle size can result in the migration of fine particles into the spaces between larger fragments, resulting in reduced air-filled porosity, increased volumetric water content at specified pressures, and increased bulk density [28].
To evaluate the results received from the water retention curves and both the pore size and particle size distributions, Table 4 was created.In Table 4, the terms introduced by De Boodt and Verdonck [24] are used to characterize the water retention capacity and aeration of substrates in a range of water tensions between 0 (saturation) and −15,000 cm (permanent wilting point).Furthermore, in the same table, the bulk density values of all the substrates studied are shown.In accordance to Table 4, the total porosity, easily available water, water buffering capacity, difficult available water, and nonavailable water were decreased in peat-perlite and coir-perlite mixtures compared to pure peat and coir, respectively.These differences were significant mainly in the case of 50% perlite percentage.On the other hand, as anticipated, substrate air space increased along with increasing the percentage of perlite in mixtures.In the case of the commercial substrates, Ps 60 :Pb 30 :P 10 had significantly greater values of total porosity, easily available water, water buffering capacity, and difficult available water, and a lower value of air space than A 70 :Ps 24 :P 6 .To evaluate the results received from the water retention curves and both the pore size and particle size distributions, Table 4 was created.In Table 4, the terms introduced by De Boodt and Verdonck [24] are used to characterize the water retention capacity and aeration of substrates in a range of water tensions between 0 (saturation) and −15,000 cm (permanent wilting point).Furthermore, in the same table, the bulk density values of all the substrates studied are shown.In accordance to Table 4, the total porosity, easily available water, water buffering capacity, difficult available water, and nonavailable water were decreased in peat-perlite and coir-perlite mixtures compared to pure peat and coir, respectively.These differences were significant mainly in the case of 50% perlite percentage.On the other hand, as anticipated, substrate air space increased along with increasing the percentage of perlite in mixtures.In the case of the commercial substrates, Ps60:Pb30:P10 had significantly greater values of total porosity, easily available water, water buffering capacity, and difficult available water, and a lower value of air space than A70:Ps24:P6.
In Table 4, the corresponding values of "ideal substrates" as identified by De Boodt and Verdonck [24], and which have been implemented and used commercially at a large scale [31][32][33], are indicated.According to the concept of "ideal substrates", water retention curves below this zone have rapid water release and low water holding capacity, while those above it have less air space and greater water holding capacity.The only substrate to meet this zone for the full range of water tension is A70:Ps24:P6, as represented in Figure 4, followed by the substrates C75:P25 and Ps50:P50.The rest of the substrates have a remarkable portion of their water retention curve falling either below (e.g., C50:P50) or above this zone (e.g., Ps, Ps75:P25, C, C50:P50, Ps60:Pb30:P10).In Table 4, the corresponding values of "ideal substrates" as identified by De Boodt and Verdonck [24], and which have been implemented and used commercially at a large scale [31][32][33], are indicated.According to the concept of "ideal substrates", water retention curves below this zone have rapid water release and low water holding capacity, while those above it have less air space and greater water holding capacity.The only substrate to meet this zone for the full range of water tension is A 70 :Ps 24 : P6 , as represented in Figure 4, followed by the substrates C 75 :P 25 and Ps 50 :P 50 .The rest of the substrates have a remarkable portion of their water retention curve falling either below (e.g., C 50 :P 50 ) or above this zone (e.g., Ps, Ps 75 :P 25 , C, C 50 :P 50 , Ps 60 :Pb 30 :P 10 ).Water 2018, x FOR PEER REVIEW 9 of 15

Figure 4.
Comparative presentation of measured water retention curves for the substrates and the "ideal substrates zone" identified by De Boodt and Verdonck [24].
In Table 5, the measured values of water content and hydraulic conductivity at saturation (H = 0 cm), as well as the predicted ones by the Mualem-van Genuchten model at water pressure heads (H = −10, −50, −100 cm) within the range of easily available water and water buffering capacity, as introduced by De Boodt and Verdonck, are presented.As shown, a sharp decrease of the Comparative presentation of measured water retention curves for the substrates and the "ideal substrates zone" identified by De Boodt and Verdonck [24].
Water 2018, 10, 722 9 of 15 In Table 5, the measured values of water content and hydraulic conductivity at saturation (H = 0 cm), as well as the predicted ones by the Mualem-van Genuchten model at water pressure heads (H = −10, −50, −100 cm) within the range of easily available water and water buffering capacity, as introduced by De Boodt and Verdonck, are presented.As shown, a sharp decrease of the unsaturated hydraulic conductivity was observed within this range of water pressure heads (0 to −100 cm).Similar results have also been reported by other researchers (e.g., [20,44,45]).Comparing the unsaturated hydraulic conductivity values of the substrates A 70 :Ps 24 : P6 , C 75 :P 25 , and Ps 50 :P 50 characterized as "ideal" (Figure 4), it is apparent that remarkable differences exist among them (Table 5).During plant growth in this study, the pressure heads varied from −10 to −50 cm between two successive irrigations (range of easily available water).For this range of pressure heads, the hydraulic conductivity of A 70 :Ps 24 :P 6 is decreased by approximately three orders of magnitude, whereas in the rest of the substrates, it is decreased by approximately two orders of magnitude.Specifically, K values ranged from 5.5 × 10 −1 to 2.4 × 10 −4 cm•min −1 , with the smallest ones observed for the substrates C 50 :P 50 and A 70 :Ps 24 :P 6 (Table 5).
The knowledge of both K(H) and θ(H), is of vital importance mainly in greenhouse cultures and would contribute to alleviating water stress conditions and improving the quality of substrates.Nevertheless, we should bear in mind that in some cases, the predicted K(H) values using the water retention curve data and the saturated hydraulic conductivity may deviate significantly from the actual K(H) values [46][47][48][49][50].
In general, during plant growth, the O 2 concentration of the substrates studied was greater than 20.6%.However, there were no significant differences among the substrates nor along the depth of the substrates (data not shown).

Plant Growth
Plants grown in Ps 60 :Pb 30 :P 10 had significantly greater both shoot dry weight and percent growth increase than the plants grown in all the other substrates, with the exception of A 70 :Ps 24 :P 6 in the case of percent growth increase (Table 6).Plants grown in pure coir (C) had the lowest percent growth increase.With regard to plant growth in peat-based (Ps, Ps 50 :P 50 , Ps 75 :P 25 ) and coir-based (C, C 50 :P 50 , C 75 :P 25 ) substrates, no significant differences were noted amongst them.The analysis of both macro-and micronutrients of plant tissues are shown in Table 7.The comparison between peat and its mixtures with perlite (Ps, Ps 50 :P 50 , Ps 75 :P 2 5) as well as coir and its mixtures with perlite (C, C 50 :P 50 , C 75 :P 25 ) showed that there are no significant differences in the macroand micronutrients measured, except in the case of K and Na for plants grown in coir-based substrates, due to the higher initial content of K and Na in coir [4].Plants grown in Ps 60 :Pb 30 :P 10 and A 70 :Ps 24 :P 6 had similar mineral concentrations with the abovementioned substrates for all nutrients, except for K and Mn, which were higher in plants grown in A 70 :Ps 24 :P 6 .Note that in all treatments, the same amount of fertilizer was added to the irrigation water during the experiment, and no apparent mineral nutrient deficiency or toxicity symptoms were observed on the plants grown in all the substrates.As presented in Table 2, the two commercial prefortified substrates (Ps 60 :Pb 30 :P 10 and A 70 :Ps 24 :P 6 ) had a significantly higher concentration of N, especially, than the other substrates used, giving an advantage at the early state of the cultivation.However, this was not realized at the plant tissue analysis (Table 7).According to Reuter and Robinson [51], the concentrations of the micronutrients Zn, Mn, and Cu are within adequate amounts.Analyses of Nelson et al. [52], which were performed on younger leaves at 5 cm width or greater size of a different Begonia variety to that of the current research, agree that only the Fe concentrations measured are low and the Mg concentrations are within adequate amounts for all the substrates studied.Moreover, the concentrations of N were low for all the substrates used and agree with the results given by Nelson et al. [52], obtained at the end of the cultivation.Amongst all the substrates studied, the observed differences in nutritional status were not so significant as to solely justify the differentiation in plant growth.Additionally, although the higher values of almost all nutrient concentrations were exhibited in plants grown in the A 70 :Ps 24 :P 6 substrate, which was also classed as a hydraulically ideal substrate, the percent growth increase had not significantly increased in relation to Ps 60 :Pb 30 :P 10 and the peat and coir mixtures.Instead, plants grown in Ps 60 :Pb 30 :P 10 that had similar nutrient concentrations with plants grown in A 70 :Ps 24 :P 6 and was classed as hydraulically nonideal, presented significant differences in both the shoot dry weight and percent growth increase with all the other substrates studied.
Therefore, it is apparent that the physical properties of the "ideal substrates" alone (Table 4) cannot constitute the only criteria for selecting growing substrates, as there are notable substrates, such as those assessed in this paper, that did not fall within the range of the physical properties for ideal substrates.It is worth noticing that the higher plant growth was achieved in the Ps 60 :Pb 30 :P 10 substrate, although it did not fall into the "ideal substrates" zone.Taking into account the hydraulic parameters of the substrates studied, the reduction of hydraulic conductivity values between two successive irrigations (at pressure heads from −10 to −50 cm) did not appear to be a limiting factor in plant growth for any substrate, "ideal" or not.In particular, K values ranged between 5.5 × 10 −1 and 2.4 × 10 −4 cm•min −1 (Table 5), or in other words, between 7920 and 3.46 mm•d −1 , and daily evapotranspiration values did not exceed 1.8 mm•d −1 , as has been determined in our previous study [14].In any case, the knowledge of K(H) is essential for the substrate evaluation and proper irrigation management.So, the "nonideal" substrate Ps 60 :Pb 30 :P 10 resulted in the higher plant growth because it had an adequate water flow rate (due to proper irrigation management).Additionally, Ps 60 :Pb 30 :P 10 had both a growth advantage, because it was charged with a higher N concentration at the first stage of cultivation, and a high oxygen diffusion rate, as has been determined in our previous study [14].
As other researchers (e.g., [53][54][55][56]) have concluded with similar results (i.e., plants grown in substrates which did not fall in the "ideal substrates" zone showed satisfactory growth), it should be noted that these physical parameters are static and should be used only as references, because the only criterion for defining them is negative water pressure head.The important role of the dynamic hydraulic properties of the substrates such as gas diffusivity, saturated and unsaturated hydraulic conductivity, their chemical composition, and the growth particularities of different plants have not been considered.
Overall, it can be stated that because the parameters of the substrates are static, they represent steady-state conditions.However, during plant growth in pots or in the field, a dynamic state is formed that constantly changes over time and space.Therefore, the method, timing, and amount of irrigation water beyond the values of the physical properties of the "ideal substrates" (Table 2) play a significant role in defining a substrate as "ideal" or not.

Conclusions
The determination of physical-hydraulic properties (e.g., water retention curves, hydraulic conductivity) of the eight different substrates examined, commercial or not, used widely in floriculture, is essential to proper irrigation management.All substrates studied presented a high percentage of total porosity, available water capacity, and saturated hydraulic conductivity.The estimation of unsaturated hydraulic conductivity, K, using the Mualem-van Genuchten model showed a sharp decrease of K values in a range of water pressure heads from 0 to −50 cm.The presence of perlite in the mixes contributed to reducing the water retention ability and increasing the percentage of large pores.Knowing the particle and pore size distribution helps to create or correct the structure of substrates that are product mixtures.Also, the comparison of the substrates studied with the "ideal substrates" of De Boodt and Verdonck [24] showed that porosity alone cannot constitute a criterion for classifying substrates as "ideal" or not.Plants grown in the Ps 60 :Pb 30 :P 10 substrate, classified as

Figure 2 .
Figure 2. Pore size distribution of the substrates.

Figure 2 .
Figure 2. Pore size distribution of the substrates.

Figure 3 .
Figure 3. Particle size distribution of the substrates.Values are the means of three replications (n = 3).

Figure 3 .
Figure 3. Particle size distribution of the substrates.Values are the means of three replications (n = 3).

6 Figure 4 .
Figure 4.Comparative presentation of measured water retention curves for the substrates and the "ideal substrates zone" identified by De Boodt and Verdonck[24].

Table 1 .
Composition on a volume basis of peat-and coir-based substrates used.

Table 2 .
Initial chemical characteristics of substrates. 1

Table 3 .
The Mualem-van Genuchten model fitting parameters α, n, θr and measured values of θs for the substrates.

Table 3 .
The Mualem-van Genuchten model fitting parameters α, n, θ r and measured values of θ s for the substrates.

Table 4 .
Physical properties of the substrates.Values are the means of three replications (n = 3).

Table 5 .
[24]ured values of water content and hydraulic conductivity at saturation (H = 0 cm) and those predicted by the Mualem-van Genuchten model at water pressure heads within the range of easily available water and water buffering capacity (H = −10, −50, −100 cm), as introduced by De Boodt and Verdonck[24], for the substrates.

Table 6 .
Shoot and root dry weights at the end of the experiment and percent growth increase (%) of begonia plants throughout the duration of the experiment as affected by the substrate type.Columns not followed by the same letter are significant (Tukey-Kramer, at p = 0.05; n = 10).

Table 7 .
Concentration of various macro-and micronutrients in plant tissues of Begonia xelatior, "The President", at the end of the experiment as affected by the substrate type.