Abstract
In current horticultural practice, potential acidity or basicity of fertilizers is estimated using Pierre's method (PM) expressed in calcium carbonate equivalents (CCE) per unit weight of fertilizer. PM was developed using mineral field soil systems and may be inaccurate for quantifying fertilizer acidity in containerized plant production given the widespread use of soilless substrates and fertigation. The PM-predicted acidity of an ammonium-based fertilizer was compared against experimental data obtained when ‘Ringo’ geraniums [Pelargonium ×hortorum (Bailey. L.H.)] and ‘Super Elfin’ impatiens [Impatiens wallerana (Hook. F.)] were grown in 70% peat:30% perlite (v:v) limed with either hydrated limestone only (HL) or a combination of carbonate and hydrated limestone (CHL). Plants in 10-cm-diameter (0.35 L) containers were top-irrigated with a total of 2.0 L over 6 weeks using a 15.2N–1.9P–12.6K fertilizer [100% of nitrogen (N) as NH4-N] applied with each irrigation at 100 mg N/L without leaching. According to PM, 61.8 meq of fertilizer acidity was applied per liter of substrate. During the experiment, the pH of the substrate decreased from 7.05 to 4.41 for the HL substrate and from 7.14 to 5.13 for the CHL substrate. A corresponding drop in substrate-pH was observed when 37.1 (HL) or 43.3 (CHL) meq of CCE from 0.5 N HCl was applied per liter of substrate in a laboratory titration of the same substrates without plants. Gasometric analysis of residual carbonate at Day 0 and at the end of the experiment quantified change in CHL substrate alkalinity with time, resulting in an estimated 30.7 meq of neutralized alkalinity. Using an electroneutrality approach that assumed anion uptake (NO3–, P2O5–) was basic, and cations (NH4+, K+) were potentially acidic, nutrient analysis of the substrate at the beginning and end of the experiment estimated that an average 48.5 meq of acidity was contributed by the fertilizer. Experimentally measured acidity values were 13.1 to 31.1 meq·L−1 of substrate lower for HL and CHL than those expected from PM, suggesting PM overestimated the amount of fertilizer acidity applied to the substrate. These results support the need for an alternative method to predict fertilizer acidity for plant production in soilless substrates.
Fertilizer type and concentration, plant species, water alkalinity, substrate components, and lime amendment dynamically affect substrate-pH during plant production. Fertilizer effect on substrate-pH occurs primarily through plant and microbial uptake of nutrients and the subsequent release of ions and/or root exudates along with other microbially induced processes such as nitrification (Marschner, 1995). Using nitrogen as an example, uptake of NO3-N can result in the release of OH– and HCO3– or coabsorption of a cation, whereas H+ ions are released with uptake of NH4-N. The microbial conversion of NH4-N to NO3-N (nitrification) also produces H+.
Pierre's method [“PM,” Pierre (1933)] is the standard used in the U.S. fertilizer industry to quantify the potential of a fertilizer to change soil acidity or alkalinity. The reported PM value for an acidic fertilizer, in units of calcium carbonate equivalents (CCE) of acidity per unit weight of fertilizer, refers to the CCE required to neutralize the acidity resulting from application of the fertilizer (Pierre, 1933). For basic reaction fertilizers, the CCE of basicity represents the basic residue in CCE left in the soil after application of the fertilizer.
Pierre's original methodology was modified and adopted by the Association of Official Agricultural Chemists (AOAC) and is recognized as the official method to estimate the equivalent acidity/basicity of complete fertilizers (Tisdale and Nelson, 1966). The final official procedure was published by AOAC in 1970 with a minor revision in 1999 (AOAC, 1970, 1999). In brief, the fertilizer mixture is ground and a 0.25- to 1.0-g dried sample is mixed with a sodium carbonate–sucrose solution and ashed in a furnace (at 575 to 600 °C) to remove carbon and nitrogen compounds. After cooling, HCl is added for lime digestion (AOAC, 1970; Hignett, 1985). Once digested, the sample is titrated to pH 4.3 using 0.5 N NaOH and compared with a similar titration of a blank sample of sodium carbonate–sucrose solution only (AOAC, 1970). The titrated difference between the blank sample and the fertilizer sample is the acidity or basicity of the fertilizer other than that contributed by nitrogen (Hignett, 1985). Table 1 shows the equivalent acidity or basicity factors for individual salt elements (Hignett, 1985), in which a positive value indicates the fertilizer sample is basic (non-acid-forming) and a negative value indicates the sample is acidic (Horat, 1939). For elements other than nitrogen or phosphorus, therefore, the acidity or basicity of a fertilizer salt is a direct measurement of its inorganic acid/base reaction in this titration and is independent of the soil or plant system.
Equivalent acidity (negative) or basicity (positive) values in units of kg of CaCO3 per kilogram of element using Pierre's method (Hignett, 1985).
Factors for nitrogen and phosphorus were based on empirical field studies by Pierre in the 1920s and 1930s that measured soil pH changes after fertilizer application, primarily using Cecil clay loam with an initial pH of 6 (Hignett, 1985). Pierre assumed that: 1) the acidifying effect is caused by all the sulfur and chlorine, one-third of the phosphorus, and one-half of the nitrogen contained in the fertilizer; 2) calcium, magnesium, sodium, and potassium are base-forming elements; and 3) ammonium nitrogen is completely nitrified; nitrate nitrogen combined with bases such as sodium or calcium will have a net basic effect (Pierre, 1933; Tisdale et al., 1999; Tisdale and Nelson, 1966).
These assumptions merit critical evaluation when applied to use of water-soluble fertilizers in soilless container crop production. Pierre's method assumes nitrogen in fertilizers is provided as nitrate or is converted to nitrate through nitrification (Pierre, 1933). Rapid nitrification does indeed occur in container substrate above pH 5.5, which is the typical growing range for most container crops (Argo and Biernbaum, 1997; Lang and Elliott, 1991). However, PM does not take into consideration acidification resulting from plant NH4+ uptake or acidity from nitrification, and therefore it underestimates the lime-equivalent values required to neutralize the acid formed by the various nutrient salts in the fertilizer (Tisdale et al., 1999; Tisdale and Nelson, 1966).
Andrews (1954) maintained that all ammonium converted to nitrate creates acidity equivalent to 3.57 kg CCE per kilogram of nitrogen (N) [calculated as mw of CaCO3/(2 × mw of N) to correct for valence differences]. Andrews also argued that 3.57 kg CCE base is leached from the soil along with every kg of NO3-N leached, and his analysis included multiple years of field trials and re-evaluation of Pierre's data. Andrews' assumptions result in differing acidity and basicity potentials for fertilizer salts than those of Pierre; he considered anhydrous ammonia to be twice as acidic as Pierre's estimate. Andrews' calculations indicate calcium nitrate is slightly acidic and potassium nitrate neutral, whereas Pierre's calculations indicate that they are both basic (Andrews, 1954; Tisdale and Nelson, 1966).
Pierre considered that NH4+ uptake was possible but the amount was insignificant under normal agricultural conditions (Pierre, 1933). However, ammonium uptake is energetically favored over nitrate uptake when both N forms are supplied (Engels and Marschner, 1995; von Wiren et al., 2001). Plant species is an important factor to consider in N preference. For instance, plant species adapted to soils that are acidic or with low redox potential tend to have a greater rate of uptake of ammonium, whereas plant species adapted to calcareous soils have greater uptake of nitrate–nitrogen (Marschner, 1995, p. 247). Ammonium follows an electrochemical gradient from apoplasm to symplasm without the requirement of metabolic energy. The exact path of NH4+ uptake is still unclear. It is believed that ammonium is taken up through an NH4+ uniport, H+-NH4+ symporter, diffusion, or ion channels through the plasma membrane (Engels and Marschner, 1995; von Wiren et al., 2001). In contrast to NH4+, NO3– requires active transport from the apoplasm into the symplasm in association with H+ fluxes through H+-NO3– symporters and H+-ATPase stimulation that maintains membrane electrical potential (Engels and Marschner, 1995). It is estimated that the absorption of 1 mol of NO3– consumes 1 to 3 mol of ATP (Touraine et al., 2001). For that 1 mol of NO3– to be used by the plant, it must first be reduced to NH4+ in a process requiring 347 kJ of energy (Lewis, 1986). Low NO3– concentration in the soil solution, low soil pH, and low soil temperature have been shown to favor NH4+ uptake over NO3– (Engels and Marschner, 1995).
The relative balance among nitrification, ammonium and nitrate uptake, and nutrient leaching, which in turn affect substrate acidity, is affected by factors such as species, irrigation method, soil microbial activity rate, plant evapotranspiration rate, and relative size of the plant and substrate volume. Pierre's assumption that only half of the N is actually acid-forming is therefore an approximation based on the experimental conditions under which his model was calibrated.
Although on a strict inorganic chemistry basis using PM, cations such as calcium, magnesium, sodium, and potassium have a basic residue (Table 1), these cations can also have acidic behavior in a soil/crop system. Displacement of protons on exchange sites by cations causes a decrease in substrate-pH (Rippy and Nelson, 2005). Cation uptake by plants is an acidic or neutral process accompanied by anion uptake, base uptake, or in association with a proton pump (Marschner, 1995). Similarly, anion uptake by plants is a basic reaction despite the acidic factors for sulfur and chlorine in Table 1. Overall, the substrate-pH effect of a fertilizer salt is a balance between inorganic chemistry and the dynamic plant/microbial/soil system.
The objectives of this study were 1) to evaluate protocols to quantify fertilizer acidity or basicity in soilless substrates; and 2) to use these protocols to compare the reported fertilizer acidity from PM with the experimentally measured acidity when plants were grown with an ammonium-based, water-soluble fertilizer in a peat/perlite substrate.
Materials and Methods
Greenhouse substrate experiments
In Spring 2007, a 70%:30% (v:v) peat:perlite substrate was mixed with dolomitic hydrated limestone [97% Ca(OH)2· MgO, 92% of which passed through a 45-μm screen; National Lime and Stone, Findlay, OH; reported acid-neutralizing value of 161% CCE] at a rate of 2.8 kg·m−3 to raise substrate-pH to 7.04. One week later, the substrate was divided with half remaining amended with hydrated lime only (HL) and half amended with additional dolomitic carbonate limestone (CHL) [CaMg(CO3)2, Lawn & Garden limestone™; Oldcastle Stone Products®, Thomasville, PA] at a rate of 2.22 kg·m−3. The additional dolomitic carbonate lime in the CHL substrate was added to provide residual carbonate and substrate-pH buffering throughout the experiment. The peat source used in the research substrates was Canadian Sphagnum peat (Sun Gro Horticulture, Vancouver, Canada) with long fibers and little dust (von Post scale 1 to 2; Puustjarvi and Robertson, 1975). The HL and CHL substrates were placed in 10-cm, 350-mL containers. Container capacity was 255 mL or 73.0% of container volume. Plugs of ‘Ringo Scarlet’ and ‘Ringo Coral’ geraniums [Pelargonium ×hortorum (Bailey. L.H.)] and ‘Super Elfin White’ and ‘Super Elfin Salmon’ impatiens [Impatiens wallerana (Hook. F.)] were transplanted into the containers. The geraniums had been seed-propagated in 105-celled plug trays and the impatiens in 162-celled plug trays.
The containers received a total of 20 irrigations in 6 weeks applied to the substrate surface. They were irrigated each time with 100 mL of an acidic fertilizer solution consisting of 100N–12.8P–83.3K–107.8S–75.8Cl mg·L−1 (100% total N as NH4-N) and an Fe-EDDHA-chelate micronutrient blend of 1Fe–0.25B–0.25Cu–0.5Mn–0.1Mo–0.5Zn mg·L−1 (Greencare Fertilizers, Inc., Kankakee, IL) in deionized water. Saucers were placed under each container to allow reabsorption of any leachate. The plants were grown in a polycarbonate-covered greenhouse for 6 weeks with average temperatures of 22.4 ± 4.3 °C and average photosynthetically active radiation light of 10.9 ± 3.0 mol·m−2·d−1 (daily mean ± sd).
The greenhouse experiment was a randomized complete block design with a factorial of two substrates (HL and CHL) × two species (geraniums and impatiens) × two cultivars (for each species). There were three blocks (greenhouse benches). Each block included one replicate (or container) of all substrate, species, and cultivar combinations for a total of three replicates (containers) per variable measured each sample date. Data were collected for all measurement variables over the course of 6 weeks from destructively sampled containers (i.e., there were no repeated measurements on the same plants). Substrate-pH and electrical conductivity (EC) were measured weekly (Weeks 0 to 6) using saturated media extraction (SME) (Rippy and Nelson, 2005; Warncke, 1986). SME solution samples were also analyzed for nutrient content by Quality Analytical Laboratories (Panama City, FL) using inductively coupled plasma (for phosphorus and potassium) and Lachat® (for NH4+ and NO3–). Changes in residual carbonate concentration were measured biweekly (Weeks 0, 2, 4, 6) using a gasometric analysis method (Huang et al., 2007). The acid drench experiment was conducted in a laboratory using the same HL and CHL substrates used in the greenhouse experiment. There were six drench rates (treatments) with four replicates each. Substrate-pH data were collected immediately before and 7 d after drenching for regression analysis. All data were analyzed using SAS PROC GLM (SAS Institute, 2001) and means were separated using Tukey's honestly significant difference (α = 0.05).
Methods for measuring fertilizer acidity or alkalinity
Pierre's method.
The acidity of each salt that comprised the ammonium fertilizer was computed and summed using the factors shown in Table 1. The fertilizer blend consisted of 7.26% monoammonium phosphate (NH4H2PO4), 68.16% ammonium sulfate [(NH4)2SO4], and 24.56% potassium chloride (KCl) by weight. To calculate the acidity of the individual salts, the following equations were used:
-
(A)
-
-
(B)
-
(A) × percent mass of fertilizer salt in fertilizer blend = kg of CaCO3 acidity or basicity per metric tonne of nutrient in fertilizer blend.
-
NH4H2PO4:
-
The negative CCE values indicate that the elements of N and phosphorus in NH4H2PO4 are acidic. Total calculated acidity from 1000 kg (metric tonne) of NH4H2PO4 was equivalent to 651.47 kg CaCO3 (217.89 kg + 433.58 kg = 651.47 kg), which contributed 47.28 kg CaCO3 acidity per tonne of 15.2N–1.9P–12.6K fertilizer (7.26% × 651.47 kg of CaCO3 = 47.28 kg).
-
(NH4)2SO4:
-
The negative CCE values indicate that the elements of N and sulfur in (NH4)2SO4 are acidic. Total calculated acidity was equivalent to 1136.61 kg CaCO3 per tonne of (NH4)2SO4 (379.56 kg + 757.25 kg = 1136.61 kg), which accounted for 774.92 kg acidity per tonne of 15.2N–1.9P–12.6K fertilizer (68.16% × 1136.61 kg = 774.92 kg).
-
KCl:
-
The net effect of KCl is therefore ≈0 kg CaCO3 acidity according to PM assumptions.
Milliequivalents CaCO3 (CCE) acidity applied per liter of substrate during each irrigation were calculated as:
-
-
(B)100 mg·L−1 N applied × 0.1 L of fertilizer applied = 10.0 mg N applied per irrigation.
-
-
-
(E) (D) × 2.86 containers per L of substrate × 20 meq per g CaCO3 = 3.09 meq CCE
Because 20 applications (2.0 L) of the 100 mg·L−1 N 15.2N–1.9P–12.68K ammonium fertilizer solution was made, a total of 61.8 meq of acidity was applied per liter of substrate during the 6-week experiment. Of this total acidity (N, phosphorus, sulfur, and chlorine = 2458.76 kg CCE), the relative percentage contribution to acidity estimated using PM for each nutrient was 24% N, 18% phosphorus, 31% sulfur, and 27% chlorine, whereas all of the basicity was contributed by potassium.
Gasometric method.
Acid drench method.
Water alkalinity is normally calculated by titrating with mineral acid to a target pH of 4.5 (Bailey, 1996). The same principle was adapted in an acid drench method to measure alkalinity in the substrate (Huang et al., 2009). Samples of 350 mL of the HL and CHL substrates were placed in plastic bags and kept at container capacity for 7 d (without plants), and substrate-pH was measured. Hydrochloric acid (0.5 N) was then added at volumes of 0, 14, 28, 42, 56, and 70 mL, resulting in 0, 20, 40, 60, 80, or 100 meq of acid per L of substrate, respectively. Substrate-pH was measured 7 d after acid application with four replicates. The change in substrate-pH (ΔpH) was analyzed using quadratic regression to generate ΔpH–milliequivalent relationships for each substrate. The ΔpH–milliequivalent regression curves were then used to estimate the equivalent milliequivalent of acidity from fertilizer in the greenhouse trial with geranium and impatiens using the measured change in substrate-pH from the beginning to the end of the crop.
Charge balance method.
One approach to quantify acidity or basicity of a fertilizer is based on the principle of electroneutrality, whereby plant uptake of positively or negatively charged nutrient ions is counterbalanced with the exchange between plant and substrate of protons (H+), hydroxyl (OH–), carbonate (CO32–), or other root exudates (Marschner, 1995). In this context, the milliequivalent of cations taken up by plant roots causes an equal milliequivalent of acidity to be added to the substrate, whereas anion uptake causes an equivalent basic reaction.
When determining the net milliequivalent of acidity created from application and uptake of fertilizer using charge balance, the acidity estimate of total nutrients applied must be corrected for those applied nutrients that are not taken up by the plant (accumulating in the substrate) or that change form (such as the NO3– product of nitrification).
For the charge balance method used in this study, acidity estimates were based only on N, phosphorus, and potassium because these three macronutrients comprised the bulk of the experimental fertilizer, are among the elements taken up in the greatest quantity by plants, and thus would be expected to have the greatest influence on acidity (Epstein and Bloom, 2005). Calcium (Ca) and magnesium (Mg) were not included in the acidity estimates because for each milliequivalent of H+ released by a milliequivalent of Ca and Mg, a milliequivalent of acid neutralizing HCO3– or CO3 was applied, thus neutralizing the Ca and Mg addition. Milliequivalents of acidity or basicity derived from the portion of nutrients applied but not taken up or nitrified can be determined using the change in nutrient concentration in the substrate. Substrate analysis has a key advantage over tissue analysis because total N uptake by plant tissue does not differentiate between nitrate and ammonium forms. These estimations include the assumption that no leaching, volatilization, or other loss of nutrients occurred. To calculate the net milliequivalent of acidity (or basicity) of fertilizer nutrients applied and taken up or nitrified (as determined by nutrient change in the substrate), the following series of calculations were made:
-
1. Calculate total milliequivalent of acidity/basicity of nutrient applied:
-
(b) (a) × L of fertilizer solution applied per container = mg applied per container.
-
(c) (b) × 2.86 containers·L−1 substrate = mg applied/L substrate.
-
-
(e) (d) × valence of nutrient salt = milliequivalent of acid (or base)/L substrate.
-
2. Subtract milliequivalent of potential acidity/basicity of nutrient not taken up (or nitrified) based on measured substrate nutrient concentration. There was no nitrate–nitrogen applied in the fertilizer, so all nitrate measured in the substrate was assumed to be equivalent to the amount of ammonium nitrogen nitrified and not taken up.
-
(a) mg·L−1 nutrient(final) – mg·L−1 nutrient(initial) = mg·L−1 change in nutrient per container.
-
(b) (a) × 2.86 containers/L substrate = mg change in nutrient/L substrate.
-
-
(d) (c) × valence of nutrient ion = milliequivalent of potential acid (or base)/L substrate from nutrient not taken up or nitrified.
-
(e) milliequivalent of acid (base) applied – milliequivalent of acid (base) not taken up = milliequivalent of acid (base) from nutrient taken up.
Statistical analysis
Fertilizer acidity data estimated from all the methods were subject to analysis of variance analysis using SAS PROC GLM (SAS Institute, 2001). Means were separated using 95% confidence interval by substrate types (for gasometric and acid drench methods) or by species (for charge balance methods).
Results and Discussion
With respect to all data measured in the greenhouse experiment, cultivar within each of the two species did not affect any treatment comparisons. Data of cultivars within each species were therefore combined within species and further analyzed for species and/or substrate influences.
Substrate-pH and electrical conductivity
Substrate-pH did not differ between species with time (P = 0.1706). As expected, residual carbonate limestone in the CHL substrate resulted in less pH change (i.e., greater pH buffering) than the HL substrate (P = 0.0001) (Fig. 1). After 6 weeks, substrate-pH decreased by 2.01 units from the starting pH of 7.14 in the CHL substrate, compared with 2.64 units below the initial pH of 7.05 in the HL substrate. Although the additional carbonate limestone in the CHL substrate provided some buffering against decreasing pH, the pH of both the CHL and HL substrates dropped low enough to result in visual micronutrient toxicity symptoms in the geraniums, which commonly occurs below pH 6 (Argo and Fisher, 2002).
Substrate-pH change with time in the CHL substrate (containing hydrated and carbonate lime) and HL (hydrated lime only) in the greenhouse trial with data averaged across geranium and impatiens plants. Error bar represents pooled 95% confidence interval and symbols represent the mean of 12 replicates.
Citation: HortScience horts 45, 7; 10.21273/HORTSCI.45.7.1099
Substrate EC differed between species (P = 0.0011) but not substrate type (P = 0.2714). Substrate EC increased with time from an average 1.7 to 3.5 dS·m−1. It finished 0.5 dS·m−1 higher for geranium than impatiens. EC levels remained within acceptable to optimum values for container plants reported by Warncke (1995).
Substrate nitrogen–phosphorus–potassium
Substrate nutrient data differed between species for NO3, NH4, and phosphorus (P) (P ≤ 0.0003) but not potassium (K) (P = 0.3531). The residual and non-residual substrates differed only in NO3 concentration (P = 0.0047). There was also a species × substrate interaction for NO3 (P = 0.0250), but substrate nutrient data for both substrates were combined for each species in the charge balance method results and calculations shown in Table 2 for a simpler illustration of the method.
Fertilizer acidity calculated by the charge balance method which assumes cations (+) are acidic and anions (–) are basic.z
Pierre's method.
Using the previously shown calculations, 3.1 meq CCE of acidity per liter of substrate was introduced into the substrates with each fertigation event. During the course of 6 weeks, an estimated 61.8 meq CCE of acidity was contributed by the fertilizer per liter of substrate (Fig. 2A).
(A) Milliequivalents (meq) of calcium carbonate (CCE) of acidity from 15.2N–1.9P–12.6K (100% N as NH4-N) fertilizer applied per liter of substrate determined by Pierre's method (PM) and meq of CCE of carbonate neutralized with time as determined by change in residual carbonate using the gasometric method. (B) Milliequivalents of residual CCE in CHL (carbonate and hydrated lime) substrate measured with time using the gasometric method used to calculate the gasometric curve in A. With each fertigation, 100 mL of 100 mg·L−1 N ammonium fertilizer was applied 20 times per 10-cm-diameter container, which according to PM provided a total of 61.8 of CCE per liter of substrate (3.09 meq per irrigation). Error bar represents 95% confidence interval of each data point and symbols represent the mean of 12 replicates.
Citation: HortScience horts 45, 7; 10.21273/HORTSCI.45.7.1099
Gasometric method.
Because the gasometric method measures residual carbonate, only the CHL substrate (which included carbonate limestone) was analyzed. There were no statistically significant differences between species (P > 0.05) at any week, so data were combined within each measurement date. Residual calcium carbonate lime measured in the substrate decreased with time (Fig. 2B). Residual carbonate CCE decreased to 27.1 meq/L of substrate from the initial 57.8 meq CCE of carbonate measured. By difference, fertilizer applications were therefore estimated to contribute 30.7 meq CCE of acidity into the substrate (Fig. 2A).
Acid drench method.
Substrate-pH decreased with increasing milliequivalent of HCl applied. Substrate-pH decreased as much as 3.49 units in the CHL substrate and 4.32 units in the HL substrate 7 d after applying the HCl acid drench (Fig. 3). Using the acidity and pH response curves shown in Figure 3, the estimated acidity required for similar pH drops observed in the greenhouse experiment were 43 and 37 meq for the CHL (2.01 units) and HL (2.64 units) substrates, respectively.
Change in substrate-pH 7 d after a single application of 0.5 N HCl acid to CHL (containing hydrated and carbonate lime) and HL (hydrated lime only) substrates using the acid drench method. A quadratic curve was fit separately to each substrate type: Change in pH of CHL = 0.0002 meq2 – 0.0574 meq + 0.0821, R2 = 0.8907; change in pH of HL = 0.0004 meq2 – 0.0839meq – 0.0812, R2 = 0.9626). Each symbol represents a single replicate.
Citation: HortScience horts 45, 7; 10.21273/HORTSCI.45.7.1099
Charge balance method.
Species resulted in different N and P (but not K) concentrations in the substrate at Week 6. The acidity estimate using the charge balance method was therefore calculated separately by species in Table 2. There was 3.5 times more NH4-N (55.3 mg·L−1 SME) and nearly twice as much P (11.5 mg·L−1) remaining in the geranium substrate compared with the impatiens substrate (15.5 mg NH4-N/L and 6.5 mg P/L) (Table 2). There was also 10 times more NO3-N remaining in the geranium substrate (33.0 mg NH4-N/L) than in the impatiens substrate (3.1 mg NH4-N/L). There was no nitrate applied through fertigation, so the presence of nitrate in the substrate was assumed to be the result of nitrification. Nitrification has been shown to occur rapidly in container substrates above pH 5.5 (Argo and Biernbaum, 1997; Lang and Elliott, 1991). The lower concentration of nitrate in the impatiens substrate indicates there was either less nitrification or faster nitrate uptake compared with geraniums.
The net estimated acidity applied from N, P, and K fertilizer was calculated as 50.6 meq (NO3– = 0.0, NH4+ = 40.8, P = –2.3, K = 12.1 meq) with 40.8 meq (80.6%) from NH4-N alone (Table 2). The charge balance method used measured substrate nutrient data to account for nutrients that were applied but not taken up by the plant. Using data collected from the geranium substrate, N–P–K acidity was 48.2 meq CCE (NO3– = 1.3, NH4+ = 38.5, P = –2.1, K = 10.5 meq) with 80.1% from NH4-N uptake alone (Table 2). Using data collected from the impatiens substrate, N–P–K acidity was calculated at 48.7 meq CCE (NO3– = –0.3, NH4+ = 40.6, P = –2.3, K = 10.6 meq) with 83.4% from NH4-N uptake.
Comparisons between Pierre's method and other methods
Experimental estimates (summarized in Table 3) indicate that PM overpredicted the potential acidity of the ammonium-based fertilizer used in this experiment. Results of the gasometric, acid drench, and charge balance methods varied by 18 meq between each method. After 20 100-mL, 100-mg·L−1 N applications of ammonium (100% NH4-N) fertilizer, a total 61.8 meq of CCE (acid) per liter of substrate was applied during the course of the 6-week experiment using PM (Table 3). Using other methods, measured CCE acidity values ranged from 30.7 to 48.7 per L of substrate. Pierre's method, therefore, overestimated the acidity of the ammonium fertilizer by as much as 31.1 meq compared with other methods.
Comparison of milliequivalents (meq) of calcium carbonate equivalent (CCE) acidity per liter of substrate estimated using different fertilizer acidity methods when a 10-cm, 350-mL container received a total of 2 L of 15.2N–1.9P–2.68K fertilizer at an N rate of 100 mg·L−1.
Conclusions
Three experimental methods were compared with PM. The charge balance method illustrates how the valence of nutrient ions can be used to estimate acidity. However, a weakness of this method was in our calculation of changes in substrate nutrient concentrations during the experiment, because an SME test with deionized water does not measure total nutrients, but only nutrients in the substrate solution. A more complete analysis using the charge balance approach could be applied to other nutrients in addition to N, P, and K, and indeed that would represent a completely new model that would require further validation. Ions contributed by the substrate itself, Ca and Mg from the lime sources, micronutrients and other ions in the fertilizer, irrigation water quality, or other contaminant sources would have some influence on pH dynamics.
The concept of measuring change in substrate alkalinity was achieved using two approaches. The gasometric method requires a carbonate source (typically limestone) in the substrate and is therefore not appropriate for all substrates (Huang et al., 2007). In addition, once all available carbonate is neutralized, the alkalinity measured with the gasometric method will remain zero even if additional acid continues to be added by fertilizer. The acid drench method would be useful across a wider range of experimental conditions than the gasometric method, and the laboratory-generated pH response curve means that only pH measurement is required in subsequent plant experiments for a particular substrate.
We conclude that an alternative estimation of fertilizer acidity or basicity to PM could be developed that more closely represents plant, microbial, and soil chemical processes in soilless substrates with the potential to improve pH management. To develop such an equation, research and validation would be required. The charge balance–plant uptake and microbial components are particularly sensitive to N, which would therefore be a focus of future research and integration of the wealth of knowledge on cation–anion preference and the N cycle. An emphasis on N is also appropriate in terms of expected importance of N in contributing to pH effects of fertilizer. Epstein and Bloom (2005) indicate that on average, crop plants take up 1000 N atoms for every 250 K, 50 P, 30 sulfur, and three chlorine atoms. This ratio differs greatly from the relative importance of ions estimated using PM for the ammonium fertilizer in this study, whereby contribution to acidity was estimated in the proportions of 24% N, 18% P, 31% sulfur, and 27% chlorine, and 100% of basicity was estimated to result from K. We did not observe differences in pH response between Pelargonium [an iron-efficient species (Argo and Fisher, 2002)] and Impatiens. However, differences between species in their total nutrient uptake ratios, cation–anion uptake ratio, and iron efficiency may mean that either species-specific parameters would need to be developed for fertilizers or, more realistically, assumptions would be needed to average across multiple species.
There are limits to the accuracy in predicting the pH effect of a fertilizer based on its chemical formulation only, because multiple factors interact once a fertilizer is applied in plant production (for example, plant species, initial soil pH, residual lime, and so on). By necessity, a series of assumptions would be needed for any estimation. However, an equation could include 1) a charge balance–plant uptake component that considers nutrient ionic charge and valence and relative proportion of uptake; 2) a microbial component, which takes into account the N cycle process such as nitrification and urea hydrolysis; and 3) a pure chemical component that accounts for pH reaction of fertilizer and other salts (for example, bicarbonate) in solution. By basing the acidity or basicity on concentration of ions in a nutrient solution rather than a dry fertilizer weight basis, such an equation would more closely represent the pH dynamics of production in soilless media or with hydroponics. Working on a solution basis would also facilitate balancing alkalinity of an irrigation water source with the acidity of the fertilizer to develop an overall neutral pH response of the overall nutrient solution. However, the same equation could also be adapted for controlled-release, slow-release, and granular fertilizers applied as solids to soilless substrates.
Literature Cited
Andrews, W.B. 1954 The response of crops and soils to fertilizers and manures 2nd Ed William Byrd Press, Inc Richmond, VA
Argo, W.R. & Biernbaum, J.A. 1997 Lime, water source, and fertilizer nitrogen form affect medium pH and nitrogen accumulation and uptake HortScience 32 71 74
Argo, W.R. & Fisher, P.R. 2002 Understanding pH management for container-grown crops Meister Publishing Willoughby, OH
Association of Official Analytical Chemists (AOAC) 1970 Official methods of analysis 11th Ed AOAC International Washington, DC
Association of Official Analytical Chemists (AOAC) 1999 Official methods of analysis 16th Ed AOAC International Gaithersburg, MD
Bailey, D.A. 1996 Alkalinity, pH and acidification 69 91 Reed D.W. Water, media, and nutrition for greenhouse crops Ball Publishing Batavia, IL
Engels, C. & Marschner, H. 1995 Plant uptake and utilization of nitrogen 41 82 Bacon P.E. Nitrogen fertilization in the environment Marcel Dekker, Inc New York, NY
Epstein E. & Bloom A.J. 2005 Inorganic components of plants 41 66 Mineral nutrition of plants: Principles and perspectives 2nd Ed Sinauer Associates, Inc Sunderland, MA
Hignett T.P. 1985 Physical and chemical properties of fertilizers and methods for their determination 284 316 Fertilizer manual Martinus Nijhoff/Dr W. Junk Publishers Dordrecht, The Netherlands
Horat, L.E. 1939 Report on acid- and base-forming quality of fertilizers Journal of the AOAC 19 289 297
Huang, J. , Fisher, P.R. & Argo, W.R. 2007 A gasometric procedure to measure residual lime in container substrates HortScience 42 1685 1689
Huang, J. , Fisher, P.R. , Horner, W.E. & Argo, W.R. 2009 Limestone particle size and residual lime concentration affect pH buffering in container substrates J. Plant Nutr. 33 846 858
Lang, H.J. & Elliott, G.C. 1991 Influence of ammonium: Nitrate ratio and nitrogen concentration on nitrification activity in soilless potting media J. Amer. Soc. Hort. Sci. 116 642 645
Lewis, O.A.M. 1986 Plants and nitrogen Edward Arnold Baltimore, MD
Marschner, H. 1995 Mineral nutrition of higher plants 2nd Ed Academic Press San Diego, CA
Pierre, W.H. 1933 Determination of equivalent acidity and basicity of fertilizers Ind. Eng. Chem. 5 229 234
Puustjarvi, V. & Robertson, R.A. 1975 Physical and chemical properties 23 38 Robinson D.W. & Lamb J.G.D. Peat in horticulture Academic Press London, UK
Rippy, J.F.M. & Nelson, P.V. 2005 Soilless root substrate pH measurement technique for titration HortScience 40 201 204
SAS Institute 2001 SAS/STAT user's guide. Version 8.2 SAS Institute, Inc Cary, NC
Tisdale S.L. , Havlin J.L. , Beaton J.D. & Nelson W.L. 1999 Soil acidity and alkalinity 38 85 Soil fertility and fertilizers 6th Ed Prentice Hall Upper Saddle River, NJ
Tisdale, S.L. & Nelson, W.L. 1966 Soil fertility and fertilizers 2nd Ed Macmillan Company New York, NY
Touraine, B. , Daniel-Vedele, F. & Forde, B.G. 2001 Nitrate uptake and its regulation 1 36 Lea P.J. & Morot-Gaudry J.-F. Plant nitrogen Springer New York, NY
von Wiren, N. , Gojon, A. , Chaillou, S. & Raper, D. 2001 Mechanisms and regulation of ammonium uptake in higher plants 61 77 Lea P.J. & Morot-Gaudry J.-F. Plant nitrogen Springer New York, NY
Warncke, D.D. 1986 Analyzing greenhouse growth media by the saturation extraction method HortScience 21 223 225
Warncke, D.D. 1995 Recommended test procedures for greenhouse growth media Recommended soil testing procedures for the northeastern United States 2nd Ed Univ. of Delaware Agricultural Experiment Station Bulletin 493 76 83