Availability of Three Phosphorus Fertilizers to Corn Grown in Limed Acid-Producing Mine Tailings

Substrate properties as affected by liming and P source are presented in Table 1. The pH was the most affected. The smallest yield was obtained at the lowest lime application rate (Ca₁). The ANOVA of treatment effects on top yield indicated significant main effects and a significant lime x P source interaction (Table 2). The lime effect was quadratic and the P rate effect was linear. The effect of P source on corn growth is presented in Table 3. In all CaCO₃ treatments, the significantly highest yield was obtained with plot receiving PSC. In general, lime application and P source stimulated corn growth.

The maximum yield is obtained at intermediate liming (Table 1). There was a highly significant correlation (P<0.001) between P uptake by corn and both tailing pH (r=-0.786) and oxalate-extractable Fe (r=0.762). In general, heavy metal cations (Zn, Cu, Fe and Mn) in corn tissues decreased with lime rate (data not shown).

The Mehlich 3 solution used as a soil test for available-plant P level in Quebec extracted higher amounts of tailing P (P_M3) than the acid ammonium oxalate method (P_ox) (Table 1). The amount of P_M3 varied from 32.7 to 49.5% of total P with an average of 42%. Inclusion of independent variables that met the 0.500 significance level for entry into multiple linear regression, e.g. P_M3, P_M3/total P ratio, and Mehlich-3 extractable Ca and Mg increased the R² value to 63.9%.
 

The evolution of P extracted by EDTA in some tailing samples is presented in Figure 1. In general, the amounts of extractable P were higher in BF-treated samples than in PSC-or MKP-treated tailings. The results of the evolution of P-EDTA extracted in treated tailings samples after 1 and 14 days are show in Table 4.

In another series of experiments, the amounts of P extracted by DTPA-TEA-CaCl₂ (DTPA) increased nonlinearly with extraction temperature (Figure 2), indicating that P desorption by DTPA was an endothermic reaction.

PSC: commercial peat-shrimp wastes compost; BF: commercial bone flour; MKP: reagent-grade KH₂PO₄
 

The greenhouse study shows that lime and P applications stimulated the growth of corn tops and roots. Maximum yield was obtained at intermediate rates. The effect of liming was more pronounced than P source, as indicated by the highly significant positive correlation between corn yield and tailing pH. Many workers [1,10,11] showed the benefit of liming acid mineral soil or acid mine tailing. The effects of low pH on plant growth could also produce nutritional imbalances [5] and influence the solubility of P compounds. Phosphates may be fixed into less soluble states as iron and aluminium phosphates, especially at low pH [5]. After heavy liming, soluble phosphates could become less available to plants due to chemical precipitation as calcium phosphates on carbonate surfaces [5]. As a result, the interaction between P source and lime rate is related to the influence of pH on P availability. Non-significant correlations between corn yield and total P or P_ox indicated that only a small proportion of the P fixed on metal-oxide surfaces accounted for the variation in corn yield. Top yield was significantly higher in PSC-treated tailing than in MKP- or BF-treated tailing samples (Table 3). Hountin et al. [12] also found that application of the same commercial PSC with lime to an acid loamy sand soil maintained high yield of barley and soil fertility. Crop response to compost could be attributable to supplemental nutrients supplied by the compost.
 

Alexander and Robertson [13] suggested that EDTA-extractable P derives mainly from Fe and Al compounds present in soils. Hence, P desorption by EDTA provides an index for the release of readily-extractable P in sulfide tailings rich in Fe compounds. The graphs of EDTA-P released from the sulfide tailing as a function of extraction time follow a parabolic shape (Figure 1). Thus, the recovery of inorganic P was time-dependent, as also found in many reports on the effect of extraction time on P desorption in mineral soils. As illustrated in Figure 1, the higher amount of extracted P was observed with BF, due possibly to mineralisable organic P compounds.

For mineral soils, Beaton et al. [14] found that the rate of dissolution of pellets containing 15 mg of P prepared from water-soluble P fertilizers increased markedly as extraction temperature was increased from 5° to 35° C. Numerous reports indicated that the optimum growth conditions for common microorganism species in the sulfide tailings such as Thiobacillus ferrooxidans is 30° C [15]. In general, the release of DTPA-extractable P in relation to temperature from tailing samples increased in the following order: MKP<BF<PSC (Figure 2). The activity of microorganisms involved in the decomposition of organic P into inorganic P or in the oxidation of iron sulfide into P-fixing iron hydroxide could be stimulated as temperature rose from 4° to 30° C. Despite those opposite reactions, the amount of DTPA-extracted P was highest in the temperature range of 20° to 30° C over all treatments.

In order to be useful, a desorption kinetic model must fit the data and yield parameters related to plant response. The kinetics of P desorption from agricultural soils was described by various kinetic models [16] such as:

First-order reaction:ln C=ln C₀+kt(1)

Second-order:1/C=1/C₀-kt(2)

Elovich-type equation: ln C=a+blnt(3)

Two-constant rate equation: ln C=lnk+blnt(4)

Parabolic diffusion law: C=C₀+kt^0.5(5)

Modified first-order desorption model: ln C=ln C₀+kt^0.5 (6)

Where C is desorbed P concentration in mg/L (same for all models); t is extraction time in hours (same for all models); C₀, a, b and k are empirical constants estimated by the least squares method. The R² and the standard error of estimate (SE) were used as criteria for comparing the equations with respect to the goodness of fit [16].

The modified first-order equation (6) provided the best fit, followed by the two-constant rate model (4) (Table 5). The slope varied among treatments due to variations in surface reactivity for various P compounds in cultivated tailing samples. Sharpley [17] related Elovich P desorption parameters to extractable Al and CaCO₃ in acidic and calcareous soils, respectively. The k values of the modified first-order model (6) was correlated to P_M3 (r=0.394, P<0.05), indicating that k values could depend on labile P levels. The k values were negatively correlated to top yield (r=-0.336, P<0.05) and root yield (r=-0.326, P<0.05). On the other hand, C₀ values were positively correlated with dry matter of corn tops (r=0.385, P<0.05) and root yield (r=0.460, P <0.05). Hence, corn yield may be predicted from P desorption measurements in acidic mine tailings.