UBC Theses and Dissertations
Soil testing for phosphorus availability to some conifers in British Columbia Curran, Michael Patrick
Two complementary investigations were conducted as a preliminary study to find an adequate method for estimating "available" phosphorus (P) in British Columbia forest soils. Eight soil materials provided a range of British Columbia forest soil properties, P forms and P levels. The 12 methods evaluated were H₂O-soluble P, Truog, modified Olsen (using polyacrylamide to remove coloured organic constituents), Bray P₁ and P₂, modified Bray P₁ and P₂ (longer extraction time), modified double-acid (0.05 N HCl + 0.025 N H₂SO₄, using polyacrylamide), NH₄OA[sub c] (ammonium acetate) at pH 4.8, modified (using polyacrylamide) NH₄HCO₃-DTPA (diethylene triamine pentaacetic acid), 0.01 _N HC1, and new-Mehlich. The first study investigated relationships among 12 soil test methods for estimating soil available P and 4 chemical P fractions. The 4 chemical P fractions were obtained by a modified Chang and Jackson procedure. The second study evaluated the test methods as indices of "tree-available P" through a greenhouse pot trial with the 8 soil materials x 2 P treatment levels (P added and control) x 3 tree species (Douglas-fir, western hemlock, and lodgepole pine). The first study revealed that most methods extract the greatest amount of P from forest floor samples (apart from a P fertilized soil) and the least from Podzolic B horizon samples. Similar to the literature, P levels obtained by a number of the methods are significantly correlated with each other. Modified Olsen values are significantly correlated with values from the greatest number of other methods. The 4 Bray methods, modified double-acid and new-Mehlich methods yielded the next largest group of significant correlations. Water-soluble P and 0.01 HCl values were seldom significantly correlated, and NH₄OA[sub c] (pH 4.8) values were not correlated, with other method test values. Contrary to agricultural soils literature, only one significant correlation existed between the soil test values and P forms: NH₄OA[sub c] (pH 4.8) extraction with the Ca-P fraction. Correlations between P fractions were also not as extensive, with only reductant-soluble P and Fe-P being significantly related for all 8 soils. Across the 5 "Podzolic" soils, Ca-P and Al-P were also significantly related (negatively). In the greenhouse study, seedling growth was best on the forest floor material, and worst on the calcareous soil material for lodgepole pine and Douglas-fir and on a Podzolic B horizon for western hemlock. All species displayed dramatic responses to P on some of the soils. In a hierarchical soil analysis screening (12 methods x 3 soils; then 3 candidate methods x 5 remaining soils) the new-Mehlich values were the most significantly correlated with foliar P concentration for all three species in the first stage. Many of the methods were significant for lodgepole pine. Modified Olsen and modified NH₄HCO₃-DTPA also appeared good for Douglas-fir, but these alkaline extractants were considered too cumbersome for routine laboratory anaysis of the study soils. The three chosen candidate methods were new-Mehlich; 0.01 N HCl, and Bray P₁. Evaluating these methods across various groups of the 8 soils for each species (i.e. without' the organic, calcareous, eluviated and P-fertilized soil materials in turn and in combinations) suggested that the new-Mehlich method may be best. Bray commonly did not correlate well with Douglas-fir foliar P. The 0.01 N HCl did not appear as good as it had in the preliminary screening, but did well across Podzolic horizons alone. All methods correlated poorly with western hemlock foliar P under certain conditions. The new-Mehlich method was evaluated for the individual treatment replicates, yielding correlations with foliar P which were consistent with prior results. Correlations for the organic and eluviated soils were highly significant for all three species. For individual soils, no significant correlations existed for all but two Podzolic soil materials. However, grouped Podzolic soils yielded strong correlations for the new-Mehlich method. The results reported here need field testing and it is recommended that the new-Mehlich and Bray P₁ soil tests still be considered together until adequate field data provide further information.
Item Citations and Data