16.3.5  Nutrient interactions

Printer-friendly version

As outlined above, agricultural production on Australia’s impoverished soils has been severely limited by widespread deficiencies of various nutrients. Twin deficiencies of N and P were ubiquitous and often accompanied by one or more additional deficiencies, so that nutrient interactions on plant growth and crop productivity have assumed unusual significance. P was the first deficiency to be recognised, and after 1882, P fertilisers were used increasingly on wheat. Years later, N deficiency rather than water deficiency was appreciated as a major restriction on crop response to added P. N fertilisers proved too expensive for wheat production, and interest focused on legume pasture leys to build up soil N reserves.

In pastures, N × P responses affect botanical composition (Section 12.3.3) as well as dry matter production. H.C. Trumble and co-workers clarified underlying principles in sand culture experiments at the Waite Institute in Adelaide with subterranean clover and rye grass. When established separately, rye grass only grew well when both N and P were added, while clover responded strongly to P irrespective of added N. When grown with clover, rye grass had little effect on clover in the absence of added N, but rye grass suppressed clover when N was added. These findings explained field observations on changes in grass and clover components in response to fertiliser P and fluctuations in soil N.


Figure 16.9 Logarithmic plots of shoot dry mass (g    plant-1) as a function of soil nutrient additions to show positive interactions between (a) P × K, (b) N × P and (c) N × K with respect to plant growth. (Based on Anderson and Thomas 1946)

In many areas of Australia, introduction of legumes and their special requirements for Rhizobium sp. plus nutrients for N2 fixation brought a host of new interacting factors and problems. In resolving these problems, A.J. Anderson and his colleagues at CSIRO in Adelaide and Canberra developed and applied innovative ways of interpreting fertiliser responses and their interactive effects on plant growth and reproductive development (Anderson 1956). They replotted Mitscherlich’s early and extensive data on oat response to N, P and K (Figure 16.9), observing for these response surfaces that any two nutrients occurring in a 2 × 2 factorial arrangement frequently showed a positive interaction. Gross imbalance between N, P or K additions did result in mild toxicity and growth was depressed by excess N, P or K, and especially when applied in combination with minimum levels of accompanying nutrients (see supraoptimal effects in Figure 16.9).


Figure 16.10 Contrasting effects, and interactions between N and Mo fertilisers on yield of (a) a legume (subterranean clover) and (b) a non-legume (flax). Small additions of Mo will substitute for large additions of N in legumes because Mo promotes N2 fixation. Flax showed little response to Mo if N was not added; moreover, flax needed Mo in order to utilise nitrate N. As a consequence, N x Mo effects are mildly interactive. Treatments are abbreviated as follows: 0 = control (no added N or Mo); Mo = molybdenum added at 0.152 mg sodium molybdate per pot of clover, and 1.21 mg sodium molybdate per pot of flax; N = nitrogen + molybdenum addition at rates specified for single elements. (Based on Anderson and Spencer 1950)

Anderson and co-workers contrasted these interactions between two independently essential nutrients with the negative Mo × N interaction they obtained with clover. In that case, clover showed a strong response to addition of Mo in the absence of N; but clover response to added Mo tended to disappear as more N was added (Figure 16.10a). A negative Mo × N interaction of this sort would be expected if Mo was concerned solely with symbiotic N fixation. By contrast, a plant such as flax (Figure 16.10b) is totally dependent on existing soil N, and showed only a mild additive response to N + Mo compared to N alone.

Anderson (1942) used the same approach to solve other problems of legume establishment including a negative Mo × lime interaction and a negative lime × Rhizobium interaction to show that responses to many tonnes of lime could be re-placed by a few grams of Mo plus small applications of lime and Rhizobium with clover seed at sowing. Large areas of southern Australia previously thought too dry and inhospitable for improvement were subsequently sown to legume-based pastures as a direct result of these research outcomes.