Wheat and Lupin Potassium Responses

Assessment of soil test and K responses for wheat and canola in southern Victoria.


04 Mar 2015

2014 Annual Report

Close up of plants in-situ

Investigation of TBK as a soil test to predict K response.
The K response experiment near Lake Bolac was selected as an opportunity to take further measurements of TBK in order to better understand the TBK variable as an indicator of reserve K.
In addition, further exploration of the data and relationships developed in Holland et al. (2014) were undertaken with a view to better understanding TBK and K dynamics overall.

Background experiment details

Location: approx. 25 km south-west of Lake Bolac
Experimental design: Randomised complete block design with 4 replicates
Treatment levels:
4 (K0 = control, K1 = 25 kg K/ ha, K2 = 50 kg K/ ha, K3 = 100 kg K/ ha)
Fertiliser applied: KCl
Crop types: wheat (Kellalac); lupins (Jindalee)
Sowing rate: 80 kg/ha (wheat), 75 kg/ha (lupins)

Table 1. Basic soil chemical properties at Lake Bolac trial site
Exchangeable cations
Ca Mg K Na
(cmol(+)/kg )
Colwell P (mg/ kg)TBK (mg/kg)

Table 2. Clay mineralogy content (percent by weight) of the soil at Lake Bolac trial siteA
A Microcline/ orthoclase, anatase and hematite were present at or less than one percent.

Australian Soil Classification
Yellow Kurosol (Great Group = Mesotrophic; Subgroup = Mottled)

The relationship between soil test values (exch. K, TBK and DGT-K) and relative yield (%) and actual grain yield response at the 2013 Lake Bolac K trial.
Intensive soil sampling in each plot was undertaken before the K rates were applied. The soil was sampled at 3 depths: 0-10, 10-20 and 20-30 cm. The soil was measured for the following K variables: exchangeable K, TBK and DGT-K.
TBK (0-10 cm)
The relationship between exchangeable K and tetraphenyl borate K (TBK)

Values for exchangeable K (exch. K) and tetraphenyl borate K (TBK) were determined and plotted against the relationship provided in Fig. 3 of Holland et al. (2014) (see Fig. 16 below). Fig. 3 of Holland et al. (2014) showed the relationship using soils from Queensland (Moody and Bell 2006) and soils from Central and Southern NSW. Here we strengthen the relationship with soils from the Lake Bolac trial as well as additional soil samples from 8 potential trial sites from different parts of south-western Victoria. This provides further confidence that exch. K and TBK can be described using a log-log relationship.

Fig. 1. Relationship between ln(exch. K) and ln tetraphenyl borate K (ln(TBK)) for each soil and fertiliser rate treatment from the study by Holland et al. 2014 (Southern and Central NSW) (Δ) compared with 37 Queensland soils (●) from Moody and Bell (2006) with the linear regression (—) and the upper/ lower 95% confidence intervals (– – –) and upper/ lower 95% prediction intervals (……); with the Lake Bolac trial soil data () and data from potential trial sites located across SW Victoria ().