Agronomic evaluation of bone char as phosphorus fertiliser after five years of consecutive application

Agronomische Bewertung nach fünfjähriger Phosphordüngung mit Knochenkohle

Journal für Kulturpflanzen, 72 (12). S. 561–576, 2020, ISSN 1867-0911, DOI: 10.5073/JfK.2020.12.02, Verlag Eugen Ulmer KG, Stuttgart

Correspondence

Dr. Kerstin Panten, Julius Kühn Institute (JKI) – Federal Research Centre for Cultivated Plants, Institute for Crop and Soil Science, Bundesallee 69, 38116 Braunschweig, Germany, email: kerstin.panten@julius-kuehn.de

Accepted

4 November 2020

This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/deed.en). Dies ist ein Open-Access-Artikel, der unter den Bedingungen der Creative Commons Namensnennung 4.0 International Lizenz (CC BY 4.0) zur Verfügung gestellt wird (https://creativecommons.org/licenses/by/4.0/deed.de).

Materials and methods

In autumn 2013, the experiment was established in a completely randomized block design with iSPTCs been already randomly available from the former experiment. Fertilizer types were assigned randomly in each iSPTC maintaining the original three blocks. The plots had a size of 5.75 m × 17.5 m (original plots were split in half), and a crop rotation of winter barley, winter oilseed rape, winter wheat, lupine and winter rye (Table 2). All straw remained on the field and was therefore not taken into account for P budgeting and uptake. Annually, chisel ploughing and ploughing to a depth of 25 cm incorpo­rated straw and crop residues before seeding. Table 2 provides an overview of agronomic measures in the period 2013 to 2018.

Three different types of P fertiliser were applied to an equivalent of 45 kg P ha^–1 once a year shortly before seeding and a control without P for comparison: bone char (BC) pyrolysed at more than 800°C, surface-modified bone char (BC^plus) with sulphur compounds from biogas streams (patent DE102011010525), and highly water soluble triple super phosphate (TSP). Fertilisers were used as provided with 95–100 % of the particles bigger than 1 mm. Total concentration of the fertilizer elements were extracted by Aqua regia (VDLUFA, 2000) and P (177.4 nm), Ca (318.1 nm), K (766.4 nm), Mg (279.0 nm), and Na (589.5 nm) were measured by Induc­tively Coupled Plasma-Optical Emission Spectro­metry (ICP-OES, icap 6000, Thermo Fisher, Cambridge, United Kingdom), As, Cd, Cr, Cu, Ni, Pb, U, and Zn by High-Resolution Inductively Coupled Plasma Mass Spectrometry (ICP-MS, Thermo Element XR, Thermo Fisher, Cambridge, United Kingdom). BC and BC^plus had much lower concentrations of As, Cd, Cr, Cu, Ni, and U than TSP fertiliser (Table 3). Further detailed information about the three fertilizers was provided by Zimmer et al. (2019).

From dried soil samples, soil pH was determined in 0.01 M CaCl₂ (10 g soil in 25 ml CaCl₂) and water soluble P (P_water) was extracted in a procedure slightly modified from van der Paauw et al. (1971) (1.5 g soil in 2 ml aqua dest. for 22 hours, addition of 70 ml aqua dest., overhead shaking for 60 min) before P was analysed colorimetrically (Specord 50, Analytik Jena, Germany) using the molybdenum blue method at a wavelength of 882 nm accord­ing to Murphy & Riley (1962). Plant available P (P_CAL) was extracted from soil with calcium acetate lactate (Schüller, 1969) and total soil P concentration (P_tot) was extracted by Aqua regia (VDLUFA, 1991). In both extracts, P was measured at a wavelength of 213.6 nm using Inductively Coupled Plasma-Optical Emission Spectrometry (ICP-OES; icap 6000, Thermo Fisher, Cambridge, United Kingdom).

From the spring samples (0–10 cm) potential acidic and alkaline phosphatase activity was determined accord­ing to Schinner et al. (1996) based on the method developed by Tabatabai & Bremner (1969) and Eivazi & Tabatabai (1977). The enzyme activities of 1 g field moist soil after 1 h of incubation at 37°C were measured as p-nitrophenol coloured by 0.5 m L^–1 sodium hydroxide; colorimetric analysis was carried out at 400 nm (Specord 50, Analytik Jena, Germany).

Mean relative grain yield and mean relative P uptake of grain in relation to TSP fertilisation (100 %) from 2014–2018 was calculated according to Eq. 1 and 2.

For the first crop rotation (2014–2018), the apparent nutrient recovery efficiency of grain (ANR; (Baligar et al., 2001)) and the P budget were calculated according to Eq. 3 and 4, respectively.

Significant effects of the iSPTC on grain yield were recorded in the 1^st, 3^rd, and 4^th year and effects of ferti­liser type only in the 1^st year (2014) of the experiment (Table 6). Interacting effects of fertiliser type with iSPTC on yield were not observed (Table 6). Summarizing the mean relative yield (Eq. 1) in relation to TSP fertilisation of all years (Fig. 1) showed significant effects of the iSPTC (P value = 0.0021) but there was no statistical evidence of the fertiliser treatments (P value = 0.2718) or the interaction of iSPTC and applied fertiliser type (P value = 0.3616). Only for iSPTC-A a yield effect of fertiliser was obvious and means of relative yield followed the order P0 (90 %) < BC (94 %) < BC^plus (95 %) < TSP (100 %) (Fig. 1).

Fig. 1. Mean relative grain yield with TSP = 100 % as affected by initial soil P-test class (iSPTC; A = severely deficient, B = deficient, C = sufficient) and type of P-fertiliser (P0, BC, BC+ (= BC^plus), TSP) during the first crop rotation of the experi­ment from 2014–2018.

P uptake by grain was affected by the iSPTC in the 1^st, 3^rd, and 5^th and by fertiliser type in the 2^nd and 5^th year of the experiment (Table 7). Even though in the 5^th year (2018, winter rye) individual significant effects of iSPTC and fertiliser type on the P uptake by grains were observed, there was no statistical evidence of an interaction of these in any of the years (Table 7). Mean relative P uptake by grain in relation to TSP fertilisation of all years (Fig. 2) showed significant effects of the iSPTC (P value = 0.0017) as well as of the fertiliser treatments (P value = 0.0004) but no statistical evidence of any inter­action of iSPTC and applied fertiliser type (P value = 0.2475) were found. Effects were most pronounced in iSPTC-A and -B, with increasing effects on the P uptake by grain in class A depending on the fertiliser treatment: P0 (81 %) < BC (88 %) < BC^plus (91 %) < TSP (100 %) and in class B: P0 (84 %) < BC^plus (90 %) < BC (92 %) < TSP (100 %) (Fig. 2).

Fig. 2. Mean relative P uptake by grains with TSP = 100 % as af­fected by initial soil P-test class (iSPTC; A = severely deficient, B = deficient, C = sufficient) and type of P-fertiliser (P0, BC, BC+ (= BC^plus), TSP) during the first crop rotation of the experi­ment from 2014–2018.

The total P uptake by grains, cumulated for the whole crop rotation, ranged between 67.6 kg ha^–1 for the unfertilised, initially severely P-deficient soil and 87.8 kg ha^–1 for the initially P-sufficient soil supplied with TSP ferti­liser (Fig. 3), leading to a P-use efficiency of 30 to 39 % (225 kg ha^–1 P applied in five years). A one-factorial ana­lysis of variance of the total P uptake by grains cumulated for one crop rotation showed that soils with sufficiently available P concentrations were not affected by the fertiliser treatments (Fig. 3). In deficient and severely deficient soil, TSP fertilisation increased the P uptake by grain significantly in comparison to the control without P fertilisation, whilst the BC and BC^plus fertilisation did not significantly increase the P uptake (Fig. 3). The total P uptake by grain of all years showed significant effects of the iSPTC (P value < 0.0001) as well as of the fertiliser treatments (P value = 0.0005) but no statistical evidence of any interaction of iSPTC and applied fertiliser type (P value = 0.3597) when analysed by a two-factorial analysis of variance.

Fig. 3. Sum of P uptake by grain as affected by in­itial soil P-test class (iSPTC; A = severely deficient, B = deficient, C = sufficient) and type of P-fertiliser (P0, BC, BC+ (= BC^plus), TSP) during the first crop rotation of the experi­ment from 2014–2018.

These results were reflected by the very low ANR results for the iSPTC-C of 0.9–1.4 % (Table 8). The ANR values were larger in soils that had lower initial P levels (iSPTC-A and -B). Both experimental factors influenced the ANR significantly but no statistical evidence of any interaction of iSPTC and applied fertiliser type were found (Table 8). Except for the control treatment without P fertilisation, in which soil P was depleted, all fertiliser applications resulted in a surplus in the P budget (Table 9). The P surplus was highest in treatments with the lowest P uptake and vice versa (Table 9).

Annual P fertiliser application (45 kg P ha^–1) was much higher than the P removal by harvested products ranging from 8 to 24 kg P ha^–1 depending on experimental treatments and crops grown (Table 5). Differences in the concentrations of plant available P (P_CAL; concentrations 2018 minus concentrations 2013) were detected during the crop rotation (Table 10) which did not correlate with the P budget (r² = 0.1501). The P_CAL concentrations were significantly different between the iSPTCs and the applied fertiliser type but there was no evidence of any interaction of both factors (Table 10). Highly water-soluble TSP fertiliser increased the P_CAL concen­trations in all three iSPTCs, whereas the BC treatment led to a decrease in iSPTC-B and -C and the BC^plus treatment to a decrease in iSPTC-C (Table 10). Only the application of TSP led to a change in soil P-test classes: iSPTC-A changed to soil P-test class B after two and iSPTC-B changed to soil P-test class C after four fertiliser applications. Similar effects were observed for the water-extractable P concentration in top soil after harvest, which increased for all three iSPTCs after five years of TSP appli­cation but decreased for all other treatments except for iSPTC-B and BC^plus fertilisation (Table 11).

In spring, P uptake under plant available P deficient field conditions (iSPTC-A and -B) tended to increase in BC^plus compared with BC treated plots (Table 4). How­ever, BC^plus did not lead to higher yields (Table 6) or P uptake at harvest (Table 7). This disagrees with results of a pot experiment conducted with the same fertilisers, reporting a statistically significant increase in P uptake through the application of BC^plus (Zimmer et al., 2019). Then again, yields of the pot and field experiment were similar, with an increase by BC^plus application in comparison to BC, however, these effects could not be confirmed statistically (Zimmer et al., 2019). In iSPTC-C, BC and BC^plus reduced P_CAL concentrations after one crop rotation. This reduction was lower in the BC^plus than BC treatment, which had similar effects as the P0 treatment, suggesting a trend towards a better P solubility in BC^plus treated soils with sufficient plant available P status. It also points towards an “in situ digestion” effect by the S component of BC^plus (Fan et al., 2002). The initial question, if additional S components of BC^plus can increase the plant availability of P in soil, yield, and P uptake by plants, can be affirmed only partly. In 2015, significantly higher P uptake by grains of winter oilseed rape in the BC fertilised plots compared with BC^plus, TSP and P0 were unexpected, since oilseed rape does not form mycorrhizal symbiosis for P acquisition and its root system architecture is hampered by a low P environment (Yuan et al., 2016). However, Leppin (2007) showed that oilseed rape was able to access P from fertilised apatite (soft ground rock phosphate) in pot experiments, confirming findings of Hoffland et al. (1989). These authors showed that oilseed rape, as a reaction to P-starvation, was able to acidify the local rhizosphere by exudation of organic acids, which is a likely possibility to solubilise P (Hoffland et al., 1989; Leppin, 2007).

Until today, statistically significant interactions between iSPTC and fertiliser type of P_water in spring or changes in P_CAL and P_water after harvest could not be established. Since pyrolysis of BC and BC^plus above 800°C mainly creates insoluble HA, solubilisation of P was expected to be slow. However, whilst in iSPTC-A both bone char treatments resulted in only minor changes in the P_CAL status of soils after harvest, BC fertiliser applied to iSPTC-B had a negative effect and BC^plus a positive effect on the P_CAL concentration (Table 10). Nevertheless, the BCs contribution to P uptake of grains at P deficient soils (Table 7, Fig. 3) indicates that the routine soil P-tests do not cover the complete pool of plant available P during a growing season. Exceptional high variation of the sum of P uptake by grain of the P0 treatment on iSPTC-A soil might also prevent statistical evidence for the severely P deficient soil (Fig. 3).The question, whether the initial P status of experimental soil affects the P solubility of BC and BC^plus, needs further evaluation.

Generally, water extractable soil P correlates well with P uptake by plants (Zehetner et al., 2018) but is also used as an indicator for potential P leaching loss from soils (Wang et al., 2015). In spring, water extractable P concentrations were marginally higher in the BC^plus than in the BC treated soil (Table 5) indicating a slightly better P supply to crops by BC^plus. Due to the overall low levels of water extractable P concentrations for BC and BC^plus, especially in P deficient soils of iSPTC-A and -B, a higher risk of P leaching losses cannot be expected.