As plants can only take
As plants can only take up P in an inorganic orthophosphate form, it was expected that the proportion of orthophosphate P in the OA would be a good indication of the availability of the P in the OA. While the P speciation of the COMP sample could not be determined by deconvolution due to the broadness of the NMR signal for this sample, we found that all three litters in this study were high in orthophosphate, which is normal for animal manures (Sharpley and Moyer, 2000). The PIG-STR sample contained the highest proportion of orthophosphate which could explain why PIG-STR resulted in a greater plant P uptake than CHK-STR and CHK-SD. Conversely, CHK-STR contained a higher proportion of orthophosphate compared to CHK-SD. However, plant P uptake was similar for both chicken litter treatments. Therefore, the proportion of orthophosphate in the OA alone is not enough to determine the availability of P in OA. After orthophosphate, the second greatest pool of P in the OA was phytate. The CHK-SD contained a higher proportion of phytate than CHK-STR, and both chicken litters contained higher proportions of phytate than PIG-STR. Chicken manures often contain a high proportion of phytate (e.g. 11–37%; Peirce et al., 2013) because of the high amount of seeds in their diet (Nelson et al., 1968) and the inability of chickens to efficiently digest phytate (Toor et al., 2005). While phytate was once thought to be stable in soils (He et al., 2006, Celi et al., 1999), there is now evidence that phytate can be rapidly mineralised to orthophosphate by a range of AP-III-a4 microbes (Doolette et al., 2010). Moreover, mineralisation of phytate may replenish the soluble orthophosphate pool faster than the rate of orthophosphate stabilisation by soil, providing a more constant pool of plant-available P. This could explain why CHK-SD resulted in similar plant P uptake to CHK-STR, despite being lower in orthophosphate P. The proportion of P in OA present as bicarbonate-extractable P may give a better indication of P availability than P species. The bicarbonate extractable method was developed for soil (as the Colwell P method), but it has been used for biochar (Hossain et al., 2010, Chan et al., 2008) and could also be used to determine plant-available P in litters and composts. However, the amount of bicarbonate-extractable P added with a treatment did not always correspond with the increase observed in Colwell P of the soil, with less than 58% of the of the bicarbonate-extractable P in OA contributing to an increase in Colwell P, compared with the 61.3% of P in the INORG-P which contributed to an increase in soil Colwell P. This could be driven by differences in P buffering capacity between the OA and the soil. While the PBI of the soil was very low (55.6–71.7), indicating that small additions of P should have a large effect on plant-available P, it is possible that the PBI of the OA (not measured) was even lower. While PBI of OA is not generally determined, there is evidence that chicken litter can stabilise P in a similar way that soil does, but over a longer time-frame than soil (months compared with days; Peirce et al., 2013). Therefore, it is likely that upon addition to soil, the bicarbonate-extractable P in the OA was quickly sorbed to soil particles and was no longer extractable by the Colwell P method. Moreover, there were differences among treatments in the increase observed in Colwell P. While 51.4–57.7% of the bicarbonate-extractable P in the PIG-STR, COMP and CHK-SD, contributed to an increase in Colwell P, only 36% of the bicarbonate-extractable P in the CHK-STR contributed to an increase in Colwell P. This corresponds with the finding that CHK-STR provided plants with similar amounts of P as CHK-SD even though it had a higher proportion of orthophosphate P and bicarbonate-extractable P. While Colwell P has been commonly used to determine plant-available P for many years, DGT P is developing into a promising technique which has been proven successful in fields fertilised with inorganic P (Mason et al., 2010), as well as more recently in pot experiments with OA (Six et al., 2014). When P is removed from the soil solution, e.g. via plant P uptake, insoluble P then often moves into solution. The DGT method seeks to mimic this process by measuring the movement of P from the soil solution into the DGT device. The Colwell P method, on the other hand, measures the pool of P that can be extracted with bicarbonate. This is an estimation of the P pool that can become available to plants, although it is not always accurate as the processes that affect soil P pools are complex. This study indicated that the DGT P method may be preferable over the Colwell P method in systems where P is applied as OA. This is for two reasons: i) DGT P had a larger range than Colwell P, potentially indicating a greater sensitivity of the DGT method to detect differences in plant-available P compared with the Colwell method; and ii) sample sizes for DGT method are greater than for the Colwell method, which can eliminate some of the variability in the results (as seen for Colwell P, Fig. 1b). This variability is likely driven by the fact that the OA were coarse and therefore could not be mixed homogeneously throughout the soil. However, it is likely that neither method correctly identifies the actual plant-available pool of P, given the range of complex processes involved in plant P uptake. While alternative P extractions exist, such as the sequential P extraction of Hedley et al. (1982), there is little evidence that these provide better results than more common soil analyses (Motavalli and Miles, 2002).