SummarySoil phosphates exchange oxygen atoms rapidly with soil water once recycled by intracellular enzymes, thereby approaching an equilibrium 18 O P signature that depends on ambient temperature and the 18 O W signature of soil water. We hypothesized that in the topsoil, phosphates reach this equilibrium 18 O P signature even if amended by different fertilizers. In the subsoil, however, there might be phosphates with a smaller 18 O P value than that represented by the isotopic equilibrium value, a condition that could exist in the case of limited biological P cycling only. We tested these hypotheses for the HCl-extractable P pool of the Hedley fractionation scheme of arable soil in Germany, which integrates over extended time-scales of the soil P cycle. We sampled several types of fertilizer, the surface soil that received these fertilizer types and composites from a Haplic Luvisol depth profile under long-term agricultural practice. Organic fertilizers had significantly smaller 18 O P values than mineral fertilizers. Intriguingly, the fields fertilized organically also tended to have smaller 18 O P signatures than other types of surface soil, which calls into question full isotopic equilibrium at all sites. At depths below 50 cm, the soil 18 O P values were even depleted relative to the values calculated for isotopic equilibrium. This implies that HCl-extractable phosphates in different soil horizons are of different origins. In addition, it supports the assumption that biological cycling of P by intracellular microbial enzymes might have been relatively inefficient in the deeper subsoil. At depths of 50-80 cm, there was a transition zone of declining 18 O P values, which might be regarded as the first evidence that the degree of biological P cycling changed at this depth interval.