Landscapes typically deemed at risk from leached losses of nitrogen (N) and phosphorus (P) are those with short subsurface hydrologic time lags. Due to the short time it takes nutrients to move from a source to an area of concern, such sites are deemed perfect to test the efficacy of programmes of measures as management changes. However, a small subset of these sites can retain nutrients in soil/subsoil layers, which in turn are leached and can be either attenuated (e.g. nitrate converted to gaseous forms or immobilised in soil and P can be mineralised) or mobilised over time. This biogeochemical time lag can have long lasting effects on water quality. In an intensive agricultural karst oxidised aquifer setting, the aim of this study was to improve understanding of P and N inputs, retention, attenuation and subsurface pathway distribution and to inform how similar sites can be managed in the future. This was undertaken for the present site by integrating existing secondary and new primary datasets for both N and P. Results showed that in the years pre-2000 slurry from an on-site integrated pig production unit had been applied at rates of 33 t ha-1 annually, which supplied approximately 136 kg ha-1 total N and approximately 26 kg ha-1 total P annually. This practice contributed to large quantities of N (Total N and NH 4-N) and elevated soil test P (Morgan extractable P), present to a depth of 1 m. This store was augmented by recent surpluses of 263 kg N ha-1 , with leached N to groundwater of 82.5 kg N ha-1 with only 2.5 kg N ha-1 denitrified in the aquifer thereafter. High resolution spring data showed greatest percentage loss in terms of N load from small (54-88%) and medium fissure pathways (7-21%) with longer hydrologic time lags, with smallest loads from either large fissure (1-13%) or conduit (1-10%) pathways with short hydrologic time lags (reaction time at the spring from onset of a rainfall event is within hours). Although soils were saturated in P and in mobile forms to 0.5 m, dissolved reactive P concentrations in groundwater remained low due to Ca and Mg limestone chemistry. Depletion of the legacy store with no further inputs (taking 25% of available mass of soil organic N as available in 1 m of soil/subsoil to be 75 kg N ha-1) would take approximately 50 years, with NO 3-N concentrations in the source area dropping to levels that could sustain groundwater NO 3-N concentrations below admissible levels within 9 years. Biogeochemical time lags (decades) are longer than hydrologic time lags on this site (months to years). Future management should target farm surpluses that maintain a legacy store at or below a soil organic N mass of ~ 20 kg N ha-1. Incorporation of biogeochemical and hydrologic time lag principles into future water quality regulations will provide regulators with realistic expectations when implementing policies.
