Abstract
The loss of phosphorus (P) from soil to waterways is often attributed to the accelerated
eutrophication of surface waters, especially in agricultural areas where phosphate fertilisers
may be used. High P concentrations were previously reported in several streams in the Teviot
Valley, Central Otago, New Zealand. This study investigated the potential sources of
phosphorus (P) to streams in the Teviot Valley using geochemical tracers including major
elements, trace elements, rare earth elements (REEs), and strontium (Sr) isotopes to assess the
influence of chemical weathering and fertiliser application on the chemistry of the surface
waters, soils, and rocks. Surface water, soil, and rock samples were collected from five streams
and one river seasonally over the span of one year and a sample of phosphate fertiliser was also
obtained. Water samples were analysed for dissolved reactive phosphorus (DRP), total
phosphorus (TP), and silicon (Si) by Flow Injection Analysis (FIA). Major elements, trace
elements and REEs in the stream water, soil, rock, and fertiliser samples were analysed by
inductively coupled plasma mass spectrometry (ICP-MS). Once Sr concentrations were
known; select water, soil, and rock sample digests were and run through resin-containing
columns to separate Sr from other elements and the 87Sr//86Sr compositions of these samples
were analysed by multicollector ICP-MS (MC-ICP-MS). Three out of the five streams
consistently exceeded the regulatory DRP guideline of 0.01 mg/L over the sampling period,
although two of these streams had higher DRP and TP than the other. This suggested a greater
input of particulate-bound P entering these streams which may be linked to natural erosion or
livestock treading and may need to be managed differently to dissolved inputs. P concentrations
in the soils were also found to be significantly correlated (p<0.05) with Fe which indicated that
the upstream Fe-rich brown soils likely have the highest P retention values, although this was
not considering other factors affecting P retention. The schist samples analysed were also found
to have a slightly greater apatite content than other Otago schist, although weathering indicators
including the Chemical Index of Alteration (CIA), mobile and immobile elements in soils, and the major element compositions of the streams showed similar weathering patterns among the
streams therefore weathering would be unlikely to contribute to significant variation in P
inputs. Mixing calculations and the Sr isotope compositions of the soils also suggested the
enrichment of P, trace elements, and certain REEs between winter and spring at some sites
could not be strongly attributed to the application of the fertiliser which was analysed but may
be due to other soil amendments or natural inputs which may be influencing soil chemistry.