Hubert Tunney(1), Isabelle Kurz(1, 2), David Bourke(1), Robert Foy(3), and David Kilpatrick(3)
(1) Teagasc, Johnstown Castle Research Centre, Wexford, Ireland, (2) current address, EPA, Johnstown Castle, Wexford; 3DARDNI, Newforge Lane, Belfast, Northern Ireland Hubert.Tunney@teagasc.ie
Introduction
A large proportion of farmland in many countries in northern Europe is grassland. In Ireland 90% of the 4.2 million ha of farmland is grassland. Phosphorus (P) deficiency limited grassland production in Ireland and this was corrected by chemical fertiliser use in the 1960s and 1970s. There was an input of about 3m tonnes of P in chemical fertiliser and purchased feed, which was over double the removals in milk and meat, over the past 50 years and this contributed to an ten-fold increase (about 0.8 to 8 mg L-1 soil, Morgan’s P) in mean soil test P (STP). There was also an increase in nitrogen and other nutrient inputs. This led to increased intensification of grassland (doubling of grass yield and of grazing animal numbers, from about 3m to over 6m livestock units) which contributed to increased P loss from grassland to water. There is little information on the relative contribution to P loss of increased chemical fertiliser use compared to increased grazing animal numbers. The main objective of this paper is to present results of the dissolved reactive P (DRP) concentrations in overland flow from six grassland (cut and grazed field plots with a range of STP levels) and the implications for mitigation of P loss to water.
Methods
The nutrient concentrations and nutrient loads from six grazed and cut field plots (0.24 to 1.54 ha each) were studied in this experiment which started in September 2000 and finished in March 2004, at Teagasc, Johnstown Castle, Wexford. The six plots, Warren 1 (Plot 1; grazed 2001 and 200, cut 2002), Warren 2 (Plot 2; cut 2001 and 2003, grazed 2002), Cowlands 1 (Plot 3: grazed 2001-2003), Cowlands 2 (Plot 4: cut 2001-2003), Diary 1 (Plot 5: grazed 2001-2002, cut 2003) and Dairy 2 (Plot 6:
cut 2001-2002, grazed 2003) had STP of 3.5, 4,8, 17.9, 16.7, 7.0 and 7.2 mg L-1 soil respectively (Figure 1).
Flow proportional overland flow samples were collected and analysed for P and N fractions; in addition some samples were analysed for potassium, and suspended solids.
Results
There were significant variations in DRP concentrations over the seasons and between the six field plots (Figure 1 and Table 1). Concentrations of DRP in overland flow varied form under 0.005 mg L-1 to over 3 mg L-1. There was a significant
(P<0.01) linear relationship between STP in the six plots and mean annual DRP concentrations in overland flow for 2002 and 2003. There was more than a ten-fold difference in mean DRP concentrations in overland flow between plots with the lowest (Plot 1, 3.5 mg P L-1 soil) and highest (Plot 3, 17.9 mg P L-1 soil) STP (0.05 versus 0.59 mg L-1 DRP in 2002 and 0.03 versus 0.72 mg L-1 DRP in 2003). This compared to a maximum of 66% increase (0.43 (Plot 5) to 0.72 (Plot 3) mg L-1 DRP for cut and grazed, respectively in 2003) that could be attributable to the presence of grazing animals.
Figure 1. The mean daily DRP concentrations in overland flow, when overland flow occurred, from the six plots from January 2002 February 2004.
The estimated annual dissolved DRP loads in overland flow from the plots ranged from 0.1 to 1.2 kg ha-1 year-1. There was a significant correlation between the three P fractions measured, 86% of total P (TP) was total dissolved P (TDP), 77% of TP was DRP and 90% of TDP was DRP. Excluding a relatively small number of summer time
overland flows, the highest DRP concentrations and loads occurred in autumn when overland flow started (combining high flows and high P concentrations, Figure 1) after an extended summer dry period (autumn/winter wash-out effect).
The difference in mean DRP concentrations in overland flow between cut and grazed plots (most evident on the high STP plots) were generally highest at this time of year (autumn/winter). In contrast there was generally no difference between cut and grazed plots in January and February when concentrations were lowest and before the grazing season started.
Table 1. Seasonal mean concentrations of DRP (mg L-1) in overland flow between paired grazed and cut plots at the Cowlands and Dairy sites, 2002 - 2004.
Site and Significance p<0.05 p<0.001, NS, p<0.05 NS p<0.05 p<0.001 Dairy (Plot 5, grazed in 2001-02, cut in 2003: Plot 6, cut in 2001-02, grazed in 2003) Grazed (Plot 5) 0.37 0.52 0.45 0.21 0.98 0.42 Cut (Plot 6) 0.33 0.54 0.31 0.17 0.83 0.15
Significance NS NS NS NS NS p<0.001
Results of soil physical measurements in subplots in these field plots showed that grazed plots had significantly lower macroporosity (57-83%) and higher bulk density (8-17%) and resistance to penetration (27-50%) than cut plots. Rainfall simulation experiments on small plots (0.5 m2) gave increased overland flow on grazed compared to cut areas.
Soil test P concentrations in soil under dung-pats were shown to increase 3 to 4 fold over the 90 day decomposition periods studied. Soil microbial biomass turnover of 51 kg P ha-1 year-1 was calculated for the Cowlands site (Plots 3 & 4).
The final reports of this work will be published in 2007 (www.epa.ie).
Conclusions
Three factors influencing DRP concentrations in overland flow were indicated in this work: a) the highest DRP concentrations were from plots with the highest STP and visa versa, b) a seasonal P cycle with high DRP concentrations in autumn/winter when overland flow commenced after the summer and decreasing over the following two months, c) relatively small difference in DRP concentrations between grazing and cutting treatments.
Overall the results indicate that high STP from surplus P inputs was a more important factor than the increased animal numbers in increasing P loss from grassland to water in Ireland. It is concluded that the P loss in overland flow from grassland can be increased by the presence of grazing animal. The main potential for P loss mitigation from grassland is to maintain soils at or near the lowest STP level
compatible with good grassland production on soils subject to P loss in overland flow and match P inputs with outputs in milk and meat.