Sources of contaminants in the Waikato-Waipa catchment Bill Vant Waikato Regional Council I’ve been asked to speak to an analysis I’ve recently completed that refines and updates earlier assessments of the sources of nitrogen (N) and phosphorus (P) in the catchments of the Waikato and Waipa Rivers. The recently-released national Policy Statement for Freshwaters requires that sources of water quality contaminants in Freshwater Management Units be accounted-for. My recent analysis is an example this type of accounting (and will be included as a Case Study in the relevant MfE Guidance Document). The underlying principle of contaminant accounting is the conservation of mass: as materials move through ecosystems they are neither created or destroyed. In theory we could track the progress of individual atoms of nitrogen say as they flowed through an ecosystem.

Water quality monitoring networks River monitoring, 20 locations (WRC, NIWA) Flow – continuous (m3/s) Concentration – monthly (g/m3) Point sources, 19 locations (“consent monitoring”) Flow – reported daily-to-monthly Concentration – daily-to-monthly Load = Σ(flow × concentration) (g/s, kg/d, t/yr) WRC routinely monitors river flows and water quality at 18 sites in the catchment of the Waikato and Waipa Rivers. NIWA operates further sites. I have chosen to use data from a total of 20 sites (18 WRC, 2 NIWA). Flows are monitored continuously; water quality is determined monthly. In addition, 19 moderate-to-large operations hold consents to discharge loads of contaminants, including loads of N and P, to waterbodies in the catchment. Each consent includes the requirement that the flow and water quality of the discharge be monitored. Flow is reported at various internals ranging from daily totals to monthly averages. Water quality is measured daily at some sites, through to monthly at others. The instantaneous “load” (or mass flow/flux) of a material is defined as the product of the flow and the concentration. Summing the individual loads over time – I’ve done this for the decade 2003-12 – gives the average load. The units are mass per time, g/s, or its equivalents. I’ve chosen to report the nutrient loads in tonnes per year, noting that 1 g/s = 31.5 t/yr.

This map shows the locations of the 20 river flow and water quality sites (in green) and the 19 point source discharges (in yellow). I’ve divided the catchment into three sub-catchments: (1) Upper Waikato River (from Taupo gates to Karapiro dam), (2) Waipa, and (3) Lower Waikato (everything else).

For example, various sites, 2003-12 Flow (m3/s) [Total N] (g/m3) Load (t/yr) River sites Waikato-Taupo 158 0.1 339 Waikato-Narrows 235 0.5 3695 Waipa-Whatawhata 88 1.1 4069 Waikato-Tuakau 402 0.8 11,193 Point sources Hamilton sewage 0.48 12 189 Horotiu meatworks 0.02 114 90 Ngaruawahia sewage 14 8 This table shows some typical results. It shows the average flow and average total nitrogen concentration during 2003-12 at four of the river sites and three of the point sources. The load of nitrogen at each of the sites is also shown. (Load calculations are as described by Vant 2011, WRC technical report 2011/06; an important point is that the loads shown here are not equal to the product of the average flow and average concentration – which is likely to provide a biased result.) Flows at the river sites are typically much higher than the flows of wastewater. Conversely, nitrogen concentrations in the wastewaters are typically much higher than those in the river, with meatworks wastes containing particularly high levels. Also, the loads carried by the rivers increase moving downstream. And larger point sources discharge larger loads of nitrogen.

Loads from point sources, 2003-12 This plot shows the average loads of nitrogen and phosphorus discharged by the 19 point sources during 2003-12, based on my analysis of WRC’s records of consent monitoring data. The upper part of the plot shows results for 11 sewage discharges, and the lower part shows results for eight industrial discharges.

Contaminant accounting (NPS-FW 2014) Determine load carried by river (A) Identify background or natural contribution (B) Add up contributions from all point sources (C) Calculate contribution from landuse, D (= A – B – C) Here’s the process I’ve used to assess the relative importance of the different sources of N and P in the Waikato and Waipa catchments (and indeed the process I used previously for the Hauraki Rivers in WRC technical report 2011/06). It’s similar to the process outlined in the (currently draft) guidance document developed by NIWA and others for the NPS-FW. Note that “D”, the contribution from land that has been developed (e.g. for pastoral agriculture, cropping or urban areas) is the extra load, over and above the background contribution (B) from that land.

For example, nitrogen, Waipa catchment A, Waipa-Whatawhata 4069 t/yr B, Background 928 t/yr (= 3093 km2 @ 0.3 t/km2/yr = 928 t/yr) C, Point sources 66 t/yr Otorohanga sewage 14 Te Awamutu sewage 11 Te Kuiti sewage 26 Te Awamutu dairy factory 15 Sum, point sources 66 D, Landuse (= A – B – C) 3075 t/yr For example, for the nitrogen loads in the Waipa catchment, as follows: A is the load carried by the river, as determined from the measurements of river flow and water quality at the Waipa @ Whatawhata monitoring site. B is estimated as the area of the Waipa catchment (3093 km2) multiplied by the specific yield, estimated to be 0.3 t/km2/yr. This value is based on the loads determined made in a small number of undeveloped catchments in the Waikato region (see EW technical report 2006/54). C is the sum of the four point sources in the Waipa catchment. D, obtained by difference, is the estimated load resulting from the development of the Waipa catchment – primarily from the development of about three-quarters of the catchment area for pasture and cropping. (Note that this land also contributes to B; in fact it contributes about three-quarters of B.)

Three sub-catchments, 2003-12 The Waipa nitrogen loads are shown here as the middle bar graphs. The corresponding results for the other two sub-catchments are also shown. Note that the results for the Upper Waikato sub-catchment include the loads flowing out of Lake Taupo (noting that this flow contributes more than half of the river flow at Karapiro dam). Note also that the results for the Lower Waikato do not include the inputs from the upstream sub-catchments, namely the Upper Waikato and the Waipa sub-catchments. The corresponding loads of phosphorus are also shown. The greatest concentration of point sources is found in the Lower Waikato sub-catchment, where they contributed 12% of the nitrogen and 31% of the phosphorus. Between 44% (Upper Waikato) and 75% (Waipa) of the load of nitrogen carried by the rivers is estimated to have come from diffuse agricultural sources. However, background and point source loads of phosphorus were somewhat more important than those of nitrogen. So lesser amounts—between 26% (Upper Waikato) and 59% (Waipa)—of the loads of phosphorus carried by the rivers are estimated to have come from diffuse agricultural sources. *Ignoring inputs from upstream catchments

Sources of nutrients, Waikato/Waipa, 2003-12 Loads in river and from point sources are measured Point sources: about 7% of the overall nitrogen load and 18% of the overall phosphorus load Background – 29% and 35%; landuse – 61% and 45% The contributions to the three sub-catchments can all be added together to identify the contributions of nitrogen and phosphorus to the Waikato/Waipa river system as a whole. That is, the contributions to the loads measured in the Lower Waikato River near Tuakau during 2003-12. Altogether, the 19 point sources contributed about 7% of the mass flow of nitrogen carried to the sea by the Waikato and Waipa Rivers during 2003–12. They also contributed about 18% of the mass flow of phosphorus. Landuse accounted for about 61% of the combined load of nitrogen and 45% of the phosphorus transported to the Tasman Sea. The outflow from Lake Taupo contributed 3% of the N and 2% of the P.

PS loads, changes during the decade Phosphorus: 30% reduction Nitrogen: 7% reduction There has been ongoing improvement in the treatment of wastewaters at several of the point sources, with good data available for 11 sites (shown here). In these cases I’ve calculated the average loads discharged up until the end of 2010, and compared them with the loads discharged during 2011-12 (noting that comprehensive records were not available for the earlier part of the decade in several cases). The combined load of phosphorus discharged from these 11 sites during 2011-12 was 30% lower than that discharged prior to 2011. The reduced load from Hamilton sewage was responsible for much of this. The combined load of nitrogen discharged during 2011-12 was 7% lower (results not shown).

E. coli loads from some point sources I’ve been asked whether corresponding information is available for the sources of E. coli. The answer is, “yes, but ...” This plot shows the loads of E. coli discharged from 13 of the point sources during 2004-13. The loads are expressed in giga-cfus per day, or billions of cells per day. However, we know that E. coli die when discharged to river waters. So it’s unclear just how much of the E. coli load discharged at Te Kuiti, for example, will be present at river sites downstream of there. So I’m reluctant to take this analysis much further.

E. coli loads from some point sources The combined point source load was equivalent to about 5% of the corresponding value for Waikato River @ Tuakau What I can say is that the combined load of E. coli discharged from the 13 point sources during 2004-13 was equivalent to about 5 percent of the average load that was carried in the river at Tuakau during the decade. Given that some of the E. coli will not have survived the journey between the discharge points and Tuakau, this means that no more than 5 percent – and possibly considerably less than this – of the load carried by the river at Tuakau has come from point source discharges. I’m unable to estimate the background or the landuse loads of E. coli.

Mangaokewa @ Te Kuiti While the point source contribution to the “whole of catchment” load is small, point source discharges can be important in particular locations. The previous slides showed that the discharge of E. coli from the Te Kuiti sewage treatment plant was large. And we know that the river into which it is discharged – the Mangaokewa – is relatively small. This combination of a large point source load being discharged to a small waterbody represents something of a “worst case” situation. So what can we say about the effect of this particular point source on river water quality immediately downstream of the discharge point? This plot shows the median E. coli concentration at the WRC monitoring site in the Mangokewa Stream at Te Kuiti, about 4 km upstream of the sewage plant. The E. coli concentration is about 500 cfu per 100 mL. The dashed line is the water quality standard for swimming in the current Waikato Regional Plan, namely a median concentration of 126 cfu/100 mL. So upstream of the sewage plant discharge point, E. coli concentrations in the river are about four times higher than the swimming standard.

Mangaokewa @ Te Kuiti, low flow Now let’s take account of the discharge from the sewage plant (as shown in slide #11). This plot shows how the large load of E. coli discharged from the treatment plant in recent years caused a marked increase in the E. coli concentrations in the Mangokewa River, downstream of the discharge point. My calculations show that at low river flow (10%ile), the concentrations in the river would have increased from about 500 to about 3700 cfu/100 mL, with all of the increase being caused by the sewage plant discharge – shown here as the red part of the bar. At higher river flows there is more dilution, so the effect was would be less pronounced. For example, at median river flow – that is, flows that are exceeded for half of the time – the downstream E. coli concentration was about half of that shown here. (Interestingly, and perhaps somewhat surprisingly, the available data for the upstream site indicates that E. coli concentrations there don’t appear to change in a predictable fashion as river flow changes, so upstream concentrations at higher river flows appear to be broadly similar to those at low flow.) However, the sewage treatment plant at Te Kuiti has recently been upgraded. In the middle of last year (July 2013), disinfection using ultraviolet (UV) light was added to the treatment process. This has produced a 100-fold reduction in median E. coli concentrations in the discharged wastewater (from 26,000 cfu/100 mL before the upgrade, to 230 cfu/100 mL after it).

Mangaokewa @ Te Kuiti, low flow So this plot also shows the situation in the months following the upgrade to the treatment process (right-hand bar, “Post-upgrade”). The E. coli concentration upstream of the discharge is still about 500 cfu/100 mL, but now the concentration downstream is only about 6% higher than this – see the sliver of red in the right-hand bar. The discharge from the sewage treatment plant is still contributing a bit to the E. coli concentration in the Mangaokewa River, but most of the contamination comes from sources upstream of the treatment plant.

Conclusions Monitoring data can be used to identify the contributions of different sources of contaminants Overall, point source discharges contribute 7% of the N and 18% of the P to the Waikato/Waipa (at Tuakau). Landuse – mostly pastoral farming – contributes about 60% and 45%, respectively. “Non point sources” include background or natural sources. These can be appreciable (c. 30%). Overall, point source discharges are a minor source of E. coli. But they can be locally important.

Proportion of annual load carried during six months Dec-May (growing season) Nitrogen Ohakuri 42% Narrows 35% Tuakau 29% Phosphorus Ohakuri 50% Narrows 41% Tuakau 40%