Managing the lakes of the Rotorua District, New Zealand

Abstract

In 2005, Burns, McIntosh and Scholes described strategies to manage the Rotorua Lakes using lake monitoring together with designated baseline Trophic Level Index values established for each lake. Continued monitoring has revealed that 9 of the 12 Rotorua Lakes have Trophic Level Index values in excess of their baseline values. Action Plans have been drawn up for the remediation of these damaged lakes that specify the excess nutrient loading to each lake and propose actions for the decrease of these loadings. Nutrient loading to various lakes has been decreased by upgrading waste treatment facilities, dosing tributary streams with alum, diverting an enriched tributary flow directly into the outflow channel of a lake, precipitating in-lake phosphorus with Phoslock™ and zeolite additions, and removal of macrophyte biomass from a lake and planting an artificial wetland at the entry point of a tributary to a lake. Where data are available, the results of these actions are explored. The similarities between the management system for the Rotorua lakes with the management systems used for two American and European lakes are described.

Key words:

• action plans
• trophic level criteria
• water quality criteria
• trophic level indexes

The twelve Rotorua Lakes lie in the middle of the North Island of New Zealand in what is know as the Central Volcanic Plateau (Fig. 1) and vary widely in character. The beauty of the lakes, together with the interesting geothermal features of the region, have resulted in the Rotorua District being one of the most important national tourist and sport fisheries area of New Zealand. The region also supports a significant amount of successful farming and forestry. The sustainable management of the Rotorua Lakes is the legal responsibility of the Environment Bay of Plenty (EnvBOP) Regional Council. Water quality of the Rotorua Lakes began to change in the early 1900s soon after European settlement of their catchments. A program of routine monitoring of the lakes was started in 1990 by EnvBOP and intensified in 1999 when strategies to halt the deterioration of all the lakes were put into place by EnvBOP (described in detail in Burns et al. 2005). Statutory legislation supporting this management strategy is contained in EnvBOP's Water and Land Plan (W&LP). This article is a sequel to the previous article (Burns et al. 2005) and describes the strategies, their implementation, their refinement with time and some results of the actions that have been taken.

Figure 1 A map of the Rotorua District, New Zealand showing the twelve lakes under management in this district. The enlarged area is shown as Figure 2.

Background

The Rotorua Lakes are considered to be a national resource in New Zealand. Widespread public concern about the degradation of many of these lakes prompted an investigation and report by the New Zealand Parliamentary Commissioner for the Environment (2006). His report endorsed the strategy for restoration of the Rotorua Lakes and led to the signing of a Memorandum of Understanding between the Crown (New Zealand Government) and the Rotorua Lakes Strategy Group, consisting of the Te Arana Lakes Trust, the Rotorua District Council and EnvBOP (EnvBOP 2007a). The Memorandum endorsed the use of the Trophic Level Index (TLI) system together with the TLI targets included in the W&LP. Subsequent to the signing of this agreement, the New Zealand government agreed to pay NZ$72 million toward the estimated cost of NZ$144 million for the planned remedial work to improve the damaged lakes.

In 2006, the Government of New Zealand signed an agreement with the Te Arawa Maori Trust Board transferring legal title to the Rotorua lakebeds to the local Maori people, while protecting public access, as a partial redress for past actions against the Maori people. This does not currently change EnvBOP's management role.

Methods

Assessment and management methods

The lake assessment methods used by EnvBOP and described in Burns et al. (2005) are summarized as follows:

1.	The procedures endorsed by the New Zealand Ministry for the Environment, and described in its publication Protocols for Monitoring Trophic Levels of New Zealand Lakes and Reservoirs (Burns et al. 2000), were adopted by EnvBOP for the assessment of monitoring results from the Rotorua District lakes. In following these Protocols, EnvBOP calculates a trophic level for each lake annually using the annual average value of the four key variables: chlorophyll a (Chla), Secchi depth (SD), total phosphorus (TP) and total nitrogen (TN), for each year for each lake (Burns et al. 1999, 2000): where TLx values = index values for Chla (TLc), SD (TLs), TP (TLp) and TN (TLn), respectively. The TLI value and its standard error are calculated for each lake and year from: These equations normalize the annual average values so that TLx values are the same for the average New Zealand lake. By doing this, individual TLx values from a lake can be compared, allowing those that deviate most to be identified. For example, a TLn value significantly lower than the TLp value indicates the lake is probably N-limited, and similar TLn and TLp values indicate lakes that are co-limited. A classification scheme for lake trophic type was developed from the TLx and TLI values (Burns et al. 1999, 2000) with each integer value designating a different type lake: for example, lakes 2.0–2.9 are oligotrophic; 3.0–3.9 are mesotrophic. The LakeWatch computer program (EarthSoft.com 2007) has been used by EnvBOP to facilitate calculation of TLI values and to prepare annual reports on monitoring results.
2.	A reference value TLI for each lake is determined using the 1994 TLI. In cases where the 1994 water quality of a lake was unsatisfactory, a reference TLI was chosen that existed when the lake was considered to be in an acceptable condition. These reference values have been written into the statutory EnvBOP Regional Water and Land Plan and are referred to as the W&LP TLIs.
3.	Prepare an Action Plan for the water quality improvement of any lake that has a 3-yr average TLI greater than its W&LP TLI by 0.2 for 2 consecutive years.
4.	Implement the Action Plans.
5.	Continue to evolve the management strategies for the lakes. In this regard, identification of algal taxa in surface samples has evolved to counting cells of different type, in particular the cyano-bacterial species. This program has intensified because cyano-bacterial counts exceeding 15,000/ml require notices be posted advising against water contact recreation. Further, in 2005, EnvBOP implemented the Submerged Aquatic Plant Index (LakeSPI; Clayton et al. 2002) surveillance program to assess lake condition, including changes in species composition and distribution of aquatic plants. Intensive surveillance is also carried out in areas where pest plant fragments are most likely to dislodge from boats, such as launching ramps, to prevent the spread of undesired plants. LakeSPI can provide an independent identification of lake trophic type for different lakes (see Table 1).

Table 1 Long-term trends and three yearly average TLI values for the Rotorua District lakes in comparison to the TLI values set in the Regional Water and Land Plan, together with LakeSPI condition.

Lake	3 yearly average TLI to 2007 TLI units	Regional Water & Land Plan TLI units	Long Term Trend** Based on analysis over many years data (>13)	Lake Type Based on Trophic Level	LakeSPI Condition
Okaro	5.5*	5.0	Definite improvement	Supertrophic	Poor
Rotorua	4.9*	4.2	Degraded, no change	Eutrophic	Poor
Rotoehu	4.6*	3.9	Degraded, no change	Eutrophic	Average (Poor)
Rotoiti	4.1*	3.5	Definite degradation	Eutrophic	Poor
Rotomahana	3.9	3.9	No change	Mesotrophic	Good
Rerewhakaaitu	3.5	3.6	No change	Mesotrophic	Average
Okareka	3.3*	3.0	No change	Mesotrophic	Average
Tikitapu	3.0*	2.7	Possible degradation	Oligotrophic	Good
Okataina	2.8*	2.6	No change	Oligotrophic	Good
Tarawera	2.8*	2.6	No change	Oligotrophic	Average
Rotoma	2.5*	2.3	Definite degradation	Oligotrophic	Good
Rotokakahi	—	3.1 ^1	—	Mesotrophic	Average

Action plan development

The development of Action Plans is a fundamental part of the management process. Prioritizing lakes for Action Plan development is based on consideration of the following factors using a formal weighting and scoring system (EnvBOP 2005):

A.	In-lake indicators: • TLI: 3-yr average for the past 2 years • Hypolimnetic dissolved oxygen • Cyanobacterial blooms • Condition of native lake plants
B.	Catchment indicators: • Land development constraints • Nitrogen and phosphorus inflows from humans • Intensity of land use in catchment
C.	Inter-lake indicators: • Risk of affecting a downstream lake
D.	Introduction of invasive macrophytes and exotic pest fish
E.	Social indicators: • Use of lake by district/region • Community concerns • Existing action for managing lake quality.

Once a lake has been selected for Action Plan development the following steps are taken:

1.	Define the existing nutrient budget through a spreadsheet model, usually using a detailed land-use map together with land-use nutrient loss coefficients (Menneer et al. 2004) to calculate diffuse source inputs, plus calculation of inputs from point sources. In this regard, a Rule exists in the EnvBOP W&LP that prohibits loss of nutrients from pastoral land at rates greater than the average level of loss during 2001–2004.
2.	Determine a sustainable level of nutrient input to maintain the lake's W&LP TLI. Methods vary depending on the information available on a lake from earlier published research. Sometimes estimates of the desired N and P loads have been obtained by initially using Nurnberg's (1984) method for estimating the TP retention coefficient (R), for a lake. This coefficient is then related to the outflow concentration (C; also assumed to be the lake annual average concentration), the annual nutrient load (M), and the annual outflow volume (Q; Hoare 1980) to give C = (1− R) M/Q, which is rearranged to give the desired annual load: Hoare's (1980) study of Lake Rotorua found very similar retention coefficients for both N and P, so this has been assumed to be the case for other lakes in the Rotorua district.
3.	Propose possible actions to achieve the desired nutrient reductions, taking a holistic approach to the remediation of the lake, considering: sewerage, landscape, current and potential land use, climate change, economics, tourism and matters that affect lakeshore residents. This planning phase takes considerable consultation with different groups to formulate and gain acceptance of decisions.
4.	Test selected possible scenarios that could achieve the nutrient reduction targets and model their effect on the lake. The DYRESM model (Imberger and Patterson 1981) is used to model the physical components of the lake and the CAEDYM model (Hipsey et al. 2004; Romero et al. 2004) to model the nutrient and biological processes to test the various proposed engineering options (Burger 2006).
5.	Select the desired lake management actions.
6.	Seek approvals and implement decisions.

Results and discussion

Status of lake management efforts through 2009

Nine of the 12 Rotorua Lakes had 3-yr average TLI values in excess of their designated W&LP values (Table 1) and thus require remediation. The Rotorua lakes are each different and as such their action plans and implementation are best described on a lake-by-lake basis.

Lake Rotorua

Lake Rotorua is the largest lake of the group with 80.8 km² area, 45 m maximum depth and 11 m average depth. It is polymictic and stratifies intermittently. The lake has 52% of its catchment in pasture, the city of Rotorua on its southern shore, and is in excess of its W&LP TLI by 0.7 units. The lake was degraded in large part by the input of treated wastewater over many years (Rutherford 1984), now partially alleviated by the diversion of the wastewater into a nearby forest (Rutherford 1996). However, much of the nitrate pumped into the forest is no longer absorbed and now enters the lake via Puarenga Stream. The lake has always received a significant load of N and P from groundwater via rivers and geothermal springs. This nitrate load is proving to be very problematic because, first, the lake tends to be N-limited (TL_p – TL_n = 0.6) and second, the nitrate concentrations in the tributary rivers and springs are increasing annually. Most of this nitrate originates from the urine of pastured animals. It has entered the deep groundwater during 40 years of grazing and is now surfacing in increasing concentrations each year (Rutherford 2003, Morgenstern et al. 2004).

The Lake Rotorua Action Plan (Proposed Lakes Rotorua and Rotoiti Action Plan, EnvBOP 2007b) has been prepared together with that of the adjacent Lake Rotoiti because this lake receives 70% of its nutrients from Lake Rotorua via the Ohau Channel (Fig. 2). In addition to excessive external loads, the Action Plan notes a large internal load of soluble P resulting from the intermittently anoxic sediments (Burger 2006). The desired loads for lake Rotorua have been taken as those present in 1960 (Rutherford et al. 1984). The Action Plan states that to reduce the present 3-yr average TLI of 4.9 to the W&LP TLI of 4.2, external nutrient inputs need to be reduced by 112 metric tons (T) of N/yr from the current load, and 224 T N/yr from the predicted load in 2055 (because stream concentrations of nitrate are increasing with time), and 10 and 25 T P/yr from current external and internal loads, respectively. Actions taken or proposed in the Plan to achieve required load reductions are:

1.	The wastewater treatment plant has been upgraded to remove an additional 15 T N/yr because the forest receiving the wastewater is only removing a small fraction of the nitrate.
2.	Utuhina Stream is being dosed with alum at 1 g/m3 of flow to achieve a load decrease of 2 T P/yr.
3.	The wastewater from eastern areas of the lake catchment will be collected and treated for a reduction of 11 T N/yr and 0.25 T P/yr.
4.	Stormwater upgrades should result in a reduction of 3 T N/yr and 0.5 T P/yr.
5.	Alum flocculation of two other streams should produce a further reduction of 6 T P/yr.
6.	Other actions under consideration are removal of high concentrations of geothermal ammonium from the Tikitere stream by diverting the flow through a nitrification plant and then through a woodchip denitrification field to reduce load by 30 T N/yr.
7.	Internal loading may be addressed via sediment capping.
8.	Hamurana Stream could be diverted into the Ohau Channel outlet for a reduction of 53 T N/yr and 6.3 T P/yr, but this diversion will increase the residence time of the lake water. The most difficult task is to effect the large load reduction of 170 T N/yr and 6 T P/yr by 2017 from nonpoint source runoff. A Landuse Futures Board has been formed to develop the detailed strategy required to reduce the nutrient loss from farms.

Figure 2 Map of Lakes Rotorua and Rotoiti showing the Ohau Channel and the Kaituna River.

Lake Rotoiti

Lake Rotiti is a large lake comprised of three basins with an area of 34.6 km², maximum and mean depths of 110 m and 31.5 m, respectively, and only 16% of its catchment in pasture. The major inflow is from Lake Rotorua via the Ohau Channel, and the outflow is the Kaituna River that is fairly closely situated to the Ohau Channel in the Okere arm of the lake (Fig. 2). The lake contains Okawa Bay, an enclosed bay that is far more eutrophic than the rest of the lake. Lake Rotoiti has a 2007 3-yr average TLI of 4.1, 0.6 units above its W&LP TLI (3.5). The Proposed Action Plan for the lake (EnvBOP 2007a) indicates a load reduction of 130 T N/yr and 19 T P/yr is needed to achieve a TLI of 3.5 units. The load reduction was estimated using Eq. 6 to estimate both the 2002-3 nutrient load from in-lake concentrations and the load that would result in the W&LP TLI concentrations.

To summarize important elements of the Plan:

1.	A wall will be built in the lake to divert the water flowing into the lake from the Ohau Channel into the Okere arm and hence directly into the Kaituna River (Fig. 2). This wall has been urgently constructed in 2008 at a cost of NZ$11 million and will divert 150 T N/yr and 15 T P/yr from the lake.
2.	Okawa Bay residents have had wastewater collected for treatment since 2007. Improved water clarity was noticed almost immediately but has resulted in an explosive growth of macrophytes in the shallow bay.

The combined nutrient reduction actions described above have resulted in a presumed load reduction of 162 T N/yr and 15.3 T P/yr, more than the N target of 130 T N/y but less than the P target of 19 T P/yr. A further reduction in the phosphorus load should occur in the deep eastern basin of the lake where there is an anoxic hypolimnion for over 3 months of the year, recycling about 50 T N/yr and 20 T P/yr. These internal loads should be reduced as a result of the diversion of the Ohau Channel inputs by the diversion wall. The effect of this action will be assessed before further management actions are planned to determine if the reduction targets have been met.

Lake Okareka

Lake Okareka has an area of 3.3 km², maximum and average depths of 33.5 m and 20 m, respectively, and a 56% pastoral catchment. It is a picturesque lake with a number of homes and cottages built around it. The lake had a 2007 3-yr average TLI of 3.3, 0.3 units higher than its W&LP TLI of 3.0. The Lake Okareka Catchment Management Plan (EnvBOP 2004) indicates the nutrient load to the lake needs to be reduced by 2.5 T N/yr and 80 kg P/yr to attain its W&LP TLI. The lake has tended to be P-limited, with TL_p–TL_n = −0.83; and the N:P ratio is increasing because 2.4 T N/yr enters the groundwater from septic tanks.

Allophane clays in the watershed retain much of the soluble phosphorus, but soils are now thought to be saturated because of excessive loading from septic systems that surround the lake. Rutherford and Cooper (2002) used Nurnberg's (1984) method to estimate the retention coefficient and Eq. 6 to estimate current loads using the average 1992–2001 in-lake concentrations. They compared these load estimates with the loads computed using the spreadsheet model and a detailed land-use map together with land-use nutrient loss coefficients (Menneer et al. 2004). They concluded that Eq. 6 yielded a more reasonable estimate of current nutrient loading to the lake. In addition, Eq. 6 was used to calculate the desired loads to give the concentrations that would yield the W&LP TLI.

The three major nutrient sources to the lake are septic tanks, nonpoint source runoff and internal loading (EnvBOP 2004). Actions underway to address these sources are:

1.	Approval and funding have been obtained to collect and treat the wastewater from all homes around the lake.
2.	Work will be undertaken to reduce the degree of pastoral farming in the catchment and to date, a large landowner is seeking approval to change from grazing land use to a parkland style of residential development.
3.	Soluble reactive phosphorus (SRP) concentrations increased in the hypolimnion while the lake was stratified. Vertical mixing during overturn resulted in a significant increase in the surface concentrations. Because the collection of wastewater had not yet taken place, Phoslock™ was used to absorb phosphorus from the lake waters and to seal the bottom sediments to reduce internal P loading. The application took place in August 2005, June 2006, and March 2007 (EnvBOP 2007c) at the rate of 118 kg/ha over the deeper half of the lake. Results to May 2008 show a suppression of sediment P release but no significant difference in the TP and SRP concentrations in the lake water since 2005. Because lanthanum is the active ingredient in Phoslock™ there was some concern pertaining to its bioaccumulation in aquatic biota. A study of fish in Lake Okareka showed that lanthanum concentrations in the flesh of crayfish and trout increased after each application of Phoslock™ but that the intervals between applications were long enough for the element to depurate from the organisms before the next dosing occurred (Landman et al. 2007).

Lake Okaro

Lake Okaro is the smallest lake in the region with an area of 0.32 km², maximum and average depths of 18 m and 12.1 m, respectively, and 91% of its catchment in pasture. The lake has received much attention as the most eutrophic lake in the region with a 3-yr average TLI of 5.5, in excess of its W&LP TLI of 5.0 by 0.5 units. The bottom waters (or hypolimnion) of this eutrophic lake are anoxic for most of the stratified period, leading to internal loads of 2410 kg N/yr and 380 kg P/yr. These loads are large in comparison with the external nutrient reduction targets of 910 kg N/yr and 20 kg P/yr, as indicated in the Lake Okaro Action Plan (EnvBOP 2006). The nutrient load reductions were estimated from the required decrease in TN and TP concentrations in the lake to achieve the W&LP TLI by Rutherford and Cooper's (2002) method.

Lake Okaro is being used as a testing ground for lake improvement methods. Because the internal loads are so large, they are being treated in tandem with the external loads. In July 2004, the lake was dosed with alum to reduce the TP concentration. In September 2007, the lake was spread with 112 T of modified zeolite, an ammonium-ion–absorbing mineral found in abundance near the Bay of Plenty. When treated commercially, an additional P-absorption and nitrate-absorption capability is established (Yang et al. 2004). As a result of these treatments, the TP concentrations have decreased by about 6.7%/yr (Fig. 3). The nitrate absorption capability of the modified zeolite is most useful in this instance because Lake Okaro tends toward N-limitation (TL_P–TL_N = −0.53). From December to May (mid-summer to the end of autumn) in 2005–2008, the nitrate concentrations ranged from 0.5 to 5.0 mgN/m³ while the SRP was higher, ranging from 0.5 to 15.0 mgP/m³, indicating an SRP surplus.

Figure 3 Average total phosphorus and annualized residual concentrations for Lake Okaro.

Various activities have been implemented to minimize the external loads from nutrient runoff from pastured portions of the watershed. A 2.3-ha wetland has been planted to intercept the streams flowing into the lake with a design flow of 30 L/sec. Substantial lengths of stream bank have been fenced off and retired as a buffer against the effects of grazing. In addition, a Herd Home™ for winter housing of animals (uncommon in New Zealand) has been installed to decrease nutrient loss from the animals by about 40%.

Lake Rotoehu

Lake Rotoehu has an area of 8 km², maximum and means depths of 13.5 m and 8 m, respectively, and 34% catchment in pasture. It is eutrophic (TLI = 4.6) and in excess of its W&LP TLI by 0.7 units. The lake tends toward N-limitation (TL_P–TL_N = 0.42) and experiences annual cyanobacterial blooms. In 1992, the lake had a TLI of 3.7, but in 1993 it suffered a long stratified period with internal loading, resulting in a TLI of 5.2. The lake has never recovered fully from this event and now has an average TLI of 4.6. In addition, large beds of hornwort (Ceratophyllum dermersum) now exist in the lake.

The Proposed Lake Rotoehu Action Plan (EnvBOP 2007d) states that nutrient load reductions of 8.9 T N/yr and 0.7 T P/yr are required to rehabilitate the lake. The required nutrient load reductions are based on the decrease in TN and TP concentrations needed to achieve the W&LP TLI. Nurnberg's (1984) method to estimate the retention coefficient and Eq. 6 were used to estimate current loads from the 2005 in-lake concentrations, and W&LP TLI concentrations used to estimate desired loads.

Another action considered to be effective in removing nutrients from the system is the annual harvest of hornwort, a macrophyte that obtains almost all its nutrients from the water, not the sediments. This procedure was trialed in April 2006 when 500 T wet-weight of hornwort was harvested in 10 days, effectively removing 1.2 T N and 0.16 T P from the system (EnvBOP 2007d). The intention is now to remove 1000 T of hornwort, and thus 2.4 T of N and 0.32 T of P, annually.

Planning is also proceeding for the replacement of septic tanks in the catchment with modern ones to effect a further reduction of 0.47 T N/yr. Alum dosing of a stream will be undertaken to remove 0.5 T P/yr. A floating wetland has been installed in the lake, and a large area of these wetland plants is being considered for stream mouths to take up incoming nitrogen and phosphorus.

Lake Rotoma

Lake Rotoma is an oligotrophic lake (TLI = 2.5) with an area of 11 km², maximum and average depths of 83 m and 37 m, respectively. An Action Plan (EnvBOP 2009) has been drafted for this lake because it exceeds its W&LP TLI by 0.2 units. In addition, this good quality lake is undergoing definite degradation, as shown in the TLI Report (Table 2), with trends of annual increases in TP, TN and Chla values, probably due to increased loading from septic tanks and pastoral land in the lake catchment. Collection of wastewater from homes around the lake is being considered together with riparian retirement and wetland enhancement to diminish the effect of pastoral runoff and reduce nutrient loading (Fig. 4).

Table 2 TLI Report for Lake Rotoma.

Lake Rotoma: 1994 to 2007 (1 Jul 1994–30 Jun 2007)
Percent Annual Change (PAC)
Lake		Chla (mg/m3)	SD (m)		TP (mgP/m3)	TN (mg/m3)		HVOD (mg/m3/day)		Avg PAC	Std Err		p-Value
Change–Units Per Year		0.03	(0.00)		0.18	2.88		(0.36)
Average Over Period		1.3	(12.25)		4.35	148.96		(19.03)
Percent Annual Change (%/Year)		2.31	0		4.14	1.93		0		1.68	0.78		0.1
Burns Trophic Level Index Values and Trends
Period	Chla (mg/m3)	SD (m)	TP (mgP/m3)	TN (mg/m3)	TLc	TLs	TLp	TLn	TLI Average	Std. Err. TL av	TLI Trend units/yr	Std. Err. TLI trend	p-Value
Jul 1994–Jun 1995	0.93	12.71	3.44	139.67	2.14	2.26	1.79	2.85	2.26	0.22
Jul 1995–Jun 1996	0.85	15.61	2.83	139.83	2.04	1.9	1.54	2.85	2.08	0.28
Jul 1996–Jun 1997		14				2.09			2.09	0
Jul 2000–Jun 2001	1.96	9.81	3.27	137.91	2.96	2.66	1.72	2.83	2.54	0.28
Jul 2002–Jun 2003	1.47	9.36	4	129.84	2.65	2.73	1.98	2.75	2.53	0.19
Jul 2003–Jun 2004	1.37	14.16	4.91	147.07	2.56	2.07	2.24	2.91	2.45	0.19
Jul 2004–Jun 2005	1.33	13.01	4.85	150.24	2.54	2.22	2.22	2.94	2.48	0.17
Jul 2005–Jun 2006	1.27	14.3	6.33	164.55	2.48	2.06	2.56	3.06	2.54	0.21
Jul 2006–Jun 2007	1.29	11.23	4.29	176.41	2.5	2.46	2.07	3.15	2.60	0.23
Averages	1.31	12.69	4.24	148.19	2.48	2.27	2.01	2.92	2.42	0.07	0.04	0.02	0.0542
SUMMARY:
PAC = 1.68 ± 0.78% per year			The guide used in the PAC average
p-Value = 0.01			p-Value evaluation is:
			p-Value Range				Interpretation
TLI Value = 2.42 ± 0.07 TLI units				P	< 0.1		Definite Change
TLI Trend = 0.04 ± 0.02 TLI units per year			0.1 <	P	< 0.2		Probable Change
p-Value = 0.0542			0.2 <	P	< 0.3		Possible Change
			0.3 <	P			No Change
ASSESSMENT:
Oligotrophic
Definite Degredation

Figure 4 Current and planned future nutrient loads to lake Rotoma.

Other Lakes

Action Plans will be developed for the other Rotorua District lakes: Lake Tarawera first, then lakes Tikitapu, Okataina, Rotokakahi, Rerewhakaaitu, and finally Rotomanhana.

Figure 5 Fluorescence and dissolved oxygen profiles on April 24, 2007 at Site 1, Lake Tikitapu (fluorescence units – μL as Chla equivalent).

Rotorua lakes management summary

The EnvBOP management scheme uses the nutrient status of the lakes together with their macrophyte and algal conditions in determining remedial actions to be taken. Other factors also have to be considered, however, such as hypolimnetic oxygen concentrations, land development intensity, introduction of invasive weeds, lake use intensity and the concerns of lakeside residents. Two important factors in achieving success in developing and implementing the proposed and accepted Action Plans by EnvBOP staff are; first, extensive consultation with the people of the Rotorua District, and second, helping them gain a knowledge of basic limnology. Educating the District farmers is particularly important because nutrient reduction from farms is essential and, in most instances, not mandatory but voluntary because the farmers understand the impact of increased nutrients on lakes.

Consultation has been, and is being, carried out in numerous meetings with different cultural groups, local officials, scientists and ratepayers. The wider understanding of lake processes by nonprofessional people is helped by annual meetings of the Lakes Water Quality Society in Rotorua in 2001, 2004, 2006 and 2008. Introduction of the TLI system in 2000, together with rapid reporting each year of monitoring results, has helped nonlimnologists grasp an understanding of the degree of change occurring in the Rotorua Lakes (e.g., EnvBOP 2008).

Comparison of management approaches

Many lakes in the world are being improved using systems that have much in common. The information presented in this report describes actions being taken in one region of New Zealand, but there are many similar examples in America and Europe. Applications under Section 303(d) of the Clean Water Act of the United States of America for approval of the Total Maximum Daily Load (TMDL) are numerous, such as the Iowa Department of Natural Resources (IDNR 2005) for Clear Lake and the South Dakota Department of Environmental and Natural Resources (SDDENR 2005) for Lake Herman. A European example of lake management is described by Schauser and Chorus (2007) for lakes Tegel and Schatlachensee. The prerequisite for any lake management program is some level of water quality monitoring before, during and after management actions are taken. The essential steps for lake management, excluding monitoring, are illustrated by the four examples discussed here and can be listed as follows:

1.	Convert narrative standards into numerical standards for the desired long term condition of the lakes (Table 3). The selection of a numerical standard for a water body is perhaps one of the most important aspects of lake management because it then becomes a simple matter to determine whether remedial work is necessary. An example of this process is described by Heiskary and Wilson (2008). Note that the Carlson TSI system (Carlson and Simpson 1996) and the Burns TLI system (Burns et al. 1999) are similar, and that one system is a multiple of the other. In comparing TSI and TLI values for the four key variables observed in Lake Mead, LaBounty and Burns (2005) found that TSI = 10.8 TLI whether the relatively high concentrations of the Inner Basin or the lower concentrations of the Outer Basin were compared. Thus the TLI target for Clear Lake would be 6.2 TLI units and for Herman Lake would be 6.8 TLI units (the LakeWatch program will produce Trophic Index Reports as in Table 2, in either TLI or TSI units).
2.	Determine the nutrient loads to each lake (Table 4).
3.	Use a lake-loading model to determine nutrient loads that would achieve the designated values for the numerical standards (Table 5).
4.	Calculate the load reduction targets for each lake by subtracting the target loads from the input loads.
5.	Determine acceptable means to achieve the load reduction targets (Table 6).
6.	Obtain the funds to implement the plans and, once implemented, carry out follow-up monitoring to gauge the success of the actions taken. If necessary, consider further management actions (Table 7).

Table 3 Comparison of 4 different sets of lake management criteria.

The Rotorua Lakes	Clear Lake (IDNR 2005)	Lake Herman (SDDENR 2005)	Lake Tegel and Schlachtensee (Schauser and Chorus 2007)
The numerical standard selected for these lakes is their W&LP TLIs and this is written into legislation.	The narrative standard of the prevention of nuisance algal blooms and the provision of water quality to support the lakes designated uses was translated into Carlson's (Carlson and Simpson 1996) Trophic State Index (TSI) values of TSP = 70; TSChla = 65; TSSD = 65 for a TSI = 66.7.	The narrative standard of providing water to satisfy the beneficial uses of contact recreation, stock watering, irrigation, and warm-water semi-permanent fish was translated into a TSP = 68; TSChla = 74.7; TSSD = 74.7 for a TSI = 73.9.	There was no narrative standard; a numerical target of achieving TP concentration of 30 ug P/L was chosen.

Table 4 Methods of determining nutrient loads to lakes.

The Rotorua Lakes	Clear Lake (IDNR 2005)	Lake Herman (SDDENR 2005)	Lake Tegel and Schlachtensee (Schauser and Chorus 2007)
Land loss coefficients (Menneer et al. 2004) were used to estimate nutrients draining from agricultural and forestry land. Total loads to some lakes can be estimated using eqn (1). Direct calculations are done for loads from geothermal sources, drains, treatment plants and internal loads.	The (Vollenweider 1982) combined OECD relationship empirical model was used to calculate the external load that would sustain the observed in-lake P concentration.	The Agricultural Non-point Source Model (Walker 1996) was used to estimate the load from a watershed that contained no significant point sources.	The loads for each lake were taken as those calculated earlier (Klein 1990 and Heinzmann et al. 1991).

Table 5 Lake loading models used to determine the nutrient loads that would achieve the designated values for the numerical standards.

The Rotorua Lakes	Clear Lake (IDNR 2005)	Lake Herman (SDDENR 2005)	Lake Tegel and Schlachtensee (Schauser and Chorus 2007)
The model developed by Hoare (1980) was used together with Nurnberg's (1984) method of estimating retention coefficients.	The Vollenweider (1982) combined OECD relationship empirical model was chosen (no reference given but the equations used are shown in the report).	BATHTUB model (Walker 1996) was used.	The Vollenweider (1976) and the modified One-Box models (Gachter and Imboden 1985, SAS 1989) were used.

Table 6 Different acceptable means to achieve load reduction targets.

The Rotorua Lakes	Clear Lake (IDNR 2005)	Lake Herman (SDDENR 2005)	Lake Tegel and Schlachtensee (Schauser and Chorus 2007)
Action Plans for the impaired lakes have been or are being developed and then implemented.	The TMDL approval application included an implementation Plan to achieve the required P load reduction.	The required reduction in the external load would be achieved by implementing a number of Best Management Practices in the watershed.	The plan was to put the entire water input to each lake through Phosphorus Elimination Plants.

Table 7 Results of actions taken to reduce loads.

The Rotorua Lakes	Clear Lake (IDNR 2005)	Lake Herman (SDDENR 2005)	Lake Tegel and Schlachtensee (Schauser and Chorus 2007)
Some planned actions have been implemented and some results have been obtained, as described above.	No published reports are known regarding the implementation and results of the load reduction plans.	No published reports are known regarding the implementation and results of the load reduction plans.	The Phosphorus Elimination Plants were built and Schlachtensee has reached the target of 30 ugP/L for some of the time with an accompanied large reduction in Chla concentrations. P and Chla concentrations have reduced considerably in Lake Tegel but the P target has not been reached because of continued internal loading.

Preserving lakes

Because so many of the world's lakes are damaged, most lake management is focused on stressed lakes. Nevertheless, it is also important to preserve or sustain lakes currently in good condition. Lake Rotoma provides a prime example of the need to be vigilant about good quality lakes. This oligotrophic lake provides a sport fishery and has excellent water quality, with an average SD of 12.6 m. There is no obvious indication of deterioration of the lake, but lake monitoring shows deterioration at the rate of 0.04 ± 0.02 TLI/yr or 0.43 ± 0.2 TSI/yr (Table 2). It has a 3-yr average TLI of 2.5, and the lake is now exceeding its W&LP TLI. Hopefully, action will be taken soon. Good quality lakes need numeric water quality baselines and monitoring if they are to be preserved.

Notes

*Lakes in excess of their W&LP TLI.

¹Rotokakahi has not been monitored in the last 3 years.

**Long term trend is based on an analysis period of over a decades worth of data for most lakes. Linear regression is performed on deseasonalised data that is averaged over the key TLI parameters.