Lake Restoration: Is There a Successful Model?

Lake Ellesmere/Te Waihora Symposium Lincoln, 15-16 November 2011

Lake restoration: Is there a successful model?

David Hamilton University of Waikato

Photo: Warrick Powrie Lake restoration: Is there a successful model?

Outline • Straight to the point – an overview of key needs for lake research/management • What are the issues globally? • A plea for models • Case studies – Rotorua lakes • So?

Photo: Warrick Powrie Point 1. Ecological processes in lakes will generally not follow linear trajectories expected from linear changes in external forcings (e.g. catchment nutrient loads)

Scheffer et al. (2004) Souce: www.nature.com Lake Waihola, Otago Regime shift in a eutrophic Rotorua lake

Ozkundakci et al. (2010), Ecological Engineering

Effects of increasing nutrients loads on chlorophyll distributions (deep lakes) Natural/low external loading

Deep chlorophyll maximum c. 1% of surface light

Increased external loading

Surface populations, ↑ buoyant species (including N-fixers) and horizontal variability

Distributions of chlorophyll: Lake Taupo

Hamilton et al. 2010, Aquatic Sciences Lake Rotoiti Point 2. Reductions in external loading are fundamental to effective control of eutrophication. Diffuse sources now represent the greatest challenge to external nutrient reductions Lake Ngaroto, Waikato Point 3. Despite many years of theoretical, empirical and modelling studies, we have often failed to adequately capture and quantify nutrient loads, particularly stormloads Changes in phosphorus concentrations in a stormflow event

Jonathan Abell University of Waikato

0.3

) 0.25

1 - 0.2 Total 0.15 phosphorus 0.1 3-

0.05 PO4

] and and ] (mg [TP] L

- 3

4 0

[PO

1:40 4:40 7:40 2:20 5:20 8:20 1:50 4:50 7:50

11:20

16:40 14:20 13:40 19:40 22:40 10:40 17:20 20:20 23:20 16:50 19:50 22:50 10:50 1/08/2010 2/08/2010 3/08/2010 4/08/2010 ‘Old age’ groundwater to Rotorua

Hamurana 110 yr Ngongataha 16 yr Awahou 61 yr Waiowhero 42 yr Point 4. Opportunities exist to virtually revolutionise temporal and spatial coverage of lake ecosystems but require lake ecologists to adopt increasingly flexible, interdisciplinary communication linkages that may not necessarily initially be fully productive. …in the old days on Ellesmere

New tools to study lakes: remote-sensing - high-frequency data - computer modelling

Mat Allan, Waikato University

Current Projects: 25 lakes Physical variability in temperate lakes: A global analysis of high- frequency instrumented buoy data from 25 temperate lakes. Science and communication

“Scientists can be most effective if they make their results accessible to interested laypersons”

Ludwig, D. 2001. The era of management is over Ecosystems 4: 758-764 So…to lakes across the globe Lake Taihu P.R. China

Taihu Algal bloom evolution in Lake Taihu

1950 1970 1980

1987 1994 2000

Big problem…expensive solution? Lake Tahoe, California, USA

Charles Goldman UC Davis Transparency in Lake Tahoe Secchi depth in m

Charles Goldman, UC Davis Obvious culprits in loss of transparency Air pollution: A major cause of loss of water transparency in Lake Tahoe

Charles Goldman, UC Davis Lake Mendota, Madison, Wisconsin, USA Ice cover in Lake Mendota: Characterised by discontinuities Ice Duration 160

140

120

100 Days

80 Ice Cover (Days) Cover Ice

40 1855 1875 1895 1915 1935 1955 1975 1998 DroughtDrought in Australia: 2008 Response to drought at Milang, Lake Alexandrina

Oct 2006 Jan 2008 Floods and drought in Australia: 2010 From drought to flood: unprecedented rainfall in eastern Australia

http://i.telegraph.co.uk/multimedia/archive/01797/australia-flood_1797071b.jpg Lake Kinneret, Israel

(c. 40% of Israel’s drinking water)

Water level (m) level Water Denmark: the challenge of non-point source pollution

Source: Erik Jeppesen Diffuse pollution – perhaps the most difficult case

“When an issue becomes highly controversial when it is surrounded by uncertainties and conflicting values… there are experts for the affirmative and experts for the negative. We cannot settle such issues…”

Simon HA. 1983. Reason in human affairs. Stanford University Press. Polarisation…

• Economism is the placing of an exceptional and inordinate emphasis upon economic values in contradistinction to all others • Scientism is the belief that science…is inherently capable of solving almost all human problems. • Technocracy represents an effort to achieve policy solutions by recourse to technological innovation or through what is sometimes called a "technological fix."

Caldwell LK. 1990. Between two worlds: science, the environ-mental movement and policy choice. Cambridge University Press. And for Te Waihora… even more difficult

• Nutrients • Mahinga kai • Water level regimes • Climate change • Wind and sediments Rotorua lakes’ Technical Advisory Group

“If you don‟t deliver then our people will be planning and writing policy without your science input”

“I need the science to record our on-ground restoration efforts”

You won‟t get it [the science] 100% but it will be necessary to use it for our policies and plans Delevan Lake, Wisconsin, USA Lake Delevan, Wisconsin, USA

Inlet extension Rotenone application, Delevan Lake Rotenone application effects, Delevan Lake Internal Loading Near Bottom Total Phosphorus Concentrations

1.2

1.0 198319831983 1984 19841984 0.8 1985 19851985 19851986 198619871986 19861987 0.6 19871988 198719891988 1994 1988 1990 19881989 19891991 0.4 1990 1993 19891992 19901991 19901993

19911992

Total P Concentrations, mg/L Concentrations,P Total

Total P Concentrations, mg/L Concentrations, P Total Total P Concentrations, mg/L Concentrations, P Total Total P Concentrations, mg/L Concentrations, P Total 1994 0.2 19921992 197219721972 19911991

0.0 AprilApril MayMay JuneJune JulyJuly AugustAugust SeptSept OctOct NovNov Alum application Lake Delevan

Annual P Loading to Delavan Lake 10 9 8

7 6

PLoad (t) PLoad

4 3 2 1 0 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 The challenge to scientists

“...the reason we expect so little from science is that we have not yet realized that our job is to predict and not to describe. We have not yet come face-to-face with our failure to do effective ecological science”

...Rigler and Peters (1995) Case study of potential effects of climate change on three NZ lakes N Lake Rotoehu Lake Rotoiti Lake Rotorua

Te Waihora Lake Rotoma (Lake Ellesmere)

10 km Lake Okataina

Lake Okareka Lake Tarawera Lake Tikitapu

Lake Rotokakahi Lake Rotomahana

Lake Okaro Lake Rerewhakaaitu 10 km

Lake Max depth (m) Trophic state Okareka 33.5 Mesotrophic Rotoehu 13.5 Eutrophic Te Waihora (Ellesmere) 2.5 Highly eutrophic PredictingScenario effects of simulationschanging nutrient load and/or climate on chlorophyll a

30 Plus 50% Year 2100 Year 2100 plus 50%

Relative Relative to 20

Day

a -

15 Present 10

% Change Change Chl % in 5

0 Okareka Rotoehu Te Waihora

Trolle, D, D. Hamilton, C. Pilditch, I. Duggan, E. Jeppesen (2011). Predicting the effects of climate change on trophic status of three morphologically varying lakes: Implications for lake restoration and management. Environmental Modelling.& Software Managing for resilience to blooms in a changing climate

“Increasing nutrients and temperatures have a synergistic effect on cyanobacterial blooms, but only when nutrient levels are high. Ultimately nutrients are the primary factor influencing bloom formation” Clearwater Lake near Sudbury, Canada A ‘reversing’ temperature in Clearwater Lake

C) o 20 Surface 18 16 14 12 Bottom 10 Y = 497 - 0.245 X 8 2

Mid-summer temperatureMid-summer ( r = =0.73, P<0.001 6 1975 1980 1985 1990 1995 2000

Tanentzap et al., Limnology & Oceanography And they’re using our models where? Lake Constance, Germany Lake Wivenhoe, Queensland Lake Mead, Nevada

Slide 57

Timelines for applying models

years before present years into future Present 10000 100 10 1 10-1 10-2 0.01 0.1 1 10 100

} } } } }

}

Hindcasting Hindcasting Hindcasting Forecasting: Simulation: Prediction: - past climate recent human for detailed - response to - lake - future climate -past land-use disturbance: temporal weather management -future land use - understanding forecast strategies eutrophication, : - algal bloom - extreme - invasive - material likelihood/ events species, fluxes -fate (weather, - pollutant fate - transport storms) - testing theory

Levels of complexity for model applications

Van Nes, E. and M. Scheffer, 2005. A strategy to improve the contribution of complex simulation models to ecological theory. Ecological Modelling 185: 153-164. The science essential to assist management at the ecosystem level will be by those that connect the disciplinary science (“join the dots”) Seasonality of nutrient limitation Box Plot of multiple variables grouped by Month Rotorua_model_output_nutrient_limitation_function 9v*2922c Median; Box: 25%-75%; Whisker: Non-Outlier Range 1.0

0.8

0.6 f(P,N) 0.4 Cyanobacteria f(P) 0.2 Box Plot of multiple variables grouped by Month f(N) Rotorua_model_output_nutrient_limitation_function 9v*2922c 0.0 7 8 Median;9 10 Box: 25%-75%;11 12Whisker:1 Non-Outlier2 Range3 4 5 6 1.0 Month f(P) 0.8 f(N)

0.6 Diatoms f(P,N) 0.4

0.2 Box Plot of multiple variables grouped by Month Rotorua_model_output_nutrient_limitation_function 9v*2922c 0.0 Median; Box: 25%-75%; Whisker: Non-Outlier Range 7 8 9 10 11 12 1 2 3 4 5 6 1.0 Month

0.8

0.6 Chlorophytes f(P,N) 0.4 f(P) 0.2 f(N) 0.0 J7 A8 S9 O10 N11 D12 J1 F2 M3 A4 M5 J6 Source: Deniz Özkundakci Month And to New Zealand lakes NZ has some of the most „oligotrophic‟, but also the highly eutrophic lakes in the world

Whakaneke, Waikare, Waiporohita, Rotorua (SI), Ellesmere

Hawea, Coleridge, Pukaki, Benmore, Tekapo, Sumner, Ohau

Abell, Özkundakci, Hamilton n = 2200, c. 40 countries + NZ Lake Rotorua – a history of water quality issues Phytoplankton species change along trophic gradients

Lake Rotoma Lake Rotorua Photo: D. Trolle

Wendy Paul University of Waikato

1000

800

600

400 mg-N/m3 . mg-N/m3 But didn’t the wastewater treatment plant 200

0 (1991) solve the problem? NH4 NO3 TKN

Nitrogen150 concentrations, Ngongotaha 1000 (1) Kaituna 68-70 catchment 800 100 control

76-77 3 - scheme 600 87-89

mg-P/m3 . mg-P/m3 50 400 91-95

mg-N/m3 . mg-N/m3 Ngongotaha Stream

mg N m N mg 02-03 200 0 2 0 DRP TP 1 Total NH4 NO3 TKN NH4 NO3 TKN Streams 150

68-70 100 (2) Advanced WWTP, 76-77 land treatment 87-89 Sewage mg-P/m3 . mg-P/m3 50 91-95 Rotorua WWTP 02-03

0 DRP TP Rutherford et al. (2004) Google Maps Changing land use, Lake Rotorua catchment ‘Old age’ groundwater to Rotorua

Hamurana 110 yr Ngongataha 16 yr Awahou 61 yr Waiowhero 42 yr Lake Rotorua modelling as a decision support tool

ROTAN catchment model Lake model

Climate model High frequency monitoring Inform One scenario: 350 t N/y removed R-0 R-350

No Dairy, less Dry Stock, more Forest, more Lifestyle

Numerous other possible scenarios with the same lake load Temperature: past and future Climate Change Scenario

Lake Rotorua annual mean temperature (oC)

Baseline (1960-1991 average) Climate change projection (2100)*

Groundwater catchment boundary Lake boundary Wei Ye University of Waikato 4-year average N load 900 R0 (without CC) N load (t/yr) 800 R250hd (with CC) N load (t/yr) R300nd (with CC) N load (t/yr) 700 R350nd (with CC) N load (t/yr) 600 R0 (with CC) N load (t/yr)

500

400

Nitrogen load (t/yr) load Nitrogen 300

200

100

1923 1930 1937 1944 1951 1979 1986 1993 2000 2007 2014 2021 2028 2035 2042 2049 2056 2084 2091 2098 1958 1965 1972 2063 2070 2077 Year DYRESM-CAEDYM Changing ‘internal’ loads of nutrients 8000 y = 4.7167x + 4472.5

7500 R² = 0.6302

DW) 1

- Changes in sediment composition 7000 are closely related to the increasing 6500 nutrient (nitrogen) load – a parameter for making future 6000 projections of internal nutrient loads

5500 Sediment TNSediment (mg N kg 5000 0 200 400 600 800 4-year average N load (t yr-1) 9,000

8,000

7,000

6,000 Modelled Observed (Pearson et al.) 5,000

Sediment Sediment TN (mg N/kg DW) 1920 1940 1960 1980 2000 2020 2040 2060 2080 2100 Year Effects of climate change on Trophic Level Index (Target TLI = 4.2)

(A) 5.2 5.05.2 4.85.0 4.64.8

TLI 4.6 4.4 4.4 4.2 4.2 4.04.0 3.83.8 19211921-29-2919711971-79 -7920012001-09 -092031-203139 -392061-692061-209169 -992091-99

R0 R0(no climate change)

(B) 5.2 5.05.2 4.85.0 4.64.8 4.6 TLI 4.4 4.4 4.24.2 4.04.0 3.83.8 19211921-29-29 19711971-79-79 20012001-09-09 20312031-39-3920612061-69 -6920912091-99 -99 R0 R0 (Hamurana diversion) R250 R300 R350 5.2 (C) 5.05.2 4.85.0 4.8 4.6 4.6 4.4 TLI 4.4 4.2 4.2 4.04.0 3.83.8 19211921-29-29 19711971--7979 20012001--0909 20312031-39-39 20612061-69-69 20912091-99-99 R0 R0 (Hamurana diversion) R250hd (Hamurana diversion) R300nd (Hamurana diversion) R350nd (Hamurana diversion) (A) 20% R0 2031-2039 R250 2031-2039 16% R300 2031-2039 R350 2031-2039 12%

8% Frequency

0% 0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5 12.5 (B) 20% Bottom water dissolved oxygen concentration (mg L-1) R0 2061-2069 R250 2061-2069 16% R300 2061-2069 R350 2061-2069 12%

8% ProbabilityFrequency distribution of bottom-water (20 m) dissolved oxygen4% concentrations in Lake Rotorua for the base scenario (R0) and three other external nutrient load 0% scenarios for 2091-2099 (C) 0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5 12.5 20% Bottom water dissolved oxygen concentration (mg L -1) R0 2091-2099 R250 2091-2099 16% R300 2091-2099 R350 2091-2099 12%

8% Frequency

0% 0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5 12.5 Bottom water dissolved oxygen concentration (mg L-1) (A) 40% R0 2031-2039 35% R250 2031-2039 30% R300 2031-2039 R350 2031-2039 25% 20%

Frequency 15% 10% 5% 0% 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 (B) 40% Cyanobacteria concentration (mg m-3) R0 2061-2069 35% R250 2061-2069 30% R300 2061-2069 R350 2061-2069 25% 20%

ProbabilityFrequency 15% distribution of cyanobacteria 10%concentrations in Lake Rotorua 5% for the base scenario (R0) and three other external nutrient load scenarios0% for 2091-2099 (includes climate change) 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 (C) 40% Cyanobacteria concentration (mg m-3) R0 2091-2099 35% R250 2091-2099 30% R300 2091-2099 R350 2091-2099 25% 20%

Frequency 15% 10% 5% 0% 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 Cyanobacteria concentration (mg m-3) (A) 5.2 5.05.2 4.85.0 4.64.8

TLI 4.6 4.4 4.4 4.2 4.2 4.04.0 Effects3.8 of3.8 land use change and inflow diversion on Trophic19211921-29-29 Level19711971-79 -79 2001Index2001-09 -092031 (Target-203139 -392061 -TLI692061-209169 = -4.2)992091-99 R0 R0(no climate change)

(B) 5.2 5.05.2 4.85.0 4.64.8 4.6 TLI 4.4 4.4 4.24.2 4.04.0 3.83.8 19211921-29-29 19711971-79-79 20012001-09-09 20312031-39-3920612061-69 -6920912091-99 -99 R0 R0 (Hamurana diversion) R250 R300 R350 5.2 (C) 5.05.2 4.85.0 4.8 4.6 4.6 4.4 TLI 4.4 4.2 4.2 4.04.0 3.83.8 19211921-29-29 19711971--7979 20012001--0909 20312031-39-39 20612061-69-69 20912091-99-99 R0 R0 (Hamurana diversion) R250hd (Hamurana diversion) R300nd (Hamurana diversion) R350nd (Hamurana diversion) Impacts of climate and nutrient load reductions on cyanobacteria concentrations

Relative change of days with cyanobacteria concentration > 20 mg m-3

Period R0 R0 (noCC) R250 R300 R350

1921-1929 0% 0% 0% 0% 0%

1971-1979 0% 0% 0% 0% 0%

2001-2009 0% 0% 0% 0% 0%

2031-2039 0% -7% -22% -39% -61%

2061-2069 0% -3% -14% -37% -56%

2091-2099 0% -6% -30% -47% -60% Inflows and outflows for Lake Rotoiti: Historical case

To Lake Rotoiti

5 m3 s-1

21 m3 s-1 To Kaituna River outlet

Photo: Environment BOP Landsat-derived chlorophyll a for Rotorua lakes

Allan et al. (2010). Int. J. Remote Sensing

Hamilton et al. (2010). Aquat. Sci.

Strong vertical and horizontal gradients necessitate the application of highly spatially resolved models Implications beyond Lake Rotorua: Ohau Channel diversion wall Image: Kohji Muraoka Inflows and outflows for Lake Rotoiti: Diversion case

Lake Rotoiti

5 m3 s-1 16 m3 s-1

21 m3 s-1

To Kaituna River outlet

Photo: Environment BOP Lake Rotorua – Rotoiti connection

Lake Rotorua Lake Rotoiti Percentage of current Ohau Channel inflow versus cyanobacteria (as μg chl-a L-1) in Lake Rotoiti

14 Model 0% Ohau 12 Model 5% Ohau 10 Model 10% Ohau

[ug/L] Model 50% Ohau 8 a Model 100% Ohau 6

4 chlorophyll 2

0 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03 Jan-04 Jul-04 Jan-05 Jul-05 Jan-06 Photo: Andy Bruere, BOPRC Satellite/aerial views of diversion wall, Lake Rotoiti

Landsat image

Photo: Andy Bruere Lake Rotoiti – pre and post-wall

Kohji Muraoka, Nina von Westernhagen University of Waikato 8/15 Numerical Lake Modelling Kohji Muraoka ([email protected])

3D model: The Estuary and Lake Computer Model (ELCOM) Centre for Water Research, University of Western Australia

Modified zeolite application, Lake Okaro, September 2009

Photo A. Bruere, EBoP

Lake Okaro constructed wetland Photo A. Bruere, EBoP Potential to apply models to major lake ecosystem perturbations: Modified zeolite application to Lake Okaro Modified zeolite application to Lake Okaro, September 2009 Photo M.Gibbs, NIWA Phosphorus concentrations in bottom watersLake of OkaroLake Okaro

1.2

1.0 Alum application 0.8 Constructed 0.6 wetland Zeolite application 0.4

0.2

0.0 13/02/02 28/06/03 09/11/04 24/03/06 06/08/07 18/12/08 02/05/10

DRP; g/m3-P TP; g/m3-P Regime shift in a eutrophic Rotorua lake

Ozkundakci et al. (2010), Ecological Engineering

The Rotorua lakes – trends in water quality

Mann-Kendall non-parametric test for trend (seasonally adjusted time series)

Total nitrogen Total phosphorus

Lake Kendall tau p-level Lake Kendall tau p-level Okareka 0.286 0.001 X Okareka 0.376 0.000 X Ōkaro -0.088 0.276 Ōkaro -0.444 0.000 √ Okataina 0.278 0.001 X Okataina 0.113 0.160 Rerewhakaaitu 0.452 0.000 X Rerewhakaaitu 0.079 0.325 Rotoehu -0.087 0.281 Rotoehu 0.182 0.024 X Rotoiti -0.374 0.000 √ Rotoiti -0.392 0.000 √ Rotoma 0.208 0.009 X Rotoma 0.187 0.020 X Rotomahana 0.209 0.009 X Rotomahana 0.345 0.000 X Rotorua -0.061 0.454 Rotorua -0.379 0.000 √ Tarawera 0.181 0.025 X Tarawera 0.039 0.631 Tikitapu -0.013 0.869 Tikitapu -0.016 0.838 The Rotorua lakes – trends in water quality

Mann-Kendall non-parametric test for trend (seasonally adjusted time series)

Total nitrogen Total phosphorus

Landsat Model

ConclusionsConclusions

––ToThe improve balance trophic of economism,status to the desired scientism level willand (i) taketechnocracy? considerable time and (2) necessitate major, sustained reductions in current nutrient loads; – Many models but some common features: – „Intervention‟ may be required in the form of sediment treatment or flocculation,Science if a more rapid improvementManage inand lake trophic state is desired; implement – “If you don‟t provide us withPolicy the results the planners will make decisions in the absence of –knowledge“If you don’t of environmental provide us with outcomes” the results the planners will make decisions in the absence of knowledge of environmental outcomes”