DevDeveelopmenlopmentt   ofof   NutrienNutrientt   WaWatteerr   QualityQuality StStandarandardsds   foforr   OhioOhio   SurfSurfaceace   WaWatteerrss

USUS   EPEPAA   NutrienNutrientt   TMDLTMDL   WoWorrkksshhoopp,,   FeFebb   1515   –– 17,17,   NeNeww   OrleansOrleans

RoRobbeerrtt   MiltnerMiltner OhioOhio   EnEnvirviroonmennmenttalal   PrPrototectionection   AgAgencyency

      

                        

                      

                        

Nutrient Pollution d i dScope and Magnitude

• Nutrient over‐enrichment is one of the top 5 causes Nutrient over enrichment is one of the top 5 causes of impairments to Ohio streams

• Many inland lakes were posted with advisories for y p harmful algal blooms during 2010

• Resurgence of anoxia in the Western Basin of Lake gErie, plus harmful algal blooms

   

 

   

Percentiles  of  IBI  scores  by  Time  Period • Prior  to  1985  <35%  of  waters  were  meeting  CWA  goal;   1  in  5  chronic  or  acute  toxicity 2009  >  65%  of  water  were  meeting  the  CWA  goal;  toxicity  rarely  observed Achieved  through  regulation  and  investment

••

60ty PERIOD

50

otic

Integrit

1999 1994 1989 1984

30

40

Index of

Bio

2009 2005 1999

CWA Goal

20

Fish

I

Toxic 1984 2009

10 0.00.10.20.30.40.50.60.70.80.91.0

Fraction of Data

       

              

       

            

   

              

                

     

Outline of Ohio Criteria Developmentof Ohio Crit DevelopmentOutline eria

• Observational study tracing effects of nutrients – N t i t t b thi hl hNutrients to benthic chlorophyll ll

• as mediate by canopy cover

– Benthic chlorophyll to dissolved oxygygenp y • 24 hour range

• absolute daily minimum

– Di l d i b d fi Dissolved oxygen to macroinvertebrates and fishh • existing WQS for D.O.

• Identify change points/thresholds at each stepIdentify change points/thresholds at each step – CART with bootstrapping, linear regression

•• Reconcile thresholds with implementation Reconcile thresholds with implementation

Ohio  Nutrient  Study ‐ Bivariate  Representation 1000

10.0

ange

(mg/

l)

Chl

a (m

g/l)

1.0

24 D

O R

a100

Ben

thic

C

10 100 1000 Benthic Chl a (mg/l)

0.1

940

0.10 1.00 10.00 DIN (mg/l)

10

6 7 8 9

DO

(mg/

l)

30

40

Ric

hnes

s

2 3 4 5

Min

imum

D

10

20

EP

T Ta

xa R

0.1 1.0 10.0 24 h DO Range (mg/l)

0 1 2M

0 1 2 3 4 5 6 7 8 9 Minimum DO (mg/l)

0

                      

            

Change Point in Benthic Chlorophyll Over Nitrogen Based on Residual Variation Following Regression toBased on Residual Variation Following Regression to

Canopy and Percent Ag Land Use

0.4

0.5

0 5

1.0Bootstrap Statistics Median = 0.44 mg/l 75th % = 1.09 mg/l 90th % = 1 56 mg/l

0.3

Proporti0 0

0.5

SID

UAL

90 % 1.56 mg/l

0.2

ion per Bar -0.5

0.0

RE

S

0.0

0.1

0 01 0 10 1 00 10 00 -1.0

0.5

0.01 0.10 1.00 10.00 Dissolved Inorganic Nitrogen (mg/l)

                      

       

Change Points and Thresholds Identified in N t i t St d f Ohi Ri d St Nutrient Study of Ohio Rivers and Streams

Drainage Areas < 1000 mi2

DIN TP Chl DO R DO Mi CDIN TP Chl a DO Range DO Min Canopy

(mg/l) (mg/l) (mg/m2) (mg/l) (mg/l) (degree open)

Protection 0.44 0.04 120 6 6 <45

M t 1 0 0 10 183 9 5 45Management 1.0 0.10 183 9 5 <45

       

                

 

            

     

Now the Hard Part ‐ Implementation Now the Hard Part Implementation

• Weight of evidence and independent applicationWeight of evidence and independent application – How to balance the two

– Reasonable potential

• Permits – Averaging period, seasonality, variance, dilution, antidegredation, WQBELs, TBELs

• TMDLs – Model selection, scope, allocations

–

 

                              

                

 

       

Reasonable PotentialReasonable Potential

• Identifying a concentration that represents aIdentifying a concentration that represents a potential for harm against a backdrop of noisy, multivariate relationships – Structural equation modeling

– Logistic regression • requires large data sets ‐ Ohio is fortunate in this regard

– Quantile Regression

– Threshold Indicator Taxa Analysis (TITAN*)Threshold Indicator Taxa Analysis (TITAN )

* Baker and King (2010)

d ect ects

       

 

 

 

                

 

     

 

         

Structural Representation Incorporating all Measured Variables

Photosynthetic Potential

C Nit

Measured Variables

Indirect Effects

Phosphorus

Canopy Nitrogen

Pheophytin Chlorophyll

Agriculture photo‐potential 24h D.O. NH3

EPT Taxa

 ‐

.085

 ‐

.108

 ‐

.085

QHEI% Urban

24 h D.O. Range NH3

Min. D.O.

.44 .25

‐.29

41% of variance explained Number of EPT Taxa

          

    

   

 

 

        

Biological Condition and Nutrient Concentration Rivers and StreamsRivers and Streams

physical habitat quality riparian quality shading

Identified Thresholds

land drainage hydrology precipitation gradient

Excellent

Cond

ition

stream size

Good

Biological

C

Poor

Fair

B

Increasing Nutrient Concentration Forest Clearing Intensive Land Use

              

Community Composition Along a Nutrient Gradient TITAN* (Threshold Indicator Taxa Analysis) TITAN (Threshold Indicator Taxa Analysis)

CP for taxon with decreasing abundance

CP for taxon with increasing abundance 20

CP for taxon with increasing abundance

10

15

core

5

10

Z S

c

0

5

Total Phosphorus mg/l

0

*Baker and King (2010)

          

Quantile Regression EPT T Ri h d T l Ph hEPT Taxa Richness and Total Phosphorus

Reasonable Potential?

2515

20

Ric

hnes

s

10

EP

T Ta

xa R

0 5

1.0 1.5 2.0 2.5 3.0

Log10(TP ug/l)

                          

                        

Logistic Regression b b l f f l h hProbability of an ICI Score < WWH as a Function of Total Phosphorus Concentrations

Available historic data, 1981‐2009. Data culled for NH3>0.1. Drainage area < 500mi2.

1.0

0.8

0.9 ai

nmen

t

0.5

0.6

0.7

of N

on-a

tta

Management Reasonable Potential?

0.3

0.4

0.5

roba

bilit

y o

g Target

Protection

Potential?

0 0

0.1

0.2Pr

0.0

Total Phosphorus

            

           

QQH

EI

Logistic Regression Probability of IBI Meeting WWHProbability of IBI Meeting WWH

Given Habitat Quality and Total Phosphorus

100100

Streams with good to excellenthabitat quality have less potential for impairment due to

90

phosphorus concentrationsexceeding background levels.

80

70

60

50

40

3030

20

10

Total Phosphorus (mg/l)

       

   

   

          

   

                                                           

                                                                                          

                           

Matrix of Reasonable Potential ThresholdsMatrix of Reasonable Potential Thresholds

ICI EPT IBI DIN (quantile) 3.2 3.1 2.91 mg/l

DIN (logistic) NS NS Varies by QHEI1 QHEI1

TP (logistic) Varies by Varies by Varies by QHEI2 QHEI3 QHEI4

TP (quantile) 0.31 0.13 0.159 mg/l

1 For QHEI < median, probability of non‐attainment > 0.5 when DIN> ~ 2.1 mg/l ; when QHEI > 74, probability of non‐attainment > 0.5 when DIN > ~ 9.5 mg/l;

2 For QHEI < median, probability of non‐attainment > 0.5 when TP > ~ 0.6 ‐ 3.0 mg/l 3 For QHEI < median, probability of non‐attainment > 0.5 when TP > ~ 0.3 mg/l 4 For QHEI < median, probability of non‐attainment > 0.5 when TP> ~ 0.159 mg/l ; when QHEI > 74, probability of non‐attainment > 0.5 when TP > ~ 0.457 mg/l

Weight  of  Evidence  Implementation  of Draft  Water  QQ ualityy  Standards

Integrating  Multiple  Benchmarks  &  Thresholds  into  a  Single  Numeric  Scale ‐ The  Trophic  Index  Criterion

   

                               

       

                   

                      

Summary and ConclusionsSummary and Conclusions

• Measurable changes to stream systems occur along aMeasurable changes to stream systems occur along a nutrient gradient – Complexity of relationship precludes adoption of a single numeric criterion and independent application

– Exceeding a threshold or change point does not equate to impairmentimpairment

– Necessitates inclusion of response indicators (e.g., benthic chlorophyll, dissolved oxygen) in standard

 

                      

       

 

   

 

     

Final ThoughtsFinal Thoughts

• Significant restoration of surface waters has occurred siince ththe 1980s1980 – Toxins and organic matter

– Regulation Regulation

– Funding

• Remaining problems are comparatively intractableRemaining problems are comparatively intractable – Nonpoint

– Habitat,, flow

– Non‐toxic (e.g., nutrients)

– Less funding

– Little or no regulation

             Nutrient Loads ‐ In Large Part a Landscape Issue Nutrient Loads In Large Part a Landscape Issue