Sources of Bacteria and their Variability in Urban Watersheds

Robert PittCudworth Professor Urban Water Systems

Department of Civil, Construction, and Environmental Engineering

The University of Alabama1

Observed Bacteria Data in the National Stormwater Quality Database, version 3 (average and COV)

Land Use (number of observations)

Fecal Coliform (CFU/100 mL)

Overall (2,154) 47,700 (5.0)Residential (505) 55,900 (5.7)Commercial (270) 26,100 (3.0)Industrial (317) 47,300 (6.1)Freeways (67) 8,600 (2.7)Open Space (7) 7,300 (1.2)

Very large variabilities and widely ranging numbers of samples2

Experimental Design - Number of Samples Needed

The number of samples needed to characterize stormwater conditions for a specific site is dependent on the COV and allowable error. For fecal coliforms, the COV values are very large (1.2 to 6.1 in the NSQD). A typical goal of 25% allowable errors would require hundreds of samples, unless a better understanding/explanation of the variability was possible. Burton and Pitt 2002

3

Fecal Coliform Bacteria by Land Use Categories

Commercial, Freeways, Industrial, Institutional, Open Space, and Residential

FC (m

pn/1

00 m

L)

1e-1

1e+0

1e+1

1e+2

1e+3

1e+4

1e+5

1e+6

1e+7

1e+8NSQD Version 3

25th to 75th percentiles of most FC monitoring locations are between 1,000 and 20,000 MPN/100 mL; Statistically, transportation lowest and residential areas highest, but wide range seen

As noted before, the differences between land uses may be somewhat helpful (freeways and open space lower than others, but limited data). Even though large overall ranges in each land use, most data are within two orders of magnitude (the 25 to 75 percentile ranges from the high hundreds to the low tens of thousands of CFU/100 mL, and most observations well above any criteria).

4

Residential Fecal Coliforms vs. Rain Depth

Not much of a trend here….

5

Monitoring of 0.4 ha Parking Lot and Surrounding Landscaped Area, Tuscaloosa, AL

Definite “first flush” of turbidity at small source area

Bacteria generally increased during rain

Bacteria also increasing during this rain event

6

Strong seasonal influences on bacteria levels, especially when comparing snowmelt with stormwater

7

Tuscaloosa, AL, “Library” File Data (flow components affect combined water quality)

Mean/COV E. Coli(CFU/100 mL)

Enterococci(CFU/100 mL)

Tap water 0 (0)

0 (0)

Spring water 2.4 (0.8)

1.0 (1.6)

Car wash water

1480 (0.07)

1213 (1.4)

Industrial wastewater

409 (2.7)

477 (2.3)

Potential baseflow components during warm weather 8

Roof runoff E. coli levels greatly affected by overstory trees that provide habitat for urban wildlife (squirrels and birds)

9

1.00E+00

1.00E+01

1.00E+02

1.00E+03

1.00E+04

1.00E+05

0 2 4 6 8 10 12 14 16Time (days)

Cel

ls/m

l

E. ColiEnterococciBackground ECBackground Ent.k = 0.18 day-1

k = 0.09 day-1

Easton, et al. 1999

Fate of bacteria after discharge investigated in receiving waters during several studies:

Die-off curves for E. coli and Enterococci during in-situ microcosm tests

10

In-situ microcosm tests of die-off curves for Giardia lamblia in urban receiving waters

0

2000

4000

6000

8000

10000

12000

14000

0 1 3 7 10 14Time (days)

Cys

ts/L

IntactDegradedTotal

k = 0.11 day-1

Easton, et al. 199911

Fate of Bacteria after Discharge:

Interstitial Water Bacteria Levels using Peepers

Ellis, et al. 1998

Delrin peepers with 7 mL cavities every 1 cm; 75-micrometer screening and 2 hr equilibrium time

Peeper being pushed partway into sediments

Bacteria levels increase near sediment interface and are elevated in interstitial water.

12

Full Factorial Experiment to Investigate Bacteria Survival on Pavement and Urban Soils: Moisture,

temperature, and light vs. time

Wilson 201113

E. coli survival on concrete for different environmental conditions

Warm, wet, and dark moderate die-off rate

Other conditions have initial rapid die-off rate, then moderate re-growth, then moderate dieoff 14

Enterococci survival on concrete for different environmental conditions

Warm, wet, and dark steady-state population

Other conditions have initial rapid die-off rate, then moderate re-growth 15

Conclusions• Stormwater bacterial levels are high and variable, requiring many

storm samples to confidently quantify.• Some differences by land use (as expected, most open space areas

have lower bacteria levels than developed areas).• Seasonal variations very important (especially comparing stormwater

with snowmelt)• Many factors affect stormwater bacteria levels: contamination by

industrial or sewage and urban wildlife (and their habitats).• Upon discharge, rapid die-off then reach equilibrium conditions.• Sedimentation with increased populations near sediment/water

interface and elevated interstitial water bacteria levels.• Survival on concrete is similar, with potential several growth and die-

off stages. Warm, dark, and moist conditions result in their greatest survival.

16

R. Pitt17