A Chemical Study of Mill Creek: Biological Oxygen Demand, Septic Tanks, and Forest Fires.

Elliot Anders

Whitman College




A year-long study of the water chemistry of Mill Creek was conducted to generate baseline data on the creek for further use.  Study parameters included: pH, turbidity, temperature, conductivity, dissolved oxygen, limiting nutrient, and biological oxygen demand.  It was found that none of the parameters varied significantly from month to month, that the creek was very clean in terms of the parameters studied, and that the limiting nutrient for each site was phosphorous.  Data concerning stream flow was also collected from the Walla Walla District of the United States Army Corps of Engineers and is included in Appendix A.  Conclusions from the data were drawn regarding the possible effects of the introduction of fire into the Walla Walla Municipal Watershed, and how local septic tank effluent might influence these effects.




The summer of 2000 was said to be “undoubtedly one of the most challenging on record. . . . As early as September, more than 6.5 million acres . . . ha[d] burned” (Managing, 2000) in the United States, and “As early as October, more than 6.8 million acres of public and private lands had burned” (Laverty and Williams, 2000).  The City of Walla Walla was very lucky during this last fire season.  One of our most precious local resources, the watershed that supplies Walla Walla its drinking water, managed to escape completely unscathed into the wetter portion of the year. 

Contamination, fire, and mismanagement are a constant threat in watersheds.  The Walla Walla Municipal Watershed has been well protected, both from contamination and fire.  The watershed itself is completely closed to entry without permit (USDA, 1918) and well watched during the fire season by the Table Rock fire lookout.  The most threatening problem the Walla Walla watershed faces is not a natural or accidental disaster, but a managerial miscalculation. The Walla Walla District of the US Forest Service, has been extinguishing all flame in the watershed since the adoption of an ordinance in 1918.  Now, 83 years later, the management approach is up for re-evaluation, and initial suggestions will require drastic management changes.

The changes in management could take many forms, but at the forefront of suggestions at this time is a proposal written by a fire expert of the Walla Walla District (USFS), Jim Beekman.  The main point of his proposal is to begin the incorporation of managed fire in the Walla Walla watershed.

This thesis completes some of the research that will be necessary if changes in management practices do occur in the Walla Walla watershed.  It does not try to guess what management changes will actually be implemented (if any), but aims to provide a useful background data set that could be used in comparing water quality analysis after a change in management has occurred. 

The scope of the research presented here covers two changes that are likely to occur over the next few years.  The two parameters of interest are the introduction of fire into the Walla Walla watershed and an increase in human population living close to the banks of Mill Creek below the watershed boundary.  In the following sections, I will illuminate the environmental consequences of both parameters, and more importantly, the environmental consequences of the synergistic effect of both parameters occurring at the same time.

This research comes at a timely point in the watershed history for multiple reasons.  As already mentioned, this is a time of management change; the Forest Service realizes that a “catastrophic fire” (Beekman, conversation, 11/2/00) may occur in the watershed in the next few years if management to prevent such an event does not begin soon.  In addition, the watershed had an extremely dry year in 2000, but through luck suffered no fire damage. Because fire was absent in the past year, it is unlikely that Mill Creek, fed by the watershed, saw much change in chemistry due to changes in the watershed caused by fire, and therefore it was a good year to create a baseline that attempts to show the water quality when unaffected by fire.  Finally, the population of the Kooskooskie area is visibly encroaching on the creek, leaving little room for domestic liquid waste disposal between house and creek.  This research provides a background data set that could link the water quality of runoff from the Walla Walla Municipal Watershed with the occurrence of forest fires, and in-stream nutrient levels with leaking wastewater systems (cesspools, septic tanks, and drain fields).

While this report will make no attempt to guess at what motivates people to live in a flood plain or to fail to change failing management practices in a timely fashion, it will make one sociologic prediction: the damage from the flood of 1996 has almost completely disappeared, and it is my belief that as that passes further into the past of people’s memories, they may begin to move to and develop the Kooskooskie area once again. 

Background chemistry:

In this section, the chemical parameters used in this study will be presented to show the relevance and importance of each parameter.


The pH of natural waters can vary greatly across the globe, but in northeastern Oregon and southeastern Washington pH generally resides between 6.5 and 8.5 (Watershed, 1999).  The pH of natural water depends partly on solution of CO2, which is dependent partly on temperature with lower temperature waters being capable of holding more CO2.  The concentration of CO2 also depends on the presence or absence of CO32- in surrounding rocks.  The first reaction is represented by:

CO2 + H2O ----> H2CO3 ----> H+ + HCO3-

Other reactions also affect the pH of natural waters; these include reactions with acids, and reactions with bases.  The most common acid reactions involve H2SO4 or HNO3 from acid rain. Reactions that can increase pH often involve the neutralization of acidic waters with local minerals such as limestone (CaCO3).

Many anthropogenic activities can also alter stream pH.  The most common human-caused pH-altering phenomenon is probably acid rain, and other forms of acid precipitation, mostly in the form of H2SO4 and HNO3 from coal burning (SO2) and vehicle emissions (NOx) respectively.  In addition, the input of ammonia-based fertilizers can also alter the pH significantly.  The environmental consequences of changing stream pH ranges greatly.  Most organisms can survive in a pH range from roughly 5 to 9, depending on species and adaptability. 

Mill Creek is a relatively uncontaminated ecosystem, and as a result tends to have a relatively constant pH. However, in the future pH values might increase with the addition of basic ash from forest fires.  Further down the creek the addition of fertilizers, detergents and the like could also affect pH values.  This baseline pH study simply provides one more parameter to the list of parameters for determining stream health over time. In situ pH measurements were conducted monthly.  Results are shown in Table 1 (Data Section).







Turbidity is a slightly outdated, but still very useful, measure of total suspended solids (TSS).  The TSS is usually measured by filtration of a known water volume, dehydration of the filtered media, and weighing of the remaining solids.  Turbidity measurements are a photometric method of measuring the opacity of water samples, and relating them to the suspended material in the water.  In Walla Walla, the turbidity of the water coming directly out of the watershed determines the usability of the water for consumption throughout the city.  When the turbidity reaches 2 NTU the city switches over to well water to service residents (Water Quality, 2001). 

Turbidity is effected only by an increase in sediment or other particulate matter in the water.  For the purpose of this research project, the turbidity also represents the added expense Walla Walla might incur when fire or logging begins in the watershed.  Every day that Walla Walla has to use groundwater shortens the time until its underground aquifer is depleted.  Walla Walla may be able to use the Mill Creek watershed for water throughout the year if it installed a filtration plant in-line with the current water treatment plant, but the expense involved has been prohibitive thus far, and in dryer years the City could encounter problems maintaining the minimum in-stream flow required by law. In situ turbidity measurements were conducted monthly.  Results are shown in Table 1 (Data Section).



Dissolved Oxygen:

Dissolved Oxygen (DO) is another useful measure of stream health because most organisms require some amount of DO to survive. The saturation level of DO in a stream at 10.0°C is roughly 11 mg/L (Sawyer and McCarthy, 1978).  DO is caused by photosynthesis of aquatic plants, diffusion of oxygen into the water, and most importantly through turbulent mixing between the stream water and atmosphere.  More important than how DO gets into water is how fast it is consumed.  Decomposition of organic material, minimal consumption by oxygen consuming organisms, and consumption by nitrogen based compounds all contribute to DO consumption.  If the DO drops close to zero mg/L the body of water is considered anaerobic, and oxygen-consuming organisms begin to die. In dying, the plants decompose and consume more oxygen.

The problems associated with this consumption of DO are two-fold.  First, any organisms that depend on oxygen for survival, no longer have an oxygen source.  Second, the influx of nutrients released back into the stream or lake provides an ideal situation for a similar cycle to begin again.  This discussion will arise again in this paper in the section on limiting nutrients. In situ DO measurements were conducted monthly.  Results are shown in Table 1 (Data Section).



Temperature was measured using the same instrument and probe as for dissolved oxygen.  Similar to conductivity, temperature of stream waters can vary through a wide range before significant effects are seen in stream habitat.  It is known that fish can survive a change in water temperature of a few degrees, but if shocked too quickly they may not recover.  For the purpose of this project, temperature serves solely as a measure of change from month to month and year to year.  There are many factors that effect the temperature of Mill Creek; most likely a decrease in plant cover throughout the watershed after a catastrophic wildfire would raise stream temperature significantly. In situ temperature measurements were conducted monthly.  Results are shown in Table 1 (Data Section).


Biological Oxygen Demand:

Biological oxygen demand (BOD) is one of chemistry’s most widely used measures of stream health.  The procedure is simple yet the implications of the results can be drastic.  The most common use of BOD is at sewage treatment facilities to measure the amount of sewage entering and leaving the facility.  The measurement is simply the change in dissolved oxygen over a period of time in a water sample at constant temperature.  BOD speaks to the availability of oxygen in the water for aquatic species. 

When organic matter such as dead plants, leaves, grass clippings, manure, sewage, or even food waste is present in a water supply, bacteria will begin the process of breaking down this waste. When this happens, some, or all, of the available dissolved oxygen is consumed by aerobic bacteria, robbing other aquatic organisms of the oxygen they need to live (stevens-tech.edu, 2000).

There are many methods of measuring BOD, ranging from complex methods such as the Winkler titration to the more simple electrode analysis (Greenberg et al., 1992).  The two most common methods of BOD analysis are outlined in the EPA Standard Methods for Water and Waste Water Analysis (1992).  These two methods are listed under Method 4500-0 Oxygen (Dissolved)(DO) (Greenberg et al., 1992).  These methods simply measure dissolved oxygen, and are the basis for the 5-day BOD test and Ultimate BOD test listed under method 5210 B and C (Greenberg et al., 1992). 

Because of the availability and speed of a dissolved oxygen electrode, it was used in this research as described in Method 4500-0 G Membrane Electrode Method (Greenberg et al., 1992).  The membrane electrode method has a few advantages over some of the titration methods.  Most notably, the procedure is much more simple and expedient.  In addition, the same sample can be used for both a BOD5 and ultimate BOD measurements. 

Measurements of biological oxygen demand describe the ability and speed of organic matter and microorganisms in water to consume dissolved oxygen.  Minear (1984) describes the theory behind BOD measurement very simply in the following reaction:


Organic Material + O2 ---------------------> CO2 + H2O + microbiological solids

Organic material deposited in rivers or lakes is decomposed by native microorganisms and in the process consumes available oxygen.  The amount of oxygen they decompose is called the biological oxygen demand (BOD). In general the BOD is measured over a period of time, and most often this period is 5 or 20 days.  These two measurements are called BOD5 and BODu (ultimate or theoretical BOD) respectively.  In this project both the 5-day BOD analysis and a modified BODu were used.  BOD5 was used to measure the amount of decomposable organic matter in Mill Creek, and BODu was used to determine the limiting nutrient and ultimate BOD. 

The electrode analysis of BOD5 is a good tool for determining stream health for two reasons.  First, it is a relatively quick process, providing results after 5 days.  Second, electrode analysis of BOD measured is very close to the BOD in the stream and is non-destructive.  Some other methods mistakenly include extra oxygen demand from non-organic sources because of side reactions with chemicals used in determining the BOD nitrogen based compounds are especially notorious for this phenomenon called chemical oxygen demand.  The five-day analysis avoids introduction of nitrogenous oxygen demand as shown in Figure 1. 













Figure 1 – Generic BOD vs Time curve

(Minear, 1984)


For the purposes of this project, BOD analysis provides a good evaluation of change because very little BOD is present naturally in Mill Creek.  This means that an increase in BOD after management practices change could mean a significant change in stream health and would be easy to detect.  It is my belief that if fire is introduced, or burns naturally through the watershed, than the increased nutrients and carbon in the run-off alone could significantly increase the BOD levels in Mill Creek.  Further exploration of this thought will be discussed in the next section.




The implications of a fire in the Walla Walla watershed are widespread.  On an environmental level, the health of Mill Creek comes into question in the form of changes in water quantity and quality after a fire.  The effects on the depletion of the aquifer alone, because of an increased use of well water in place of surface water, are large enough to warrant a thorough study of potential impacts of such a fire. 

Even if a catastrophic fire never occurs in the watershed, changes in management might include the introduction of fire, or logging, both of which have been absent for years and could have effects similar to a catastrophic fire.  Whether natural or human caused, any changes to the watershed ecosystem will upset the current ability of the watershed to purify our local water supply and could cause catastrophic effects in terms of our ability to use the municipal watershed in the future without the addition of a water filtration plant.

In one long-term study of fire-induced ecosystem changes, Bormann and Likens (1979) conducted research on an ecosystem in the White Mountains (Hubbard Brook Watershed).  Their research focused on the ecosystem wide effects of clear-cutting an ecosystem but is relevant to possible changes due to fire as well.  They began by taking baseline measurements and then cutting areas and leaving all the debris on the ground.  With this type of cutting it is easy to draw the similarities between clear-cutting and forest fire. 

In attempting to create a steady state map of inputs and outputs, they found that “during early regrowth after clear-cutting net removal by grazing animals may be a factor of some importance” (Bormann and Likens, 1979).  This regrowth or “aggradation phase” begins roughly “15 years after clear-cutting, when the total biomass curve . . . shifts from net loss to net gain” (Bormann and Likens, 1979).  In the Mill Creek watershed, recovery times of this length could have catastrophic effects on Walla Walla’s ability to consume runoff water directly.  There is a relatively large elk population that lives within the watershed, and if fire wiped out a large portion of the vegetation, it is possible that the elk population could hinder its regrowth, by eating new plants as quickly as they grow.  Also, if the elk were to die because of lack of food, there might be enough biomass and bacterial growth to further contaminate the water supply.   At this point, not only is it obvious that a set of baseline data needs to be collected on Mill Creek before any changes are made, but also that any changes made should be considered very carefully in terms of their immediate effects and their long term consequences.

Environmental Impact Statements (EIS) are necessities now before any changes in government policy is implemented. In this case the Walla Walla District of the United States Forest Service will be required to finish an EIS and associated data collection before it can move forward with a management change regarding the Walla Walla watershed.  Jim Beekman and Tom Wordell of the Walla Walla District of the USFS both noted the pressing need for changing the management of the watershed in recent interviews, and the need for an EIS before these changes can take place.  Creating an EIS can be a lengthy process, and thus it is my hope that this project can compose part of the EIS when it is eventually put together. 

As Bormann and Likens (1977) note in their experimental clear-cutting, parameters such as soil erosion should be studied and quantified before initiating change.  “To simplify this situation, we decided to limit our discussion to one starting point, a northern hardwood stand that had not been clear-cut. . . .  in fashioning our ideas on secondary development we limited our consideration to clear-cut sites with relatively little erosion . . . to evaluate the effect of forest destruction in the absence of mechanical disruption” (Bormann and Likens, 1977).  This focus on finding undisturbed sites is very important in ecosystem and environmental analysis since it is similar to running a blank reaction in the lab.  The goal of this project was exactly that, to create a blank for further studies on Mill Creek water quality.

This thesis suggests that there is a close link between wastewater systems and biological oxygen demand in Mill Creek.  Thus far, I have covered the purpose of a BOD baseline analysis for Mill Creek and how these levels might change if nutrient concentrations increased.  As shown in the limiting nutrient analysis, phosphorous is the limiting nutrient in Mill Creek, and if more phosphorous found its way into the creek, we could see a significant increase in BOD levels.  The direct connection between Mill Creek and wastewater systems is visible in Figure 2.












Figure 2 – Some homes in Kooskooskie are immediately adjacent to Mill Creek.

The small town of Kooskooskie, Washington sits literally on the creek bank, and any effluent exiting wastewater systems would drain almost directly into the creek providing a source of nutrients and BOD.  If, in addition to the influx of nutrients from Kooskooskie, the watershed also burned and dumped a large amount of organic material and nutrient rich ash into the creek, there could be a large algal bloom, and an ensuing eutrophication of the creek.  The environmental implications of such a catastrophic change are large.  This brings us full circle, back to why this baseline research is necessary, and to my hypothesis, which is that a baseline study needs to be completed quickly, before the fire season of 2001, so that changes can be studied after the watershed is changed by fire.

Materials and Methods:


Research for this thesis was conducted from the summer of 2000 through March of 2001.  It consisted of an almost year-long monitoring of specific water quality characteristics of Mill Creek between the lower edge of the watershed and the Walla Walla wastewater treatment plant.  Parameters monitored on a monthly basis were pH, turbidity, dissolved oxygen, conductivity, temperature, and especially biological oxygen demand.

Conclusions throughout this thesis will be drawn from this research regarding the local environmental consequences of fire, both managed and catastrophic, in the Walla Walla Municipal Watershed, and from increases of population near the banks of Mill Creek. 

The greater part of this thesis will focus on biological oxygen demand (BOD) as an indicator of upstream changes; in addition other water quality parameters will be used to monitor the background chemical stability of Mill Creek. 

The Sites:

Five sites were chosen along Mill Creek for BOD and chemical analyses.  The sites were spaced from the base of the watershed to a site just below the Walla Walla waste treatment plant.  The five sites were located: (1) just above the confluence of Mill Creek and Tiger Canyon roughly 1/4 mile from the base of the watershed, (2) just above the bridge in Kooskooskie roughly four miles downstream from the watershed, (3) on the downstream side of the Five-Mile Road bridge roughly 13 miles downstream from the watershed, (4) at the Yellowhawk Creek branch from Mill Creek at the Army Corps office roughly 17 miles downstream from the watershed, and (5) on the upstream side of the bridge just downstream from the Walla Walla waste treatment plant roughly 21 miles downstream from the watershed (Figure 3). The sites at Yellowhawk Creek (Army Corps) and at the Five-Mile Road bridge were chosen because of USGS monitoring stations located at those sites, that collect daily stream flow data, located in Appendix 1.  This data will become significant when future studies of the creek chemistry are done for comparison of baseline data. 

Figure 3 – Locations of the five sampling sites, including the entire watershed.

Figure 3






The pH of Mill Creek was measured using an Orion 290 pH meter.  The probe was calibrated using pH 4 and pH 7 buffered solutions, and measurements were taken until the meter reached a stable measurement, per the instructions provide with the probe.


Turbidity was measured using a HACH 2100P turbidimeter using the standard calibration procedure, and sampling technique as described in the HACH instruction manual.

Dissolved Oxygen:

Dissolved oxygen (DO) was measured using a YSI 55 dissolved oxygen type DO meter.  The probe was allowed to equilibrate for 2 minutes or until the reading became constant. 


Conductivity is a measure of the dissolved ions ion a water sample.  Generally, fresh water conductivity is relatively low, and not of consequence.  It has been included here simply as another means of generating data for a base line for comparison in future studies. 

Conductivity was measured using a HACH CO150 conductivity meter, and was allowed to equilibrate for 2 minutes or until the reading became constant.


In this project, each of the five sites were analyzed separately and in duplicate.  The first site was assumed to have the least BOD because of its placement at the base of the watershed; it was tested both with and without added nutrients, for comparison.  All other sites were tested with additional nutrients, as described in method 5210 B (Greenberg et al., 1992).  Each sample from each site was given 1 mL 0.018N ferric chloride solution, 1 mL 0.25N calcium chloride solution, 1 mL of 1.5N phosphate buffer solution, and 1 mL 0.41N magnesium sulfate solution, per liter of sample.  The samples were then placed in 300 mL BOD bottles. The dissolved oxygen was measured with an Orion 290 A pH meter and an Orion 97-08-99 oxygen electrode.  The bottles were sealed with water and parafilm and set in a 20 ± 0.1°C incubator for the next 5 days for 5-day BOD analysis, and 20 days for ultimate BOD and limiting nutrient analysis.  For 5-day BOD analysis, the samples were tested again for DO after 5 days.  For Ultimate BOD and Limiting Nutrient analysis the samples were tested again after 1 day, 2 days, 5 days, 10 days, and 20 days.



Phosphate buffer solution, 1.5N: Dissolve 207 g sodium dihydrogen phosphate, NaH2PO4•H2O, in water.  Neutralize to pH 7.2 with 6N KOH and dilute to 1 L.

Ammonium chloride solution, 0.71N: Dissolve 38.2 g ammonium chloride, NH4Cl, in water.  Neutralize to pH 7.0 with KOH.  Dilute to 1.0L; 1mL = 10 mg N.

Calcium chloride solution,  0.25N: Dissolve 38.2 g CaCl2 in water and dilute to 1.0L; 1 mL = 10 mg Ca.

Magnesium sulfate solution, 0.41N: Dissolve 101 g MgSO4•7H2O in water and dilute to 1L; 1 mL = 10 mg Mg.

Ferric chloride solution,  0.018N: Dissolve 4.84 g FeCl3•6H2O in water and dilute to 1 L; 1 mL = 1.0 mg Fe.

Potassium hydroxide solution, 6N: Dissolve 336 g KOH in about 700 mL water and dilute to 1 L.

Acid solutions, 1N: Add 28 mL conc. H2SO4 or 83 mL conc. HCl to about 700 mL water, Dilute to 1 L.

Alkali solution, 1N: Add 40 g NaOH to 700 mL water. Dilute to 1 L.

(from Method 5210 D) (Greenberg et al.,1992).


Two BOD bottles per site. 

An Orion 290 A pH meter and an Orion 97-08-99 oxygen electrode (or equivalent DO probe).

Parafilm or plastic wrap and rubber bands.


Procedure for BOD5:

Add 1 mL each, phosphate buffer, FeCl2 solution, CaCl2 solution, and MgSO4 solution per liter of sample.

Fill an airtight bottle with nutrient rich sample to overflowing.

Measure dissolved oxygen with a DO probe.

Seal bottle, fill to overflowing, drop in the ground glass lid, make sure to remove all air bubbles, fill reservoir around lid with more sample if necessary, cover with plastic wrap or parafilm seal.

Incubate at 20°C ± 1°C for 5 days.

Measure DO with a DO probe again after 5 days.

 Blanks should be made for at least one site, follow steps 2 through 6 without the addition of nutrients.

Procedure for Ultimate BOD:

Fill an airtight bottle with sample to overflowing.

Measure DO with a DO probe.

Seal bottle as outlined in BOD5 procedure.

Incubate at 20°C ± 1°C.

Measure DO again with a DO probe after 1, 2, 5, 10, 15, and 20 days.

Procedure for Limiting nutrient:

Add 1 mL of phosphate buffer, FeCl2 solution, CaCl2 solution, and MgSO4 solution, to separate 1 Liter portions of sample. 

Fill one airtight bottle with each solution and one with pure sample (no additives).

Measure DO with a DO probe.

Seal Bottles  as outlined in BOD5 procedure.

Incubate at 20°C ± 1°C.

Measure DO again after 1, 2, 5, 10, 15, and 20 days.

All light was excluded from all samples during incubation.  Samples were kept no longer than 2 hours from collection at room temperature or 6 hours at 4°C.  Samples were brought to 20±3°C before analysis, and were shaken vigorously to saturate with oxygen before the first DO measurement.  When chlorine contamination was suspected, samples were placed in a window while they were brought up to temperature, and were swirled various times to allow chlorine gas to escape or photodissociate.  All procedures were taken and modified from Greenberg et al. (1992).


The data collected for this project falls into three main categories.  The first category is the baseline data collected for future comparison studies; this includes pH, turbidity, DO, conductivity, and temperature, and is summarized in Table 1. The second catagory, is the BOD related data, BOD5, BODu, and Limiting Nutrient calculations; these are included in Tables  2 and 3 and are analyzed in Figures 4 - 9.  Finally, the water level data collected from the US Army Corps of Engineers is included in Appendix 1, for comparison in future years; this last data category will remain undiscussed, and unchanged from its original format.

Parameter data:

MC = Mill Creek just above the Tiger Canyon confluence.

Koos = Kooskooskie site, at the bridge at Kooskooskie road.

Five-Mile = Five-Mile Bridge site.

Army = Army Corps site, at the Yellowhawk creek Garrison / diversion.

WWTP = Walla Walla Wastewater Treatment Plant just below the treatment facility.

Key to all Data that follows:











BOD5 Data

The first noticeable trait of Figure 4 (BOD5 data) is the missing data on 8/26 from the first site (MC).  This is missing because that was the month that the limiting nutrient study was done, and data were collected for each individual nutrient but in the same sample.  Also, note that all of the BOD5 concentrations are relatively small, approximately 1 or 2 mg/L.  Typical sewage water can range up to 150 mg/L, thus showing that Mill Creek is quite clean.






BODu Data

The ultimate BOD for all five sites along Mill Creek (Table 2) is quite low, in fact the BODu is only slightly higher than the BOD5, suggesting that most of the decomposition of organic material is done within the first five days, and that during the remaining 15 days in the BODu period, very little oxygen consumption is occurring.  The site below the wastewater treatment plant had the largest BOD, for both the five-day and ultimate tests, but still remained well inside the category of a healthy stream.  In an attempt to explain the higher BOD for site 5, an influx of nutrients from the waste treatment plant and perhaps the farms nearby may be increasing the BOD by providing more of the limiting nutrient and allowing the microorganisms that decompose organic material to decompose more material already present in the water before dying.  My reason for this conclusion comes from the Limiting nutrient data in Table 3 (Limiting Nutrient data).

Table 2 – Ultimate BOD



mg/L BOD





















Limiting Nutrient Data

The data analysis presented below shows that phosphorous is the limiting nutrient in all five sites.  The wastewater treatment plant site showed the most significant increase when extra phosphorous was added (Table 3).  This suggests that there was more BOD in this water than at the other four sites, and that even though the ultimate BOD of this site is already higher than at the other sites, there is still a significant amount of undecomposed organic matter at that site than at the others.

Table 3 – Limiting Nutrient Analysis


















MC - no nut








MC - P








MC - FeCl2








MC - CaCl2








MC - MgSO4








Koos - no nut








Koos - P








Koos - FeCl2








Koos - CaCl2








Koos - MgSO4








5Mile - no nut








5Mile - P








5Mile - FeCl2








5Mile - CaCl2








5Mile - MgSO4








Army - no nut








Army - P








Army - FeCl2








Army - CaCl2








Army - MgSO4








WWTP - no nut
















WWTP - FeCl2








WWTP - CaCl2
















Below, in Figures 5 through 7, are plots of the modified BODu experiment for determining the limiting nutrient.  Notice how on each plot, that the phosphorous enhanced sample significantly exceeded the other samples in BOD.  Less remaining oxygen implies more BOD.



Finally, I have included the discharge data from the US Army Corps of Engineers Walla Walla district.  This includes water flow measurements from the Yellowhawk Garrison site (Army Corps site), the Five-Mile bridge site, and from the bridge in Kooskooskie. It is included because after a fire in the watershed the amount of runoff may increase (or decrease), and it will be interesting to see the changes in comparison to the changes that occur in the BOD as well.

Rate constant calculation:

In addition to calculating the limiting nutrient and ultimate BOD of Mill Creek, a rate constant or ‘k’ value was calculated for the microbial rate of degradation of organic matter.  BOD is a pseudo first-order process (Dunnivant, 1997) in which the concentration of dissolved oxygen can be estimated at any time t by using the Thomas Slope method:

L = Lo e-kt

Therefore, k can be calculated as follows:

k = -(ln(L/ Lo)/t

For Mill Creek, ks were calculated using the Thomas Slope method on the nutrient free BODu measurements, and were averaged over the 20 day period, as shown in Table 4

Table 4 – Rate Constants for Mill Creek.





Mill Creek




5 Mile


Army Corps





The calculated values for k (Table 4) are close to the accepted value of 0.17 (Dunnivant, 1997).  This relation simply implies that Mill Creek re-aerates at roughly an average rate, showing that it is a healthy creek.  In addition to k being near the accepted value, plots of linearized Thomas slope method calculations are almost linear as theoretically demonstrated by:

(t/y)(1/3) = (Lok)-(1/3) + (k(2/3)/6Lo(1/3))t

where t is time, y is the BOD in mg/L at time t, Lo is the ultimate concentration of biodegradable organic matter, and k is the rate constant (Dunnivant, 1997) (Figure 10).


In determining the limiting nutrient for BOD in Mill Creek an important discovery was made.  The limiting nutrient is phosphorous (Figure 9, limiting nutrient for Site 5). This discovery is important because it implies that if the watershed were to burn over and donate a significant amount of phosphorous and other nutrients to the creek, BOD might increase significantly.  Bormann and Likens (1979) found that after clear-cutting the Hubbard Brook watershed:

The highly predictable relationship between streamflow and concentrations of dissolved substances that is characteristic of the aggrading forest ecosystem was rapidly and markedly altered by devegetation. . . . For most ions, stream-water concentrations reached a peak during the second year after cutting and declined during the third year.

(Bormann and Likens, 1979)


If anion levels were to increase significantly, the creek would see a large jump in BOD, and possibly a subsequent eutrification in less aerated areas.  During the limiting nutrient tests on Mill Creek it was found that the addition of any nutrients (Fe, Mg, P, NH3) increased the BOD significantly over normal levels, but the addition of phosphorous could drive the BOD up as much as 4 mg/L or roughly 40 – 50%, while most of Mill Creek is well aerated and could handle a spike in BOD, there are portions that might become significantly choked with plant growth if supplied with enough nutrients.   In addition, in the Bormann and Likens (1977) study the debris from the clear-cutting was left to decompose on the ground, allowing nutrients to seep slowly into the environment.  In the case of a fire burning through the Walla Walla watershed, the debris might be deposited directly into the creek, or might wash in more quickly than larger woody debris.  This quick influx of nutrients and carbon could have two consequences Bormann and Likens (1979) didn’t experience.  First, the influx of ash, and carbon-based material directly into the stream could provide even more nutrients than Bormann and Likens (1979) observed, and in addition could also provide even more carbon-based material for decomposition, essentially multiplying the BOD inputs.  Second, the influx of nutrients and carbon might reach the stream significantly faster than it did in Bormann and Liken’s (1979) study.  The ash and charcoal, being more mobile than the woody clippings, would not only increase turbidity in the creek, but would also reach the stream more quickly.  This means that while the Hubbard Brook watershed took two years to peak for nutrient influx, the Walla Walla watershed might take much less time if a catastrophic fire were to burn through it.  It might peak within the first year, and drop off much more quickly, thus allowing for faster recovery, but it might also flood the creek with more carbon and nutrients than it can decompose initially, and might end up clogging the creek more than if the watershed were just clear-cut. 

Additional Research:

One of the next steps in characterizing Mill Creek would be to sample the stream from just below the watershed boundary to just above the Yellowhawk-Garrison diversion, taking samples every 100-300 meters, and analyzing them for total organic carbon (TOC).  The TOC measurements would give an estimate of the total amount of organic matter that is currently in Mill Creek, and measuring every 100-300 meters would enable identification of sites of gross organic contamination.  Agricultural runoff, septic tank leakage, organic chemical dumping, and ranch waste runoff would all show up in such an analysis.  After identification of problematic sites,  BOD analysis can be conducted and effects of dumping can be quantified. 

Additionally, it would be ideal for all measurements made for this project to be repeated at least once after a change in management or catastrophic fire.  If the study is done once in the year after such a change and then again a few years later, after Mill Creek has appeared to have returned to an equilibrium state, the comparisons would be very informative, as to the effects of major changes in municipal watersheds due to fire, and, for the City of Walla Walla, it might serve as a warning about how much such a problem could cost.


Jim Beekman -- Walla Walla District of the US Forest Service.

Tom Wordell -- Walla Walla District of the US Forest Service.

David Reese – US Army Corps of Engineers.

Robert Gordon – Walla Walla water treatment facility, City of Walla Walla, 527-4380, last contacted 4/24/01.

The Work and Hurlow endowments funded by Whitman College Alumni, for research funding.

Frank Dunnivant – Whitman College, thesis reader, and idea man.

Bob Carson – Whitman College, thesis advisor and reader.



Literature Cited:

1.    Beekman, Jim, A Wildfire And Fuel Treatment Risk Analysis For The Mill Creek Municipal Watershed, Washington Institute, Inc.  Colorado State University, CO, Draft release, April 14, 2000.

2.    Bormann, F. H., Likens, G. E., BioGeoChemistry of a Forested Ecosystem, Springer-Verlag NY, 1977.

3.    Bormann, F. H., Likens, G. E., Pattern and Process in a Forested Ecosystem, Springer-Verlag NY, 1979.

4.    Greenberg, A., Clesceri, L., Eaton, A., Standard Methods For the Examination of Water and Wastewater, 18th edition, American Public Health Association, New York, 1992.

5.     Laverty, L. and Williams, J. Protecting People and Sustaining Resources in Fire-Adapted Ecosystems, A Cohesive Strategy, 2000, available online at http://www.fs.fed.us/pub/fam/Cohesive-Strategy-00oct13.pdf.

6.    Managing the Impact of Wildfires on Communities and the Environment, A Report to the President In Response to the Wildfires of 2000 September 8, 2000, available online at http://www.whitehouse.gov/CEQ/firereport.html.

7.    Meyers, R. Encyclopedia of Environmental Analysis and Remediation, John Wiley & Sons, Inc. 1998.

8.    Minear, R A. Keith, L. H. Water Analysis Volume III: Organic Species, Academic Press Inc. NY, 1984.

9.    Mill Creek Hydrologic Data Flow measurements from US Army Corps Walla Walla Office, available from David Reese, 509-527-7283.

10. Ordinance No. 506 – from the Forest Supervisor’s Files, Walla Walla, WA, US Forest Service, 1918 – contact Tom Wordell.

11.  Sawyer, C. McCarthy, P., Chemistry for Environmental Engineering, Third Edition, McGraw-Hill Book Company, New York, 1978.

12. Watershed Professionals Network, 1999 Oregon Watershed Assessment Manual, June 1999, Prepared for the Governor’s Watershed Endangerment Board Salem, Oregon.



Websites and computer programs:

  1. Dunnivant, F. Enviroland 3.0 -- available at http://edusolns.com/enviroland/ -- produced in conjunction with Hartwick College, 1997.
  2.  http://www.coe.ttu.edu/ce/TRM/en09002.htm  -- last visited 12/1/00.
  3.  http://k12science.stevens-tech.edu/curriculum/waterproj/bod.html -- last visited 12/1/00.
  4. http://courses.ncsu.edu:8020/classes/wps460001/lec298/sld001.htm -- last visited 12/1/00.
  5. http://stream.rsl.psw.fs.fed.us:80/streamnt/jan98/jan98a4t.html -- last visited 12/1/00.