I spent the last 15 years of my career collecting water quality samples for the Texas Commission for Environmental Quality in the 22 county area of Northeast Texas. The two biggest water quality issues in that part of the State were bacteria and low dissolved oxygen (DO). In contrast, we hardly ever have low DO in the Hill Country because we have flowing streams compared to sluggish bayous.
How to interpret results
E.coli is found in the digestive system of all warm blooded animals. (It is not found in your urinary tract. If it is there, you are in trouble.) E.coli is used as an indicator organism in water quality sampling. There are only a few strains that are pathogenic. (Think of the recent romaine lettuce recalls because of E.coli 0157:H7.) The vast majority of strains are helpful, even necessary for digestion. So why test for it? It turns out that E.coli is a good indicator of the presence of many other pathogenic bacteria. When E.coli concentrations are high, the incidence of stomach related illness goes up in swimmers, but most likely because of pathogens other than E.coli. Water Quality Standards are compared to individual samples as well as for long-term averages. A single sample should not exceed 394 colonies/100 milliliters. But the more important regulatory sample result is the geometric mean (different from a simple average) of at least 10 samples collected over a relatively long period of time, at least one year.
When samples should be collected
It becomes very important when to collect samples for this geometric mean. It is very easy to bias the results. A random sample design is needed. If you target samples to one or two days after a rain, you will get more exceedances than if you were to randomly select the sample days. Everywhere in the state, bacteria concentrations spike after any rainfall event that causes runoff. Bacteria attach themselves to sediment particles. As long as the sediment is suspended in the water column, they are subject to being ingested by an individual. With time, these sediment particles settle and are no longer a concern. There is a group of individuals that collect bacteria samples in Wimberley on the first Monday after Market Days. This is random sample design and that is good.
Limitations of source tracking
Source tracking has its limits. It is expense and it is only as good as the regional “library” of data with which to compare. It can only tell you presence or absence, not relative importance. To my way of thinking, livestock and wild animal sources should be expected after a rainfall in just about every rural waterbody in Texas. Not so with human sources. They do not belong in the creek anytime. Properly run septic systems as well as city owned sewer systems should not contribute to the bacterial waste load of any waterbody.
Some common sense suggestions
Numerical standards are OK, but they give you yesterday’s data. The water you are swimming in today is miles downstream by the time the samples are read the next day. So it’s an educated guess as to the real water quality today. Young children are more likely to accidently ingest water while out on the river than are adults and therefore are at greater risk than adults. There are no guarantees in life, but keep the kids out of the river for 48-72 hours after a significant rainfall event and you should be good.