My research forms part of a larger program designed to quantify the environmental benefits of agricultural best management practices in use on cattle ranches in Southern Alberta.  The basic research question this course project attempts to address is:

Can reductions in E. coli concentrations be seen in water quality measured at the farm-scale when we fence cattle out of a stream?

Figure 1. Regional landscape surrounding the project
site, Mixed grass natural subregion, Alberta.

Figure 2.  Cattle use of riparian pastures and
streams, Lower Little Bow River, Alberta.



Ranchers are being asked to change a variety of practices to reduce their environmental impacts and the protection of riparian areas is one best management practice (BMP) that seems to make sense.  Cattle impacts to streams are visible, so removing cattle should improve water quality by removing those impacts.  Numerous studies have shown watershed-scale benefits to the use of such practices, and smaller scale studies across the United States have shown improvments to agricultural stream water quality, but is this the case in Southern Alberta?

Water quality response to the installation of a riparian fencing treatment was monitored with weekly and biweekly water quality samples collected from 2004 to 2007.  An 800m stream reach within a cattle pasture served as the experimental treatment and was fenced on both banks to prevent cattle access to the channel.  An upstream channel reach of equal length was used as a control treatment and provided background water quality of the stream channel against which water quality of the treatment area could be compared.  Samples were only collected on one stream; no replicates of the treatment were used. 

E.coli concentrations collected between June and November were compared across years between treatment and control samples using annual means and weekly differences.  Data was normalized by a log transformation of E.coli concentration and transformed data was compared between fenced and unfenced treatments in each year using Welch's t test with unequal variance. The resulting p values (p>0.05) failed to support the rejection of the null hypothesis (a=0.05), suggesting that there may be no treatment effect of fencing on E.coli concentration; however, major flaws in the experimental design limit this conclusion.



Website created by:  JL Scott, AScT, BSc    jlscott@ualberta.ca

Note:  This website was created as a course project for RenR 480, Department of Renewable Resources, University of Alberta. 
Data used is modified or randomly generated.