Looking for local water quality benefits of a small-scale cattle exclusion fencing project in Southern Alberta  

Results

The mean E.coli concentration in the reach treated with fencing (M=223, SD=268) was lower than that in the unfenced, background reach (M=350, SD=1861)  in 2004 and 2007 (fenced: M=182, SD=179; unfenced: M=218, SD=269). In 2005, the mean E.coli concentration in the reach treated with fencing (M = 250, SD=167) was higher than that in the unfenced, background reach (M=223, SD=239). In 2006, the mean E.coli concentration in the reach treated with fencing (M =146 , SD=255) was very slightly higher than that in the unfenced, background reach (M=143, SD=1978).  

 

Table 1.  Results of Welch's two sample t test comparing log10 transformed annual mean E.coli concentrations between equally sized fenced and non-fenced treatments (Ho = no difference between treatment means, n=24, p<0.05).

 

 

Interpretation: t test results

Welch's two sample t test was performed comparing the 2004 mean log E.coli concentration for the fenced treatment condition (M=2.349, SD=0.419) with that for the background/unfenced condition (M=2.544, SD=0.828) at an alpha level of 0.05.  This test was found to be statistically non-significant, t(19) = 1.34, p >0.05, indicating that the treatment and background E.coli concentrations are not different (a=0.05).

Welch's two sample t test was performed comparing the 2005 mean log E.coli concentration for the fenced treatment condition (M=2.397, SD=0.496) with that for the background/unfenced condition (M=2.349, SD=0.516) at an alpha level of 0.05.  This test was found to be statistically non-significant, t(16) = 0.187, p >0.05, indicating that the treatment and background E.coli concentrations are not different(a=0.05).

Welch's two sample t test was performed comparing the 2006 mean log E.coli concentration for the fenced treatment condition (M=2.163, SD=0.272) with that for the background/unfenced condition (M=2.154, SD=0.822) at an alpha level of 0.05.  This test was found to be statistically non-significant, t(22) = 0.045, p >0.05, indicating that the treatment and background E.coli concentrations are not different(a=0.05).

Welch's two sample t test was performed comparing the 2007 mean log E.coli concentration for the fenced treatment condition (M=2.259, SD=0.514) with that for the background/unfenced condition (M=2.338, SD=0.446) at an alpha level of 0.05.  This test was found to be statistically non-significant, t(22) = 0.403, p >0.05, indicating that the treatment and background E.coli concentrations are not different(a=0.05).

The data do not support the rejection of the null hypothesis and suggest that the mean log E.coli concentration in the stream after the fencing treatment is not different from that of the background, unfenced reach in any of the sampled years.