I looked at the methodology-standard ANOVA and non parametric KR tests to account for non-normality & bonferroni adjustments for multiple comparisons. Seems fairly standard but I don't have the weed science background to know if the metrics and concentrations used made since or if the methodology is in line with the social norms and practices of this kind of literature.
This may turn out to be a solid study, even so all it concludes is that there are additional costs/externalities associated with these very commonly used and important technologies. It sheds no light on the real question-are these technologies employed optimally-do the benefits (and even positive externalities) outweigh these additional costs. The default position often is to assume not- and assumes we should have a new policy to do something about it. Similar to the climate change debate, milking the science for more than what it's worth.
The Salon piece contained a lot of politically charged meaningless terms and other shibboleths such as 'factory farms' 'big Ag' and 'industrial agriculture' plus this often repeated but misunderstood lyric:
"Eighty percent of the antibiotics sold in the U.S., after all, including 70 percent of those that are important to human medicine-go to livestock"