ummm, you can't use observational data to draw conclusions like these. period. you can use them to hint at causations and effects of changes, but you absolutely cannot make conclusive statements like those made in this paper. REALLY bad science.