Assuming that correlation establishes causation is a common practice in "bad science". Explain what it means and why it is "bad science".

The question is confusing and poorly worded I think, but correlation does not equal causation in normal science. Just because there is correlation between two variable, you cannot just assume one variable causes the other variable.  Hence, I guess bad science here means it's conclusion with logical fallacy.

dna is made up of six smaller molecules -- a five carbon sugar called deoxyribose, a phosphate molecule and four different nitrogenous bases (adenine, thymine, cytosine and guanine).

sheep is a herbivore animal whereas grasshopper belongs to the class chordates but both of them feed on greens.  

if grasshopper and sheep feeding on a pasture land, i do not think both of them will have any impact on each other. however, the number of the species can make difference. since sheep feed on greens, there is a possibility there left no pasture land for the grasshoppers or they may get extinct.