In a PCA scores plot, the axes are new "latent" variables that are made up of weighted combinations of the variables you measured. So if points are far away from eachother in the scores plot, they are different. If points are close together, they are more similar. The way the PC axes are created from weighted combinations of measured varialbes ("loadings") are represented in a loading plot. Usually arrows are used and the length of the arrow corresponds to the strength of the correlation between a variable and the PC axes. The direction of the arrow shows which direction and with which axis the variable is correlated. A biplot just puts these two together (the score plot and loading plot). So in your example, sample 6 is really different than the other points. It has a much higher value along axis 2. Axis 2 is positiviely correlated with Primary_Sludge and negatively correlated with urine. So that means sample 6 is probably higher in Primary_Sludge and lower in urine compared with other samples. Point 5 is off in the lower left corner, opposite of the direction many of the arrows are pointing. That means its probably low in Soil, Colon, River, etc, since those variables are all positively correlated with both axes.

Additionally, you can use the loading plot (the arrows) to come up with a sense of what the PCA axes "mean". Axis 2 seems like a tradeoff between urine and primary_sludge, LB_broth, and CombinedSewageOverflow (I have no idea if that makes any sense, and you should only interpret it that way if it makes sense to you.)