Two problems with long stored samples. First, depending on what you store them in, they can leach additional substances from the bottle. This can lead to either detection of these substances, or to unpredictable reactions changing the solution composition.
Second, if they are allowed to exchange gases with the atmosphere, amount of CO
2 dissolved can change. In your case it looks like you have lost some of the CO
2:
Ca
2+ + 2HCO
3- CaCO
3(s) + H
2O + CO
2That would be consistent with all three observed changes - rise of pH and lowering of concentrations of Ca
2+ and HCO
3-. Settled down CaCO
3 precipitate can be difficult to spot, as it can just leave a thin, matte (I hope that's a correct word) layer on the bottle surface. (This is the same reaction that is responsible for karst feature development).