When a wave travels from a less dense medium to a denser medium, a small portion of the wave's energy is transmitted into the denser medium, but most of the wave is reflected. The reflected wave does indeed go through a 180 deg phase shift for this situation.
For your drawing, the water's surface acts as a closed end on the air tube, as most of the sound striking the water's surface reflects off of that surface. In order to have resonance, there must be a node at the water's surface (closed end of the pipe) and an anti-node at the open end of the pipe. Because there is a node at the water's surface, there is very little oscillation of air molecules at that point. Due to this, it may not be relevant to talk about a phase shift for the reflected wave at the water's surface, particularly since a phase shift is important for wave interference situations, and this situation involves resonance, not interference.