So, what you have here is a mixed lighting situation. The background and star notes are being lit via flash (check out their shadows. Nice and soft and from top to bottom. The main bulb cluster is on the left and yet it has no impact on those shadows) and the bulbs themselves are being lit...by themselves :-).
(There's probably a speedlight with a soft box on it or some other softening light modifier and it's placed north of the shooter, aiming down. When looking at lighting, always look to the shadows to get an idea of where it came from and how soft it was)
In these types of images, the flash is controlled by:
- its power level
The amount of light captured by the bulbs is controlled by:
- shutter speed
As you can see, there's overlap there but also two unique controls: the bulbs are the only thing affected by shutter speed while the flash is the only thing affected by changes to its power level.
So, in changing the exposure from f/2.8 to f/3.2 — the shooter brought down the amount of light processed from the flash (you can see that the highlights on the notes are less hot in the second example). They brought it down by 2/3 of a stop.
Now, that would also bring down the light captured from the bulbs by 2/3 of a stop. BUT, they also brought the shutter speed from 1/3 to 3s (~just over 3 stops), increasing the the total amount of light captured from the bulbs by ~2.5-3 stops.
So, in the end, the amount of flash captured was brought down while the amount of ambient (the bulbs) was brought up.
See technique 2 in my answer here. It's a similar mixed lighting shot and another idea for you to test mixing flash and bulbs. Mixed lighting is a whole subset of lighting technique, there is a lot to it, but damn if it isn't a lot of fun.