Every cover figure in your Entry carries its sampling Standard Error — its SE — and every station carries a floor. They come from different places and they are not combined.
The sampling SE comes from a table. Fifty random points on a frame that is 30 % coral will give you somewhere around 30 % give or take 6 points, and that is true before you look at any reef — it is arithmetic, not observation. Look up the row nearest your cover, read across for your point count, write it down. It is one standard error, not a confidence interval, and the Entry labels it as such. A class you saw at only one or two points gets reported as "1 of 50", not as a percentage with a ±, because the ± would be a lie in both directions.
The floor is the smallest difference a later visit can call change, and it is the larger of two figures.
The first comes from the same table, and it is called the sampling threshold. Two honest scores of an unchanged reef, from two independent sets of points, do not come out equal — they differ by sampling alone, and the threshold is the difference below which sampling still explains everything. At 50 points and 30 % cover it is about 18 percentage points; at 25 points, about 26. That is a large number, and it is the truth of the method: nothing smaller than that can be told from sampling, whoever took the pictures. It is derived, so a diver returning in ten years computes it the same way.
The second you measure. On Dive 3 the same photo point is re-captured — by you, or by the other Team if you swapped stations — and scored again, and the difference between the two figures is the repeat-capture difference. It contains the chart placement, the camera height, the light, and — be clear about this — a second helping of sampling error, because the second frame was scored at its own points. So it is one draw of how much the capture can vary, not a clean measure of anyone's hands. The larger of your two photo points' differences is the station's observed difference.
The floor is whichever is larger. Most of the time it will be the threshold, and your observed difference will be a lucky draw underneath it — that is fine, and you write both down. When the observed difference is the larger, something about the capture varied more than sampling explains, and that is the floor. Either way the floor is the same for both photo points at a station, and both of your Level 1 statements use it.
Your Level 1 statement says all of it in one sentence:
Hard coral cover at photo point 1 was 32 % (sampling SE ± 6 pp, 50 points); repeat capture after 4 h 40 min by the same Team differed by 4 pp; the floor at this station is 18 pp.
The interval is in the sentence on purpose, and so is who made the return capture. An observed difference taken four hours apart on the same day, in the same light, with the charts still fresh in your memory, is the easiest repeat there is. A return next year will meet things that repeat never did — different water, a new chart batch, another camera, you a year older — and the course does not pretend to turn those into a number. Whoever makes that return lists them, states the floor they cleared and by how much, and lets the reader weigh the rest.