Update: More info if interested...
I played with the reef.org data some more and warm water temperatures (especially 54F+) and depth (>80ft) both correlate strongly with sixgill sighting frequencies. Perhaps the seasonality seen in the graph above might simply be a reflection of this pattern: they come shallower in response to the warm summer waters for some reason. Abnormal "heat event" years with warmer than average water temperatures also correlate though and I'm not sure yet if this is reflecting the same thing (I'm using both reef.org and NOAA water temp data). I wonder if there could be an indirect relationship with temperature like food availability or O2 (I haven't explored further).
Regardless of temperature, depth appears to have the strongest correlation with SF%. That SF% bump in March, for example, is the result of a rather deep cove 2 survey dive. One disclaimer is that this reef.org data doesn't indicate what depth the surveyor actually sees the sixgill- just the max depth of that dive. This could skew SF values slightly but I suspect it is still directionaly accurate:
Again - all this is based on a vary small sample of sightings and I haven't calculated error margins for any of the groups so they may not all be statistically significant differences. That said, the nice thing about reef.org data is that it also effectively includes how many dives did NOT see sixgills. This is really important since data that only tracks sightings often only reflects the fact that an unusual amount of dives occurred at a particular location/time/depth/etc.
It doesn't look like the reef.org sample sizes will be large enough to draw conclusions regarding abundance by locations. Perhaps this is motivation for us to do more reef surveys at some other key sites? Also don't see any obvious patterns by time of day: like location this could mean there is no pattern or just that it is small and there isn't enough data to see. Diver reported current (none/slight/strong) also doesn't show much difference. I haven't added any NOAA tide or current station data yet to my data model but someday plan to do so for a unrelated project related to visibility prediction.