It didn't come together like a granny knot, which I understand to be just a square knot with the orientation of one half flipped. The knot I learned wrapped the free end around the base of a loop and pulling a section of that free end through it to create another loop. It was unbalanced for the same reasons as a granny knot though and probably very similar.
The knot I tie now is basically a square knot where the "top" half is formed from two loops. Admittedly the knot I tie now, would have been much more difficult for toddler fingers than the knot I learned as that toddler.
It's a misleading legend, but the note at the bottom tries to clear it up a bit. This map seems to more be like "We took the range maps of 238 species of fish and overlaid them. The red area is where practically all of those range maps of each 238 species of fish overlapped." Of course there are other fish, but they were not included here because the map maker didn't have the right kind of dataset for them. To me that seems to indicate that this map isn't so much a map of actual biodiversity measured, but the potential for biodiversity of the region. Given that it's fish, I guess we shouldn't be surprised that this area is somewhere between/near the northern continent's biggest river, a large gulf, and ancient mountain range, and a coast with a strong warm current (for now...).