I think the reason is, as you saw, a lot of it is connected with that particulate fraction that will settle, but also, water is added by the tributaries downstream. If the downstream tributaries don't have much cadmium in them, they're going to dilute the upstream sources.
I think the only interpretation I could make of that entire pattern is that there's a considerable variation in sources of cadmium. There must be small deposits of high-cadmium soils in some of those tributaries. One thing that stood out, though, is that there is no clear relation to industry, as there is for some of the other metals.
I think those high values, regardless of the source, are of some concern. They're at levels where there has been demonstrated toxicity to aquatic invertebrates, for example.