Maybe not a solution for you, perhaps, but back when (mid-90s) we were growing shrimp (commercially) in large, outdoor saltwater ponds, we had to pay very close attention to alga blooms (due to their huge effect on DO). We also would have liked an NPK monitoring device, but no economical, in-situ solution existed. And, of course, the problem was aggravated as the shrimp grew larger.
So, to do this, we closely monitored pond water pH since the nutrient loads (from excess food and shrimp waste) made the algae very happy and they predictably "bloomed" vigorously. We wanted some, of course, since they were daylight DO producers (and pushed the pH up). But too much was disastrous (because at night they were voracious DO users and pushed the pH down).
Over time, it worked well enough that we could reasonably predict, from the pH levels (plus time of day), a pending bloom "crash" and crank up extra aeration (and water exchange to a limited degree) in anticipation of them. Of course, we could also simply look at the pond water color and color density. But the blooms could, literally, explode overnight and we needed to know beforehand.
My point is that the state and efficacy of the nutrient uptake by the algae (judging by the pH and intermittent chemical analysis as proof) gave us an idea of the relative reduction of NPK levels in the water.
It was a crude system, largely predicated on empirical data, very large nutrient loads, hyper shrimp concentrations (300/m[SUP]2[/SUP]) and essentially fixed water exchange rates (15%/day).
Presumably yours is a freshwater system. This ought not make that big a difference. Scale may be an issue as well: our pond's surface areas were no less than 2, and up to 7, acres.
But pH is an excellent predictive water condition (and hence, biological processes) value if the processes going on can overcome the normal buffering capabilities of the water in which these processes are happening.