Ammonia is significantly more water soluble than ammonium, but as pH increases ammonia becomes more toxic. At lower pH, ammonia binds with HCL to form ammonium(I am shooting form the hip right now so I might not be sayhing this exactly right).
Ionized NH4(ammonium) is MUCH less harmful to aquatic life.
Here is an article that does a fairly good job of explaining this in terms simple enough even I can understand it... You should not have any problem...
The laboratory method used for analysis of water measures ammonium-N plus ammonia-N. It is very difficult to directly determine the activity of aqueous ammonia, so instead the surrogate of ammonium-N plus ammonia-N is used, and then tabled values of ammonium-N plus ammonia-N are used to determine if a measured concentration will provide ammonia at a level that is detrimental to aquatic organisms, for acute and chronic conditions. These tabled values are a surrogate since the measured concentration is a total of the ammonium-N plus ammonia-N, and the concentrations in the tables for chronic or acute levels are set to reflect back to likely concentrations of ammonia-N for specific water pH and temperature.
A little more complicated, but addressing toxicity:
The acute and chronic criteria for “ammonia” have been established for Iowa streams designated for aquatic life uses (Chapter 61, Iowa Administrative Code; tables 3a, 3b and 3c). One has to carefully use the tables as the listed concentrations are for ammonium-N plus ammonia-N, not ammonia-N (the header to the tables says “ammonia”). As expected, chronic criteria (ammonium-N plus ammonia-N concentration) are higher for low pH and low temperature water (ex. pH 6.5 at 0 degrees C is 6.67 mg N/l, early life stages present) and lower for high pH and high temperature water (ex. pH 8.0 and 26 degrees C is 1.16 mg N/l). Similarly, acute criteria are higher for low pH water (ex. at a pH of 6.5 the criteria for class B (WW1-3) and B(LW) is 48.8 mg N/l) and lower for high pH water (ex. at a pH of 8.0 is 8.4 mg N/l).
So... what does this all mean.... Lots...
I am very suspicious that something else contributed to the ammonia test reading. Certainly the osmocote could have contributed, but it seems like you did everything right. 1 cup is a lot, but if it is buried, it isn't that much. LIke I said, Tom Barr even suggests coating the entire bottom pain of glass with it... That is a lot!
Sorry for possibly over-complicating this, but with a little digging we can find some really useful info here. The tests sara did are very useful, as they show the difference in a few variables. One test I think she should compare is a cup full of the same water with the osmocote below a substrate. different substrates provide remarkably different results.