Haven't nitrate levels always been high in Iowa and other Midwestern states surface waters?

No.

"Always" is the key word in this question. A paper by Keeney and Deluca (J.E.Q. 22:267-272 (1993)) indicates that nitrate levels in the Des Moines river today are not much different today than they were in 1945. However, the authors use this evidence to suggest that row crop agriculture, in general, and not just nitrogen fertilizer application are the reasons for the high nitrate concentrations.

When soils are cultivated and left bare for a good part of the year mineralization (changing organic forms of nitrogen to inorganic forms) is increased and soluble nitrate is produced. Under native plant ecosystems (prairie or forest) this process is slower and the nitrate that is present is rapidly immobilized by plant uptake or microbial activity. Thus, native ecosystems are said to have a closed or tight nitrogen cycle while present-day agricultural ecosystems are open and leaky.

The Iowa landscape has been leaking nitrogen to surface waters since the native prairie and native forests were cleared and replaced by cultivated fields, roads and urban areas. Buffers provide a small, strategically located perennial plant community which can accept a given amount of the leaking chemical load of the rest of the landscape and process it before it reaches the surface and ground waters. The research challenge is to identify the size of the buffer community that is needed to accept and reduce to an acceptable level the NPS pollutants from the unbuffered portion of watersheds.