Because the existing piers were reused, the span should be the same for all three bridges: 448 feet. Interestingly, 192 feet is the distance between a pier expansion joint and another expansion joint at the center of the span of the present bridge. Not sure of the exact design details of the 1946 bridge maybe if it was a cantilever truss without suspended span, they were for some reason measuring the length of each cantilever arm. I encounter a similar problem with swing bridges. Swing Bridge Spans should be measured from fixed pier to fixed pier, but many NBI entries measure from swing pier to fixed pier, although this seems like a "span" it is not the proper way to measure a swing truss span. Nor should cantilever trusses be measured by each arm, but would not be surprised to see it happen.
Weird...not sure why the NBI says this bridge had a 192' main span. Pretty sure this was similar to its successors which would mean a main 448' span.
Then again, it's the NBI...and they seem to have two identical data sheets for the 1946 and the 1963 bridges each as well.
I suspected that 192' was the cantilever arm length as well. A similar scenario that I've noticed here and there is when the total length of a continuous or cantilever truss is reported as the span length. I think there was a bridge in Tennessee that used to be reported that way until somebody fixed it a few years back.