Well, this is going to generate a long thread.
I think he means shuffle the cars around: during the summer (AT lower ridership, other routes high) some cars will be transferred from the AT to another LD route.
The summer is one of the peak periods for the AutoTrain! I have to check a full year of monthly reports, but July and August are among the busiest months of the year for the AT. There is also not that much seasonal variation in the total numbers for the AT. The train may run light in one direction for certain periods as people travel to and from FL, but the AT has to run back the other way to get the equipment to run in the busy direction.
So, let's see. The proposal is to kill or seriously damage the LD train that:
-pulls in the most revenue of all the LD trains at $68.6 million in FY11. (Actually pulled in more revenue than any of the corridor services other than the Acela and NE Regionals).
-without crunching the numbers, pulls in the most revenue per passenger of all the LD trains.
-has the highest cost recovery of the LD trains and comes the closest to breaking even of all the LD trains.
-The LD train that might possibly break even if gas prices stay has high and might get close to selling out on a regular basis in both directions if we enter an era of sustained >$4 to $5 a gallon gas.
Sorry, but this is not the best idea I've seen floated here. The proposal is based on incorrect ideas of AT ridership patterns and revenue that are easily checked in the Amtrak reports on their website.