Discussion: two things

My first thought was that *of course* Reed has a track that is gravel and has a hill. Then I realized it wasn't a 400m track and it wasn't so Reedy.

As for the times, maybe your position was off to the left? Then clockwise turns would have a bigger radius than counter-clockwise turns. But that's a wild guess, I have no idea if that actually happens like that.

It could be that you just haven't been doing so much regular speedwork in the last few months, so don't have a really good sense of pace yet. The Garmin isn't going to be off by multiple seconds in its timing! One possibility is that it was a windy day, and that one side of the track was basically shielded from the wind because of a hill, trees, or a building, while the other side of the track was situated so that it was a headwind one direction and a tailwind the other. The old Cornell track around the football field was a bit that way sometimes with a southwest wind where the stands were sort of shielding the west side of the track, but channeling the wind northwards along the east side- fortunately a less common wind direction. If a hill is very steep on one side but gradual on the other side reversing direction might make a couple of seconds per lap difference, but that is pretty unlikely for a loop providing anything approximately related to a time running on the flat. Usually a small hill that has only a percent or two grade so that it doesn't really much affect running stride going either up or down it will make a surprisingly small difference in the running time over the same distance on the flat.

Maybe one of your legs is longer than the other and you have a natural bank into the turns going one way, but are running off-camber in the other. Try putting a 1" thick insole into one of your shoes!

More likely coriolis effect. Can be very difficult to compensate for unless the correct spins are included with each lap. Best bet is to reverse direction each lap, neutralizing the effect.