If you keep the sensor powered all the time, it will be destroyed in under a month.I learned that from experience.
I saved the destroyed headers, but I could not find them right now to show you the remains.After less than a month only the pcb remained, the copper including plating was gone.
The dissolved copper could probably even damage the plant you are monitoring.
ACTIVE CONSTITUENT: 93g/L COPPER (Cu) present as Copper Ammonium Complex
Platinum, gold, nickel or stainless steel would be much better than tin plated copper, or a copper wire.
Sad that all they'd need to do is spring for ENIG and the sensors would stop destroying themselves...
I think you can do moisture sensing by measuring capacitance. Then you don't have to expose metals to the soil.