Hi, this is miningpoolhub.
Your invalid shares are generated from previous block's hashing.

As new block is found from litecoin network, hashing miner needs to reset their current work and start hashing with new base numbers.
When the pool send new number for next hashing, all following shares generated from previous blocks are set to be invalid shares.

There must be little time between block propagation on network.
It cannot be zero as fast running car can't stop immediately.

As you are using NiceHash, which relay new hashing base number to different kinds of miners again, tends to need more time than normal miner.

It depends time to time, but it's normal.