Unfortunately, this LoRA falls into the same trap as many on here. The style has not truly been learned because of some combination of undertraining and/or including key style information in the captioning for the training data. It won't give you anything close to a Watership Down look unless you include "watercolor painting, classic 2d illustration" and at that point it's not clear how much the LoRA is really doing.