The way I interpret part b, we are looking for the probability that Landis doesn't catch a fish in 2 hours. It makes sense to me then that:
Pr[no fish in 2 hr] = 1-Pr[Fish in two hours or more]=1-UpperBound(Pr[Fish caught in two hours])
If my logic is correct, then the answer should just be a simple application of the Markov inequality.
