Over any timeframe of course. If you refuse to accept zero or negative population change as growth, the smallest possible growth is the addition of one individual. Anything less does not count as growth, to you. Thus the limit applies for any timeframe you choose, because anything less is not growth - as you define it.
(Of course, this line of argument is unutterably silly. Population levels may remain stable or decline. Thus to model the change in population levels over time we really must acknowledge the possibility of zero or negative growth. Even if you find the idea aesthetically displeasing.)