The Shannon or Shannon-Weaver index is open-ended, so it is theoretically possible to calculate a value as high as 5 or greater. In practice, such high values are very rare. I have seen invertebrate communities in streams with a diversity > 5, but I have never seen, nor heard of, any community with a Shannon index as high as 9 -- I have difficulty imagining any community so diverse. If you have calculated values of 5 or above, I would urge you to check your calculation.
The maximum of Shannon-Weaver index equal Hmax=Ln(1/S) there S - number of species. If you want to compare habitats with different number of species you can use Pielou index E=H/Ln(1/S).
The shannon vales also varied with the distribution of species in all study sites and also depends on the ratio between the total score of an individual species and total no. of all species. In some cases a particular individual species have greater no. compared to other associate species with it, which make diffrence for resulted values of Shannon diversity.
The shannon diversity is based on total sum of ratio between the individual species and all species in a particular habitat, so it may vary according to the species individual and their sum values. But it is generally accepted taht Shannon value are below 4. May be ur data is different according to the species ditrsbution at different habitats.
Shannon-Wiener diversity index has been widely discussed this February. Here can you find valuable information: https://www.researchgate.net/post/If_I_obtain_the_Shannon-Weiner_diversity_index_as_285_what_can_I_interpret_from_this_about_the_diversity .
I was also amazed using vegan package from R to find that was possible to have a value higher than 5. In fact reading on literature most of research have value not higher than 5. So I was surprise, but I think that some time natural experiences have wide range of outcome. So I could say it is possible to have Shannon index higher than 5.