Perl has a very large toolkit called BioPerl (http://www.bioperl.org/wiki/Main_Page) that is useful in biology, including botany and plant science, especially for plant genomics.
Both PERL and PYTHON are interpreted scripting languages, which means they can be used to write code for just about anything, in any discipline. These two languages are used by scientists in all areas of the life sciences, medical sciences, mathematics, physics, chemistry, engineering, etc, etc. If you can conceive of it, you can probably code it in either of them (although potentially other languages would be preferable, depending on exactly what you are trying to do).
It depends of what are your tasks. Python is an object oriented programming language: it is easy to learn and really versatile. Also Perl is easy to learn, but the less stringent sintax rules renders more difficoult to understand code written by others. On the other hand Perl is suitable for the retrieval of any kind of information from any plain text file (it is no coincidence that regular expression syntax is often Perl derived). Both of these programming languages offer wide libraries of code apt to solve bioinformatic issues. On the other side both of them are high level languages, and this means that are not suitable for number crunching tasks.
My collaborators who worked in a field of plant genomics used both of them. BioPerl and BioPython are great libraries with a lot of useful procedures for bioinformatical analyses. Which one to chose? One of my tutors once said these languages are equally useful and it is a matter of preferences.
As Dario mentioned Python is object oriented. In Perl there are not typical classes but kind of pseudo-objects. If interested in object oriented programming chose Python, if do not need object oriented programming chose what seems more comfortable. In my opinion Perl is easier and more ordered. Another advantage of Python it its capacity to handle procedures written in C++ (much quicker) or with R ( and its great packages to environmental data computations).
Never worked with Pearl, but I use Python all the time for geosciences. I've actually done some work in precision agriculture using geostatistics.
From the previous answers I take it that this is probably not your intended field, but if by any chance you'll want to study crop scattering over large fields or exploring spatial behaviour for plant growing you should consider the following tools:
Scipy (python library) has a few interpolation methods (deterministic):