In the context of scientific writing, a discussion involves explaining, clarifying, and justifying what you have learned from your research project. Discussions are also where you examine whether what you have learned fits, or does not fit, with what you know about ecological theories and processes.
Remember that statistically significant results are not always ecologically meaningful. The magnitude of the correlation, or the difference in means, are often small, even if statistically significant. In the opposite case, you may get nonsignificant results when there really is a difference in population parameters you measured (like means) or your result shows a weak relationship, but it turns out to be one that is ecologically important. Often, these problems are due to having too small a sample size (i.e., one that is too small for a statistical test to reject the null hypothesis). Also keep in mind that there could be one or several other factors influencing the system you are studying that you did not measure.