One way to start writing the discussion of results in a journal manuscript is to compare your results to previously published research. Start with one variable of interest. Are your results similar to what other researchers found? Or are they different? If they are different - why? This is a way to start interpreting the data and pulling in other studies to support your research results and theories you have to explain your results. You can also write about how your results variables may be linked. For example, "Feed intake by animals was greater for treatment "A" which likely explains improved weight gain by this treatment."
For a detailed explanation on the best way to write the discussion part of an article, I refer you to the following link. I hope that you find it useful.
I think that the best way is to begin by stating your research hypothesis one by one and then to compare the result with it then either to accept the hypothesis or to reject it. If you are not dependent on the hypothesis then you can begin by stating your study aims and compare it with the final results.
The purpose of the Discussion is to state your interpretations and opinions. Before beginning you should try to develop an outline to organize your thoughts in a logical form. To make your message clear, the discussion should be fully stating, supporting, explaining, and defending your answers and discussing other important and directly relevant issues. Some steps are suggesting:_
1 Organize the Discussion from the specific to the general.
2 Use the same key terms, the same verb tense.
3 Begin by the hypothesis you were testing.
4 Describe the patterns, principles and relationships shown by each major finding.
5 Discuss any unexpected findings.
6 When identifying limitations and weaknesses, avoid using an apologetic tone.
7 Summarize concisely.
8 In your writing of the Discussion, discuss everything but be concise, and specific.