Bad way to ask a question. When asking a question, you need to provide all the details you can (without giving away the subject matter). This will help others get a better grasp of where your question might be.

When collecting plant samples, you will need to record information about the traits of the plants (this is also necessary for other research in the next plan) and ask an experienced plant taxonomist for help in identifying if they are the same species.

After the sequencing is completed, it is necessary to manually align the sequences after the auto alignment by software to extract the relatively consistent sequences for the phylogenetic tree construction. If yours is sanger sequencing, you should be able to see some of the sequencing errors that need to be removed, such as indeterminate bases, polyA structures, misshapen sequence fragments.

The construction of the phylogenetic tree can only be carried out after it has been determined that the alignment of the gene sequences is correct. The tandem method of tree construction may cause differences from the parallel method, so you need to determine for yourself which one is the most reliable. Secondly, you need to determine the base substitution model when building the tree, different software may not get the same structure, which you need to pay attention to. Secondly, the software for tree construction also has different treatment for INDEL, for example, RaXML will choose to delete all insertion deletions and only retain polymorphic sites; MEGA will choose to retain single-base deletions as a feature to build a tree. At the end of the tree building, you can also see the support information so you know how much confidence you have in your data.

Hopefully this information will help you find the problems you are experiencing. Good luck!

Maybe it is not an error. Sometimes samples from different regions are closely related to other region plant samples because of a lack of genetic structure (best case scenario) or a lack of resolution of the markers you're using (worst case scenario). Do you have any literature to support a hypothesis of them being differentiated?

Give it a try with different phylogenetic software, although I would assume that might not be the problem.

Good luck!

Zhuangwei Hou I will provide you with a detailed description of the samples, all of them are authenticated and are indeed the same plant samples I am working with. The sequencing I used was Sanger sequencing, and I have trimmed the regions having errors and also the 5' and 3' of the genes. secondly, after multialignment I manually checked the sequences and then proceeded to form the phylogenetic tree.

Thank you for the answer.

Jose Alberto Lopez-Aleman thank you for the answer, yes I think what you are saying is mostly probably the correct answer, "Sometimes samples from different regions are closely related to other region plant samples because of a lack of genetic structure". As I have a reference paper that mentioned the same, "The UPGMA dendrogram based on Nei’s genetic distance did not show any particular geographic patterns. The low level of genetic diversity suggested that there should be an urgent emphasis on the conservation study of this species", paper name, (Genetic Variation and Population Structure of Dendropanax morbifera Lev. (Araliaceae) in Korea).