PREreview of Implications of gene tree heterogeneity on downstream phylogenetic analyses: A case study employing the Fair Proportion index

Review of the preprint- Effects of phylogenetic variation on prioritization of species for conservation DOI: https://doi.org/10.1101/2023.01.21.525012

Discussion led by: Gohil Prapti (10.03.2023) Notes taken by: Sakshi Ravindra Tripathi

Review written by: Journal Club (Prapti Gohil, Sakshi Ravindra Tripathi, Daniel Linke, Mar Repulles, Phil Honle, Petr Klimes, Pavel Matos, Irena Kleckova,

 Pedro Ribeiro, Karolina Hruba, Alena Suchackova, Kristian Grell)

Overview

The authors collated nine datasets of multilocus phylogenies from the literature, with a broad taxonomic scope extending from fungi to primates. Their aim was to calculate the Fair Proportion (FP) index as a measure of evolutionary distinctiveness of a species, and to compare how much the FP index rank changes depending on the used molecular dataset (single genes vs. species trees). They have found that for some taxa, there is relatively stronger correlation among the calculation methods (e.g., fungi, mammals), while in other groups the correlation is weaker and results are more variable with regard to the FP index (dolphins, primates). Overall, the paper’s message is apparently that the FP index based on single gene trees performs poorer than based on species trees. Although this work is of interest in the field of evolutionary and conservation biology, we have some suggestions for improvement. These are outlined below.

Positive feedback

We believe that the manuscript is good at providing background for conservation biologists on the subject of phylogenetic diversity indices (specifically, the FP index). The simplicity of the analyses is beneficial for those not familiar in phylogenetics to understand the methods of measuring and ranking of phylogenetic diversity. The authors did a good job in assembling a very broad and interesting dataset, spanning multiple taxa. In addition, they provided a link to the Data and Scripts available in GitHub, which increases the reproducibility of their results.

Concerns

We provide below both major and minor concerns with the aim of improving the message of the article. We follow the structure of the paper:

Title: The title sounds somehow misleading because the article deals with methodological effects of single genes vs species trees, but does not really focus on how to prioritize species in conservation. If the authors prefer to expand the analyses to include for example, correlations among indices, conservation statuses, then the title would be fine. Otherwise, perhaps the title could be edited to something in the lines of: “Effects of gene-tree species-tree discordance on the estimate of unique evolutionary lineages obtained by conservation indexes”.

Abstract: Overall, the text is fine, but we provide minor feedback to improve clarity of certain passages. For example, we believe that the first section of the abstract, where authors explain the “background”, is too long and perhaps it would be better to remove the sentence that goes from “Several” to “input” to make it concise.

The FP index apportions the total diversity of a phylogenetic tree among all leaves in such a way that each species receives a “fair proportion" of its ancestry. - this sentence is too methodological and specific for the paragraph, and it is fairly explained in the previous sentence.

It  would be useful to add more information about methods and results instead of a rather long introductory section with background information. 

Also, we would suggest rephrasing the last sentence, since starting with “our aim” breaks the flow. Maybe “we conclude” or something similar would increase readability; from results (previous sentence) to conclusion.

Introduction: The introduction seems too specific/targeted for a very specific audience. We believe it would gain clarity by broadening the explanation and definition of some terms such as: gene trees and species trees, the FP index based on the target audience (probably people working in conservation). Also, if FP is a “popular” measure, it would be useful to state a few of them with their respective citations to provide a broader background. Perhaps giving it a whole paragraph starting with why you need this index in conservation, then its definition and finally its relation to phylogenetic discordance.

The fifth paragraph, starting with “Phylogenetic”, might be better positioned at the beginning of the introduction since it contains the gap of the knowledge that the authors are trying to tackle. 

Methods & Results: In the species tree reconstruction, the molecular clock was enforced (time calibrated) but it is not clear if the authors used calibration points or particular substitution rates.

Can the authors test whether those taxa having long branches and stable positions in the phylogeny have more consistent estimates of FP indices across gene trees? If that’s the case, then the estimated FP indices should be highly correlated among gene trees and species trees.

Are there certain genes for each group which give an overall good resolution for the phylogeny while others might be omitted? What would give conservationists / policy makers the best result for the least effort / cost?

Overall, we think the figures are adequate but it might be valuable to provide in the Appendix similar figures for each study group as currently provided in Fig. 2 (dolphins only). It can be useful to add the status of the endangered taxa, where available, to the figure. Importantly, there is no information whether those taxa that converge well in FP and rank high in Fig. 2 contain most of the endangered species, or whether those are the lineages that are most unique in the phylogeny. We are not sure if this is possible to assess, but such a direction would strengthen the conservation focus of the study, and make it more valuable for practical use, beyond a methodological note paper.

In addition, we have the following minor notes: 

• Can the authors compare the efficiency and usefulness of the FP index with other indices used in conservation biology?

• Selection of datasets: Four out of nine datasets are from mammals and there is no representative from invertebrates. We are also confused why the yeast dataset is included, since the paper is focusing on conservation. It’d be informative if the authors could explain their reasons behind selection of these datasets.

• We believe the figure 1 contains a mistake in the branch leading to taxon x2. The length should be “1”. Similarly, the branch leading to taxon x3 should have a length of “2”

Discussion:

Overall, the authors wrote a good and concise discussion. However, we see a potential for improvement, and we provide some suggestions below.

We think that a concrete recommendation should be made by the authors about the use of the FP index - should we avoid entirely using gene trees? - Especially if the target audience is policy makers, whose background might not necessarily be in phylogenetics or evolutionary biology.

In general, we recommend supporting crucial information with references. For instance, the first line of the first paragraph “The FP index is a popular tool for prioritizing species for conservation…” is a very broad statement, indicating that the FP index has effectively been employed in species protection, but no studies are provided.

At the end of the second paragraph, are there examples of studies or conservation projects specifically using gene trees to delimit species that are in urgent need of conservation? If yes, please add them.

Are there cases in which a gene tree would be preferred over a species tree? This point is very important for a critical assessment of the FP index.

We suggest providing, if possible, a concrete conservation-relevant example in the conclusion, which will highlight the relevance of the study.

Competing interests

The author declares that they have no competing interests.