Development of an Advanced-Generation Multi-Objective Breeding Population for the 4th Cycle of Chinese Fir (Cunninghamia lanceolata (Lamb.) Hook.)
, Qihang Feng 1, Jinzhang Wu 1, Xuefeng Zhang 1,2, Renhua Zheng 3, Xueyan Zheng 4, Zhiyuan Yang 5, Zhiqiang Chen 6, Harry X. Wu 6,7 and Jisen Shi 1
Round 1
Reviewer 1 Report
In the forest genetics literature, there are a few papers reviewing progress in ongoing advanced generation breeding programs, revisiting the progress in the past and proposing future directions. This is very useful to researchers and practitioners since choices are complex and balancing long-term tree breeding strategies and ensuring long-term genetic diversity is a difficult compromise. Therefore, a key topic in breeding strategies is how best to cope with complex breeding objectives (several traits) as well as deciding how to weight them against diversity. Not always straightforward.
The fact that the Chinese fir breeding program is in its 4th generations is really impressive and certainly worth a review as the authors seem to be attempting. So, the topic of the paper is important and relevant to readers. Unfortunately, the structure of the paper, the methodological approach and some problems with the Methodology make the paper unpublishable in its present form.
The key topic of the paper is how to balance genetic gains for the traits of economic importance, with genetic diversity avoiding excessive build-up of coancestry and inbreeding. This is an important topic, but one which has been thoroughly studied in both animal and plant breeding contexts, under the general designation of Group Selection or Optimum Genetic Contribution theory. In forestry, the contributions of Dag Lindgren and co-workers, over the years, have resolved most of the theoretical considerations.
On the other hand, the program’s breeding objective includes growth and wood density (and even seed yield). It is a multi-trait objective. The problem is a classic in Quantitative Genetics, and it is resolved by some sort of combined index selection or multivariate BLUP analysis.
The authors seem to ignore the extensive literature on these two topics and opted to tackle it using a “step-wise pruning procedure”, a sort of tandem selection, which is a bit naïve and clearly sub-optimal. It is not wrong as such (in fact it is quite useful in a genetic conservation context, for example to help choose which individuals to pick for breeding, to ensure the broadest diversity is captured), but it is not the way breeding decision under diversity constrains or under multi-trait selection ought to be addressed.
Another important caveat of the study is the lack of details regarding the estimates of breeding values used. Apparently, the estimates are based on 3 generations of trials and measurements, from where the 230 candidates were now selected for the 4th generation. But there are no details or information on how this analysis was performed, based on what models and in relation to what baseline. For example, is there any some previous published work on 1st, 2nd or 3rd generation genetics for growth and density? This is a fundamental element of the study and has to be presented.
In Table 3, gains in DBH are given in cm. This implies all trials must have been based on measurements of DBH of the same age and size, and hopefully the same variance. Was this the case? Ot are the measurements behind this result taken from just one recent trial which includes several generations of breeding?
Other problems in the paper include:
1. Introduction goes too much in detail about the various breeding strategies in forestry. It should focus on the objective and motivation for this paper
2. Pg 5, ln 146: “estimated breeding values” need to be clearly explained and detailed. Otherwise we would not know what are they
3. Pg 5, ln 181: This test is not reported in the results (as far as I could see)
4. Tables 2 and 3 are repetitive and too long for the message they contain. Possibly make them a Figure (as a function of selection intensity) would make it simpler
5. Figure 4 seems to be wrong. Kinship coefficients in the diagonal should be =1. And looking at the cluster dendrogram, trees that are closely clustered (bottom right of the matrix) must have a higher kinship coefficient
6. Pg 11: The discussions about the proposed diagrams of the future breeding population structures are too long and detailed (especially Figure 6). This paper is not about comparing breeding structures (otherwise some type of simulation would be needed to prove this structure would be preferable than, say, a single tier population, or a rolling front, for example).
7. The Discussion is also too long (3 pages long) covers many topics, from other programs, and not directly related with the results itself (inbreeding depression, size of the nucleus, single versus multiple populations or sublining).
In the end, the paper should focus on the actual progress of the breeding program (in dbh, density and diversity) over 3 generations, detect potential issues regarding genetic diversity and recommend how best to move towards the 4th generation, and what would be the corresponding impacts on gains and diversity. It is a long way from that aim.
Quality of English is appropriate and clear, although some improvements could be made
Author Response
Please refer to the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
This paper proposes a method for the rational division of the fourth-round breeding population of Chinese fir, based on SNP molecular markers and key phenotypic traits. It also aims to balance the trade-off between genetic diversity and genetic gain. The research findings not only summarize the current stage and future planning of Chinese fir breeding, but also provide a valuable reference for the multi-generational breeding of similar tree species.
In addition, the following details require further confirmation in the article:
1. Lines 198-205: It appears unnecessary to reiterate the fundamental properties of the Rad-seq data, as they have already been published in another journal.
2. Line 206: The 13 proposed core subpopulations were denoted as the S collection in Materials and Methods, and were presented as ‘S-’ in Table 2. Therefore, describing them by S collection would provide greater clarity.
3. The core subpopulations were selected based on their HGW, DBH, and WBD from the largest to the smallest gain. It is typically the case that a lower selection rate results in a greater genetic gain. However, in Table 2, the △G increased with an increase in the number of selected individuals. It would be preferable to provide additional details regarding the calculation method of △G, particularly for the S collections.
4. Table 2: Could you please clarify why PN+△G is precisely equal to 100%?
5. In Figure 3, the core population (n=50) was created through the intersection of four subpopulations. However, each individual being classified only within one of the subpopulations in Table S3. Could you please elaborate on this in greater detail?
6. In Section 4.2, it would be advantageous to reference additional literature to contribute to a comprehensive discussion.
7. It would be advantageous to include additional information regarding the kinship and coancestry rate of the core population.
Author Response
Please refer to the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
I have reviewed the article titled "Development of an advanced-generation multi-objective breeding population for the 4th cycle of Chinese fir (Cunninghamia lanceolata (Lamb.) Hook.)" by Zhao et al., which aimed to improve the breeding population by implementing a more advanced and structured approach, moving away from the single, unstructured breeding strategy used in previous cycles.
I would like to express the following relatively major concerns:
Table 3. Very low genetic gains have been achieved. Indeed, negative values are evidenced. Authors must explain about it. Additionally, explain why with high selection rate (e.g. 10% in DBH) delta genetic gain is negative, and with 45% positive. It seems inconsistent.
Line 51. Check grammar and idea: "The breeding population is composed of elite trees to be mated for the next breeding cycle and provide the genetic basis for long-term selection and breeding." This should not refer a particular breeding population but in general breeding scheme”. For example, "Breeding populations are composed of elite trees that are mated for subsequent breeding cycles, providing the genetic foundation for long-term selection and breeding."
Line 67. Eliminate “breeding populations for” because is repetitive in the sentence. Avoid repetitions in all text.
Line 74. Change to “Breeding strategy commonly used in forest tree breeding programs”.
Table 1 needs references (add column with the number of references).
Line 151. In this line block effect appears for the unique time. Experimental design is lacking. Authors must detail it.
Line 152. linear mixed “mode”, change to linear mixed “model”.
Line 155. What do fixed effects? Detail.
Line 129. Correct Jing et al. (2022). I think is 2023, and should be put in the journal style. The same for the line 86.
In the Conclusion section, additionally I have some major concerns:
Lack of Hypothesis Linkage: The conclusion should ideally connect back to the hypothesis or research questions stated at the beginning of the MS. This helps readers understand how the study's results address the initial objectives.
Emphasizing Key Findings: While the conclusion mentions the genetic diversity and population structure, it could highlight the most important and significant findings of the study to provide a clearer sense of its contributions to the field. In this sense, the article must distinguish itself from the recent study published by the same authors (Jing et al.); this raises serious doubts about its scientific novelty. Review this issue in all the text.
Avoiding Speculative Statements: The statement "may be related to the genealogy and breeding generation" is speculative and should be avoided. If possible, provide more concrete evidence or suggest avenues for further research.
Implications and Applications: The conclusion could benefit from discussing the implications of the study's results. How could these findings be relevant to practical applications or future research in the field of Chinese fir breeding or other related areas?
Also importantly, please highlight all the changes in the revised document and address/answer my concerns and suggestions.
Author Response
Please refer to the attachment.
Author Response File: 
Author Response.pdf
Round 2
Reviewer 1 Report
This new version has improved and has addressed a few of my earlier concerns, but it did not resolve the key issues I raised.
As it stands, it continues to include extensive considerations about breeding strategies in both the Introduction and Discussion sections (which are not related to the Results presented), while it describes very poorly how the genetic merit of the trees (which is a key result in the strategy) were calculated (they are considered known prior to the study).
In the end, the paper is mostly about (given the 233 selected tree’s genetic merit) what are the consequences in diversity (or coancestry) of picking the best 10, 15, 20%, etc for the various breeding traits. As such, this is very similar study to the one published earlier by the same authors:
Jing, Y.; Bian, L.; Zhang, X.; Zhao, B.; Zheng, R.; Su, S.; Ye, D.; Zheng, X.; El-Kassaby, Y.A.; Shi, J. Genetic diversity and structure 621 of the 4th cycle breeding population of Chinese fir (Cunninghamia lanceolata (lamb.) hook). Front. Plant Sci. 2023, 14, 1106615, 622 doi:10.3389/fpls.2023.1106615
Author Response
Please refer to the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
In the updated manuscript, the authors have carefully taken into consideration all of my feedback.
