L165ff: This section describes supervised and unsupervised ML methods. Various soil parameters are used for PCA, such as BD, C, N, P, pH, COCO3 and others. The methods section does not describe the analysis of these parameters, nor does it provide references to other publications where these parameters have been analysed. These data are also presented in the Supplementary Table S1 for the different soil types and also for the horizons. However, it remains unclear where the data comes from or how they were measured.

When comparing the values for the different horizons, it is very remarkable that they are similar, sometimes down to the decimal place. From my experience  with soil analysis I am not familiar with such results. These parameters are not only used for the ML, but also used for the discussion, as they have a significant influence on the work presented. Since a significant part of the work presented is based on these values, the analysis of it should be presented in a comprehensible way.

Figure 2: pleae adapt the figure,  as it is hard to read the content (too small). Please indicate what the black dots represent (outlier or means? as the bars are quite small, it is also hard to distinguish between dots or letters)

This MS focuses on the research regarding the impact of wheat cultivar development on biomass production and carbon input in tillage-eroded soils. Through experiments setting different wheat cultivars and soil erosion gradients, it is found that soil erosion can affect wheat biomass and carbon input, and both newly developed and earlier cultivars have their own advantages and disadvantages in this regard. This conclusion provides a reference for cultivar selection and the sustainable development of agriculture. However, there are still some issues in this article:

1. When expounding on the research background, the writing logic is rather jumpy. It directly transitions from the global soil organic carbon content and the impact of tillage erosion on the carbon cycle to the acceleration of tillage erosion by agricultural machinery innovation, lacking necessary connections. For example, between "Annually, tillage erosion displaces around 0.5 Pg of this carbon pool (Quinton et al., 2010). Beyond the significant impact of nutrient loss through lateral movement (Alewell et al., 2020; Quinton et al., 2010), tillage erosion disrupts soil structure..." and "The drive for greater operational efficiency in crop production has led to machinery innovations that have accelerated the pace of tillage erosion...", the relationship between machinery innovation and the previous content is not explained, resulting in a logical break.
2. Although the introduction mentions multiple research background information, it fails to clearly show how this study fills the gaps in existing research. For instance, the article points out problems such as the uncertainty in estimating root carbon input and the scarcity of carbon balance studies considering tillage erosion, but does not clarify the unique perspective and specific entry points of this study in solving these problems. As a result, the research purpose is not clear and prominent enough, which is mainly reflected in the content elaborating on the deficiencies of existing research.
3. The description of the background information on the soil carbon cycle and tillage erosion in the introduction part is too long, and some content has a weak direct relevance to the research question, making the overall logic less compact.
4. Although the article details the measurement methods of indicators such as root biomass, there are problems in actual operation. The minirhizotron scanning may not be able to cover all roots completely. At the same time, some unconsidered interfering factors, such as differences in the particle composition and pore structure of different soils, may affect the morphology and distribution of roots, and thus influence the measurement of root biomass, leading to a certain degree of underestimation of the measured root biomass. In addition, when calculating the root biomass, it is assumed that the carbon content of above-ground and below-ground biomass is both 45%, and this assumption may not be consistent with the actual situation, affecting the accurate assessment of carbon input.
5. The figures suffer from issues such as excessive information and lack of clarity. The axis labels, legends, etc. are not clear and explicit enough, making it difficult for readers to accurately understand the data and trends. Moreover, the labeling of some data points and the representation of error bars in the figures are not standardized, reducing the readability and credibility of the data.
6. When analyzing the impacts of different soil types and wheat cultivars on biomass and carbon input, only descriptive statistical analysis is carried out, lacking in - depth exploration of the internal mechanisms. For the correlations between some results, no further causal analysis is conducted, making it difficult to reveal the underlying biological processes and ecological mechanisms.
7. In the experiment, the soil moisture was measured, but the monitoring of soil nutrient dynamics was relatively scarce. The availability of soil nutrients directly affects plant growth and biomass allocation, and thus influences soil carbon input. At the same time, when simulating the impact of soil moisture on root biomass, some complex soil and plant physiological processes may be simplified, affecting the accuracy of the simulation results.
8. Root exudates, as an important medium for the interaction between plants and soil microorganisms, have a significant impact on soil carbon input and soil health. However, the experimental design of this article does not measure root exudates, which may lead to an underestimation of root carbon input.
9. When discussing the research results, only partial consistencies with previous studies are briefly mentioned, and the similarities and differences between this study and other similar studies are not systematically compared. For example, when analyzing the impact of soil erosion on wheat biomass, the differences between the results of this study and those of other studies under the same or similar conditions are not compared in detail, making it difficult to highlight the value of this study.
10. The discussion part has relatively few discussions on the limitations of the research, and does not fully explain the deficiencies of the research and their impacts on the results. For example, this study was only conducted at one experimental site with a limited sample size. Future research should increase the sample size and experimental sites to verify the universality of the results.