The authors used niche and fitness differences of modern coexistence theory to understand how interactions between two herbivorous mites on wheat change through time. As more and more ecologists are interested in modern coexistence theory, the topic seems timely. However, I could not understand the motivation and results of this study very well.

Major comments:
1. Coexistence theory is to understand how multiple species coexist despite the competitive exclusion principle. However, the authors seem not interested in the question. Rather, they tried to understand how species interactions changed dynamically and utilized niche and fitness differences for the purpose. This should be carefully explained as it is very confusing and hard to understand. Readers may need to read the original paper that defined the specific niche and fitness differences used in this study (Spaak & De Laender 2020) to truly understand what Fig. 4 shows. It may be more straightforward to parameterize intra- and interspecific competition coefficients of the Lotka-Volterra model for different time periods or employ the Empirical Dynamic Modeling (e.g., Deyle et al. 2016 Proc. R. Soc. B) for understanding time-changing interspecific interactions.

1. The authors proposed that biotic resources (here wheat) produced the complex dynamics of niche and fitness differences, but there seems no direct support for this argument. It is not clear how wheat responds to consumption, and mites facilitate each other. What about constructing competition models with abiotic and biotic resources and calculate the two indices to show the reaction of resource to consumption produces the complex dynamics of niche and fitness differences? Also, it may be possible to apply their methods to existing datasets and show that results in previous studies with abiotic resources do not produce the complex dynamics of niche and fitness differences.

2. Pastore et al. (2021) Nat. Ecol. Evol. also showed that how niche and fitness difference change through time theoretically in the process of ecological character displacement. The authors need to carefully compare their results to this theoretical study and discuss the implications.

3. Please add line numbers. It is hard to make specific comments without them.

Specific comments:
Title: How do niche and fitness differences explain higher-order interactions? It would be great to clarify it more.
P4
L2: How is biodiversity driven by the ability of species to coexist? Please clarify it.
L2-3: "MCT explains this" is very vague and it is hard to understand what the authors mean. Please clarify it more.
L5: It would be great to clarify how consumption rates change over time. I first thought it may be due to rapid evolution or phenotypic plasticity of consumers, but in this manuscript, it is due to reaction of plant resources to consumption, right?
L8: It may be better to mention that the authors considered (non-standard) niche and fitness differences recently proposed by one of the authors.
L10: I guess most readers wonder whether it is possible to consider facilitation in the context of coexistence theory. It would be great to add more explanations.
L10-12: Again, it would be great to clarify that MCT is used not for understanding coexistence mechanisms, but for understanding dynamic interspecific interactions here.
P5
L7: Why did the authors cite Sutherland et al. (2013)? It is a paper for applied ecology (conservation), not for basic ecology (coexistence theory).
L7: Why 200 years? The topic is not so old.
L17: "Differences in adaptation" sounds very vague. Please clarify it more.
P6
L4: Li Shao-Peng → Li
L5: Please clarify that Terry et al. (2021) considered apparent competition.
L5-6: It may be better to clarify the definition of basal species and why bacteria and yeasts are not basal species.
L9: For example, the classical paper on competition and coexistence (Rothhaupt 1988 Nature) considered competition of rotifers on algae. What about applying MCT to the data of Rothhaupt (1988)? Similarly, Beninca et al. (2009) Ecol. Lett. considered two zooplankton species competing for two algal species. Why is the authors' focus limited to MCT?
L14: What about citing the paper on HSS hypothesis? In the context of HSS hypothesis, the world is green and there is no resource competition between herbivores? Also, a recent theoretical paper suggests that defense against consumers may promote stable coexistence (van Velzen 2020 Ecology) as originally suggested by Pimentel (1968) Science.
L25-2: Please explain the mechanism behind it more carefully.
P7
L3-12: I guess this paragraph on higher-order interactions can be moved to Discussion to clarify the focus of this manuscript.
L16-17: It would be great to add more explanations on the relationship between invasibility and stable coexistence.
L19-22: This point should be explained in Abstract as well.
P8
L1-2: The first question is not clear. What is the difficulty? When can we say we failed to apply the methods of MCT?
L3: evolve → change
L5-6: The authors need to explain that niche and fitness differences defined by Spaak & De Laender (2020) are different from the original and standard definition. It will be also great if the authors can explain the current situation of the new definition (e.g., how many papers are using it so far).
L12: Please add more explanations on obligatory mutualism. Also, it would be great if the authors can clarify the relationship between their method and those proposed by Johnson (2019) "Coexistence and competitive exclusion in mutualism" Ecology.
L19-25: It may be better to mention Box 1 around here.
P12
L14-16: I was confused: fitness differences are defined in the context of interspecific interactions, right? Is this contradicting with the "negative intrinsic growth rate", which is a characteristic of a single species?
L20: r and mu are used in the previous section. What about using different notations?
L24: two the → the two
P14
L2: Typo? And Fig. 2 is not a semi-log plot, right?
L5-7: This part may be moved to Discussion.
L15: What does "mal-adapted" mean?
L18-20: What is the objective definition of the three phases? Why "Therefore"?
P15
I feel this section is hard to understand because I am (and I guess most readers are) not familiar with the niche and fitness differences defined recently by one of the authors. What about showing intraspecific and interspecific competition coefficients instead of the very original niche and fitness differences?
L2: Where are the definitions of N and F?
L3: What is "the previously qualitatively observed mechanism"? What is the definition of mechanism here?
P16
L11: Please add explanations on Letten & Stouffer (2019) as well as Spaak et al. (2023).
L18: by (Spaak & De Laender 2020) → by Spaak & De Laender (2020)
L20-22: This was shown by recent theoretical studies (van Velzen 2020 Ecology, Pastore et al. 2021 Nat. Ecol. Evol., Yamamichi et al. 2022 Ecol. Lett.).
P17
L1-2: Is it possible to quantify the strength of facilitation without using niche and fitness differences?
L8: (Spaak et al. 2021a) → Spaak et al. (2021a)
L8-13: This part is inconsistent and hard to understand.
L14-16: This was done by Pastore et al. (2021) Nat. Ecol. Evol.
L20: It may be better to avoid the subjective expression, "we were surprised".
P19
L13: Spaak (2021b) → Spaak et al. (2021b)
P29
Box 1: The authors did not mention Box 1 in the main text.
P35
Figure 5: Where is the coexistence region in this figure? The authors mentioned this figure just once in the main text, so it may be possible to move this to Appendix.

The manuscript was improved through revision, but there are still weak points to be addressed.

Major comments:
1. I still think Fig. 4 and its explanations (P15) are hard to understand and may not be the best way to see how intra- and interspecific interactions change dynamically from facilitation to competition. Although the authors replied that EDM cannot be used to transient dynamics, but there are several applications (e.g., Gibert et al. 2022 Funct. Ecol.). Also, there are many studies parameterizing dynamic models based on time-series data (e.g., Turchin et al. 2003 Ecology, Rosenbaum & Fronhofer 2023 Ecosphere), so I could not get the authors' point. Thus, I am not convinced that using the "dynamic" niche and fitness differences, which is hard to interpret, is the only way for understanding dynamic interactions.

1. It would be better to revise the manuscript intensively to avoid the confusion around coexistence theory and interaction dynamics. The authors are not interested in stable coexistence and focus on dynamically changing interactions. However, the authors still explain Modern Coexistence Theory, niche and fitness differences, etc., which causes confusions. Abstract is especially hard to understand for readers who are not familiar with this topic. The authors did not explain "higher-order species interactions" in Abstract, although it comes first in Title! The authors did a better job in P7 and 20, so Abstract should be carefully revised (and probably lengthened). I even feel that the authors should not use the terms "niche" as their definitions are very different from the traditional terms. Their "niche" differences can even represent predation (P12), but what does "niche" mean in this context?

2. Because of the complex context of this study, I feel this manuscript can be lengthened to fully explain the background, methodology, and implications. Thus, I feel Ecology Letters may not be the best outlet of this study. Journals with longer formats will be better to explain the complexity very well.

Specific comments:
Title and Abstract: It would be better to explain the important term "higher-order species interactions" in Abstract.
P4
L4: MCT focuses → MCT tends to focus
L5: Or "functional responses are constant over time"?
L8: This part needs to be clarified more carefully. The authors explained that niche and fitness differences are calculated based on invasion growth rates in L3, but then the authors calculated them over time. This can be very confusing.
L8-10: Again, this needs more explanations. The authors explained that MCT explains species coexistence, but then the authors used niche and fitness differences for inferring species interactions over time.
L10-13: It is not clear how biotic resources and non-constant consumption rates affected the species interactions in this study.
P5
L4: biodiversity species → biodiversity, species
L9-10: May (1972) and Allesina & Tang (2015) are papers for the complexity-stability relationship in food web dynamics and can be removed from here.
L21-22: This is confusing. Not only the conversion efficiency, but also the consumption rate and mortality rate can affect the outcome of competition.
L26: This is partly due to Hutchinson's original idea and the limited number of resources for plants and phytoplankton.
P6
L4: What are biotic resources for bacteria and yeasts?
L24-25: I was confused: do zooplankton species "consume the same phytoplankton" (L24) or "different phytoplankton" (L25)? Please clarify it carefully. The authors may be interested in De Roos et al. (2008) PNAS as well.
P7
L4-7: The logic here is hard to follow. Temporally changing per-capita interaction strength should be considered not as "a consequence" (L6) of "one possible formulation" (L4), but as a reality in nature.
L13: positive invasion growth rate → positive invasion growth rates of competing species
L14-20: I still think this point should be explained in Abstract.
L21: Wait, the authors' "main interest is not to predict the coexistence per se" (L16), right? Please check the logic more carefully.
L25-26: Unfortunately, the authors did not address my previous comment: "The first question is not clear. What is the difficulty? When can we say we failed to apply the methods of MCT?"
P8
L10-11: It would be great if the authors can clarify the relationship between positive and negative frequency-dependence and facilitation.
L23: 1996 → 1966? Please check P23, L7.
P12
L2: Priority effects due to stronger interspecific interactions than intraspecific interactions can also arise from competition for common resources. It would be great if the authors can explain (ii) in a better wording.
L4-5: Personally, I feel the authors should not use the term "niche" for describing predation. This is very confusing.
L6-12: I feel this is also confusing as many people may think "fitness difference" as the difference between competing species, but here it seems the difference between the actual invasion growth rate and hypothetical invasion growth rate.
P13
L12-15: How did the authors define the three phases? It works for the WCM trajectories (Fig. 2b, d), but the stable phase is not clear in the CRM trajectories (Fig. 2a, c). From day 20 to day 25, population densities were still increasing.
L15-16: Again, what is the definition of difference here? Fig. 3a and 3c (and Fig. 3b and 3d) are qualitatively similar, I think.
L19-20: This may be partly due to the initial decline (especially for the WCM monoculture, Fig. 2d). Are there any potential explanation for the decline? What will happen when the authors remove the data around day 5?
P14
L3-4: How did the authors know this? The outcome of competition can appear after a certain time-delay. Indeed, the population growth rates started to decline around day 10-15 (Fig. 3).
L9-10: Where is the data of the amount of resource?
L22-23: Where should readers check in Fig. 4 to see the intraspecific facilitation?
P15
L3-21: I still feel this is the most difficult part to understand in the manuscript. It would be great if the authors can graphically show how interactions temporally change (intra- and interspecific facilitation and competition) through time. And again, the authors did not mention Fig. 4e and f. They may be moved to Supplementary Information.
P16
L6-7: What does "such an assumption" mean?
L8: It would be better to clarify how biotic resources result in higher order interactions here as well.
L8-10: Do "non-constant species interactions" imply "higher order interactions"?
L17-19: This needs more explanations. The authors showed it theoretically with some assumptions, right?
L19-21: I was wondering whether Fig. 2 truly shows facilitation or not. Does facilitation include Allee effect here?
P17
L9-11: Pastore et al. (2021) considered so-called "rapid evolution" where ecological and evolutionary timescales are not distinguished (see Fig. 1 and P333, right column).
P20
L2: What about Terry & Armitage (2023) bioRxiv?
L17-19: This part should be explained in Abstract.
P30
L3: What does "Gause n.d." mean?
L4: What about citing Simha et al. (2022) Am. Nat.?
L16: 2020 → 2021
L24: it's → it is
P38
L10-11: N was not always larger than F, but two species coexisted. How can we explain the pattern?