Climate-Induced Shift in the Population Dynamics of Tortrix viridana L. in Ukraine

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper summarizes data on the occurrence of outbreaks of one important insect “pest” species of forest oak trees, collected over decades in Ukraine. The data appear to show that the area affected by serious outbreaks has diminished substantially over time, and that climate change effects may have contributed to this trend by increasing the phenological mis match between oak bud break and hatching of caterpillars from the eggs.

I am convinced that the authors, as experts on their subject, tell a “true story” here. However, I am seriously concerned as to the reproducibility of the study, since many data are not shown to the readers and/or underlying concepts are not fully disclosed (see below for details). I also would like to stress that the analysis of time series data requires the use of appropriate statistical models and tools, and NOT elementary frequentist hypothesis test (which all axiomatically expect the stochastic independence of the individual data points). This latter stochastic rationale is severely violated when using, for example, OLS regressions or Kruskal Wallis tests to analyse time series data.

Therefore, I urge the authors to (a) show all required basic data (following the principle of “Open Science” and (b) to carefully re-analyse their data using appropriate time-series statistics.

Once this is done, I expect this paper to become an important regional contribution to understanding climate change effects on forest pests in Europe.

L 56: does oviposition occur singly or in clusters? Please write explicitly.

L 106: what is the scientific rationale behind dividing the data in these two arbitrary time slices?

L 132: typo, should read petraea

L 187: in the meantime, version 5.2.2 of this software has been distributed

L 193 and below: what is your criterion for an emergence of the TV to represent an 'outbreak'? Please explain & specify. Since ‘outbreaks’ are at the heart of various of your evaluations and conclusions, it must be made crystal-clear what constitutes an outbreak, and what does not.

L 213: in this Figure 1 nothing can be seen about timing and duration of outbreaks! Where is that shown to the reader? What is a 'reliable decrease' or increase?? Please explain! I have never come across these terms in literature about moth population dynamics.

L 217: I presume, this graph shows “Annual area…” – please write explicitly.

L 220 ff: can you support this putative 'association' over time, for example through calculation of a correlation coefficient and its associated test for statistical significance? As it stands, the reader is forced to ‘believe you’, no data are shown on this postulated association between insecticide application and TV populations. Without clear supporting data, this is just speculation – which does NOT belong in the results section.

L 227: where is that claim shown in your paper? How did you quantify synchronicity? I did not find anything about measuring synchronicity in your methods section.

L 298: your statement about inverse proportionality needs to be validated with numbers! Where is the correlation coefficient, was that significantly more negative than zero (this would be required to support your claim).

L 320: the nomenclatural valid species epitheton for Phigalia is pilosaria, not pedaria. See B Müller et al (2019) The geometrid moths of Europe, vol. 6, Ennominae II (Boarmiini, Gnophini, additions to previous volumes). EJ Brill, Leiden.

L 322: since six of the listed geometrids do NOT hibernate as eggs, but rather as pupae, your statement that they ‘mainly’ hibernate as eggs is not true.

L 329 and below: in this figure and in the subsequent paragraphs, how was ‘prevalence’ measured? This concept has not yet been explained in the methods section, so no reader could cross-check or even comprehend your statements here.

L 391: how was this ‘contribution’ of individual factors assessed? To my understanding, this would require a multivariate regression model, and the "contribution" would then be the R² value accounted for by each predictor variable. Where is this regression model to be found (i.e. variance partitioning) for the reader? I could not find it, and in the methods section, this was also not explained.

Figure 15: to me this analysis appears completely arbitrary. Obviously, you have tons of most valuable data behind your statements, but as it stands the analysis and presentation is inappropriate. WHY dividing the time series into two segments? Why assuming a linear trend (as you implicitly do when calculating an OLS regression)? Clearly these fascinating time series data need to be analysed using statistics developed for data showing temporal non-independence, also accounting for potentially non-linear responses or trajectories.

L 444-446: this sentence is interpretation of your findings and does NOT belong into the results section, but rather into the methods section. Where did you measure nitrogen content of oak foliage?

Comments on the Quality of English Language

Overall, the paper is easy to read but I noted minor errors in grammar and style here and there, thus a careful editing will improve the presentation.

Author Response

Dear Reviewer 1,

thank you very much for your beneficial comments and advices.

 

The manuscript "Climate-induced shift in the population dynamics of Tortrix viridana L. in Ukraine" is corrected according to them. All corrections are shown in Edit mode or marked in yellow.

We tried to improve all parts of the manuscript. New references added.

 

Reviewer:

I am convinced that the authors, as experts on their subject, tell a “true story” here. However, I am seriously concerned as to the reproducibility of the study, since much of the data is not shown to the readers and/or underlying concepts are not fully disclosed (see below for details).

Answer.

– Supplement Table S1.xlsx contains: the data on T.viridana outbreaks area in Ukraine in 1947-2025; outbreak parameters by regions of Ukraine (Number of outbreaks, Incidence of outbreaks; Duration of outbreaks, years; Specific area, ha/ thous. ha of oak stands; Intervals between outbreaks, years); flush-feeders complex (larvae per 100 shoots) for 1979–1989, 2001–2008 and 2018; month temperature, continentality index, and the dates of stable transition of air temperature over 5 and 10°C for 1978–2025.

Reviewer:

 I also would like to stress that the analysis of time series data requires the use of appropriate statistical models and tools, and NOT elementary frequentist hypothesis test (which all axiomatically expect the stochastic independence of the individual data points). This latter stochastic rationale is severely violated when using, for example, OLS regressions or Kruskal Wallis tests to analyse time series data. Once this is done, I expect this paper to become an important regional contribution to understanding climate change effects on forest pests in Europe.

Answer:

– We used time-series statistics to analyze population dynamics of foliage-browsing insects in the 1990s. We found common rhythms in the dynamics of solar activity (W index), meteorological elements (1894 –1999), and areas of the 10 most prevalent foliage browsing insects. We calculated the start years of the next outbreaks and confidence limits, and our predictions were confirmed.

The results were published in (Meshkova, V. History and geography of foliage browsing insect pests outbreaks. Majdan: Kharkiv, 2002. 244 p (In Ukrainian) [8] and several papers also in Ukrainian. In subsequent years, the negative impact of phytophages in young man-made plantations, and xylophagous insects in old forests attracted our attention more. However, phenological observations continued, and data have accumulated. Foliage-browsing insects in oak stands very rarely formed foci since the early 2000th, and the role of Tortrix viridana in the complex was low.

Now our goal was not to predict, but to explain the possible reasons for the sharp decrease in all indicators of TV population dynamics. Therefore, we compared two periods of the same length and analyzed the reasons for the changes.

 

L 56: does oviposition occur singly or in clusters? Please write explicitly.

– added: “...eggs are laid in pairs for a total of 50-60 per female....[1, 3]”.

L 106: what is the scientific rationale behind dividing the data in these two arbitrary time slices?

– According to publications and archival data, from 1853 to 2000, 32 TV outbreaks occurred in Ukraine with an average interval between outbreaks of 10±2 years (from 8 to 12 years). Since outbreaks are cyclical, the period for evaluating the outbreak parameters should be longer than the duration of the cycle of this insect species noted in previous years. We chose the same intervals for two consecutive periods, the first with cyclic outbreaks and the second almost without them.

L 132: should read petraea

– corrected

L 187: in the meantime, version 5.2.2 of this software has been distributed.

– Thank you, I’ve downloaded 5.2.2 and tested for our tasks.

L 193 and below: what is your criterion for an emergence of the TV to represent an 'outbreak'? Please explain & specify. Since ‘outbreaks’ are at the heart of various of your evaluations and conclusions, it must be made crystal-clear what constitutes an outbreak, and what does not.

 

Answer: added to Introduction

1. viridana, like some other foliage-browsing forest insects, is characterized by a cyclical population dynamic [...]. Periods of sudden and significant increase in the population above its general equilibrium level are called outbreaks. Outbreaks of TV, like some other foliage-browsing forest insects, begin in the most attractive forest subcompartments, which are called foci. During an outbreak, TV causes considerable foliage damage, which is usually predicted by current stand defoliation, egg mass counts, and considering forest features (tree species composition, age, relative density of stocking, etc.). The area of subcompartments with a risk of defoliation over 25% is summarized for forest enterprises and at the regional level. Insecticide treatment is planned for subcompartments with predicted defoliation over 50% [..].

 

L 213: in this Figure 1 nothing can be seen about timing and duration of outbreaks! Where is that shown to the reader? What is a 'reliable decrease' or increase?? Please explain! I have never come across these terms in literature about moth population dynamics.

 

Answer: Sorry, it is Figure 2, a technical mistake.

 

L 217: I presume, this graph shows “Annual area…” – please write explicitly.

Thank you, corrected.

 

L 220: can you support this putative 'association' over time, for example through calculation of a correlation coefficient and its associated test for statistical significance? As it stands, the reader is forced to ‘believe you’, no data are shown on this postulated association between insecticide application and TV populations. Without clear supporting data, this is just speculation – which does NOT belong in the results section.

 

Answer

Facts of the prolongation of outbreaks as a result of insecticide use are known in different regions. For example: Liebhold, A. M. (2012). Forest pest management in a changing world. International Journal of Pest Management, 58(3), 289-295.

“...In an even worse scenario, the application of pesticides to check population growth may interfere with the density-dependent mortality normally inflicted by natural enemies and actually prolong an outbreak”

The respective text has been moved to the discussion.

 

L 227: where is that claim shown in your paper? How did you quantify synchronicity? I did not find anything about measuring synchronicity in your methods section.

 

Answer: In this paper, we did not study synchronicity. We tried to find the reasons for the change in the outbreak area in the late 1980s. Many publications discuss the synchronicity of insect outbreaks in various regions [..]. After replacing the aerial insecticide treatment with a ground one, particular foci were not treated simultaneously, and individual populations responded differently, which affected the total foci area.  

The respective text has been moved to the discussion.

 

L 298: your statement about inverse proportionality needs to be validated with numbers! Where is the correlation coefficient, was that significantly more negative than zero (this would be required to support your claim).

 

 

Answer:

In Materials and Methods:  

Outbreak duration (years) was calculated as the ratio of outbreak years and the number of outbreaks for the respective period. The mean interval between outbreaks (years) was calculated as the ratio of the duration of the investigated period and the number of outbreaks during this period.

 

 

L 320: the nomenclatural valid species epitheton for Phigalia is pilosaria, not pedaria. See B Müller et al (2019) The geometrid moths of Europe, vol. 6, Ennominae II (Boarmiini, Gnophini, additions to previous volumes). EJ Brill, Leiden.

 

Answer: Thank you, corrected

 

L 322: since six of the listed geometrids do NOT hibernate as eggs, but rather as pupae, your statement that they ‘mainly’ hibernate as eggs is not true.

 

Answer: In Ukraine five of the mentioned geometrids hibernate as pupae: Alsophila aescularia (Denis & Schiffermüller, 1775), Apocheima hispidaria (Denis & Schiffermüller, 1775), Biston strataria (Hufnagel, 1767), Lycia hirtaria (Clerck, 1759), Phigalia pilosaria (Fabricius, 1787).

Moths emerge early in the spring when the forest litter thaws, at approximately 5°C, and eggs hatch at almost the same time as those of species that lay eggs in the fall.

The text was improved:

Some of these species hibernate as eggs. Several geometrids hibernate as pupae, swarm, and oviposit in early spring (approximately, on the dates of stable temperature transition over 5°C).

 

L 329 and below: in this figure and in the subsequent paragraphs, how was ‘prevalence’ measured? This concept has not yet been explained in the methods section, so no reader could cross-check or even comprehend your statements here.

 

Answer: added to the Materials and Methods:

Larvae were counted on oak branches in May, and the number of individuals of species or groups of species (leaf rollers, geometrids) was recalculated per 100 shoots of the current year. The prevalence of individual species or groups was evaluated as a percentage in the complex.

 

L 391: how was this ‘contribution’ of individual factors assessed? To my understanding, this would require a multivariate regression model, and the "contribution" would then be the R² value accounted for by each predictor variable. Where is this regression model to be found (i.e. variance partitioning) for the reader? I could not find it, and in the methods section, this was also not explained.

 

Answer:

A point assessment of the attractiveness of stands for individual species of forest defoliators, including Tortrix viridana, has been developed [] and later tested in various regions. (To regret, our publications in Russian or Ukrainian were not known to our foreign colleagues, and now the journals recommend minimizing self-citation of authors). The potential area of subcompartments in which an outbreak may occur is evaluated using the forest arrangement database. Tree species composition, age, and relative density of stocking in some subcompartments change over time due to natural or anthropogenic factors. The proportion of coppices decreases, mature stands are felled, etc. Then the total area in which conditions are favorable for the outbreaks of certain insects also changes. It can increase or decrease. We calculated changes in forest area in four forest enterprises, considering each of the analyzed indicators and their combined changes in individual subcompartments. The contribution of each indicator to the change in the total area attractive for the formation of outbreaks was evaluated as the percentage of shifted area due to changes in age, density of stocking, etc.

The databases of forest arrangement for each forest enterprise contain hundreds of subcompartments, and their boundaries are not always similar in the two inventories. Therefore, it is possible only to calculate the area with certain forest characteristics for the whole oak forest in each forest enterprise.

Some phrases are added to methods.

 

Reviewer

Figure 15: to me this analysis appears completely arbitrary. Obviously, you have tons of most valuable data behind your statements, but as it stands the analysis and presentation is inappropriate.

WHY dividing the time series into two segments?

Answer:

We added the data for the whole period: (c) 1978–2025.

If we consider the two periods together (Figure 15 c), we can notice a tendency for a faster decrease in the date of the bud flushing than the egg hatching, and close dates of these events in the early 2000s.

We divide the time series into two segments of equal length because the second period showed more pronounced changes for outbreak dynamics (see Figure 2).

 

Reviewer1

Why assuming a linear trend (as you implicitly do when calculating an OLS regression)?  Clearly these fascinating time series data need to be analysed using statistics developed for data showing temporal non-independence, also accounting for potentially non-linear responses or trajectories.

 

Answer: The regression line is not intended for forecasting; it only visualizes the trend. Indeed, in the second period, a trend of shifting in the ratio of oak bud flushing and TV egg hatching is visible.

 

L 444-446: this sentence is interpretation of your findings and does NOT belong into the results section, but rather into the methods section. Where did you measure nitrogen content of oak foliage?

Answer: Thank you. I agree. This sentence is placed in the Discussion.

 

Resume

We tried to explain the reasons for the decrease in the dynamics of TV outbreaks, which occurred in all regions of Ukraine over the past 20 years, using the most complete field data from the Kharkiv region.

We confirmed the continuous decrease in the percentage of TV in the flush feeders’ complex.

We compared data on the structure of stands in four forestry enterprises of the Kharkiv region for two periods between forest inventories (over 26 thousand subcompartments in each).

We analyzed the shifts in the dates of temperature rise over 5 °C and 10 °C in spring, oak bud flushing, and TV egg hatching periods for two periods.

Our calculation and analysis of publications shows that our assumptions are valid. Undoubtedly, we will deepen and improve the application of modern statistical methods in further studies, focusing on the biological meaning of the relationships between various events in forest ecosystems.

 

Reviewer: Comments on the Quality of English Language

Overall, the paper is easy to read, but I noted minor errors in grammar and style here and there, thus a careful editing will improve the presentation.

 

Answer: We tried to do it.

 

Thanks once more

correspondence author

Valentyna Meshkova

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The paper is interesting, considering a very long period in which biological changes can be observed. It might be useful to mention the levels of parasitisation by oophagous parasitoids, which I understand the authors have already taken an interest in. Also, Tortrix viridana is an elective prey of tits, and it would be interesting to know if there are any local publications dealing with the breeding of these birds in nestboxes. If there are none locally, it would be interesting to mention the topic and search the web for results on tits breeding and climate change over periods of at least 30 years.

The authors made a comparison between two periods (1978-2001 and 2002-2025), but could also have examined the whole period (1978-2025) to observe the slope of the line of both bud flushing and ovideposition of T. viridana. I would suggest that they do this and give this information; statistics will certainly show that the trend (=bud flushing) is statistically significant.

Suggestions of taxonomic importance: in some cases, square brackets could be used to indicate that the year of description was changed later from the original (see Archips crataegana); square brackets indicate that the year of description is different from the year of publication, or that the date is uncertain or changed. If the year of description is only an approximation or has not been precisely determined, the square bracket may be used to indicate that the year has been estimated (Hübner, [1796-1799]).

Finally, please check some notes written along the text.
 

 

Comments for author File: Comments.pdf

Author Response

Dear Reviewer 2,

thank you very much for your beneficial comments and advices.

 

The manuscript "Climate-induced shift in the population dynamics of Tortrix viridana L. in Ukraine" is corrected according to them. All corrections are shown in Edit mode or marked in yellow. We tried to improve all parts of the manuscript. New references added.

 

Reviewer

The paper is interesting, considering a very long period in which biological changes can be observed. It might be useful to mention the levels of parasitisation by oophagous parasitoids, which I understand the authors have already taken an interest in.

 

Answer.

Among the reasons for TV population decrease, we do not consider the shift in the role of entomophages because of the lack of corresponding surveys in recent years.

In 1978–1984, parasitoids of TV were studied in the Kharkiv region by S. Gamayunova.  

 Gamayunova, S.G. Bioecological features of parasitoids of green oak leaf roller. Forestry and Forest Melioration 1988, 76, 51-53.

A total of 64 parasitoid species were identified, including 38 primary ones. No oophagous parasitoids were found. Females of Ascogaster rufidens (Wesmael 1835) and Ascogaster quadridentata (Wesmael, 1835) (Hymenopera: Braconidae: Cheloninae) laid eggs in the host eggs, and larvae developed in caterpillars. Larval parasitoids Macrocentrus linearis (Nees, 1812), Oncophanes minutus (Wesmael, 1838) (old name Oncophanes lanceolator (Nees, 1834) and the specialized pupal parasitoid Dirophanes invisor (Thunberg, 1822) (old name Phaeogenes invisor (Thunberg. 1824)) were the most numerous. However, the infestation of TV larvae rarely exceeded 40%. It was concluded that the complex of parasitoids cannot be the main factor limiting TV population growth, due to the absence of oophagous, dominance of polyphagous, multivoltine species, and many secondary and primary-secondary entomophages.

Polyphagy of egg parasitoids is described in various regions. At high numbers of phytophagous insects, their infestation with parasitoids increases, but after the outbreak collapses, parasitoids survive due to polyphagy. Text is added to Discussion.

 

Reviewer 2.

Also, Tortrix viridana is an elective prey of tits, and it would be interesting to know if there are any local publications dealing with the breeding of these birds in nest boxes.

 

Answer: A. Chapligina [Chaplyhina Thesis] studied the dynamics of spring detection, reproductive period, and diet structure of 36 species of dendrophilic Passeriformes (including Parus major) in 1982–1991 and 1991–2017 in the area of our research in the Kharkiv region. The first detection of 26 species of dendrophilic Passeriformes occurred earlier by 2–14 days. Parus major feeds fledglings with caterpillars of available phytophages, including Lepidoptera (Tortricidae, Geometridae, etc.). Fledglings appear in the 3rd decade of May. (VM – when T. viridana larvae are in the late instars).

 

Reviewer 2

If there are none locally, it would be interesting to mention the topic and search the web for results on tits breeding and climate change over periods of at least 30 years.

 

Answer:

An analysis of publications shows that tits do not exhibit high selectivity to the species composition of food. When the number of one phytophage species is high, its individuals predominate in the diet of insectivorous birds. When the number is low or one species is absent, the birds change their food preferences [36].

 

References  79–85 and the text added.

For 1976–2019, in France, egg laying date of blue tits (Cyanistes caeruleus, Passeriformes: Paridae) significantly advanced by about 14 days (Biquet et al., 2022).

Notable mismatch has been seen between the great tit (Parus major), which feeds its offspring on winter moth (Operophtera brumata) and oak leaf roller (Tortrix viridana) caterpillars in the Netherlands (Visser, Holleman & Gienapp, 2006). Peak caterpillar biomass of these two species has advanced earlier in the season, so this bird species may need to breed earlier to align with peak caterpillar abundance or find other available foods.

A long-term study of Parus major population in the Netherlands (1973 to 2020) and considering predicted climate change show, that prey phenology will advance faster than great tit phenology in the coming decades, and therefore that long-term global warming will intensify phenological mismatches (Visser et al., 2021).

Marsh Tits Poecile palustris in Białowieża National Park (Poland) commenced egg laying earlier in warmer springs (Wesołowski 2023). Egg laying commenced and breeding finished ahead of most other local bird species. Marsh Tits reproduced successfully even in years of very low caterpillar abundance. However, caterpillar abundance became important only in colder springs, when higher brood reduction rates and delayed fledging were observed.

Using a 45-year data set, it was shown that vegetation phenology at very local scales is the most important predictor of the timing of breeding within years, suggesting that birds can fine-tune their phenology to that of other trophic levels (Hinks et al., 2015).

VM:  In our opinion, it is easier for birds to adapt to other types of food in the absence of TV caterpillars than for monophagous caterpillars to adapt to another host tree.

 

Reviewer 2

The authors made a comparison between two periods (1978-2001 and 2002-2025), but could also have examined the whole period (1978-2025) to observe the slope of the line of both bud flushing and oviposition of T. viridana. I would suggest that they do this and give this information; statistics will certainly show that the trend (=bud flushing) is statistically significant.

 

Answer:

– We added the data for the whole period: (Figure 15 c) 1978–2025.

If we consider the two periods together (Figure 15c), we can notice a tendency for a faster decrease in the date of the bud flushing than the egg hatching, and close dates of these events in the early 2000s.

We divide the time series into two segments of equal length because the changes in the outbreak dynamics were most pronounced in the second period (see Figure 2).

The regression line is not intended for forecasting; it only visualizes the trend. Indeed, in the second period, a trend of shifting in the ratio of oak bud flushing and TV egg hatching is visible

 

Reviewer 2

Suggestions of taxonomic importance: in some cases, square brackets could be used to indicate that the year of description was changed later from the original (see Archips crataegana); square brackets indicate that the year of description is different from the year of publication, or that the date is uncertain or changed. If the year of description is only an approximation or has not been precisely determined, the square bracket may be used to indicate that the year has been estimated (Hübner, [1796-1799]).

 

Answer: corrected

 

Finally, please check some notes written along the text.

Answer: Thanks, we considered them.

 

line 66: added species ballooning

added: ... Larvae of some species (for example, Lymantria dispar (Linnaeus, 1758), Biston strataria (Hufnagel, 1767), Erannis defoliaria (Clerck, 1759), Operophtera brumata (Linnaeus, 1758) can migrate by ballooning [12, 29] lines 75-78

 

line 82 – corrected

 

A legend is added to Figure 1.

 

line 221: Insecticides used in this way destroy all phytophagous, predators and parasitoids, causing long-term damage to the forest food web.

 

Answer:

Yes, insecticides were widely used in the former USSR; they are used now in many countries with large outbreak forest areas. However, after 1990, insecticides were not used in the oak forests of Ukraine. Regulation of TV by phytophagous insects and birds may cause a decrease in outbreak severity, but we do not have supporting facts. (See earlier answers on oophagous parasitoids and birds, as well as Liebhold, 2012)

 

Figures 8–11. corrected.

lines 316, 320: names of moths corrected

 

lines 433–435. Regrettably this does not allow to conclude that TV hatching changed. Why not search for a change in a long-term period (1978-2025)?

figure (c) is added. See earlier answer.

 

lines 467–469: authors of species in some cases must be in parentheses and never abbreviated (only Linnaeus may be abbreviated into L.); I suggest do not write authors, being already written above

 

Answer: Agree. Corrected.

 

Thanks once more

 

correspondence author

Valentyna Meshkova

 

 

 

Author Response File: Author Response.pdf