Bioclimatic modeling of Avenella flexuosa (L.) Drejer distribution in connection with its possible polemochoral origin on the territory of Central Russia

Размещено на http://www.allbest.ru/

Bioclimatic modeling of Avenella flexuosa (L.) Drejer distribution in connection with its possible polemochoral origin on the territory of Central Russia

E.O. Korolkova1, 2, Ya.E. Vasilkov1

Moscow Pedagogical State University,

Moscow, 119991, Russian Federation

National Research University «Higher School of Economics», Moscow, 105066, Russian Federation

For the first time, the method of bioclimatic modeling was used to determine the boundaries of the range of occurrence of a potential polemochoral species Avenella flexuosa (L.) Drejer. The simulation was performed by the method of maximum entropy (MaxEnt). It is shown that the conditions for the growth of this species are initially suitable (the probability of finding 0.54 and higher) in the territory of Central Russia. For comparison, the same bioclimatic modeling approach was applied for the occurrence data of Carex brizoides L. species, the polemochoral origin of which is not in doubt. It is shown that its growing places are beyond the limits of the climatic optimum (the probability of finding 0.08) in the territory of Central Russia.

Acknowledgments. The work was supported by the Russian Foundation of Basic Research Grant 18-04-01206 «Analysis of modern dynamics of the flora of Western European Russia on the example of migration of polemochore species (plants introduced during the Second World War)».

Key words: polemochore plants, alien plant species, MaxEnt, biological diversity, WWII

Биоклиматическое моделирование распространения Avenella flexuosa (L.) Drejer в связи с ее возможным полемохорным происхождением на территории Центральной России

Впервые использован метод биоклиматического моделирования для определения границ ареала потенциального вида-полемохора ЛуепеНа Иехиоза (1_.) Оге_)'ег. Моделирование выполнено методом максимальной энтропии (МахЕгЦ. Показано, что на территории Средней России условия для произрастания этого вида изначально подходящие (вероятность нахождения 0,54 и выше). Для сравнения был взят вид Сагех ЬгиоШеБ _., полемохорное происхождение которого не вызывает сомнений. Показано, что его места произрастания на территории Средней России находятся за границами климатического оптимума (вероятность нахождения 0,08).

Благодарности. Работа поддержана грантом РФФИ 18-04-01206 «Анализ современной динамики флоры запада Европейской России на примере миграции видов-полемохо- ров (растений, занесенных во время Второй мировой войны)».

Ключевые слова: растения-полемохоры, заносные виды растений, МахЕгЦ биологическое разнообразие, Вторая Мировая война

Introduction

The polemochoral is the type of invasive plants, penetrated to a particular territory as a result of military actions. For a number of reasons [Shcherbakov et al., 2013] this topic was not the subject of any noticeable attention of botanists in Russia. Only recently, thanks to the efforts of A.N. Sennikov (2012), it attracted the interest of researchers.

Unfortunately, the study of this group of plants is becoming more and more difficult: over 70 years passed since the Second World War, and it is very hard to separate the plants included in the Central Russia as a result of military actions at that time from those who got there earlier (in particular, as a result of appeared parks and gardens) or later (with seed material, exotic species, etc.). Moreover, as we move from North to South, this complexity increases.

The identification of such plants is not so difficult today in the Murmansk region and Karelia, where the front line ran away from the main roads, and the area occupied by agricultural land is extremely small. It is applicable both to species listed by the Wehrmacht and its allies and to the plants listed by the Red Army. Therefore, the lists of the polemochoral plants are quite extensive there [Sennikov, 2009, 2012].

On the contrary, identification of the polemochoral plants is much more complicated in the Russian non-chernozem region. First, there are many manor parks or their remnants with the Western and Central European plants, many of which were used in the landscaping in the XIX century. Secondly, many plants from Southern and South-Eastern Russia, as well as from the Trans-Urals, came there with seed material and forage in the second half of the XX century. Therefore, at present, it is almost impossible to separate the species that entered the non-chernozem region in the post-war period from the drifts associated with the conduct of military operations by the Red Army. Clarification of a “polemochore” nature of a species related to maneuvers of the German troops requires the use of special “filters” [Shcherbakov et al., 2013], a thorough analytical work, including military nature, good knowledge of the history of that period, and sometimes just good luck [Reshetnikova, 2014].

Avenella flexuosa (L.) Drejer. can be found in the most areas of non- tropical Eurasia, North Africa, North America, in the highlands of tropical Africa and Asia and in southern South America. In Europe, it is distributed geographically from Scandinavia to Gibraltar, except for the forest-steppe and steppe parts. In the territory of Russia the species is distributed in the Arctic regions of Europe and Siberia, in the Caucasus, in the NorthWestern European part, more rarely found in the far East, in Eastern and

Western Siberia [Zhukova, 1980]. It is considered as Holarctic-Alpine species with a very broken distribution [Tsvelev, 1974]. Presumably it could have originated in the pre-glacial era, and live in post-glacial coniferous and birch forests.

This type of plant is characterized by the following type of soil: coarse sandstones, poor acidic soils, gravel, sandy shale, rarely found on clay soils and completely absent on calcareous soil [Larin et al., 1950]. Most often, Avenella flexuosa is found on poor, leached sandy loam and sandy soils of mainland meadows, upper transitional swamps, and pine forests. It is worth noting that nitrates are almost absent and the level of calcium carbonate is very low in such soils.

This type of plant prefers a soil with an average level of humidity, such as forest zone, highlands, forest tundra, etc. Places with a high degree of moisture can adversely affect it. It calmly reacts to a decrease in the level of humidity in the soil, can not tolerate long-term drought, which is why this plant can be found on the slopes and developed peat bogs with a high probability.

А flexuosa is also common in cuttings, in high-altitude and continental meadows where it grows, flowers and fruits during the period of maximum illumination [Zhukova, 1980]. The meadow has the ability to resist the wind, so it can safely occupy small-to-medium sized hills, such as tree stumps, boulders, slopes. Also, these types of habitat contribute to faster development due to the absence of a forest litter and a blackening. The species is able to withstand relatively long extremely low temperatures (up to -50°).

We often found A. flexuosa next to other species introduced by the German army in the territory of Central Russia [Shcherbakov et al., 2017; Reshetnikova et al., 2018, 2019]. At the same time, after a careful analysis, some of the finds can be attributed to location of an old manor (https://plant.depo.msu.ru/ module/itempublic?d=P&openparams=%5Bopen-id%3D140984645%5D). Thus, we set our goal to find out whether the points of this species found by us are located within the boundaries of the natural areal or outside of it.

Materials and methods

Georeferenced occurrence data for A. flexuosa were accessed from the online database GBIF.org: all finds of this species with coordinates, total 461 617 records (GBIF Occurrence Download https://doi.org/10.15468/ dl.3j9zv0 on 28th October 2019), preserved specimen only, total 4736 records (GBIF Occurrence Download https://doi.org/10.15468/dl.ywqlg8 on 28th October 2019). Georeferenced occurrence data for Carex brizoides L. were accessed from online database GBIF.org: all finds of this species with coordinates, total 53 837 records (GBIF Occurrence Download https://doi. org/10.15468/dl.kyv2wz on 31th October 2019), preserved specimen only, total 508 records (GBIF Occurrence Download https://doi.org/10.15468/ dl.acq19w on 31th October 2019).

These datasets were used to identify the actual and potential areas of distribution of Avenella flexuosa. That was done in order to understand whether this species findings in non-chernozem regions are related to its possible polemochoral origin or it is rather the natural process of the species distribution to a new suitable habitats. As a reference we used presence data of Carex brizoides which is a polemochoral plant without a reasonable doubt [Sennikov, 2009; Reshetnikova et al., 2019].

We built a model of a probable area distribution using MaxEnt v3.4.1 which is a program for modeling a species distributions from presence-only records. The approach of modeling used in MaxEnt is based on minimizing the relative entropy between probability densities of presence data and environmental characteristics of presence sites in covariate space which is equivalent to maximizing entropy in geographical space [Elith et al., 2010]. This approach works well when presence-only data is available (no absence data) and the potential distribution area can be modeled with a limited number of points of registered species findings.

The distribution of suitable habitats of the species was evaluated by MaxEnt according to the principles of maximum entropy [Phillips et al., 2006; Phillips, Dudik, 2008]. The software can be downloaded at. A detailed description of the principles and procedures is available online in Russian.

The parameters selected were: `Auto features', percentage of test sample = 25%, maximum number of iterations = 50 000, cross-validation procedure, background at random in the entire study area. We considered the model runs to be accurate based on AUC statistic value achieved (0.973 for Avenella flexuosa and 0.988 for Carex brizoides), which represents capability of a model to distinguish between classes, habitat suitability of geographical point in our case.

Eleven climactic variables with 30 arc-min resolution from IPCC were utilized as environmental layers in MaxEnt 3.4.1. [Phillips et al., 2006]:

- diurnal temperature range, annual (dtr6190_ann);

- frost frequency, annual (frs6190_ann);

- precipitation, annual (pre6190_ann);

- precipitation, January (pre6190_I1);

- precipitation, April (pre6190_I4);

- precipitation, July (pre6190_I7);

- precipitation, October (pre6190_I10);

- mean temperature, annual (tmn6190_ann);

- minimum temperature, annual (tmp6190_ann);

- maximum temperature, annual (tmx6190_ann);

- vapour pressure, annual (vap6190_ann).

The modeling outcome is presented as heat maps which display a probability of species distribution across the globe. Warmer colors represent better conditions for a species which can be interpreted as a higher probability of presence. The colors display the calculated probabilities according to the scale: warmer colors show areas with better predicted conditions; from 0,69 (yellow) to 0.92 and more (orange and red); green represents conditions similar to those in which the species was collected on the territory of Central Russia (~0.5-0.6); blue shades indicate unlikely conditions.

Results and discussion

The relative contribution of environmental variables to the MaxEnt model was estimated as a result of the modeling process (Table 1). To determine the percentage of contribution, in each iteration of the training algorithm, the increase in regularized gain is added to the contribution of the corresponding variable or subtracted from it if the change to the absolute value of lambda is negative. For the permutation importance, for each environmental variable, in turn, the values ofthat variable on training presence and background data are randomly permuted. The model is reevaluated on the permuted data, and the resulting drop in training AUC is shown in the table, normalized to percentages. So, the most influential variables (with correction on the permutation importance) are “mean temperature, annual” (tmn6190_ann) and “maximum temperature, annual” (tmx6190_ann).

To determine the importance of environmental variables on Avenella flexuosa potential distribution, another statistical approach was used. The following picture (Figure 1) shows the results of the jackknife test of variable importance [Efron, Stein, 1981].

The environmental variable with the highest gain, when used in isolation, is “mean temperature, annual” (tmn6190_ann), which therefore appears to have the most useful information by itself. The environmental variable that decreases the gain the most when it is omitted is “maximum temperature, annual” (tmx6190_ann), which therefore appears to have the most information that isn't present in the other variables. These results on variable importance should be interpreted with caution though since these variables are correlated.

Table 1 Analysis of variable contributions for Avenella flexuosa models, showing: Percent contribution (PC) and Permutation importance (PI),discussed in the text

polemochoral avenella flexuosa geographic

The results of bioclimatic modeling for Avenella flexuosa are presented in the figure 2, and for Carex brizoides in the figure 3. It is shown that the conditions for the growth of Avenella flexuosa are initially suitable in the territory of Central Russia (the probability of finding 0.54 and higher). For Carex brizoides it is shown that its growing places are beyond the limits of the climatic optimum in the same territory (the probability of finding 0.08).

White dots show the presence locations, purple dots show the records which used for program training. The probability of the species presence is reflected by logarithmic color scale, see lower right corner. On the inset in the lower left corner, a yellow rectangle indicates the area of finds in Central Russia.

There is a number of indications that currently Avenella flexuosa can be attributed to progressive species that are expanding their natural areal [Seregin, 2015]. The same can be attributed to the growth of this species outside the boundaries of the areal, taken as natural, where it can be introduced in various ways, including polemochoral. For example, during the work on the compilation of the floral check-list in the National.

Park “Orlovskoe poles'e”, Avenella flexuosa was fixed only at one point (https://plant.depo.msu.ru/module/itempublic?d=P&openparams=%5 Bopen-id%3D140984661%5D). In 2001, only one small group grew on the sandy side of a forest road. In the summer of 2019, we re-examined this place, and found that not only the roadside, but also the nearby hillfort of the Verkhneoksky Balts Radovishche (https://orlpolesie.ru/sokhranyat/ unikalnye-ob-ekty/istoriko-kulturnoe-nasledie) covered with A. flexuosa almost entirely, and the total population area was about 0.5 hectares (E.O. Korolkova, L.L. Kiseleva, observation). The most dense plants were located on the top, where regular archaeological excavations were carried out in the late 2000s. During which the living ground cover was removed, surfacing the sand from the excavations. All this is in good agreement with the previously described [Zhukova, 1980] ecological preferences of this species: poor sandy acidified soils, and also confirms its role as a plant of the initial stages of succession.

Thus, the conducted bioclimatic modeling confirms that the conditions for the growth of A. flexuosa are initially suitable (the probability of finding 0,54 and higher) in the territory of Central Russia, including the places of occurrence with alleged polemochoral origin.

References

1. Efron B., Stein Ch. The Jackknife Estimate of Variance. The Annals of Statistics. 1981. Vol. 9. No. 3. Pp. 586-596.

2. Elith J., Phillips S.J., Hastie T. et al. A statistical explanation of MaxEnt fore- cologists. Diversity and Distributions. 2010. Pp. 1-15.

3. Larin I.V. et al. Forage plants of hayfields and pastures of the USSR [Кормовые растения сенокосов и пастбищ СССР / Ларин И.В. и др. М.; Л., 1950-1951. Т. 1-2.]. Moscow; Leningrad, 1950-1951. Vol. 1-2. (In Russ.)

4. Phillips S., Anderson R., Schapire R. Maximum entropy modeling of species geographic Distributions. Ecological Modelling. 2006. Vol. 190. Pp. 231-259.

5. Phillips S.J., Dudik M. Modeling of species distributions with Maxent: New extensions and a comprehensive evaluation. Ecography. 2008. Vol. 31. Pp. 161-175.

6. Reshetnikova N.M. The road to German headquarters: The emergence of some Western European types on the territory of Kaluga region. Invazivnaya biologiya: sovremennoe sostoyanie i perspektivy: Materialy rabochego soveshchaniya, Moskva, 10-13 sentyabrya 2014 g. Moscow, 2014. Pp. 134-140. (In Russ.) [Решетникова Н.М. Дорога в немецкую штаб-квартиру: появление некоторых западноевропейских типов на территории Калужской области // Инвазивная биология: современное состояние и перспективы: материалы рабочего совещания, Москва, 10-13 сентября 2014 г. М., 2014. С. 134-140.]

7. Reshetnikova N.M., Shcherbakov A.V., Fadeeva I.A. On the flora of “Krasny Bor”, a unique protected area of the Smolensk region. Vestnik Tverskogo gosudarstvennogo universiteta. Seriya: Biologiya i ekologiya. 2018. No. 4. Pp. 160-190. (In Russ.)

8. Reshetnikova N.M., Shcherbakov A.V., Korolkova E.O. Central European species in vicinity of Kobelevo village (Kaluga region) as the traces of the Great Patriotic War. Botanicheskij zhurnal. 2019. Vol. 104. No. 7. Pp. 1122-1134. (In Russ.)

9. Sennikov A.N. Ado Haae (1934-2008), a prominent Estonian naturalist in Russia, and his Theory of Wonderglades. Memoranda Soc. Fauna Flora Fennica. 2009. Vol. 85. Pp. 61-67.

10. Sennikov A.N. The bitter memory of the earth: plants-paleohori in Eastern Fennoscandia and Northwestern Russia. Problemy izucheniya adventivnoy sinantropnoy flory Rossii i sosednih stran: Materialy IV Mezhdunarodnoy nauchnoy konferentsii (Izhevsk, 4-7 dekabrya 2012 g.). Izhevsk, 2012. Pp. 182-185. (In Russ.) [Сенников А.Н. Горькая память земли: растения-палеохоры в Восточной Фенно- скандии и северо-западе России // Проблемы изучения адвентивной синантроп- ной флоры России и соседних стран: Материалы IV Международной научной конференции (Ижевск, 4-7 декабря 2012 г.). Ижевск, 2012. С. 182-185.]

11. Shcherbakov A.V., Kiseleva L.L., Panasenko N.N., Reshetnikova N.M. Plants - live the traces of group of armies “Centre” in the Russian land. Flora i rastitelnost Tsentralno-Chernozemnogo regiona - 2013: Materialy mezhregionalnoy nauchnoy konferentsii (Kursk, 6 aprelya 2013 g.). Kursk, 2013. Pp. 198-202. (In Russ.) [Щербаков А.В., Киселева Л.Л., Панасенко Н.Н., Решетникова Н.М. Растения - живые следы группы армий «Центр» на российской земле // Флора и растительность Центрально-Черноземного региона - 2013: Материалы межрегиональной научной конференции (Курск, 6 апреля 2013 г.). Курск, 2013. С. 198-202.] Shcherbakov A.V., Korolkova E.O., Shchepkina E.P. Polemochore plants in the flora of Spas-Demensky district of Kaluga region. Environment and Human: Ecological Studies. 2017. No. 2. Pp. 78-93. (In Russ.)

12. Seregin A. 7807a. Avenella flexuosa (L.) Drej. Schaedae ad Herbarium Florae Rossicae et civitatum collimitanearum, ab Instituto Botanico Academiae Scientiarum Rossicae editum. Т. 30. Fasc. 154-157 (№ 7651-7850). G.Yu. Konechnaya (ed.). St. Petersburg, 2015. Pp. 97-98. (In Russ.)

13. Tsvelev N.N. Poaceae (Graminea). Flora evropeyskoy chasti SSSR. Leningrad, 1974. Vol. 1. Pp. 117-368. (In Russ.) [Цвелев Н.Н. Poaceae (Graminea) // Флора европейской части СССР. Л., 1974. Т. 1. С. 117-368.]

14. Zhukova L.A. Wavy hair-grass (Deschampsia flexuosa). Biologicheskaya flora Moskovskoy oblasti. Moscow, 1980. Vol. 5. Pp. 46-57. (In Russ.) [Жукова Л.А. Луговик извилистый // Биологическая флора Московской области. М., 1980. С. 46-57.]

15. Ekaterina O. Korolkova - PhD in Biology; Associate Professor at the Department of Botany of the Institute of Biology and Chemistry, Moscow Pedagogical State University; Associate Professor at the Department of Cell Biology of the Faculty of Biology and Biotechnologies, National Research University Higher School of Economics, Moscow, Russia

16. Королькова Екатерина Олеговна - кандидат биологических наук; доцент кафедры ботаники Института биологии и химии, Московский педагогический государственный университет; доцент кафедры клеточной биологии факультета биологии и биотехнологии, Национальный исследовательский университет «Высшая школа экономики», г. Москва

17. Yaroslav E. Vasilkov - undergraduate at the Department of Botany of the Institute of Biology and Chemistry, Moscow Pedagogical State University

18. Васильков Ярослав Евгеньевич - магистрант кафедры ботаники Института биологии и химии, Московский педагогический государственный университет

19. Е.О. Королькова - общее руководство направлением исследования, планирование исследования, выбор модельных видов, участие в подготовке текста статьи

20. Я.Е. Васильчиков - построение биоклиматических моделей, интерпретация полученных данных, участие в подготовке текста статьи

21. Contribution of the authors

22. E.O. Korolkova - general direction of the research, planning of the research, selection of model species, participation in the preparation of the text of the article Ya.E. Vasilkov - bioclimatic modeling, results interpretation, participation in the preparation of the text of the article

Размещено на Allbest.ru