MASARYKOVA U N I V E R Z I T A P Ř Í R O D O V Ě D E C K Á F A K U L T A Klimatická změna v suchých trávnících: Kdo zvítězí a kdo prohraje? Diplomová práce B RANIS LAV KOPČAN Vedoucí práce: Mgr. Kryštof Chytrý, Ph.D. Ústav botaniky a zoologie Program Botanika Brno 2026 M U N I SCI KLIMATICKÁ ZMĚNA V SUCHÝCH TRÁVNÍCÍCH: K D O ZVÍTĚZÍ A KDO P R O H R A J E ? Bibliografický záznam Autor: Název práce: Studijní program: Vedoucí práce: Konzultant: Rok: Počet stran: Klíčová slova: Branislav Kopčan Přírodovědecká fakulta Masarykova univerzita Ustav botaniky a zoologie Klimatická změna v suchých trávnících: Kdo zvítězí a kdo prohraje? Botanika Mgr. Kryštof Chytrý, Ph.D. Mgr. Irena Axmanová, Ph.D. 2026 97+11 klimatická změna, mikroklima, suché trávníky, sezónnost, fenologie, růst, změna biomasy, sucho KLIMATICKÁ ZMĚNA V SUCHÝCH TRÁVNÍCÍCH: K D O ZVÍTĚZÍ A KDO P R O H R A J E ? Bibliographic record Author: Title of Thesis: Branislav Kopcan Faculty of Science Masaryk University Department of Botany and Zoology Climate change in dry grasslands: Who are the winners and the losers? Degree Programme: Botany Supervisor: Mgr. Kryštof Chytrý, Ph.D. Mgr. Irena Axmanová, Ph.D. 2026 97+11 Consultant: Year: Number of Pages: Keywords: climate change, microclimate, dry grasslands, seasonality, phenology, growth, biomass change, drought KLIMATICKÁ ZMĚNA V SUCHÝCH TRÁVNÍCÍCH: K D O ZVÍTĚZÍ A KDO P R O H R A J E ? Abstrakt Narůstající teploty a s nimi spojený stres způsobený suchem postupně nabývají na významu, zejména v suchých biotopech, jako jsou třeba temperátní suché trávníky. Porozumění jejich vlivu na jednotlivé druhy je však obtížné, protože jednotlivé druhy odrážejí sezónní dynamiku výkyvů počasí. Některé druhy se proto mohou novým podmínkám přizpůsobit posunem fenologie nebo posunem do mikroklimaticky příznivějších stanovišť. Abychom mohli posoudit, které druhy se stanou v kontextu klimatické změny relativními vítězi a které naopak poraženými, se zohledněním těchto faktorů, založili jsme v roce 2021 dlouhodobý terénní monitoring trvalých ploch ve třech typech suchých trávníků na jihovýchodě České republiky. Na základě mikroklimatických teplotních dat a monitoringu vegetace s dvouměsíční frekvencí j sme pomocí binomických smíšených modelů detekovali optimální teploty růstu jednotlivých druhů. Následně jsme z pomocí projekcí makroklimatického vývoje teplot vypočítali budoucí mikroklimatické podmínky a využili je k simulaci počtu růstových dní pro jednotlivé druhy. Naše mikroklimatické modely odhalují asymetrický vztah mezi makroklimatem a mikroklimatem, s lokálně tlumenými teplotami v zimě, a naopak amplifikovanými teplotami v létě. To vede k postupnému zhoršování podmínek v létě což způsobuje, že se pro většinu druhů zkracuje období vhodné pro růst. Nej výraznější poklesy jsme zaznamenali ve středním scénáři vývoje změny klimatu (SSP2-4.5). V tomto scénáři prodloužení růstového období na jaře a na podzim nedokáže u většiny druhů kompenzovat letní ztráty. Pokles růstu byl konzistentní pro většinu růstových strategií, ale lišil se mezi vegetačními typy. Tyto výsledky naznačují, že prodloužení a intenzita letního stresu se v brzké budoucnosti stane hlavní hrozbou pro suché trávníky ve střední Evropě a může vést k tomu, že se druhy ocitnou v podmínkách, na které nej sou adaptovány, a mohou být proto vystaveny zvýšenému konkurenčnímu tlaku. KLIMATICKÁ ZMĚNA V SUCHÝCH TRÁVNÍCÍCH: K D O ZVÍTĚZÍ A KDO P R O H R A J E ? Abstract Rising temperatures and related drought stress are gradually gaining importance, especially in dry habitats such as temperate dry grasslands. The understanding of their effects on individual species is, however, challenging as species reflect seasonal dynamics of weather fluctuations. Some of the species might be, therefore, able to adapt to new conditions by adjusting their phenology or by escaping to microclimatically more suitable areas. To make an assessment about which species might eventually become winners and which losers while accounting for these factors, we established in 2021 a long-term permanent-plot monitoring of three types of dry grasslands in southeastern Czech Republic. Using microclimate temperature data and bimonthly vegetation records, we fitted Binomial Mixed-effect Models to describe optimal growth temperatures for individual species. We then derived future microclimatic temperatures from future macroclimate projections and used them to predict the number of growth-days for individual species. Our microclimate models reveal an asymmetric macroclimate-microclimate relationship, with buffered winter temperatures, and amplified summer temperatures. That makes summers becoming increasingly harsher, causing the majority of species to shorten their period of growth in summer. Interestingly, the most profound declines were observed under the intermediate scenario (SSP2-4.5). Under this scenario, prolongation of the growth period in spring and autumn fails to compensate for summer losses for most species. Growth declines patterns were consistent across most plant strategies but differed between vegetation types. These results suggest that the prolongation of summer stress will become the primary threat in Central European dry grasslands, potentially forcing species into novel seasonal environments and under competitive pressures. U N I SCI UJISARHDVA liniVEDZ [TA FHIRODOVE DECKA F H K U L T Jl tO T LA R S KA 2, 511 37 BRNO IC: DOĚ16Ě24 QIC: CZQOĚ1B22I ZADÁNÍ D I P L 0 M 0 V É PRÁCE Akademicky rok: 202&202S Ústav: Ústav botaniky a zoologie Student: Be. Branislav Kopčan Program: Botanika Specializace: Ekologie rostlin Ředitel ústavu PřF MU Vám ve smyslu Studijního a zkušebního řádu MU určuje diplomovou práci s názvem: Nazev prace: Klimatická zmena v suchých trávnících: Kdo zvítězí a kdo prohraje? Název práce anglicky: Climate change in dry grasslands: Who are the winners and the losers? Jazyk práce: angličtina Oficiální zadání Cílem diplomové práce je stud ium dopadů klím atické zmeny na vybrané d runy rosllin skal nich stepí, suchých trávní ků a pěchavovych trávníků v Pavlovských vrších na jižní Morave. VetTech typech suchých trávníků zde od roku 2021 přebíhá pravidelný monitoring, kdy je zapisováno druhové složení cévnatých rosllin včetně odhadů jejich pokryvností a dalších charakteristik. Monitoring probíhá několikrát ročné pro podchycení sezónní variability vegetace. Student na tento monitoring naváže a po následující dva roky v nem bude pokračovat (koordinace terénních prací, do kterých bude zapojeno více kolegů a dobrovolníků, vlastni zápisy: digitalizace dat). Získaná data pak bude analyzovat s využitím teplotních čidel, která jsou umístěná v každé z ploch. Z a pomoci lineárních modelů odvodí závislost mezi růstem jednotlivých druhů a mikroklimatickými podmínkami. V tomto ohledu je důležité se soustředit právě na mikroklimatické podmínky, které mohou byt dosti odlišné od makroklimatickych podmínek díky specifickým topografickým situacím. Na nékterych místech tak může docházet k pufrování teplotních extrémů, zatímco na jiných mohou byt tylo extrémy amplifikovány. Vztah mezi makroklimalHm a mikroklimatem je v krajinách bez dlouhodobé snehové pokrývky poměrně staticky. Dostupná mikroklimatická měření se teoy dají použít i k odvození mikroklimatických podmínek budoucích. Student za tímto účelem využije různé RCP (representanve concemratlon pathways) scénáře globálního oteplováni a s jejich pomocí odvod! průměrné mikroklimatické podmínky studovaných trávníků do budoucna, napit, pro období kolem roku 2100. Takto odvozená mikroklimatická data následně použije k predikováni rúslu vybraných druhů rostlin. Predikce bude provedena tak, aby umožnila jednotlivým druhům temporální adaptaci na změnu klimatu, tedy např. přesunout svoji vegetační sezenu z vrcholného léta do jara. Student nakonec porovná, jak se sludovanélypy suchých trávníků a různé druhy rosllin liší v predikované odpovedi na klimatickou změnu, a tyto výsledky srovná s různými adaptačními strategiemi rosllin (např. sukulence, úzkolistost nebo jednoletost versus vytrvalost). Analýzy budou zpracovány v jazyce R nebo Python. Samotná práce bude strukturovaná podobně jako standardní vědecký článek s rozšířeným úvodem obsahujícím reserSi na téma predikce změn klimatickézményv různých biotopech a adaptace rosllin na ní. STRADA 1 Z I Literatura: Doležal J . , AHman J . , Jandová V , Chytrý M., Conti L.: Méndez-Caslro R E . , KlimeSová J . , Zelený D., Otlaviani G . (2022): Climate warming and extended droughts drive establishment and growth dynamics in temperate grassland plants. Agricultural and Forest Meteorology 313: 103762. httpsy/doi.org/10.101 6/jagrformet2021.108762 Fischer F.M., Chytrý K., Chytrá H., Chytrý M. & Tesitel J . (2023) Seasonal beta-diversity of dry grassland vegetation: divergent peaks of aboveg round bio mass and species richness. Journal of Vegetation Science 34: e13182. https^/doi.org/10.1111/jvs.13182 Mazal la L , Diekman M. & Dupré C. (2022): Microclimate shapes vegetation response to drought in calcareous grasslands. Applied Vegetation Science 25: e l 2672. httpstfdoi.org/10.1111/avsc.12672 Wessely J., Essl F., Fiedler K, Gattringer A., HOIber B., Ignateva C . : Moser D., Rammer W., Dullinger S. & Seidl R. (2024): A climate-induced tree species bottleneck for forest management in Europe. Nalutre Ecology and Evolution 8:1109-1117. hittpsjf /doi.org/10.1O33/s415M-024-024O6Vedoucí práce: Mgr. Kryštof Chytrý, Ph.D. Konzultant: Mgr. Irena Axmanová, Ph.D. Datum zadáni práce: 23. 10. 2024 V Brně dne: 27. 4. 2026 Zadání bylo schváleno prostřednictvím IS MU. Be. Branislav Kopcan, 17. 4. 2026 Mgr. Kryštof Chytrý, Ph.D., 17. 4. 2026 Mgr. Ivela Hodová. Ph.D., 17. 4. 202G KLIMATICKÁ ZMĚNA V SUCHÝCH TRÁVNÍCÍCH: K D O ZVÍTĚZÍ A KDO P R O H R A J E ? Declaration I declare that I have prepared my thesis independently under the mentorship of the thesis supervisor and using the information sources provided in the thesis. I declare that I have used AI tools in accordance with the principles of academic integrity and that I appropriately cite the use of these tools in the work. After using the AI tools, I have verified the content and assume full responsibility for it. Branislav Kopčan KLIMATICKÁ ZMĚNA V SUCHÝCH TRÁVNÍCÍCH: K D O ZVÍTĚZÍ A KDO P R O H R A J E ? Acknowledgements Special thanks go to everyone who contributed to the vegetation sampling over the nearly five-year period: Paula Acsová, Irena Axmanová, Manuele Bazzichetto, Simone Eusebio Bergó, Gianmaria Bonari, Dariia Borovyk, Maryna Burlaka, Helena Chytrá, Milan Chytrý, Vincent Chytrý, Josep Padullés Cubino, Lucia Cahojová, Mirjana Cuk, Pavel Daněk, Peter Deván, Will J. Eden, Anni Eichhorn, Joshua Erkelenz, Moritz Fallgatter, Nina Fahs, Felicia M . Fischer, Martin Franc, Michal Friedl, Klárka Friesová, Hanka Hrdličková, Norbert Helm, Monika Hrubanová, Jaroslav Hruban, Petr Hubatka, Susanna loni, Natália Janišová, Anni K. Jašková, Martin Kadlec, Daniel Kadaš, Veronika Kalníková, Marina Kedrova, Ilona Knollová, Monika Kolényová, Michaela Krejčová, Zdeňka Lososová, Corrado Marcenó, Emma S. Mendez, RicardaPátsch, JulianPleyer, SafýiaPraleskouskaya, ŠtěpánkaPustková, Kristýna Ratajova, Barbora Reslová, Jaroslav Rohel, Alexandra Šmolková, Vojtěch Sobotka, Marta G. Sperandii, Daniel Szokala, Jakub Těšitel, Dorota Těšitelova, Tamara Těšitelova, Pavlína Tichá, Agathe Toumoulin, Gonzalo Velasco, Radka Vaňáčková, Martin Večeřa, Branislav Vrábel, Denys Vynokurov. Special thanks go to Kryštof Chytrý for inviting me almost four years ago to the sampling on Pálavá, introducing me to the topic, and giving me the opportunity to work on the project he established together with Helena Chytrá. I thank him for the initial idea, discussions on statistical analyses, results and valuable comments on the manuscript. I also appreciate our discussions on almost anything, as well as having patience with me throughout the whole process. I would also like to thank Irena Axmanová for her co-supervision, valuable ideas and proofreading of the manuscript. As well as group of friends who became an inseparable part of the fieldwork, even under harsh weather conditions, and for their support. TABLE OF CONTENTS Table of Contents Glossary 21 List of Appendices 22 1 Introduction 23 1.1 Main Aims of the Study 29 2 Study Area Characterisation 30 2.1 Study Area 30 2.2 Climate 31 2.3 Investigated Grassland Types 32 3 Methods 35 3.1 Data Collection 35 3.1.1 Vegetation and Species Data 35 3.1.2 Microclimate Data 38 3.2 Data Preparation and Analysis 38 3.2.1 Species Growth 39 3.2.2 Temperature Data 40 3.2.2.1 Data Cleaning and Calibration 40 3.2.2.2 Gap-filling via Random Forest Regression 41 3.2.2.3 Microclimate Change Projections 42 3.2.3 Species Responses 43 3.2.3.1 Growth Probability Models 43 3.2.3.2 Defining Species Growth Temperature Ranges 44 3.2.3.3 Identifying Winners and Losers 45 3.2.3.4 Seasonal Growth Dynamics 46 3.2.3.5 Biomass Change Projections 46 4 Results 48 4.1 Microclimate Change 48 4.1.1 Macroclimate -> Microclimate Transformation 48 19 TABLE OF CONTENTS 4.1.2 Future Macroclimate and Microclimate Temperature Trajectories and Seasonal Shifts 52 4.2 Species Responses 55 4.2.1 Probability of Growth Curves and Temperature Growth Ranges 55 4.2.2 Identifying Winners and Losers 58 4.2.3 General Seasonal Growth Dynamics 63 4.2.4 Biomass Change Estimations 64 5 Discussion 68 5.1 Macroclimate/Microclimate Change 68 5.2 Intensification of Summer Stress 70 5.3 Limits of Seasonal Compensation 71 5.4 Differences Between Functional Groups and Vegetation Types 73 6 Conclusion 78 References 79 20 G L O S S A R Y Glossary AU C - Area Under the Curve CMIP6 - Coupled Model Intercomparison Project Phase 6 CHMU - Český hydrometeorologický ústav EMMs - Estimated Marginal Means GCM - Global Climate Model GLM - Generalized Linear Models GLMM - Generalized Linear Mixed-Effect Model IAMs - Integrated Assessment Models NCCS - NASA Centre for Climate Simulation NEX-GDDP-CMIP6 - NASA Earth Exchange Global Daily Downscaled Proj ections RCP - Representative Concentration Pathways ROC - Receiver Operating Characteristics SSP - Shared Socioeconomic Pathway Definition of terms: In this study (Methods, Results, Discussion), the terms microclimate and macroclimate, and their variations (microclimate temperature, microclimate soil temperature, soil temperature, and macroclimate temperature, macroclimate air temperature, air temperature) are used across the Methods, Results, Discussion, and Conclusion. Microclimate temperature refers to soil temperature measured at 10 cm belowground. Macroclimate temperature refers to air temperature at 2 m aboveground, obtained either from a weather station or from downscaled projections derived from Global Climate Models. 21 LIST OF APPENDICES List of Appendices Text Appendices Appendix A Species-specific growth probability model performance 98 Appendix B Monthly precipitation sums under current and future conditions 101 Appendix C Residuals of observed and predicted microclimate temperatures 102 Appendix D Macroclimate-microclimate relationships 103 Appendix E Species Log-ratio Growth Index across vegetation types 104 Appendix F Growth intensity across vegetation types 108 Others Appendix G Species probability of growth curves .pdf Appendix H Species seasonal growth dynamics .pdf Appendix I Biomass change estimations for all species .pdf Appendix J R scripts and data .zip 22 INTRODUCTION 1 Introduction Importance of Grasslands and their Threats Grasslands are fundamental components of the earth's biosphere, covering approximately 30-40% of the global terrestrial surface (White et al. 2000, Suttie et al. 2005), and specifically in Europe, they cover roughly 13% of the total land area (Eurostat 2011). They are important parts of landscape, encompassing a high diversity of vegetation types and, especially in Europe, they are primary drivers of regional vascular plant diversity (Partei et al. 2005, Večeřa et al. 2021). Temperate grasslands in Central Europe are even recognized as fine-scale coexistence hotspots (Wilson et al. 2012, Dengler et al. 2014). Central European grasslands form the westernmost outpost of the vast forest-steppe biome that stretches across Eurasia (Wesche et al. 2016, Erdös et al. 2018, Chytrý et al. 2022a). Situated at the crossroads between temperate, (sub-)Mediterranean and continental zones, they host not only high species diversity but also a unique mixture of biogeographical elements (Chytrý et al. 2022b) and harbour many threatened and endangered taxa (Erdös et al. 2018, Staude et al. 2023). Apart from grassland's value for biodiversity, they encompass various nature contributions to people, including food production, soil erosion prevention, water runoff regulation, and support of complex trophic networks (Török et al. 2018, Bengtsson et al. 2019). Diverse grassland communities have also been reported to exhibit higher resilience against plant invasions (Kennedy et al. 2002) and, additionally, represent important carbon sinks (Bai & Cotrufo 2022). Beyond these ecological and functional benefits, Central European grasslands represent a unique cultural heritage. During the last centuries, their biodiversity has been shaped and maintained by traditional management practices, tying human history with landscape biodiversity (Janišová et al. 2021). Within this context, Central European dry grasslands represent a highly idiosyncratic and valuable study system. However, even though they possess high cultural and conservation value, they have experienced a severe deterioration in recent decades, manifested especially through loss of habitat specialists (Wesche et al. 2012, Hilpold et al. 2018, Harásek et al. 2023, Klinkovská et al. 2024a, Klinkovská et al. 2024b). Most of these changes are 23 INTRODUCTION attributable to the cessation of traditional management and broader land-use changes, factors which have also been identified globally as the primary driver of biodiversity loss (Diaz et al. 2019). In Central Europe, this has manifested in two contrasting trajectories of grassland development, with an increasing intensification of agriculture on productive grasslands and abandonment of non-productive ones (Tórók et al. 2018). The socioeconomic shifts and increasing food demands over the last 200 years have led to conversion of former pastures and meadows into arable fields (Hejcman et al. 2013), while the remaining grasslands have been subjected to intensive management. This intensification was often accompanied by eutrophication arising both from atmospheric nitrogen deposition and runoff from agricultural areas (Millenium Ecosystem Assessment 2005). Conversely, in topographically heterogeneous areas, the cessation of traditional management practices has led to abandonment of sites that were difficult to cultivate with heavy machinery, which, in many cases, resulted in their deterioration and notable shrub encroachment (Valkó et al. 2018, Klinkovská et al. 2024b). Overall, eutrophication and successional changes have been identified as the two most significant threats to natural habitats in Central Europe (Chytrý et al. 2019). While up to now, processes related to land-use changes and eutrophication are indeed considered to be the primary reasons of the recent biodiversity crises, the recent anthropogenic climate change is rapidly gaining importance (Diaz et al. 2019). However, climate change does not act in isolation. Instead, it often confounds and modifies the effects of other drivers of the changes in vegetation. For example, warming together with nitrogen deposition has been shown to reduce community stability (Wu et al. 2020), and decrease grassland carbon sequestration in grassland (Kiibert et al. 2019), while multiple global change drivers acting together may alter the carbon cycling more strongly than individual factors alone (Reich et al. 2020). Rising temperatures and frequent extreme climatic events may therefore accelerate ongoing biodiversity crisis and shifts in species composition, while also favouring expansive species already profiting from the cessation of traditional management (Axmanová et al. 2024) or support spread of non-native species (Bradley et al. 2024). Overall, climate change not only directly disrupts ecosystem functioning and affects biodiversity, it also intensifies the effects of already existing threats (Pórtner et al. 2023). 24 INTRODUCTION Understanding how grassland ecosystems respond to future climate scenarios is therefore essential for ensuring their persistence in an increasingly unstable world. Box 1: Climate change and Radiative forcing In a stable environment, the Earth's climate system is in a dynamic equilibrium, where the energy received from the sun is balanced by reflecting and absorbing by atmosphere and surface. However, this equilibrium can be disrupted by radiative forcings, including volcanic activity, anthropogenic emissions of greenhouse gases (carbon dioxide CO2, methane CH4, nitrous oxide N2O), aerosols, and modulated by feedbacks like cycles of water vapour. In other words, the Radiative Forcing (RF) framework provides a way to quantify the disruption of energy balance relative to a historical baseline (i.e. pre-industrial state of atmosphere) (Shine et al. 1990, National Research Council 2005, Cornell et al. 2012). Since the mid- 18th century human activities have significantly increased the atmospheric concentration of greenhouse gases. Between 2011 and 2019, these concentrations continued toriseand reached the following global levels: CO2 (410 ppm), CH4 (1866 ppb), N2O (332 ppb). Consequently, the global surface temperatures for the 2011-2020 period were 1.09 °C higher than the 1850-1900 baseline (IPCC 2021). To better understand various possible trajectories of changing climate, different climate change scenarios have been developed over the recent decades. Early scenarios focused mainly on increase of greenhouse gases and aerosols concentrations. Prior to 2011, the scientific community developed Representative Concentration Pathways (RCPs) that represent various trajectories of emission concentrations leading to specific radiative forcing states (RCP 2.6, 4.5, 6.0, 8.5). However, these pathways did not directly incorporate socioeconomic, demographic, technological assumptions, rather they modelled them separately (Moss etal. 2010). To address this, the CMIP6 (Coupled Model Intercomparison Project Phase 6) framework was established. It coordinates global institutions and ensures consistency in climate simulations (Eyring et al. 2016). One of the central activities of CMIP6 is the Scenario Model Intercomparison Project (ScenarioMIP) that develops future pathways derived from Shared Socioeconomic pathways (SSPs) (O'Neill et al 2016, Meinshausen et al. 2020). The SSPs describe various socioeconomic developments, such as economic growth, population growth, investments in technology, education and inequality (O'Neill et al. 2016). These narratives are subsequently translated into land-use changes, greenhouse gas emissions and emissions of other air pollutants using Integrated Assessment Models (IAMs) (Riahi et al. 2017). The results are converted into concentration pathways and radiative forcing levels that serve as drivers of climate change (Shine et al. 1990). These values are used as inputs in the Global Climate Model (GCM) simulations that project temperature, precipitation and other climatic variables. Four high-priority SSP scenarios, that are also evaluated in the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC 2021) are the following: SSP1-2.6: 'sustainability scenario'. This scenario assumes relatively successful mitigation and adaptation to climate change across the globe through sustainable practices. Shifts towards sustainability are coupled with high economic growth, investments in health, institutions, education and reduced inequality between countries. These efforts lead to an increase in global mean temperature by approximately 1.8 °C between 2081-2100 relative to 1850-1900. 25 INTRODUCTION SSP2-4.5: 'middle of the road scenario'. This scenario expects medium challenges in mitigation and adaptation to climate change across all socioeconomic sectors leading to development roughly following historical patterns. This pathway projects an increase in global mean temperatures by approximately 2.7 °C by the end of the century relative to 1850-1900. SSP3-7.0: 'regional rivalry scenario'. This scenario is characterized by major inequality across regions, nation-centred politics and limited international cooperation, leading to increased inability to mitigate and adapt to climate change. The estimated global mean temperature increase will be approximately 3.6 °C by 2081-2100. SSP5-8.5: 'fossil-fuel development scenario'. Under this scenario the economy will register massive growth mainly due to continued exploitation of fossil fuels, this will be coupled by low ability to mitigate climate change. On the other hand, the high economic growth will support technological progress, investments into education and health, which might result in local adaptations to environmental challenges. This scenario predicts warming by approximately 4.4 °C by 2081-2100. Based on O'Neill et al. (2016), Riahi et al. (2017), IPCC (2021) Ecological responses to climate change As climate change is a highly complex process, it can be expected that various species will respond to it in many different ways. Among the most likely responses are shifts in species distributions, altered growth patterns, and phenology. The (in)ability to copy with the changes through one of these mechanisms makes some of the species "climate change winners" and others "climate change losers". Some of the losers may not only suffer abundance declines, but even local extinctions (Panetta et al. 2018). The most straightforward way for organisms to track the changing climate is altering their distribution ranges through migration. Indeed, it has been documented that species have already exhibited shifts in their distributions over recent decades (Parmesan & Yohe 2003). As the climate warms, many plant species find new opportunities for colonizations towards previously cooler regions, either in poleward and upward directions (Pauli et al. 2003, Engler et al. 2009, Chen et al. 2011). Such shifts are yet best documented from high mountains (Grabherr et al. 1994, Pauli et al. 2012), where steep elevation-dependent temperature gradient can act as an escalator into new suitable areas (Gottfried et al. 2012, Yue et al. 2026). In contrast, latitudinal shifts are documented much less. One of the main reasons for that is the long history of landscape exploitation and related habitat fragmentation in lowlands, which make such shifts nearly impossible. Correspondingly, the relative proportion of species that might turn out as "winners" is likely smaller in lowland areas than 26 INTRODUCTION in high mountains and, moreover, their increased competitive ability may lead to further impoverishment of species richness of local communities (Olsen et al. 2016, Kapitany et al. 2025). While some species with good dispersal and colonization abilities migrate to new areas (Graae et al. 2018), others may simply persist in situ and tolerate new conditions, either because they have broad-enough climatic niches, or they are competitively strong and/or phenologically plastic (Lenoir et al. 2010). In addition, in heterogeneous environments some species may be able to utilize microhabitats with suitable conditions that will enable them to survive even under suboptimal macroclimates (Scherrer & Körner 2011, Jimenez-Alfaro et al. 2024, Kemppinen et al. 2024). Nevertheless, the need for such adaptations make it likely that some species will simply not cope with the new changes and the reduced pool of winners may therefore lead to overall biotic homogenization (Britton et al. 2009, Pauchard etal. 2016). Box 2: Macroclimate vs microclimate in climate change modelling Macroclimate represents large-scale climatic conditions that are typically measured at 1-2 m aboveground, using weather stations or modelled by Global Climate Models over grid scales of a few kilometers. In contrast, microclimates are local climatic conditions shaped by topography, soil types, and vegetation (Bramer et al. 2018). While most climate change projections and scenarios are based on macroclimate data of varying spatiotemporal resolution, these may not reflect the conditions actually experienced by organisms. For ecological applications it may therefore be more important to consider proximal microclimate, that is, the climatic conditions species actually experience, rather than simply using the finest available spatial resolution (Klinges et al. 2024). These conditions may differ from macroclimate not only in magnitude, but also in their temporal dynamics, being either more stable or more variable, and can vary over short spatial distances. Because species respond to the conditions they actually experience, such variation may influence growth patterns, phenology (Kankaanpää et al. 2018, Ward et al. 2018, Vitasse et al. 2021), seed germination (Espinosa del Alba et al. 2025), and overall community structure and dynamics (Kemppinen et al. 2024). It may also help to explain why vegetation responses differ among sites, and whether some locations may function as microrefugia under climate change (Moritz et al. 2013). Overall, microclimate conditions are essential in understanding species responses to climate change and are important for predicting vegetation shifts in warming environments. Apart from migrations, some organisms can cope with altered climatic conditions, through adjusting the timing of their phenology. It has been estimated by global syntheses that phenology is recently advancing by 2.3 days per decade across species (Parmesan & Yohe 27 INTRODUCTION 2003). In many systems, warmer spring temperatures induce earlier flowering and earlier onset of growth (Cook et al. 2012). However, accurately predicting temporal shifts in phenology and growth remains challenging, as the phenological responses observed in manipulative warming experiments often do not match responses from long-term observational studies (Wolkovich et al. 2012). Although, remote-sensing techniques enable effective identification of broad community changes in vegetation phenology across large spatial areas, they often lack the fine-scale resolution needed to distinguish what species are responsible for these changes and how their responses translate into ecosystem shifts (Fitchett et al. 2015, Piao et al. 2020). In addition, local microclimatic variability may further influence species phenological responses, as individuals from the same locality may experience different conditions, these microhabitats may provide buffered conditions allowing species to maintain their phenological timing even under warming climate (Kankaanpaa et al. 2018). Simulations of future climates suggest a widespread earlier onset of growth in Europe. However, this does not necessarily translate into a longer growing season everywhere. While alpine and mediterranean regions might indeed experience prolonged periods of growth, in Central Europe, the summer droughts might, in turn, shorten its overall length (Chang et al. 2017). While the environmental changes may potentially increase the global terrestrial carbon uptake, through increased growing season length and vegetation productivity (Piao et al. 2020), the increasing frequency and intensity of climate extremes can offset these effects by suppressing photosynthesis and plant growth, while maintaining or accelerating respiration and decomposition (Reichstein et al. 2013, Frank et al. 2015). However, these processes may differ between species, habitats and regions depending on their sensitivity, productivity and resilience to stress. Consequently, many European regions may act as carbon sinks under lower increased temperatures, but these sinks are predicted to saturate and potentially become carbon sources under more extreme climate scenarios (Chang et al. 2017). Such contrasting projections suggest that vegetation and plant responses to climate change are strongly context dependent and shaped by regional climatic conditions, species pools, and ecological interactions. Thuiller et al. (2005) suggested that species from different 28 INTRODUCTION European regions will have contrasting responses to climate change, with species from drought tolerant environments being better adapted to future temperature increase. In addition, previous studies in grasslands have shown that extreme events can alter species composition (Stampfli & Zeiter 2004) or reduce productivity (Xu et al. 2019). This indicates that climate change effects cannot be fully understood solely at the broad community or biome levels. Responses may vary among individual species and are driven by proximal microclimatic conditions rather than by broad macroclimatic trends. Predicting species-specific responses in grassland systems may therefore provide valuable insights into which species are likely to benefit from future changes in climate and which may be at risk of decline. 1.1 Main Aims of the Study Using dry grasslands in Southern Moravia as a model system, this study aims to address the following questions: i) How will projected climate change affect the growth activity of individual dry grassland species? ii) Do species responses to climate change differ among functional and ecological groups? Hi) How will climate change affect the seasonal distribution of growth? Can species adapt temporally to a warming climate? 29 STUDY A R E A CHARACTERISATION 2 Study Area Characterisation 2.1 Study Area The study area is located in southern Moravia (Czech Republic), with nine study sites (Figure 1) distributed within and around the Pálavá Protected Landscape Area (Pálavá PLA). The highest point in the region is Děvín (554 m a.s.l). The area represents the westernmost outposts of the forest-steppe biome stretching across Eurasia and belongs to the Pannonian phytogeographical province (Fekete et al. 2016, Erdós et al. 2018). The geology and topography of the area is quite heterogeneous and is reflected in local vegetation patterns. The sites Milovická stráň, Vysoký roh and Liščí vrch are situated in areas primarily composed of tertiary sediments and the bedrock is flysch (layers of sandstones and claystones), which is frequently overlaid with loess-derived soils. This bedrock supports the development of relatively deep soils, predominantly chernozems and pararendzinas. In contrast, the rest of the sites are in areas characterized by prominent limestone rocky outcrops composed of Jurassic and Lower Cretaceous limestone. The soils in this terrain are often shallow and underdeveloped rendzinas (Mackovčin et al. 2007, Česká geologická služba 2026). 30 STUDY A R E A CHARACTERISATION Figure 1 Map of the study area (OpenStreetMap, QGIS Development Team (2025)). Study sites are represented by points coloured based on vegetation type. 2.2 Climate All study sites are situated in the warm and dry part of southern Moravia with a subcontinental climate with a rather high difference between winter and summer temperatures and precipitation peak in summer. Based on meteorological data recorded between 2000 and 2025 (weather station Brod nad Dyji) the area exhibited a mean annual temperature of 10.6 °C. Within this period, the warmest year was 2024 with a mean annual temperature of 12.3 °C, while the coldest was 2010 with 9 °C. The warmest month was July, with mean temperatures reaching 20.9 °C while the coldest month was January with mean temperatures around -0.1 °C. Mean annual precipitation was 494 mm, with a minimum of 331 mm in 2015 and a maximum of 700 in 2002 (CHMU 2026). 31 STUDY A R E A CHARACTERISATION The temporal variability in temperature and precipitation is quite common (Fischer et al. 2020), but the variation is not only expressed temporally but also spatially, since the topographically diverse landscape creates microsites that are either sheltered or exposed to solar radiation and sites with better water holding ability. 2.3 Investigated Grassland Types Our dataset consists of nine study sites that can be roughly classified into three relatively distinct vegetation types, based on an expert field assessment. The syntaxonomic units provided in the description of vegetation types correspond to the concepts used in Chytrý et al. (2007). Rocky dry grasslands Localities: Svatý Kopeček, Kotel, Děvín This vegetation type occurs on shallow, skeletal soils on limestone or dolomitized limestone bedrock, primarily on exposed, south-facing slopes. It represents open canopy grasslands with high prevalence of annuals as well as various stress-tolerant species. The grasslands are dominated by tussock-forming graminoids such as Festuca csikhegyensis, Stipa capillata, Stipapulcherrima and Festuca valesiaca. Common perennials include Potentilla incana, Fumana procumbens, Silene otites, Teucrium montanum, Sanguisorba minor and short-lived perennials like Centaurea stoebe or Acinos arvensis. The annuals, include Arenaria serpyllifolia, Medicago minima, Cerastiumpumilum agg., Holosteum umbellatum, Veronica praecox alongside more rare species such as Viola kitaibeliana or Erophila spathulata. The vegetation roughly corresponds to the alliance Bromopannonici-Festucion pallentis, locally transitioning towards the alliance Festucion valesiacae (class Festuco- Brometea). Loess dry grasslands Localities: Milovická stráň, Vysoký roh, Liščí vrch Open dry grassland vegetation on deeper loess deposits. The community is dominated by narrow-leaved graminoids such as Festuca valesiaca, Festuca rupicola, Stipa pulcherrima, 32 STUDY A R E A CHARACTERISATION Carex humilis, Koeleria macrantha, Bothriochloa ischaemum. The perennials are represented by Veronica spicata, Astragalus onobrychis, Dictamnus albus, Eryngium campestre, Crambe tataria, Galatella linosyris. Of the three vegetation types studied, this one is the most variable including both relatively open, species poor stands and more productive, species-rich sites. The vegetation roughly corresponds to the alliance Festucion valesiacae (class Festuco-Brometea). Dealpine Ses/er/'a-dominated grasslands Localities: Děvín, Svatý kopeček, Stolová hora Compared to the previous types, these grasslands are located on cooler north-facing slopes on limestone bedrock and represent a relatively closed community dominated by dealpine species Sesleria caerulea. Other dealpine species found here include Biscutella laevigata, while additional graminoids include Festuca csikhegyensis or Carex humilis. Common perennials are for example Centaurea triumfetti, Dianthus lumnitzeri, Dorycnium germanicum, Primula veris, Campanula persicifolia or Thesium linophyllon. Due to the relatively dense nature of this vegetation and high bryophyte cover, this type does not support as many annual species as the previous two vegetation types. The vegetation falls under the alliance Diantho lumnitzeri-Seslerion (class Festuco-Brometea). 33 STUDY A R E A CHARACTERISATION Figure 2 The nine study sites: F l = Svatý Kopeček (rocky dry grassland), F2 = Svatý Kopeček (dealpine Selena-dominated grassland), F3 = Milovická stráň (loess dry grassland), F4 = Děvín (rocky dry grassland), F5 = Děvín (dealpine Selena-dominated grassland), F6 = Vysoký roh (loess dry grassland), F7 = Kotel (rocky dry grassland), F8 = Stolová hora (dealpine .Ses/ena-dominated grassland), F9 = Liščí vrch (loess dry grassland). Photos by Kryštof Chytrý (Fl, F2, F3, F4, F5, F6, F9) and Branislav Kopčan (F7, F8). 34 METHODS 3 Methods 3.1 Data Collection 3.1.1 Vegetation and Species Data The so-called Fenomor** monitoring was designed by Kryštof Chytrý and Helena Chytrá in order to expand the phenological monitoring of Děvín (Fevin*) originally designed by Jakub Těšitel, Kryštof Chytrý, Helena Chytrá & Felícia Fischer and have been conducted on Děvín (Pavlov Hills) since 2019 (Fischer et al. 2020; see Box 3). The Fenomor sampling sites were established in May 2021 across nine sites selected by local botanists and researchers from Masaryk University and University of Vienna (Kryštof Chytrý, Helena Chytrá, Hana Hrdličková (née Galusková) and Petr Hubatka). These sites represent the most preserved fragments of three distinct vegetation types: (1) rocky dry grasslands, (2) loess dry grasslands and (3) dealpine Sesleria-dommated grasslands (for detailed description, see Study Area Characterisation: Investigated Grassland Types). Each vegetation type is represented by three sites. Between May 2021 and May 2024, the sampling was organized by Kryštof Chytrý and from July 2024 by Branislav Kopčan. Box 3: Fevin* and Fenomor** monitoring datasets The Fevin and Fenomor datasets represent long-term vegetation monitoring in dry grasslands of the Pálavá region. The origins of this dataset date back to 1993, when Jin Danihelka established seven permanent plots on the summit Děvín (Pavlov Hills). These plots were sampled annually for over 30 years, creating a unique dataset that was used to study the interannual variability in dry grassland vegetation and its responses to weather fluctuations and extreme climatic events (i.e. droughts) (Fischer et al. 2020). In 2019, this annual monitoring was expanded into the Fevin dataset. Permanent subplots (six per original plot) were established to capturefine-scaletemporal dynamics. During the first year, the vegetation was sampled monthly. Subsequently, the sampling was restricted to every two months. In 2021, the monitoring was further extended into the surrounding dry grasslands through the establishment of the Fenomor dataset. The aim of Fenomor is to capture the variability in vegetation dynamics across different dry grassland types and under heterogeneous microclimatic conditions. As of 2026, the Fevin and Fenomor datasets have been running for more than seven and nearly five years, respectively. 35 METHODS A standardized design has been used across all sites. At each of the sites, a 4 x 4 m2 plot was established, oriented parallel to the slope, containing six 0.5 x 0.5 m2 subplots. The four subplots were placed in the corners of the main plot, with the remaining two positioned perpendicular to the corner ones (see Figure 3). We monitor the vegetation dynamics through long-term bimonthly resurveys. After the initial sampling in May 2021, we have conducted resurveys roughly every two months, with each resurvey lasting one or two days. At each subplot, we estimated the percentage cover of total herb layer, moss layer (including lichens), bare soil, protruding rocks, gravel and litter (including standing dead). We recorded the cover of vascular plants only for rooting individuals using Braun-Blanquet nine-degree cover abundance scale (Westhoff & van der Maarel 1978), while we did not identify bryophytes and lichens at the species level. Additionally, we recorded height and phenology status (seedling, vegetative, flowering, fruiting) for every species present. During each resampling we also took a vertical photo of every subplot to ensure data consistency. During each survey, surveyors were equipped with species lists from the last three surveys. While this practice is known to bias the number of species found during resurveys, we assume that high frequency of our sampling and small size of our plots levels these biases out. Moreover, with the help of photographs of sampled communities, we apply systematic post-hoc checks of taxonomic determination of the recorded species composition to minimize sampling errors. This extensive bimonthly monitoring is carried out with the help of researchers and volunteering students from the Department of Botany and Zoology at Masaryk University, Department of Biodiversity Dynamics and Conservation at University of Vienna, local botanists from the Pálava Protected Landscape Area Administration Office and many volunteers from various international universities and research institutes. To date (March 2026), 65 participants contributed to the field data collection (see Acknowledgements for the full list of contributors). The field data collection is always carried out in pairs, with at least one field botanist experienced with the monitoring present. 36 METHODS future Figure 3 Schematic overview of the study design and workflow. The field experiment consists of nine sites, each with six subplots, that were sampled every two months from 2021 to 2025. For each species, we estimated cover and height, which were transformed into growth estimates. At each site, microclimate temperature was measured, and macroclimate data from weather stations were used to establish the macroclimate-microclimate relationship. These relationships were then applied to future macroclimate projections to estimate future site-specific microclimates. Using growth data and future projected microclimates, we fitted Generalized Linear Mixed-effect Models (GLMMs) and calculated species probabilities of growth. These probabilities were subsequently used to identify winners and losers and to predict seasonal dynamics under climate change. 37 METHODS 3.1.2 Microclimate Data To account for microclimatic variability, at each of the sites we installed an EMS Minikin data soil temperature logger in the centre of the plot during the initial site establishment in May 2021. The loggers were buried approximately 10 cm belowground to record soil temperature. To obtain a more precise microclimate profile, we installed additional data loggers at the upper edge of each plot in November 2024. These included EMS MicroLog SP3 sensors that measure belowground temperature as well as soil water potential, and TOMST TMS-4 sensors (Wild et al. 2019) that measure soil moisture, belowground, ground surface and aboveground temperature. In order to avoid situations when logger failure leads to data losses, we repeatedly read-out all the loggers at least twice a year. 3.2 Data preparation and Analysis We performed all calculations and data wrangling in R (R Core Team 2025) using the packages: tidyverse (Wickham et al. 2019), readxl (Wickham & Bryan 2025), ranger (Marvin & Ziegler 2017), gridExtra (Auguie 2017), ImerTest (Kuznetsova et al. 2017), pROC (Robin et al. 2011), glue (Hester & Bryan 2024), patchwork (Pedersen 2025), performance (Liidecke et al. 2021), emmeans (Lenth & Piaskowski 2025), zoo (Zeileis & Grothendieck 2005), broom (Robinson et al. 2025),factoextra (Kassambara & Mundt 2020), cluster (Maechler et al. 2025), sf (Pebesma 2018), rnaturalearth (Massicotte & South 2026), rnaturalearthdata (South et al. 2024), ggspatial (Dunnington 2025). All of our following projections of species responses to climate change are based on temperature change. Although precipitation is a key component of climate change research, we have not explicitly incorporated it into our modelling framework due to limitations in the translation of broad scale precipitation projections into microclimate conditions. Nevertheless, macroclimatic projections suggest that while total precipitation may change moderately, its seasonal distribution and variability will likely shift. In our studied region, the main seasonal peak of precipitation in summer remains relatively unchanged, but is slightly shifted towards late spring, with some precipitation increase in the colder seasons 38 METHODS (see Appendix B). However, future precipitation regimes are expected to become more variable, with an increasing proportion of rainfall occurring in extreme events (IPCC 2021). Additionally, in systems that are characterized by high evaporation and evapotranspiration and lie on shallow soils limiting water retention, such as dry grasslands, the warming might further increase these processes and reinforce summer drought stress (De Boeck et al. 2007, Bastos et al. 2020). 3.2.1 Species Growth We entered all the field observations into digital spreadsheets and semi-automatically and manually checked for inconsistencies in cover, height and phenology status to account for clear observer errors. We excluded taxa that remained unidentified even after several surveys or were identified only at genus level (less than 1% of the total occurrences), this usually concerned tiny individuals that did not survive beyond seedling stage. We standardized nomenclature and taxonomic concepts according to Kaplan et al. (2019). For this study, the dataset was restricted to a four-year period from May 2021 to May 2025. To ensure temporal consistency, we treated the rare occasions when subplots within the same main plot were recorded on different dates during the same sampling event, by averaging sampling dates. During the field surveys we recorded only species presences, thus we introduced 'zeroes' into the dataset wherever the species was missing during a specific sampling event but has been recorded at that subplot at least once during the study period. We converted species covers from the Braun-Blanquet cover abundance scale to percentage using the centres of individual intervals. We calculated a biomass volume estimate for each species in every subplot using the following formula: B = cover[%] x Vheight[Cm]. Using the interval between two consecutive samplings, we calculated the change in biomass for each species. Finally, for subsequent analysis we summed these subplot-level values to determine total biomass change at the main plot level. 39 METHODS To model growth as a function of temperature, we firstly converted the biomass change values into a binary response: growth was recorded as 1 if the change in biomass between consecutive samplings was > 0, and 0 if the change was < 0. To ensure robustness of further models we filtered the species dataset based on the following criteria: i) The species occurred at two or more sites. ii) The species were present in at least 20 site-time intervals. iii) Species growth events were recorded at least five times. Based on these criteria, we initially selected 89 species for modelling. 3.2.2 Temperature Data 3.2.2.1 Data cleaning and Calibration Initially, we plotted raw temperature measurements from each temperature logger and visually inspected them for any malfunctions, such as battery failures or displaced sensors. During 5 years of continuous measurements, we encountered 4 events of battery failures, which resulted in complete data gaps and 10 events of sensor displacement, which manifested themselves by increased daily temperature fluctuations. We removed these false measurements (see Table 1) to have them later replaced with modelled microclimatic data (see 3.2.2.2 Gap-filling via Random Forest Regression). Table 1 provides a summary of data gaps in the temperature time series. It includes specific intervals of temperature logger malfunctions (Gap Start and Gap End) and the total day durations of missing data (No. of Days in Gap) for each temperature data logger. Logger ID Gap Start Gap End No. of Days in Gap F l 01/03/2022 08/03/2022 8 F2 08/03/2022 08/03/2022 1 F2 04/10/2022 24/10/2022 21 F2 25/12/2023 25/12/2023 1 F5 29/04/2024 29/04/2024 1 F7 15/05/2021 30/05/2021 16 F7 24/12/2023 30/05/2024 159 F8 15/05/2021 30/05/2021 16 F8 28/05/2025 28/05/2025 1 40 METHODS Logger ID Gap Start Gap End No. of Days in Gap F9 15/05/2021 29/05/2021 15 F9 01/02/2025 21/03/2025 49 EMS Minikin loggers provided the primary temperature measurements, but three sensors stopped working in November 2024. For these three sites, temperature measurements continued with EMS MicroLog SP3 sensors. Since the temperature sensor of the EMS MicroLog SP3 logger is placed slightly higher in the soil profile than the one of the EMS Minikin, we had to perform a post-hoc temperature calibration. For each pair of loggers, we used a linear model with harmonic terms to account for observed temperature offset between sensors and seasonal variation: (2n.DOY\ /Ztz.DOY\ Tmmkin = fio + PiTSp3 + Pi s i n ^ — J + P3 cos y — J + £ DOY represents the day of the year and TMinikin, TSP3 represent the recorded temperature from EMS Minikin logger and EMS SP3 logger, respectively. This allowed us to correct EMS MicroLog SP3 recordings for the period between November 2024 and December 2025, ensuring a smooth transition between these two sensors. 3.2.2.2 Gap-filling via Random Forest Regression Gaps in the temperature time series that resulted from logger malfunctions have been filled using a site-specific Random Forest based approach. In the context of our work, gap-filling refers to estimations of missing values based on existing patterns in our dataset. We have decided to use the Random Forest approach as it is used as a standard machine learning technique for gap-filling meteorological data and can capture complex non-linear interactions (Lalic et al. 2024, Boomgard-Zagrodnik & Brown 2022, Tang & Ishwaran 2017). For each site (i.e. F l , F2, F5, F7, F8, F9), we created a training dataset, using periods where temperature data was available for both the target site and the remaining study sites. The temperature at a target site was used as the response variable, while the temperatures from the other sites served as predictors. Specifically, for site F l the training dataset included 34 366 observations, with each time step comprising seven predictor sites. The site F2 was 41 METHODS trained on 34 558 observations with seven predictor sites, while sites F5, F7, F8, and F9 were each trained on 34 558 observations using six predictor sites. We fitted a separate Random Forest model for each site. We parameterized each site-specific model with a forest size of 500 decision trees (trees = 500). To ensure model diversity and to prevent overfitting we limited each split to a random subset of three predictors (mtry = 3), and each terminal node was required to have at least 10 observations (minn = 10). After the model training, the models were used to predict missing values at each of the target sites. This spatial imputation approach does assume a relatively high-degree of crosscorrelation between site microclimates at any given hour. While the models' performance might decrease if multiple sites experience simultaneous data loss, we ensured that the missing data structure in our dataset left enough predictor sites for reliable estimations. 3.2.2.3 Microclimate Change Projections We acquired the future climate data projections from the NASA Centre for Climate Simulation (NCCS). We specifically used the NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP-CMIP6). This dataset contains many Global Climate Models (GCMs) that are built based on the Coupled Model Intercomparison Project Phase 6 (CMIP6) framework. The GCMs provide downscaled and bias corrected daily climate projections at approximately 0.25° resolution (-27.8 km) (Thrasher et al. 2021, Thrasher et al. 2022). All of our study sites were situated within one grid cell. In our study we are using the high-resolution G C M model MPI-ESM-1-2-HR developed by the Max Planck Institute for Meteorology (Müller et al. 2018) and downscaled and bias corrected by NCCS. For further analysis we are using the four high-priority SSP scenarios, that are also evaluated in the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC 2021): SSP1-2.6, SSP2-4.5, SSP3-7.0, SSP5-8.5 (see Box 1: Climate change and Radiative forcing). We transformed the macroclimate future temperature projections onto site-specific microclimate temperatures, based on our measured microclimatic soil data collected 42 METHODS between May 2021 and December 2025. This approach involved two steps. First, we established a relationship between observed microclimate soil temperatures and macroclimate air temperatures taken from the weather station Brod nad Dyji (CHMU 2026). For each site, we built a linear mixed-effect model that estimated overall microclimate/macroclimate relationship, while allowing for month-specific intercepts (seasonal offsets) and slopes (sensitivity to macroclimate): Tmicroj.j = {.Pq £ t 0.j ) (Pi u l j ) ^macroA.j £ i,j Where Tmicro and Tmacro represent the microclimate and macroclimate temperature for observation i in month j. The u0j and ux;- represent the random intercepts and slopes for each month, respectively. After we fitted the linear mixed-effect models for every study site, we extracted the conditional and marginal R2 , as well as daily residuals to validate the models' accuracy in transformation of macroclimate temperatures into site-specific microclimates. Second, we used this trained model to transform the MPI-ESM-1-2-HR macroclimate projections for each SSP scenario into site-specific microclimate temperatures. We got a continuous microclimate temperature dataset for each of our study sites from 2021 to 2100. 3.2.3 Species Responses 3.2.3.1 Growth Probability Models We fitted a Generalized Linear Mixed-Effect Models (GLMM) with binomial distribution for each species to estimate the probability of growth (89 models in total; see Appendix G). To account for the non-linear response of growth to temperature, we used second order polynomials. logitipu) = p0 + pju + faTl + m 43 METHODS Where ptj is the probability of growth for the observation i at sitej, Tis the mean interval temperature and u;- represents the site-specific random effect. In cases where the random effect showed zero variance (singular fit), we refitted the data using Generalized Linear Models (GLM) without the random intercept. To evaluate the goodness of fit, we extracted the coefficient of determination (R2 ) for each species. For GLMMs we calculated the Nakagawa marginal R2 (representing variance explained by fixed effects), and for GLMs we used the McKelvey-Zavoina R2 (see Appendix A) We calculated the Receiver Operating Characteristics (ROC) curves and Area Under the Curve (AUC) to further evaluate models' performance for each species. We visually inspected the probability of growth curves and excluded 14 species {Dianthus lumnitzeri, Erysimum diffusum, Jovibarba globifera, Poa bulbosa, Veronicapraecox, Minuartia rubra, Minuartia setacea, Carex humilis, Medicago minima, Melica ciliata, Reseda lutea, Senecio jacobaea, Seseli osseum, Poa angustifolia) that showed unrealistic growth-temperature relationships and low AUC/R2 values. This step left us with 75 species for which we predicted the probability of growth under current (2021-2025) and all four future (2096- 2100) climate scenarios (SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5). Additionally, to convert the probabilities back to binary growth/no-growth responses, we calculated speciesspecific thresholds using Youden's Index that maximizes the sum of sensitivity (true positive rate) and 1-specificity (false negative rate). In further analyses of species growth responses, we constrained the results to just three SSP scenarios, two which are considered the most plausible (SSP2-4.5, SSP3-7.0) and the most extreme scenario (SSP5-8.5) for comparison (Hausfather & Peters 2020). The SSP1-2.6 scenario was excluded as unrealistic for our study, as the future predictions have already been reached by current climate measurements. 3.2.3.2 Defining Species Growth Temperature Ranges We applied the converted probabilities of growth to a temperature grid, representing the temperature ranges measured during our study, extended by 10 °C beyond the observed 44 METHODS values. Because some species (especially annuals) did not exhibit a typical bell-shaped response to temperature and often predicted growth at extreme below zero temperatures, we applied a bottom threshold, where we treated all probabilities as no-growth at temperatures < 1 °C (this threshold was defined arbitrarily based on information in literature, with similar results when using a 0 °C threshold; Peacock 1976, Wingler & Hennessy 2016). We further classified the temperature ranges based on their minimum (start) and maximum (end) temperatures of predicted growth ranges. Prior to classification we converted the temperature values into a matrix of Euclidean distances. We then performed hierarchical clustering using Ward's method and determined the final number of clusters with silhouette width analysis. 3.2.3.3 Identifying Winners and Losers We used the predicted binary growth responses under current (2021-2025) and future (2096-2100) climate scenarios to calculate the number of growth-days for each species. To identify the 'winners' (species gaining growth-days in the future) and 'losers' (species losing growth-days in the future) we calculated a Log-ratio Growth Index (LGI): To characterise the functional and ecological drivers of these shifts, we fitted separate linear regression models with Gaussian distribution for each climate scenario. The LGI for each of three climate scenarios was used as the response variable (SSP2-4.5, SSP3-7.0, SSP5-8.5) to test for significant differences across following categorical predictors: life history strategy (annuals, short-lived perennials, perennials), biogeographical range (mediterranean, submediterranean, continental, temperate and central European species), leaf functional anatomy (succulent, scleromorphic, mesomorphic leaves, as well as their intermediate categories) (Pladias 2026) and IUCN conservation status (Grulich 2017). The life history strategies were obtained from Fischer et al. (2020) and Fischer et al. (2023), while categories of biogeographical range follow classification used in Chytrý et al. (2022b). For each model, we estimated marginal means (EMMs) for all levels within each categorical predictor and 45 METHODS tested whether they differed significantly from zero (zero represents no change in growing season length). We assessed the statistical significance of each model using 95% confidence intervals and p-values derived from models (plotted EMMs). 3.2.3.4 Seasonal Growth Dynamics To analyse whether species will shift their growing seasons under climate change, we aggregated the daily growth predictions across all species. For each day of the study period (~4 years), we calculated the total number of species that showed growth. We also calculated the number of species exhibiting growth for each vegetation type separately, to compare the seasonal dynamics between vegetation types (see Appendix F). Additionally, to determine individual response in vegetation season shifting, we projected the growth/no growth response for each species separately across the whole study timeframe (see Appendix H). 3.2.3.5 Biomass Change Projections To estimate changes in biomass dynamics, we modelled daily biomass change for each of the 75 studied species. We modelled the changes in biomass using a two-part approach that captures the probability of occurrence and the magnitude of biomass change. First, we modelled the probability of positive and negative biomass change using two separate binomial generalized linear mixed-effect models (GLMMs). The response variable was the occurrence of positive (increase) or negative (decrease) daily biomass change (indicated as 1/0), mean daily temperature as a second-order polynomial was a predictor and sites were included as a random effect. Second, we modelled the magnitude of biomass change using two linear mixed-effect models. We fitted separate models for positive and negative biomass changes, with log-transformed daily biomass change as the response variable, daily mean temperature with second-order polynomial as the predictor, and sites as a random effect. To obtain the expected daily net biomass change, we combined the predicted probabilities and magnitudes. We calculated the expected positive biomass change as the probability of positive change multiplied by its predicted magnitude, while the expected negative biomass change was calculated similarly. We back-transformed the predictions from the linear mixed-effect models prior to any further calculations. Finally, 46 METHODS we estimated the overall expected daily net biomass as: expected daily net biomass change = expected positive change - expected negative change. We applied this method to both current (2021-2025) and future (2096-2100) climate scenarios. To get mean daily biomass change for each species, we averaged the expected daily values across the studied years. To assess the potential for seasonal compensation under future climate warming, we used the modelled expected daily net biomass change to calculate the cumulative positive biomass accumulation for two distinct periods, the first half of the year (January to July) and the second half of the year (August to December). For this analysis, we excluded species that did not exhibit the typical two-peak growth pattern and instead maintained positive biomass change during summer period (e.g. Dorycnium germanicum, Galatella linosyris, Lamium amplexicaule, Seseli osseum, Stipa capillata and S. pulcherrima). For each remaining species and climate scenarios, we calculated the total seasonal biomass accumulation by summing only the predicted positive daily net biomass changes. To compare future climate scenarios against the current baseline, we calculated the relative biomass accumulation for each SSP scenario, as Relative biomass accumulation = Future positive net biomass accumulation/Current positive net biomass accumulation. 47 RESULTS 4 Results 4.1 Microclimate Change 4.1.1 Macroclimate -> Microclimate Transformation The linear mixed-effect models of the relationship between macroclimate air temperature measured by local weather station and on-site-measured soil temperatures showed high explained variance across study sites. The conditional R 2 values (representing the variance explained by both fixed and random effects, i.e., individual months) exceeded 0.90 on all but one site (F7) which had a value of 0.81. This indicates that macroclimate air temperature can be relatively accurately translated to microclimate temperature based on a rather simple linear relationship (Figure 4). Additionally to R2 , the extracted daily residuals were mostly centred around 0 °C (mean residual across sites = 5.46 x 10"1 5 o C, standard deviation of residuals across sites = 1.68 °C). The 95% range of residuals (2.5%-97.5% quantiles) across sites was -3.2 to 3.49 °C (see Appendix C). The scatter plots and month-specific regression lines revealed that microclimate buffering is both temperature dependent and seasonally asymmetric. During colder seasons (spring, autumn and winter months) when macroclimate air temperatures are low, the microclimate temperature remains significantly warmer than macroclimate temperature (flatter slopes). On the other hand, when macroclimate temperature warms during these months the microclimate temperature remains cooler. In contrast, during summer months, the microclimate temperature is almost consistently warmer than macroclimate temperature, with the exception of dealpine grasslands, where microclimate temperature remains lower compared to high macroclimate temperature (see also Appendix D). 48 RESULTS F1: Svatý Kopeček F4: Děvín F7: Kotel -10 0 10 20 30 -10 0 10 20 30 -10 0 10 20 30 F2: Svatý Kopeček F5: Děvín F8: Stolová hora -10 0 10 20 30 -10 0 10 20 30 -10 0 10 20 30 Macroclimate temperature (°C) Figure 4 The scatter plot of daily macroclimate (Tmacro) and microclimate (Tmicro) temperature across nine study sites, facetted by vegetation type. The points are coloured by month. The dashed line represents the identity line (macroclimate = microclimate). The regression lines represent the site-specific linear mixed-effect models: Tmicro~ T m a c r o + (Tmacro(month)). The microclimate soil temperature was more stable with fewer extreme fluctuations compared to macroclimate air temperature. In rocky and loess dry grasslands, which are 49 RESULTS predominantly situated on south-facing slopes, the daily soil temperatures were almost consistently higher than the macroclimate air temperature. In contrast, the dealpine Sesleriadominated grasslands, situated on north-facing slopes exhibited lower microclimate temperatures than the surrounding macroclimate. Our downscaled temperature predictions track observed soil temperatures effectively across study sites. However, seasonal differences are present, where the transformed values follow the microclimate soil temperatures more closely during summer months than in colder seasons (Figure 5). 50 RESULTS F1: Rocky grassland F9: Loess grassland 30 O o 0) 20 2 10 a. .© 0 30 20 10 0 F5: Dealpine grassland • .f I in A 1 j j ^ us Jan 2022 Apr 2022 Jul 2022 Oct 2022 Jan 2023 / • / Micro T (measured) / - / Micro T (modelled) Macro T (weather station) Figure 5 Comparison of observed microclimate soil temperature (Micro T (measured)), modelled microclimate soil temperature (Micro T(modelled)) and macroclimate air temperature (Macro T (weather station)) across three vegetation types for the year 2022. Each vegetation type is represented by one particular study site. 51 RESULTS 4.1.2 Future Macroclimate and Microclimate Temperature Trajectories and Seasonal Shifts Macroclimate projections across all scenarios showed a consistent warming trend, where under the SSP1-2.6 scenario, macroclimate temperatures increased at a rate of 0.037 °C per decade, reaching 10.7 °C in 2100. Under the moderate SSP2-4.5 scenario, the rate increased to 0.18 °C per decade, with temperatures reaching 13.4 °C in 2100. The SSP3-7.0 and SSP5- 8.5 scenarios exhibited the steepest warming, with rates of 0.43 °C and 0.6 °C per decade, respectively, both reaching 14.5 °C by the end of the century (Figure 6). When comparing these broad macroclimate trends with microclimate conditions we observed a clear offset, which was dependent on vegetation type. Specifically, loess and rocky dry grasslands showed a positive offset, with microclimate temperatures remaining higher than macroclimate mean throughout the century. However, this positive offset decreased under more extreme scenarios. On the other hand, the dealpine Sesleria-dommated grasslands exhibited an opposite trend, they maintained a negative offset that intensified with more pronounced macroclimate warming. The mean annual microclimate temperatures showed pronounced interannual fluctuations across all climate scenarios, with a relatively clear divergence becoming visible after the second half of the century. While fluctuation patterns were consistent across vegetation types, the temperatures reached by 2100 differed significantly (see Table 2). The loess and rocky dry grasslands maintained the highest annual microclimate temperatures throughout the century, with loess sites being consistently the warmest, whereas the dealpine Sesleriadommated grasslands remained substantially cooler (Figure 7). Table 2 Microclimate offsets and projected microclimate temperatures across vegetation types under all four climate change scenarios. The Mean offset (°C) represents the mean deviation of microclimate soil temperature from macroclimate air temperature across the studied period 2021- 2100. The Mean 2100 T (°C) represents the mean annual microclimate soil temperature for each vegetation type under each climate scenario by the end of the century. Scenarios Vegetation types Mean offset (°C) Mean 2100 T (°C) SSP1-2.6 Dealpine -0.58 10.32 SSP1-2.6 Loess 2.56 13.43 52 RESULTS Scenarios Vegetation types Mean offset (°C) Mean 2100 T (°C) SSP1-2.6 Rocky 2.11 13.01 SSP2-4.5 Dealpine -0.88 11.49 SSP2-4.5 Loess 2.35 15.01 SSP2-4.5 Rocky 1.91 14.59 SSP3-7.0 Dealpine -1.24 11.96 SSP3-7.0 Loess 2.08 15.67 SSP3-7.0 Rocky 1.64 15.27 SSP5-8.5 Dealpine -1.44 11.95 SSP5-8.5 Loess 1.94 15.63 SSP5-8.5 Rocky 1.50 15.21 Regarding the seasonal temperature dynamics, the microclimate projections indicated shifts both in timing and intensity of warming across the year (Figure 8). We identified two trends. Firstly, the warmer temperatures were projected to extend into earlier spring and later autumn, with the expansion being most pronounced under SSP3-7.0 and SSP5-8.5 scenarios. Secondly, while winter mean minimum temperatures remained relatively stable (ranging from 0.15 °C to 1.35 °C across scenarios), the summer mean maximum temperatures increased substantially with climate warming, where the summer peak under SSP5-8.5 reached 29.1 °C compared to 25.1 °C under current conditions. 2040 2060 2080 2100 2040 2060 2080 2100 Year Year 53 RESULTS Figure 6 Projected mean annual macroclimatic temperature (°C) for the period 2021-2100 across all four SSP scenarios, as projected for the 025° * 0.25° grid cell. The points and solid lines represent mean annual temperature, while the thicker line shows thefive-yearfloatingaverage. Dashed lines indicate the linear trend over the studied period. SSP1-2.6 SSP2-4.5 E 2020 2040 2060 2080 2100 2020 2040 2060 2080 2100 - SSP3-7.0 SSP5-8.5 2020 2040 2060 2080 2100 2020 2040 2060 2080 2100 Year Vegetation type: Rocky grasslands Loess grasslands H Dealpine grasslands Figure 7 The mean annual microclimatic temperatures from 2021 to 2100 for each vegetation type (rocky, loess and dealpine dry grasslands), facetted by climate change scenarios. The points and solid lines represent mean annual temperature, while the thicker line shows thefive-yearfloating average. Dashed lines indicate the linear trend over the studied period. 54 RESULTS SSP5-8.5- SSP3-7.0- 1 s— c SSP2-4.5- o CO SSP1-2.6- Current^ . y _ K * & 0 ° C>N ~ Q>N Date Figure 8 Heatmap of mean daily temperatures across the year (days 1-365). The plot compares the current climate (2021-2025) with four future climate scenarios (2096-2100). 4.2 Species Responses 4.2.1 Probability of Growth Curves and Temperature Growth Ranges Studied species showed distinct temperature growth ranges, which were classified based on their minimum and maximum growth temperatures into four groups (Figure 9). The first group (cluster 1, 37 species) can be characterized as thermal generalists. This group represents the species with a broad thermal range, characterized by a mean start temperature of 3.7 °C and a mean end temperature of 18.7 °C. This group is dominated by perennials but also includes short-lived perennials and annuals that extend their growing season into late May, such as Acinos arvensis and Arenaria serpyllifolia. 55 RESULTS The second group (cluster 2, 13 species) can be described as cold-adapted specialists. This group consists mainly of early spring annuals like Arabis auriculata, Cerastium pumilum agg., Holosteum umbellatum, Saxifraga tridactylites. These species show growth at low temperatures (mean start temperature of 2.9 °C) and stop growing around 12.1 °C. This group also includes short-lived perennials (e.g. Alyssum alyssoides) and a few perennials like Festuca csikhegyensis and Dianthus pontederae that initiate their growing season at relatively low temperatures. The third group (cluster 3, 11 species) can be described as heat-adapted generalists. They are characterized by the highest thermal growth thresholds, where the mean start temperature is 9.6 °C and the mean end temperature is 23.9 °C. These species are predominantly scleromorphic perennials (Asperula cynanchica, Stipa capillata, Fumana procumbens) adapted to maintain growth into the warm summer temperatures. There are also species of dealpine grasslands (Thalictrum minus) and loess grasslands (Galatella linosyris, Securigera varia, Thymus pannonicus). The fourth group (cluster 4, 14 species) can be described as warm adapted specialists. This group shows intermediate thermal ranges (mean start temperature = 8.8 °C, mean end temperature = 17.6 °C). While species from this cluster start growing at similar temperatures as cluster 3, they have overall narrower ranges and stop growing sooner. This group consists mainly of perennials with scleromorphic to mesomorphic (Sanguisorba minor, Potentilla incana, Teucrium chamaedrys, Fragaria viridis, Euphorbia cyparissias) leaves suggesting higher sensitivity to peak summer temperatures. 56 RESULTS Vincetoxicum hirundinariaVeronica prostrataTeucrium chaimaedrysStachys recta Sanguisorba minorSalvia pratensisPotentilla incana • Galium verum Galium glaucumFragana viridis Euphorbia cyparissiasCentaurea stoebeBupleurum falcatumAvenula pubescensThymus pannonicusThalictrum minusStipa capillataSecurigera variaMedicago falcataGalatella linosyrisFumana procumbensDorycnium germanicumCarex supinaAsperula cynancnicaAnthericum ramosumViola tricolor Saxifraga tridactylitesMicrothlaspi perfoliatumLamium amplexicauleHolosteum umbellatumFestuca csikhegyensisDianthus pontederaeCerastium pumilum agg.Aster amelTusArabis sagittata Arabis auriculataAnthyllis vulnerariaAlyssum alyssoidesVeronica spicataVeronica chamaedrys agg. Verbascum lychnitisThymus praecoxTeucrium montanumStipa pulcherrimaSesleria caeruleaSedum albumScorzonera austriacaScabiosa ochroleucaPoa badensisPlantago mediaPhleum phleoidesPeucedanum cervariaMuscari tenuiflorumLinaria genistifoliaKoeleria macranthaIris pumilaHypericum perforatumFestuca valesiacaFestuca rupicolaErysimum odoratumEryngium campestreEJymus hispidusEchium vulgareCentaurea triumfetti Campanula persicifoliaCamelina microcarpaAstragalus onobrycnisArrhenatherum elatiusArenaria serpyllifoliaAlyssum montanumAllium senescens subsp. montanumAllium flavumAdonis vernalisAcinos arvensisAchillea collina- Cluster 1 10 20 Temperature °C Figure 9 Predicted temperature growth ranges for 75 dry grassland species across a temperature grid (-8.4 °C to 35.6 °C). Each horizontal bar represents the interval between the minimum and maximum temperatures where the probability of growth exceeds the species-specific threshold (Youden's index). A minimum physiological threshold was set to 1 °C for all species. Species are coloured into four groups based on the results of Ward's hierarchical cluster analysis. 57 RESULTS 4.2.2 Identifying Winners and Losers Projected changes in the number of growth-days are highly species-specific, however a general decrease in growth-days was predicted across the community under future climate scenarios (Figure 10). Under the scenario SSP2-4.5, the community experienced the widest decrease in the number of growth-days. In this scenario only three species are projected to gain growth-days (Stipa capillata, Asperula cynanchica, Vincetoxicum hirundinarid). The gain is most substantial for Stipa capillata, followed by Asperula cynanchica. For the remaining 72 species the LGI values are negative. The most significant relative losses under this scenario have been predicted for Anthyllis vulneraria, Aster amellus, Dianthus pontederae, Festuca rupicola, and Salviapratensis. As the warming intensifies under the SSP3-7.0 and SSP5-8.5 scenarios more species take advantage of the warming of the cooler seasons. Under SSP3-7.0, 23 species show positive LGI values. The species gaining the most growth-days relative to their current baseline include Arabis sagittata, Asperula cynanchica, Stipa capillata, Sanguisorba minor, and Galium glaucum. In contrast, the primary losers under this scenario are Saxifraga tridactylites, Aster amellus, Arabis auriculata, Viola tricolor and Alyssum alyssoides. A similar trend can be observed under the most extreme scenario, SSP5-8.5, where 17 species gain growth-days, including Arabis sagittata, Galium glaucum, Sanguisorba minor, Asperula cynanchica, and Potentilla incana. The species with the most severe losses of growth-days are represented by Galatella linosyris, Saxifraga tridactylites, Arabis auriculata, Alyssum alyssoides, and Fumanaprocumbens. 58 RESULTS SSP2-4.5 SSP3-7.0 ssps-a.5 Stipa capillata Asperula cynanchica Vincetoxicum hirundinaria Carex supina Allium senescens subsp. montatium Erysimum Ddoratum Bupleurum falcatum Thalictrum minus Arrhenalherum elatius Sesleria caerulea Anthericum ramosum Galatella linosyris Galium verum Elymus hispidus Thymus praecox Alyssum alyssoides Scorzonera austriaca Muscari tenuiflorum Avenula pubescens Festuca valesiaca Arabia auriculata Dorycnium germanicum Astragalus onobrychis Stipa pulcherrima Veronica chamaedrys agg. Veronica spicata Hypericum perforatum Verbascum lychnitis Iris pumila Linaria genistifolia Lamium amplexicaule Microthlaspi perfoliatum Poa badensis Holostsum umbellatum Achillea collina Medicago falcata Sedum album Alyssum montanum Peucedanum cervaria Scabiosa ochroleuca Arabis sagittata Saxifraga tridactylites Camelina microcarpa Adonis vernalis Securigera varia Koeleria macrantha Cerastium pumilum agg. Teucrium chamaedrys Teucrium montanum Allium flavum Eryngium campestre Festuca csikhegyensis Potentilla incana Echium vulgare Sanguisorba minor Stachys recta Euphorbia cyparissias Fragaria viridis Viola tricolor Acinos arvensis Veronica prostrata Arenaria serpyllifolia Centaurea triumfetti Thymus pannonicus Fumana procumbens Centaurea stoebe Galium glaucum Phleum phleoides Campanula persicifolia Plantago media Salvia pratensis Festuca rupicola Dianthus pontederae Aster amellus Anthyllis vulneraria Arabis sagittata Asperula cynanchica' Stipa capillata Sanguisorba minor Galium glaucum Potentilla incana Avenula pubescens Veronica prostrata Vincetoxicum hirundinaria Veronica chamaedrys agg. Erysimum odoratum Centaurea stoebe Thalictrum minus Bupleurum falcatum Galium verum Euphorbia cyparissias Fragaria viridis Anthericum ramosum Allium senescens subsp. montanum Festuca csikhegyensis Elymus hispidus Thymus praecox Salvia pratensis Medicago falcata Muscari tenuiflorum Carex supina Astragalus onobrychis Festuca valesiaca Securigera varia Stipa pulcherrima Scorzonera austriaca Linaria genistifolia Arrtienatherum elalius Iris pumila Koeleria macrantha Dorycnium germanicum Poa badensis Echium vulgare Galatella linosyris Scabiosa ochroleuca Sesleria caerulea Adonis vernalis Campanula persicifolia Achillea collina Peucedanum cervaria' Teucrium chamaedrys Thymus pannonicus Centaurea triumfetti Veronica spicata Hypericum perforatum Verbascum lychnitis Teucrium montanum Stachys recta Alyssum montanum Sedum album Eryngium campestre Acinos arvensis Allium flavum Camelina microcarpa Dianthus pontederae Plantago media Phleum phleoides Fumana procumbens Arenaria serpyllifolia Festuca rupicola Holosteum umbellatum Cerastium pumilum agg. Lamium amplexicaule Microthlaspi perfoliatum Anthyllis vulneraria Alyssum alyssoides Viola tricolor Arabis auriculata Aster amellus Saxifraga tridactylites Arabis sagittata Galium glaucum Sanguisorba minor Asperula cynanchica Potentilla incana Centaurea stoebe Salvia pratensis Festuca csikhegyensis Astragalus onobrychis Euphorbia cyparissias Veronica prostrata Erysimum odoratum Stachys recta Jnaria genistifolia Aster amellus Muscari tenuiflorum Echium vulgare Elymus hispidus Adonis vernalis Dianthus pontederae Veronica chamaedrys agg. Fragaria viridis Anthyllis vulneraria Galium verum Koeleria macrantha Carex supina Peucedanum cervaria Thymus praecox Festuca valesiaca Allium senescens subsp. montanum Campanula persicifolia Alyssum montanum Camelina microcarpa Acinos arvensis Sedum album Verbascum lychnitis Poa badensis Arrhenatherum elatius Eryngium campestre Securigera varia Iris pumila Allium flavum Phleum phleoides Teucrium chamaedrys Scabiosa ochroleuca Centaurea triumfetti Holosteum umbellatum Scorzonera austriaca Cerastium pumilum agg. Arenaria serpyllifolia Hypericum perforatum Thymus pannonicus Festuca rupicola Stipa pulcherrima Achillea collina _amium amplexicaule Anthericum ramosum Medicago falcata Sesleria caerulea Microthlaspi perfoliatum Veronica spicata Teucrium montanum Viola tricolor Thalictrum minus Avenula pubescens Stipa capillata Bupleurum falcatum Dorycnium germanicum Plantago media Vincetoxicum hirundinaria Fumana procumbens Alyssum alyssoides Arabis auriculata Saxifraga tridactylites Galatella linosyris Index 0 1 0 u -0.1 -0.2 -0.1 00 -0 100.058 0* V -0.1 0.0 0.1 Growing index (log ratio of growing days) Figure 10 Log-ratio Growth Index (LGI) for 75 species across three selected climate scenarios (SSP2-4.5, SSP3-7.0, SSP5-8.5). The LGI represents the relative change in the number of future growth-days compared to the current baseline. Bars are coloured by the direction of change (blue indicates the loss of growth-days, while red indicates a gain in growth-days). Species are ordered along the y-axis from the most negative to the most positive LGI values. 59 RESULTS In addition to general trends, we evaluated species LGI values across three vegetation types (rocky dry grassland, loess dry grassland, and dealpine Sesleria-dominated grassland). Rocky and loess grasslands showed similar patterns in both the number and identity of species gaining growth-days across studied climate scenarios. The highest number of gaining species was projected under SSP5-8.5 (e.g. Sanguisorba minor, Centaurea stoebe, Galium glaucum, Potentilla incana), while only a few species gained growth-days under SSP2-4.5 (e.g. Stipa capillata, Asperula cynanchica or Arabis sagittata). In contrast, dealpine grasslands exhibited different patterns. A relatively high number of species was projected to gain growth-days under SSP2-4.5 and SSP3-7.0 (up to 33 species), including Asperula cynanchica, Dorycnium germanicum, Thalictrum minus, Vincetoxicum hirundinaria. However, under the most extreme scenario (SSP5-8.5), the number of species gaining growth-days (e.g. Festuca csikhegyensis, Aster amellus, Arabis sagittata, Asperula cynanchica) was lower compared to rocky and loess dry grasslands (see Appendix E). The linear model analysis of LGI revealed consistent negative values of the log-ratio growth index (LGI) across all trait groups, indicating an overall reduction of growing season length relative to current conditions (see Figure 11). However, the significance of these responses varied across climate change scenarios and trait categories. Under the SSP2-4.5 scenario, most of the studied trait groups showed significantly negative estimated marginal means. In contrast, responses under the more extreme scenarios (SSP3-7.0 and SSP5-8.5) were more variable, with an increasing number of trait categories exhibiting not significant deviations from zero. Regarding life history strategies, all categories (i.e. annuals, perennials and short-lived perennials) showed significantly negative LGI values under the SSP2-4.5 scenario. However, under scenarios SSP3-7.0 and SSP5-8.5 annuals and perennials maintained significantly negative values, while short-lived perennials exhibited not significant declines (p-value > 0.05). Biogeographical elements followed a similar trend as life history strategies under the SSP2-4.5 scenario, with all elements exhibiting significantly negative LGI values. Under the SSP3-7.0 scenario, Mediterranean and Sub-mediterranean species retained significantly negative values, while Temperate, Continental and Central European species 60 RESULTS showed no significant declines. Under the most extreme scenario (SSP5-8.5), most groups showed significantly negative values, with the exception of Central European species, which remained not significant. Species classified as Least Concerned (LC) consistently exhibited significantly negative LGI values across all climate scenarios. Other categories (NT, VTJ, EN) showed significantly negative values under SSP2-4.5, but their responses were not significant under SSP3-7.0. Under SSP5-8.5, all categories except Vulnerable (VU) showed significantly negative values. Finally, under SSP2-4.5 all leaf anatomy groups except succulent species exhibited significantly negative LGI values. Under SSP3-7.0, species with succulent and succulent/mesomorphic leaves showed no significant responses, while the other groups remained significantly negative. Under SSP5-8.5, species with succulent and succulent/mesomorphic, and mesomorphic/hygromorphic leaves exhibited not significant responses, while the rest of the groups continued to show significantly negative values (Figure 11). 61 Estimated marginal mean Succulent Succulent Mesomorp Scleromorph 2 Scleromorph < Mesomorph Mesomorph Mesomorph Hygromorph Succulent Succulent Mesomorp > 3 Scleromorph 2 Scleromorph < Mesomorph Mesomorph Mesomorph Hygromorph 1 (rt SSP2 W r i (rt SSP2 i * i *- in * * W • Í i * i M W • * i L Estimated marginal mean o ó ó o o _i LJ. b b b Cn O CH O Cn u s-u SSP3- i [ u s-u SSP3- ¥ h i + * ~ * -j bh i + * ~ * h i + * ~ * h i + * ~ * Estimated marginal mean c O Z . (I < p Al Estimated marginal mean Annuals Perennials Short-lived perennials * SSP 1 2-4.5 ; • 7J m CO ci— —i CO Estimated marginal mean Annuals Perennials Short-lived perennials — 1 — SSP w 4 1 3-7.0 4 1 • 01 Estimated marginal mean Estimated marginal mean Estimated marginal mean Estimated marginal mean 0) Succulent Succulent Mesomorp Scleromorph 9 Scleromorph •< Mesomorph Mesomorph Mesomorph Hygromorph 1 zs Ü1 in 1SSP5-8.5 • 1 - 1 zs Ü1 in 1SSP5-8.5 • 1 ft- f zs Ü1 in 1SSP5-8.5 * * 1 p1 cn 1 c o a < (Q > o L SS i • P5-8.5 i • i ? P5-8.5 • <ň 9 1 * Central European - CD a cn CD O to m Continenta Mediterranean • ' Submediterranean • Temperate 3 C 1 •J Ii 1 SSP5-8 - i * SSP5-8 1 * 1 in W : 1 * IÍ * 1 * Annuals ~ Perennials o -3 Short-lived perennials 1 X —I— SSP X 1 cn1 Co in M *l •1 RESULTS Figure 11 Estimated marginal means (EMMs) of the log-ratio growth index (LGI) derived from linear models with Gaussian distribution. Separate models were fitted for each trait category and climate change scenario, with LGI as response variable and trait category as the predictor. Points represent the estimated marginal means, and error bars 95% confidence intervals. The horizontal dashed line marks zero, corresponding to no change in growing season length relative to current conditions. The Estimated marginal means are on logarithmic scale, positive values indicate increase in the number of growth-days under future conditions relative to current conditions, while negative values indicate a decrease. Statistical significance of differences from zero are indicated by asterisks (p < 0.001 = '***'; p < 0.01 = '**'; p < 0.05 = '*'; p > 0.05 = 'n.s'). 4.2.3 General Seasonal Growth Dynamics Under the current climate conditions, the species showed a distinct bimodal pattern of growth activity. The primary growing season started in February and persisted until midJune, followed by a summer non-growing period. The secondary growing season started in September and gradually faded off in December. As the climate warms this pattern is projected to shift. Under the SSP2-4.5 scenario, the growing patterns remained comparable to the current baseline, with one exception being the extension of the summer non-growing period further into October. However, under the more extreme scenarios (SSP3-7.0 and SSP5-8.5) the seasonal dynamics shift more. While under these pathways, growth events increased during winter months, this trend was offset by an earlier end of primary growing season, which faded in May. Furthermore, the autumn growing season did not start prior to October, widening the summer gap. Sporadic growth events could be still observed during the summer months under the current temperature conditions but were absent under the most extreme warming scenarios (Figure 12). While the general seasonal trends remained consistent across the studied area, individual vegetation types maintained unique differences under their specific microclimatic conditions (see Appendix F). In rocky dry grasslands, the spring growing season started earlier than in colder dealpine grasslands, though it ended sooner. Additionally, the summer non-growing period was the most pronounced in these grasslands, closely followed by loess dry grasslands. These site-specific differences were to a large degree wiped out under the future warming scenarios. Under the most extreme scenario (SSP5-8.5) the differences across 63 RESULTS vegetation types were practically eliminated. The timing of growing and non-growing periods became largely homogenized across all habitats. Figure 12 The heatmap of growth intensity across the year for the current climate and three future climate scenarios (SSP2-4.5, SSP3-7.0, SSP5-8.5). The growth intensity is defined as the mean sum of species predicted to grow on a given day across all study sites. 4.2.4 Biomass Change Estimations Daily biomass change showed a consistent seasonal pattern across most studied species, characterized by two distinct periods of biomass accumulation. The first peak occurred during winter and spring months, while the second peak was observed in autumn. In contrast, during the summer months the negative biomass change was most pronounced, indicating net biomass loss. A subset of species deviated from this general pattern and maintained positive biomass change during summer (e.g. Dorycnium germanicum, Galatella linosyris, Lamium amplexicaule, Seseli osseum, Stipa capillata and S. pulcherrima). Across the climate change scenarios, several consistent shifts in biomass dynamics could be observed. For most species, the end of the spring biomass accumulation period occurred earlier under 64 RESULTS future climate conditions, which is in line with the projected shifts in the growing season length. At the same time, biomass change during cooler months remained predominantly positive. Negative biomass change during summer months was already estimated for the current conditions, but its magnitude increased under future climate scenarios. This effect was intensified with increasing warming, where it was least and most pronounced under the SSP2-4.5 and SSP5-8.5 scenario, respectively (see Appendix I). SSP2-4.5 SSP3-7.0 SSP5-8.5 0.00 -0.05 -0.10 - - _ *** - • 1 - - I1 i i T V ~~I I I I I I I r~ ^^^V^^A6 V>V° ^V° Date Scenario: />/ Current SSP2-4.5 f / SSP3-7.0 SSP5-8.5 Figure 13 Estimated daily net biomass change across the year for three representative species differing in life history: Acinos arvensis (short-lived perennial), Arenaria serpyllifolia (annual), and Teucrium montanum (perennial). The grey line represents the response under current climate conditions, while the coloured lines show projected responses under future scenarios (SSP2-4.5, SSP3-7.0, and SSP5-8.5). The dashed horizontal line indicates zero net biomass change. The analysis of relative biomass accumulation (Figure 14) revealed contrasting trajectories between the two seasonal periods and across climate change scenarios. For the winter-spring period (January-July) species projections under SSP2-4.5 were generally around the current 65 RESULTS baseline, with species showing slight declines. Under the more extreme scenarios (SSP3-7.0 and SSP5-8.5), the relative biomass accumulation exceeded the current baseline, with the most pronounced increases under SSP5-8.5. However, a different pattern was observed during the autumn period (July-August). Under the SSP2-4.5 scenario, relative biomass accumulation fell below the current baseline (1) across nearly all species, representing a higher reduction than that observed during the winter-spring period. While the relative biomass accumulation values were projected slightly below 1 under SSP3-7.0 scenario, under SSP5-8.5, they were projected above the current baseline. Despite these late-season increases under more extreme warming, the magnitude of accumulation during this period was lower than that observed during winter-spring period. 66 RESULTS ö • ) o • ) 0 • > • ) o • ) 0 • C c « > • ) o • ) 0 • C c « > • ) o • ) 0 • C c « o o o o • • I o o o o • • I o o o o • • I o o o • o • n A > •o o o • o • n A > • o o rv i o o o • o • n A o o rv i o o o • o • n A o o rv i I •I • oo • > •o c o o oo • > •o c o o oo • > •o c o o o o • D• o_ o o • D• o_ o o • D , o o Q O • o o • A l n A •, o o Q O • o o • A l n A •, o o Q O • o o • A l n A , o o Q O • o o • A l n A o o < > o * • CO o o < > o * • CO o o < > o * CO •CO • O 0 • oo •O 0 • oo •O 0 • oo o o • 0 o • CO • 30 • •o . • O 0 • oo o o • 0 o • CO • 30 • •o . •a o o o • 0 o • CO • 30 • •o . •a o o o • 0 o • CO • 30 • •o . •a o o o • 0 o • CO • 30 • •o . o o • 0 o • CO • 30 • •o . oo o o • 0 o • CO • 30 • •o . oo o o • 0 o • CO • 30 • •o . ° o • o • o • o • o o o< n o • o • o • o • o o o< n o • o • o • o • o o o< n o • o • o • o • o o o< nJ V o • n d • J V o • n d o •<-> o c o •<-> o c •o o • O 0 • o 'o • o o • O 0 • o 'o • o o • O 0 • o 'o • •CO 1 0.8 1.0 1.2 1.4 1.6 o O i • O i o o • o O i • O i o o • •o O i • O i o o • •o O i • O i o o • /-\ ' ("1 •» O O , § v i w o o« O O1 • o Cm • o 6 o o« O O1 • o Cm • o 6 o o« O O1 • o Cm • o 6 o o« O O1 • o Cm • o 6 o o« O O1 • o Cm • o 6 O O 9 • o O ' O ö w O O 9 • o O ' O ö w o o' • (-} o ' •» o O i «o O i « O O ' O /"\ 1 / - A•m n 1 V v. ü 1 o o • o o# O A • o o# O A •W 1 o cm o 1•O W 1 o cm o 1 o q> o o • o q> o o • o q> o o • *-> U i • J us O Oi • O O • o c* < » , o p • J us O Oi • O O • o c* < » , o p • J us O Oi • O O • o c* < » , o p • J us O Oi • O O • o c* < » , o p • J us O Oi • O O • o c* < » , o p • J us O Oi • O O • o c* < » , o p • J us O Oi • O O • o c* < » , o p •O i 9 <-> U f • 0 (9l A O 1• O O • /—v /m O O • /—v /m k O A \ U f r~i A o •o • \_/ WW O 1 W ^ 1 v 1 (J v-W O i0» o o« • oo 1 o o1 • o «0 , o»o , o o • O • ' o o • o 6 m em (J v-W O i0» o o« • oo 1 o o1 • o «0 , o»o , o o • O • ' o o • o 6 m em (J v-W O i0» o o« • oo 1 o o1 • o «0 , o»o , o o • O • ' o o • o 6 m em (J v-W O i0» o o« • oo 1 o o1 • o «0 , o»o , o o • O • ' o o • o 6 m em Q (J v-W O i0» o o« • oo 1 o o1 • o «0 , o»o , o o • O • ' o o • o 6 m em (J v-W O i0» o o« • oo 1 o o1 • o «0 , o»o , o o • O • ' o o • o 6 m em (J v-W O i0» o o« • oo 1 o o1 • o «0 , o»o , o o • O • ' o o • o 6 m em (J v-W O i0» o o« • oo 1 o o1 • o «0 , o»o , o o • O • ' o o • o 6 m em (J v-W O i0» o o« • oo 1 o o1 • o «0 , o»o , o o • O • ' o o • o 6 m em (J v-W O i0» o o« • oo 1 o o1 • o «0 , o»o , o o • O • ' o o • o 6 m em (J v-W O i0» o o« • oo 1 o o1 • o «0 , o»o , o o • O • ' o o • o 6 m emn (J v-W O i0» o o« • oo 1 o o1 • o «0 , o»o , o o • O • ' o o • o 6 m em— 1 Relative biomass accumulation (Jan-Jul) 0.8 1.0 1.2 Relative biomass accumulation (Jul-Dec) 1.4 Scenario: o SSP2-4.5 o SSP3-7.0 • SSP5-8.5 Figure 14 Relative biomass accumulation across studied species for the winter-spring (January-July) and autumn (July-December) periods. Points represent cumulative biomass accumulation (area under the curve of positive daily biomass change) relative to current conditions and are coloured by climate scenario (SSP2-4.5, SSP3-7.0, SSP5-8.5). The vertical dashed line indicates no change relative to current state (ratio =1). 67 DISCUSSION 5 Discussion Our results suggest that climate change in central European dry grasslands will primarily manifest through prolongation and intensification of the summer arid period that is unfavourable for plant growth. Using species-specific growth models calibrated against future microclimate temperature projections SSP2-4.5, SSP3-7.0 and SSP5-8.5, we showed that the majority of studied species are going to lose growth-days under all future climate scenarios, with losses driven predominantly by amplified summer microclimatic warming. Interestingly, the most severe decline was projected under the intermediate SSP2-4.5 scenario, revealing a non-linear relationship between warming intensity and growing season length. This surprising finding can be explained by the limits of seasonal growth compensations. While amplified summer microclimate temperatures are rising faster, winter ones are buffered and therefore rising slower. Consequently, while under the intermediate scenario, the slight increase of temperatures in spring and autumn does not compensate for losses in summer, in the strongest scenario, those losses will partly get compensated. Nevertheless, these new opportunities will most likely still fail to fully compensate for summer losses for most species. The response to changing climate systematically varied across functional groups, with heat-adapted scleromorphic species showing the smallest losses and cold-adapted early-spring annuals facing the greatest reductions in growing season length. 5.1 Macroclimate/Microclimate Change Across all grassland types, we observed an amplified warming during summer months and much slower temperature increase in winter months. The macroclimate-microclimate relationship in our study was seasonally asymmetric. During colder seasons, microclimate soil temperature was lower than macroclimate air temperature, indicating a buffering effect, suggesting strong temperature dependency. As air temperature increases, buffering weakens, and during summer months the microclimate temperatures at our sites were consistently warmer than air temperatures, suggesting an amplification of high temperatures. 68 DISCUSSION Rather than a uniform increase, temperature difference between current and future conditions across vegetation types showed similar seasonal patterns across all scenarios, with differences among scenarios occurring mainly in magnitude. Among the analysed scenarios, warming was weakest under SSP2-4.5 and stronger under SSP3-7.0 and SSP5- 8.5 (please note that we removed SSP1-2.6). Across these scenarios, the largest increase in temperature generally occurred in late summer and early autumn (August and September), while a secondary peak in spring was consistently observed either in April under SSP2-4.5 and SSP3-7.0 or in March under SSP5-8.5. A third peak in late autumn (November) was only present under more extreme scenarios (SSP3-7.0 and SSP5-8.5). These general patterns were consistent across vegetation types, but the magnitude of late summer warming was more pronounced in loess and rocky dry grasslands and less so in dealpine grasslands. These differences likely reflect the variation in site conditions, where open vegetation, shallow soils, and south-facing slopes promote heat accumulation, whereas cooler, north-facing sites retain reduced amplification of high temperatures. These findings indicate that while winter warming may be locally buffered, limiting the extent to which temperatures increase during colder seasons, summer warming can be locally amplified, potentially intensifying thermal stress and constraining plant responses. Climate change projections for Central Europe indicate warming across all seasons (IPCC 2021). Some studies have suggested that in some regions the warming might be more pronounced during winter months, potentially affecting species responses (Kreyling et al. 2019). However, these projections describe broad macroclimatic trends and may not reflect the conditions species actually experience (Klinges et al. 2024). Many studies highlight that microclimates, shaped by local environmental conditions, can buffer or amplify macroclimate signals and are therefore critical for understanding species responses to climate change (Lenoir et al. 2017, De Frenne et al. 2021, Zellweger et al. 2020, Haesen et al. 2023). In particular, a recent study in grasslands showed that microclimate buffering can be strongly temperature dependent, with microclimates providing a strong buffering effect up to a certain temperature threshold, beyond which they amplify the macroclimatic temperatures (Ashe-Jepson et al. 2025). Our results are consistent with these findings and further demonstrate that this temperature dependent buffering and amplification leads to 69 DISCUSSION pronounced seasonal asymmetry. While colder season temperatures were to some extent buffered relative to macroclimate, summer temperatures were consistently amplified, particularly in more exposed habitats. 5.2 Intensification of Summer Stress Across all scenarios, we found the largest increases of temperatures to be projected to occur during late summer, coinciding with a reduction in the number of predicted growth-days for most species. This suggests that warming does not simply extend the growing season but rather prolongs the duration of environmentally limiting conditions into late spring and early autumn. The majority of species were projected to lose growth-days under all future climate scenarios, supporting the argumentation that increasing summer stress might outweigh the potential gains in other parts of the year (Hoover et al. 2014, Huang et al. 2024, Bastos et al. 2014, Bastos et al. 2020). The negative effects of increasing temperatures during summer season on plant productivity are not only about whether species grow, but they also include physiological stress, reduction of photosynthesis efficiency and alteration of biomass allocation, where species often tend to favour root over shoot growth (Crafts-Brandner & Salvucci 2000, Zhou et al. 2020, Ma et al. 2021). In ecosystems limited by water, such as dry grasslands, high temperatures and soil moisture are strongly coupled, with elevated temperatures increasing evapotranspiration and reducing soil moisture availability (Bastos et al. 2020), thereby intensifying drought stress and its negative effects on productivity (Hoover et al. 2014, Huang et al. 2024). Consistent with these findings, our simulation of biomass change indicates that biomass loss during the summer months is likely to intensify under future climate change scenarios, particularly under the more extreme ones, suggesting that both duration and severity of summer stress will be critical in determining plant performance. Additionally, the effects of summer droughts are not limited only to the period of stress itself but may extend beyond it through legacy effects. It has been shown that summer droughts reduced net primary production of grassland ecosystems well beyond their immediate 70 DISCUSSION occurrence (Griffin-Nolan et al. 2018). Such effects not only affect aboveground but also belowground processes, where drought can alter soil microbial communities and plant-soil feedbacks and drive changes in community composition (Kaisermann et al. 2017). Plant responses to drought are further shaped by differences in resistance and recovery of species. These responses might differ between functional groups, where perennials might exhibit higher resistance to drought, maintaining biomass production more effectively during stress periods, whereas annuals tend to be less resistant but show faster post-drought recovery (Ruppert et al. 2015, Fischer et al. 2020) and may complete their life cycle before or after the arid period. In this context, plant performances under prolonged drought periods may not only depend on the resistance of species under such stress but also on their ability to recover once favourable conditions return. 5.3 Limits of Seasonal Compensation The intensification of summer stress raises the question of whether species can compensate by temporally shifting their growth activity into colder seasons. Our results show that the most widespread decline in growth activity was observed under the intermediate scenario (SSP2-4.5), indicating a non-linear response of growing season dynamics to warming. Under these conditions, most species may already experience temperatures that exceed their optimal ranges during summer, while temperatures during colder seasons will not yet surpass the thresholds required to substantially increase growth activity. As a result, the potential seasonal compensation may be limited. Under more extreme scenarios, these colder season thresholds may be crossed, enabling some species to increase their growth activity outside of the summer stress period. However, even under more extreme conditions, the majority of studied species were still projected to lose the overall number of growth-days. This is consistent with general projections for Central Europe that suggest that the growing season may not become longer despite its earlier onset (Chang et al. 2017). 71 DISCUSSION Consistent with this, our results show that many dry grassland species already use colder seasons under current conditions, either through early spring growth, sprouting or by maintaining some level of activity during relatively mild winters. This is supported by our field observations, where seedlings of annuals and short-lived perennials persist throughout winter, and some perennials remain partially active. This phenomenon is also well reflected in the relatively high number of species in our study plots during winter months, and the peak in species richness recorded in March (see also Fischer et al. 2023). Importantly, growth activity alone does not capture the magnitude of plant responses. To address this, we used a simplified modelling approach to estimate biomass changes under warming conditions. While these models are limited and do not account for many key factors in biomass modelling and climate research (Borer et al. 2017, Lawlor et al. 2024), such as water availability, nutrient depositions or rising CO2 concentrations, they provide rough preliminary approximation of how growth intensity may respond to changing temperature. Ignoring these limitations, our results suggest that species may still be able to partially compensate for the shorter growing season through increased growth intensity during colder periods. Under the SSP2-4.5 scenario, most species were projected to reach slightly lower or similar biomass change levels compared to current conditions. In contrast, under more extreme scenarios (SSP3-7.0 and SSP5-8.5), many species were projected to even exceed current biomass change levels during the winter-spring period. This indicates that warmer conditions during these periods may enhance growth rates, even if the growth activity remains limited. Although warming is often expected to enhance early-season productivity (Parmesan 2006, Rustad et al. 2001), evidence for this process is mixed. Experimental studies have already shown that warming during winter and early spring can increase biomass production in grasslands (Kreyling et al. 2019, Grant et al. 2017). However, such increases may not translate into effective seasonal compensation. Increased early-season productivity may lead to higher evapotranspiration and contribute to earlier reduction of soil moisture availability, potentially exacerbating summer drought stress (Bastos et al. 2020). 72 DISCUSSION Additionally, such biomass responses during colder seasons may be not only constrained by environmental factors, such as temperature or precipitation, but may have physiological limitations. In particular, light availability, shorter photoperiod and reduced radiation can further limit photosynthesis and plant development (Hay 1990, Wingler & Hennessy 2016). Consequently, increases in temperature may not directly translate into increases in biomass accumulation, even under otherwise favourable thermal conditions. For most species in studied grasslands the cycle of biomass productivity follows a two-peak seasonal pattern, with primary peak in spring followed by a decline during summer, and a secondary peak in autumn. The autumn period therefore represents another potential window for seasonal compensation. However, its effectiveness depends not only on environmental conditions but also on the physiological state of plants following summer stress. Experimental studies have shown that extreme temperatures coupled with drought events can significantly reduce aboveground productivity (Hoover et al. 2014), although some grassland systems exhibit high resilience following summer stress (Hossain et al. 2023). For those species that are able to persist or recover after summer stress, our analyses suggest that autumn conditions may remain relatively favourable. However, under SSP2-4.5 most species are projected to show lower biomass change levels compared to current conditions, while under more extreme scenarios the biomass change levels are comparable with current ones. Overall, the contribution of autumn growth to seasonal compensation is therefore likely to depend on the balance between summer losses and autumn gains, as well as on the species' resilience to repeated extreme events (Zhou et al. 2020). 5.4 Differences Between Functional Groups and Vegetation Types Species responses to climate change varied according to their thermal niches. Species adapted to higher temperature ranges were more likely to maintain or even slightly increase their growth activity, whereas species with cooler or broader thermal ranges showed consistent declines across climate scenarios. These could be described as scleromorphic or 73 DISCUSSION succulent species that are already adapted to higher thermal and drought stress (Griffiths & Males 2017). However, when tested explicitly, almost all different leaf anatomies showed significant declines in growth activity, with the exception of succulent or succulent-like plants where the slight decline was not significant. Previous studies (Compagnoni et al. 2021, Morris et al. 2008, Lin et al. 2010) have shown that the response of species to climate change might be different across different plant strategies, however, in our study, all life history strategies showed declines in growth activity throughout the year and across climate scenarios, except for short-lived perennials under SSP3-7.0 and SSP5-8.5, where the decline was not significant. The analysis of biogeographical elements further supports this pattern, as declines in growth activity were observed across all groups. Even species with centres of their distribution ranges in warmer regions, such as (sub-)Mediterranean showed reductions in their growthdays. This contrasts with the expectation that such species would benefit from warming conditions and suggests that increasing stress intensity may exceed the limits of tolerance even for drought-adapted taxa. Similar patterns have been observed in experimental studies, where populations of drought-tolerant species also exhibited decline in biomass under warming (Butof et al. 2012). 74 DISCUSSION Figure 15 Selection of species occurring at our study sites. Species are grouped by life strategies: annuals (top row), perennials (middle row), short-lived perennials (bottom row). Photos by Kryštof Chytrý. 75 DISCUSSION Responses to climate change also differed among vegetation types. Rocky and loess dry grasslands showed similar patterns, with summer no-growth periods under current conditions further prolonged under all warming scenarios. In these grasslands, increasing temperatures extended the duration of summer stress into both earlier spring and late autumn. In contrast, dealpine grasslands showed more growth activity during summer periods under current conditions. The projected increase in temperature dealpine grasslands is less extreme, potentially allowing for positive growth responses of some species even under more severe scenarios (SSP2-4.5, SSP3-7.0). This may reflect the effects of warming during colder seasons and less pronounced amplification of summer stress compared to rocky and loess dry grasslands. However, under the most extreme scenario, this difference was largely reduced, suggesting a convergence in seasonal dynamics across vegetation types. This may indicate that climate change can diminish existing environmental gradients, leading to more homogeneous responses across habitats. Across all vegetation types warming will probably impose increased stress. In rocky and loess dry grasslands, this will most likely occur through intensified and prolonged summer drought period and an increased frequency of extreme events, which may put pressure on many of the current species and may lead to favouring species characterised by fast recovery, drought tolerance and competitive ability, potentially leading to shifts in community composition (Dreesen et al. 2015). The summer stress may also increase gaps in vegetation for potential invasions (Van Peer et al. 2001, Graae et al. 2018). In contrast, the north-facing slopes with dealpine grasslands, as the colder microhabitats, may act as local microrefugia under warming climate (Ashe-Jepson et al. 2025), potentially providing suitable environments for colonization by species from increasingly warmer surrounding landscapes. However, dealpine grasslands support fewer heat-adapted species compared to rocky and loess grasslands, which may make them more vulnerable to warming. Species in these habitats are adapted to colder and more mesic conditions (e.g. Sesleria caerulea, Primula veris, Campanula persicifolia) and may therefore experience increased stress under warming. At the same time, these communities may also face additional pressure from species from adjacent rocky grasslands, seeking suitable microrefugia, as well as from novel species better adapted to changing climate conditions. These processes may increase local 76 DISCUSSION species richness (Steinbauer et al. 2018) but may also lead to local extinctions of habitat specialists (Panetta et al. 2018) and result in species turnover and shifts in community composition, potentially leading to reductions in the distinctiveness of vegetation types (Stewart et al. 2018, Zhu et al. 2026). 77 CONCLUSION 6 Conclusion Our results indicate that climate change in dry grasslands is likely to be driven by the intensification and prolongation of summer drought rather than by an extension of favourable growing conditions into colder seasons. Under increased warming, summer stress reduces growth activity across most dry grassland species and limits their ability to maintain their current growth dynamics. Under future conditions, species are projected to follow a non-linear pattern, where intermediate warming is likely to be insufficient to enable effective seasonal compensation, as temperatures exceed optimal thresholds during summer while remaining too low in colder seasons to support increased growth. Under more extreme warming, this threshold may be partially crossed, allowing for some redistribution of growth into spring and autumn. However, this compensation may be insufficient for many species and may be further constrained by other environmental and physiological limitations. Such shifts imply that species may increasingly operate under altered seasonal regimes, characterized by differences between temperature and water availability. As a result, traits associated with current communities may become less effective under changing climate, while strategies associated with fast recovery and higher tolerance to extremes may become more favourable. Overall, climate change may not only affect species performance but also broader community dynamics, reshaping competitive relationships and driving shifts in community composition, with these changes likely to be complex and very difficult to predict. 78 R E F E R E N C E S References During the preparation of this work, the author used Google Gemini (Gemini 3 Flash; https://gemini.google.com) and OpenAI ChatGPT (GPT-5; https://chatgpt.com) for the following: i) R script development: mainly for writing custom functions, help with statistical analysis, and advanced data wrangling. ii) Manuscript structure and formatting: providing suggestions on the structure and formatting of the manuscript. All ideas, arguments, data interpretations, and conclusions presented in this thesis were developed by the author and/or supervisors mentioned in the Acknowledgements and have not been generated by AI. Ashe-Jepson E., Turner E. C. & Bladon A. J. (2025): Local microclimates can both amplify and mitigate extreme temperatures associated with climate change. - Frontiers in Biogeography 18. Auguie B. (2017): gridExtra: Miscellaneous Functions for "Grid" Graphics. R package version 2.3. URL: . Axmanová I, Chytrý K., Boublík K., Chytrý M., Dřevojan P., Ekrtová E., Fajmon K., Hájková P., Härtel H , Hejda M., Horáková V., Jongepier J. W., Kalníková V., Kaplan Z., Koutecký P., Lustyk P., Pergl J., Prach K., Pyšek P., Sádlo J., Vojík M. & Těšitel J. (2024): Catalogue of expansive plants of the Czech Republic. - Preslia 96: 299-327. Bai Y. & CotrufoM. F. (2022): Grassland soil carbon sequestration: Current understanding, challenges, and solutions. - Science 377: 603-608. Bastos A., Ciais P., Friedlingstein P., Sitch S., Pongratz J., Fan L., Wigneron J.P., Weber U , Reichstein M., Fu Z., Anthoni P., Arneth A., Haverd V., Jain A.K., Joetzjer E., Knauer J., Lienert S., Loughran T., McGuire P.C., Tian H , Viovy N. & Zaehle S. 79 R E F E R E N C E S (2020): Direct and seasonal legacy effects of the 2018 heat wave and drought on European ecosystem productivity. - Science Advances 6: eaba2724. Bastos A., Gouveia C. M., Trigo R. M . & Running S. W. (2014): Analysing the spatiotemporal impacts of the 2003 and 2010 extreme heatwaves on plant productivity in Europe. - Biogeosciences 11: 3421-3435. Bengtsson J., Bullock J. M., Egoh B., Everson C , Everson T., O'connor T., O'Farrell P. J., Smith H. G. & Lindborg R. (2019): Grasslands—more important for ecosystem services than you might think. - Ecosphere 10: e02582. Boomgard-Zagrodnik J. P. & Brown D. J. (2022): Machine learning imputation of missing Mesonet temperature observations. - Computers and Electronics in Agriculture 192: 106580. Borer E. T., Grace J. B., Harpole W. S., MacDougall A. S. & Seabloom E. W. (2017): A decade of insights into grassland ecosystem responses to global environmental change. - Nature Ecology & Evolution 1: 0118. Bradley B.A., Beauty E.M., Gallardo B., Ibáňez L, Jarnevich C , Morelli T.L., Sofaer H.R., Sorte C.J.B. & Vila M . (2024): Observed and Potential Range Shifts of Native and Nonnative Species with Climate Change. - Annual Review of Ecology, Evolution, andSystematics 55: 23-40. Bramer I., Anderson B. J., Bennie J., Bladon A. J., De Frenne P., Hemming D., Hill R. A., Kearney M. R., Körner C , Korstjens A. H , Lenoir J., Maclean I. M. D., Marsh C. D., Morecroft M . D., Ohlemüller R., Slater H. D., Suggitt A. J., Zellweger F. & Gillingham P. K. (2018): Advances in monitoring and modelling climate at ecologically relevant scales. - In: Bohan D. A., Dumbrell A. J., Woodward G. & Jackson M. (eds): Advances in ecological research, pp. 101-161. - Academic Press. Britten A. J., Beale C. M., Towers W. & Hewison R. L. (2009): Biodiversity gains and losses: evidence for homogenisation of Scottish alpine vegetation. -Biological conservation 142: 1728-1739. Bütof A., von Riedmatten L. R., Dormann C. F., Scherer-Lorenzen M . , Welk E. & Bruelheide H. (2012): The responses of grassland plants to experimentally simulated 80 R E F E R E N C E S climate change depend on land use and region. - Global Change Biology 18: 127- 137. Česká geologická služba (2026): Geovědní mapy 1 50 000. URL: https://mapy.geology.cz/geocr50/ (accessed 16/03/2026) Chang J., Ciais P., Viovy N . , Soussana J. F., Klumpp K. & Sultan B. (2017): Future productivity and phenology changes in European grasslands for different warming levels: implications for grassland management and carbon balance. - Carbon balance and management 12: 11. Chen I. C , Hill J. K., Ohlemüller R., Roy D. B. & Thomas C. D. (2011): Rapid range shifts of species associated with high levels of climate warming. -Science 333: 1024- 1026. ČHMÚ (2026): Historická data. URL: https://opendata.chmi.cz/meteorology/climate/historical_csv/ (accessed 25/03/2026). Chytrý K., ProkešováH, DuchoňM., Grulich V., Chytrý M. & Divíšek J. (2022b): Substrate associated biogeographical patterns in the north-western Pannonian forest-steppe. -Preslia 94: 215-232. Chytrý K., Willner W., Chytrý M., Divíšek J. & Dullinger S. (2022a): Central European forest-steppe: An ecosystem shaped by climate, topography and disturbances. -Journal of Biogeography 49: 1006-1020. Chytrý M . (ed.) (2007): Vegetace České republiky. 1. Travinná a keříčková vegetace [Vegetation of the Czech Republic. 1. Grassland and Heathland Vegetation]. Academia, Praha. Chytrý M., Hájek M., Kočí M., Pešout P., Roleček J., Sádlo J., Šumberová K , Sychra J., Boublík K , Douda J., Grulich V., Härtel H , Hédl R., Lustyk P., Navrátilová J., Novák P., Peterka T., Vydrová A & Chobot K. (2019): Red list of habitats of the Czech Republic. - Ecological Indicators 106: 105446. Compagnoni A., Levin S., Childs D.Z., Harpole S., Paniw M., Römer G., Burns J.H., CheCastaldo J., Rüger N , Kunstler G., Bennett J.M., Archer C.R., Jones O.R., SalgueroGómez R. & Knight T.M. (2021): Herbaceous perennial plants with short generation 81 R E F E R E N C E S time have stronger responses to climate anomalies than those with longer generation time. - Nature Communications 12: 1824. Cook B. I., Wolkovich E. M . & Parmesan C. (2012): Divergent responses to spring and winter warming drive community level flowering trends. -Proceedings of the National Academy of Sciences 109: 9000-9005. Cornell S. E., Prentice I. C , House J. I. & Downy C. J. (eds) (2012): Understanding the earth system: global change science for application. - Cambridge University Press, Cambridge. Crafts-Brandner S. J. & Salvucci M . E. (2002): Sensitivity of photosynthesis in a C4 plant, maize, to heat stress. - Plant physiology 129: 1773-1780. De Boeck H.J., Lemmens C.M.H.M., Gielen B., Bossuyt H , Malchair S., Carnol M . , Merckx R., Ceulemans R. & Nijs I. (2007): Combined effects of climate warming and plant diversity loss on above- and below-ground grassland productivity. Environmental and Experimental Botany 60: 95-104. De Frenne P., Lenoir J., Luoto M., Scheffers B.R., Zellweger F., Aalto J., Ashcroft M.B., Christiansen D.M., Decocq G., Pauw K.D., Govaert S., Greiser C , Gril E., Hampe A., Jucker T., Klinges D.H., Koelemeijer LA., Lembrechts J.J., Marrec R., Meeussen C , Ogee J., Tyystjarvi V., Vangansbeke P., Hylander K. (2021): Forest microclimates and climate change: Importance, drivers and future research agenda. - Global Change Biology 27: 2279-2297. Dengler J., Janisova M., Torok P. & Wellstein C. (2014): Biodiversity of Palaearctic grasslands: a synthesis. - Agriculture, Ecosystems & Environment 182: 1-14. Diaz S., Settele J., Brondizio E. S., Ngo H. T., Agard J., Arneth A., Balvanera P., Brauman K. A., Butchart S. H. M., Chan K. M . A, Garibaldi L. A., Ichii K., Liu J., Subramanian S. M., Midgley G. F., Miloslavich P., Molnar Z., Obura D., Pfaff A., Polasky S., Purvis A., Razzaque J., Reyers B., Chowdhury R. R., Shin Y., VisserenHamakers I., Willis K. J. & Zayas, C. N. (2019): Pervasive human-driven decline of life on Earth points to the need for transformative change. - Science 366: eaax3100. 82 R E F E R E N C E S Dreesen F. E., De Boeck H. J., Horemans J. A., Janssens L A. & Nijs L (2015): Recovery dynamics and invasibility of herbaceous plant communities after exposure to experimental climate extremes. -Basic and Applied Ecology 16: 583-591. Dunnington D. (2025): ggspatial: Spatial Data Framework for ggplot2. R package version 1.1.10. URL: https://CRAN.R-project.org/package=ggspatial. Engler R., Randin C. F., Vittoz P., Czáka T., Beniston M., Zimmermann N. E. & Guisan A. (2009): Predicting future distributions of mountain plants under climate change: does dispersal capacity matter? -Ecography 32: 34-45. Erdós L., Ambarh D., Anenkhonov O. A., Bátori Z., Cserhalmi D., Kiss M., Króel-Dulay G., Liu H., Magnes M . , Molnár Z., Naqinezhad A., Semenishchenkov Y. A., Tólgyesi C. & Tórók P. (2018): The edge of two worlds: A new review and synthesis on Eurasian forest-steppes. - Applied Vegetation Science 21: 345-362. Espinosa del Alba C , Fernández-Pascual E. & Jiménez-Alfaro B. (2025): Microclimatic variation regulates seed germination phenology in alpine plant communities. -Journal of Ecology 113: 249-262. Eurostat (2011): Europe in figures - Eurostat yearbook 2011: Agriculture, forestry and fisheries. - Office for Official Publications of the European Communities. Eyring V., Bony S., Meehl G. A., Senior C. A., Stevens B., Stouffer R. J. & Taylor K. E. (2016): Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. - Geoscientifw Model Development 9: 1937— 1958. Fekete G., Király G. & Molnár Z. (2016): Delineation of the Pannonian vegetation region. - Community Ecology 17: 114-124. Fischer F. M., Chytrý K , Chytrá H., Chytrý M. & Těšitel J. (2023): Seasonal beta-diversity of dry grassland vegetation: Divergent peaks of above-ground biomass and species richness. - Journal of Vegetation Science 34: el3182. Fischer F. M., Chytrý K , Těšitel J., Danihelka J. & Chytrý M. (2020): Weather fluctuations drive short-term dynamics and long-term stability in plant communities: a 25-year study in a Central European dry grassland. - Journal of Vegetation Science 31: 711- 721. 83 R E F E R E N C E S Fitchett J. M., Grab S. W. & Thompson D. L (2015): Plant phenology and climate change: Progress in methodological approaches and application. -Progress in Physical Geography 39: 460-482. Frank D., Reichstein M., Bahn M., Thonicke K., Frank D., Mahecha M.D., Smith P., van der Velde M., Vicca S., Babst F., Beer C , Buchmann N., Canadell J.G., Ciais P., Cramer W., Ibrom A., Miglietta F., Poulter B., Rammig A., Seneviratne S.I., Walz A., Wattenbach M . , Zavala M A . & Zscheischler J. (2015): Effects of climate extremes on the terrestrial carbon cycle: concepts, processes and potential future impacts. - Global Change Biology 21: 2861-2880. Gottfried M., Pauli H., Futschik A., Akhalkatsi M., BarancokP., Benito Alonso J.L., Coldea G., Dick J., Erschbamer B., Fernandez Calzado M.R., Kazakis G., Krajci J., Larsson P., Mallaun M., Michelsen O., Moiseev D., Moiseev P., Molau U., Merzouki A., Nagy L., Nakhutsrishvili G., Pedersen B., Pelino G., Puscas M., Rossi G., Stanisci A., Theurillat J.-P., Tomaselli M., Villar L., Vittoz P., Vogiatzakis I. & Grabherr G. (2012): Continent-wide response of mountain vegetation to climate change. -Nature Climate Change 2: 111-115. Graae B.J., Vandvik V., Armbruster W.S., Eiserhardt W.L., Svenning J.-C, Hylander K., Ehrlen J., Speed J.D.M., KlanderudK., BräthenK.A., Milbau A., Opedal 0.H., Alsos LG., Ejrnses R., Bruun H.H., Birks H.J.B., Westergaard K.B., Birks H.H. & Lenoir J. (2018): Stay or go - how topographic complexity influences alpine plant population and community responses to climate change. - Perspectives in Plant Ecology, Evolution andSystematics 30: 41-50. Grabherr G., Gottfried M . & Pauli H. (1994): Climate effects on mountain plants. - Nature 369: 448. Grant K , Kreyling J., Beierkuhnlein C. & Jentsch A. (2017): Importance of seasonality for the response of a mesic temperate grassland to increased precipitation variability and warming. - Ecosystems 20: 1454-1467. Griffin-Nolan R.J., Blumenthal D.M., Collins S.L., Farkas T.E., Hoffman A.M., Mueller K.E., Ocheltree T.W., Smith M.D., Whitney K D . & Knapp A.K. (2019): Shifts in 84 R E F E R E N C E S plant functional composition following long-term drought in grasslands. - Journal of Ecology 107: 2133-2148. Griffiths H. & Males J. (2017): Succulent plants. - Current Biology 27: R890-R896. Grulich V. (2017): Červený seznam cévnatých rostlin ČR [The Red List of vascular plants of the Czech Republic]. - Příroda 35: 75-132. Haesen S., Lenoir J., Gril E., De Frenne P., Lembrechts J. J., Kopecký M., Macek M., Man M., Wild J. & Van Meerbeek K. (2023): Microclimate reveals the true thermal niche of forest plant species. - Ecology Letters 26: 2043-2055. Harásek M., Klinkovská K. & Chytrý M . (2023): Vegetation change in acidic dry grasslands in Moravia (Czech Republic) over three decades: Slow decrease in habitat quality after grazing cessation. - Applied Vegetation Science 26: el2726. Hausfather Z. & Peters G. P. (2020): Emissions-the 'business as usual' story is misleading. -Nature 577: 618-620. Hay R. K. M . (1990): The influence of photoperiod on the dry matter production of grasses and cereals. -New Phytologist 116: 233-254. Hejcman M., Hejcmanova P., Pavlů V. & Beneš J. (2013): Origin and history of grasslands in Central Europe-a review. - Grass and Forage Science 68: 345-363. Hester J. & Bryan J. (2024): glue: Interpreted String Literals. R package version 1.8.0. URL: . Hilpold A., Seeber J., Fontána V., Niedrist G., Rief A., Steinwandter M., Tasser E. & Tappeiner U. (2018): Decline of rare and specialist species across multiple taxonomie groups after grassland intensification and abandonment. - Biodiversity and Conservation 27: 3729-3744. Hoover D. L., Knapp A. K. & Smith M . D. (2014): Resistance and resilience of a grassland ecosystem to climate extremes. - Ecology 95: 2646-2656. Hossain M. L., Li J., Lai Y. & Beierkuhnlein C. (2023): Long-term evidence of differential resistance and resilience of grassland ecosystems to extreme climate events. - Environmental Monitoring and Assessment 195: 734. 85 R E F E R E N C E S Huang Y., Lei H. & Duan L. (2024): Resistance of grassland productivity to drought and heatwave over a temperate semi-arid climate zone. -Science of the Total Environment 951: 175495. IPCC (2021): Climate Change 2021: The Physical Science Basis. - Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. - Cambridge University Press, Cambridge and New York. Janišová M., Iuga A., Ivascu C. M . & Magnes M . (2021): Grassland with tradition: sampling across several scientific disciplines. - Vegetation Classification and Survey!: 19- 35. Jiménez-Alfaro B., Fernández-Pascual E., Espinosa Del Alba C. & Marcenó C. (2024): Spatiotemporal patterns of microclimatic buffering in relict alpine communities. -Journal of Vegetation Science 35: el 3242. Kaisermann A., de Vries F. T., Griffiths R. I. & Bardgett R. D. (2017): Legacy effects of drought on plant-soil feedbacks and plant-plant interactions. - New Phytologist 215: 1413-1424. Kankaanpáá T., Skov K., Abrego N , Lund M . , Schmidt N . M . & Roslin T. (2018): Spatiotemporal snowmelt patterns within a high Arctic landscape, with implications for flora and fauna. - Arctic, Antarctic, and Alpine Research 50: el415624. Kapitány E., Wrbka T. & Dullinger S. (2025): Local-scale increase masks landscape-scale loss of species richness in managed Pannonian grasslands. - Landscape Ecology. Kaplan Z., Danihelka J., Chrtek J., Kirschner J., Kubát K , Stech M. & Štěpánek J. (eds) (2019): Klíč ke květeně České republiky [Key to the flora of the Czech Republic], Ed. 2. - Academia, Praha. Kassambara A. & Mundt F. (2020): factoextra: Extract and Visualize the Results of Multivariate Data Analyses. R package version 1.0.7. URL: . Kemppinen J., Lembrechts J.J., Van Meerbeek K , Carnicer J., Chardon N.I., Kardol P., Lenoir J., Liu D., Maclean I., Pergl J., Saccone P., Senior R.A., Shen T., Slowiňska S., Vandvik V., von Oppen J., Aalto J., Ayalew B., Bates O., Bertelsmeier C , Bertrand R., Beugnon R., Borderieux J., Bruna J., Buckley L., Bujan J., Casanova- 86 R E F E R E N C E S Katny A., Christiansen D.M., CollartF., DeLombaerdeE., DePauwK., DepauwL., Di Musciano M., Diaz Borrego R., Díaz-Calafat J., Ellis-Soto D., Esteban R., de Jong G.F., Gallois E., GarciaM.B., GillerotL., Greiser C , Gril E., Haesen S., Hampe A. , Hedwall P.-O., Hes G., Hespanhol H., Hoffrén R., Hylander K., Jiménez-Alfaro B., Jucker T., Klinges D., Kolstela J., Kopecký M., Kovács B., Maeda E.E., Máliš F., Man M., Mathiak C , Meineri E., Naujokaitis-Lewis I., Nijs I., Normand S., Nuňez M., Orczewska A., Pena-Aguilera P., Pincebourde S., Plichta R., Quick S., Renault D., Ricci L., Rissanen T., Segura-Hernández L., Selvi F., Serra-Diaz J.M., Soifer L., Spicher F., Svenning J.-C, Tamian A., Thomaes A., Thoonen M., Trew B., Van de Vondel S., van den Brink L., Vangansbeke P., Verdonck S., Vitkova M., Vives-Ingla M., von Schmalensee L., Wang R., Wild J., Williamson J., Zellweger F., Zhou X., Zuza E.J.& De Frenne P. (2024): Microclimate, an important part of ecology and biogeography. - Global Ecology and Biogeography 33: el3834. Kennedy T. A., Naeem S., Howe K. M . , Knops J. M . , Tilman D. & Reich P. (2002): Biodiversity as a barrier to ecological invasion. - Nature All: 636-638. Klinges D.H., Baecher J.A., Lembrechts J.J., Maclean I.M.D., Lenoir J., Greiser C , Ashcroft M., Evans L.J., Kearney M.R., Aalto J., Barrio I.C., De Frenne P., Guillemot J., Hylander K , Jucker T., Kopecký M., Luoto M., Macek M., Nijs I., Urban J., van den Brink L., Vangansbeke P., Von Oppen J., Wild J., Boike J., Canessa R., Nosetto M., Rubtsov A., Sallo-Bravo J. & Scheffers B.R. (2024): Proximal microclimate: Moving beyond spatiotemporal resolution improves ecological predictions. - Global Ecology and Biogeography 33: el3884. Klinkovská K , GlaserM., Danihelka J., Kaplan Z., Knollová I., Novotný P., Pyšek P., Řezníčková M.,Wild J. & Chytrý M . (2024b): Dynamics of the Czech flora over the last 60 years: winners, losers and causes of changes. - Biological Conservation 292: 110502. Klinkovská K , Sperandii M . G., Trávníček B. & Chytrý M . (2024a): Significant decline in habitat specialists in semi-dry grasslands over four decades. -Biodiversity and Conservation 33: 161-178. 87 R E F E R E N C E S Kreyling J., Grant K., Hammerl V., Arfin-Khan M.A.S., Malyshev A.V., Pefiuelas J., Pritsch K., Sardans J., Schloter M., Schuerings J., Jentsch A. & Beierkuhnlein C (2019): Winter warming is ecologically more relevant than summer warming in a cooltemperate grassland. - Scientific reports 9: 14632. Kübert A., Götz M., Kuester E., Piayda A., Werner C , Rothfuss Y. & Dubbert M . (2019): Nitrogen loading enhances stress impact of drought on a semi-natural temperate grassland. - Frontiers in Plant Science 10: 1051. Kuznetsova A., Brockhoff P. B., Christensen R. H. B. (2017): lmerTest Package: Tests in Linear Mixed Effects Models. - Journal of Statistical Software 82: 1-26. Lalic B., Stapleton A., Vergauwen T., Caluwaerts S., Eichelmann E. & Roantree M. (2024): A comparative analysis of machine learning approaches to gap filling meteorological datasets. - Environmental Earth Sciences 83: 679. Lawlor J.A., Comte L., Grenouillet G., Lenoir J., Baecher JA., Bandara R.M.W.J., Bertrand R., Chen I.-C, Diamond S.E., Lancaster L.T., Moore N., Murienne J., OliveiraB.F., Peel G.T., Pinsky M.L., Rolland J., Rubenstein M., Scheffers B.R., Thompson L.M., van Amerom B., Villalobos F., Weiskopf S.R. & Sunday J. (2024): Mechanisms, detection and impacts of species redistributions under climate change. - Nature Reviews Earth & Environment 5: 351-368. Lenoir J., Gegout J.-C, Guisan A., Vittoz P., Wohlgemuth T., Zimmermann N E . , Dullinger S., Pauli H , Willner W. & Svenning J.-C. (2010): Going against the flow: potential mechanisms for unexpected downslope range shifts in a warming climate. Ecography 33: 295-303. Lenoir J., Hattab T. & Pierre G. (2017): Climatic microrefugia under anthropogenic climate change: implications for species redistribution. - Ecography 40: 253-266. Lenth R. & Piaskowski J. (2025): emmeans: Estimated Marginal Means, aka Least-Squares Means. R package version 2.0.0. URL: . Lin D., Xia J. & Wan S. (2010): Climate warming and biomass accumulation of terrestrial plants: a meta-analysis. - New Phytologist 188: 187-198. 88 R E F E R E N C E S Lüdecke D., Ben-ShacharM. S., Patil I, Waggoner P. & Makowski D. (2021): performance: An R package for assessment, comparison and testing of statistical models. - Journal of Open Source Software 6. Ma H., Mo L., Crowther T.W., Maynard D.S., van den Hoogen J., Stocker B.D., Terrer C. & Zohner C M . (2021): The global distribution and environmental drivers of aboveground versus belowground plant biomass. - Nature Ecology & Evolution 5: 1110-1122. Mackovčin P., Jatiová M., Demek J. & Slavík P. (eds.) (2007): Brněnsko. In: Mackovčin, P. (ed.): Chráněná Území CR 9. - Agentura ochrany přírody a krajiny ČR & Ekocentrum Brno, Praha. Maechler M., Rousseeuw P., Struyf A., Hubert M . & Hornik K. (2025): cluster: Cluster Analysis Basics and Extensions. R package version 2.1.8.1. URL: https://CRAN.Rproj ect.org/package=cluster. Massicotte P. & South A. (2026): rnaturalearth: World Map Data from Natural Earth. R package version 1.2.0. URL: . Meinshausen M., Nicholls Z. R. J., Lewis J., Gidden M . J., Vogel E., Freund M., Beyerle U , Gessner C , Nauels A., Bauer N , Canadell J. G., Daniel J. S., John A., Krümmel P. B., Luderer G., Meinshausen N , Montzka S.A., Rayner P. J., Reimann S., Smith S. J., van den Berg M., Velders G. J. M., Vollmer M . K. Wang R. H. J. (2020): The shared socio-economic pathway (SSP) greenhouse gas concentrations and their extensions to 2500. - Geoscientific Model Development 13: 3571-3605. Millenium Ecosystem Assessment (2005): Ecosystems and Human Well-being: Synthesis. - Island Press, Washington, DC. Moritz C. & Agudo R. (2013): The future of species under climate change: resilience or decline? - Science 341: 504-508. Morris W.F., Pfister C A . , Tuljapurkar S., Haridas C.V., Boggs C.L., Boyce M.S., Bruna E.M., Church DR., Coulson T., Doak D.F., Forsyth S., Gaillard J.-M., Horvitz C.C., Kalisz S., Kendall B E . , Knight T.M., Lee C T . & Menges E.S. (2008): Longevity can buffer plant and animal populations against changing climatic variability. Ecology 89: 19-25. 89 R E F E R E N C E S Moss R. H., Edmonds J. A., Hibbard K. A., Manning M . R., Rose S. K., Van Vuuren D. P., Carter T. R., Emori S., Kainuma M., Kram T., Meehl G. A., Mitchell J. F. B., Nakicenovic N., Riahi K., Smith S. J., Stouffer R. J., Thomson A. M., Weyant J. P. & Wilbanks T. J. (2010): The next generation of scenarios for climate change research and assessment. - Nature 463: 747-756. Muller W. A., Jungclaus J. H., Mauritsen T., Baehr J., Bittner M., Budich R., Bunzel F., Esch M., Ghosh R., Haak H., Ilyina T., Kleine T., Kornblueh L., Li H., Modali K., Notz D., Pohlmann H., Roeckner E., Stemmler I., Tian F. & Marotzke J. (2018). A higher-resolution version of the max planck institute earth system model (MPIESM1. 2-rTR). - Journal of Advances in Modeling Earth Systems 10: 1383-1413. National Research Council (2005): Radiative forcing of climate change: Expanding the concept and addressing uncertainties. - National Academies Press, Washington, D C Olsen S. L., Topper J. P., Skarpaas O., Vandvik V. & Klanderud K. (2016): From facilitation to competition: Temperature-driven shift in dominant plant interactions affects population dynamics in seminatural grasslands. - Global Change Biology 22: 1915— 1926. O'Neill B. C , Tebaldi C , Van Vuuren D. P., Eyring V., Friedlingstein P., Hurtt G , Knutti R., Kriegler E., Lamarque J.-F., Lowe J., Meehl G. A., Moss R., Riahi K. & Sanderson, B. M . (2016): The scenario model intercomparison project (ScenarioMIP) for CMIP6. - Geoscientific Model Development 9: 3461-3482. Panetta A.M., Stanton M L . & Harte J. (2018): Climate warming drives local extinction: Evidence from observation and experimentation. - Science Advances 4: eaaql819. Parmesan C. & Yohe G. (2003): A globally coherent fingerprint of climate change impacts across natural systems. - Nature 421: 37-42. Parmesan C. (2006): Ecological and evolutionary responses to recent climate change. - Annual Review of Ecology, Evolution, and Systematics 37: 637-669. Partel M . , Bruun H. H. & Sammul M . (2005): Biodiversity in temperate European grasslands: origin and conservation. - Grassland Science in Europe 10: 1-14. 90 R E F E R E N C E S Pauchard A., Milbau A., Albihn A., Alexander J., Burgess T., Daehler C , Englund G., Essl R, Evengárd B., Greenwood G.B., Haider S., Lenoir J., McDougall K., Muths E., Nunez M A . , Olofsson J., Pellissier L., Rabitsch W., Rew L.J., Robertson M . , Sanders N. & Kueffer C. (2016): Non-native and native organisms moving into high elevation and high latitude ecosystems in an era of climate change: new challenges for ecology and conservation. - Biological invasions 18: 345-353. Pauli H , Gottfried M . & Grabherr G. (2003): Effects of climate change on the alpine and nival vegetation of the Alps. - Journal of mountain ecology 7: 9-12. Pauli H , Gottfried M., Dullinger S., Abdaladze O., Akhalkatsi M., Alonso J.L.B., Coldea G., Dick J., Erschbamer B., Calzado R.F., Ghosn D., Holten J.I, Kanka R., Kazakis G. , Kollár J., Larsson P., Moiseev P., Moiseev D., Molau U., Mesa J.M., Nagy L., Pelino G., Pusxas. M . , Rossi G., Stanisci A., Syverhuset A.O., Theurillat J.-P., Tomaselli M., Unterluggauer P., Villar L., Vittoz P. & Grabherr G. (2012): Recent Plant Diversity Changes on Europe's Mountain Summits. - Science 336: 353-355. Peacock J. M . (1976): Temperature and leaf growth in four grass species. - Journal of Applied Ecology: 225-232. Pebesma E. (2018): Simple Features for R: Standardized Support for Spatial Vector Data. The R Journal 10: 439-446. Pedersen T. (2025): patchwork: The Composer of Plots. R package version 1.3.2. URL: . Piao S., Wang X., Park T., Chen C , Lian X., He Y., Bjerke J.W., Chen A., Ciais P., Tommervik H , Nemani R.R. & Myneni R.B. (2020): Characteristics, drivers and feedbacks of global greening. - Nature Reviews Earth & Environment 1: 14-27. Pladias (2026): Pladias - Databáze české flóry a vegetace. URL: https://pla-dias.cz/. Pórtner H. O., Scholes R. J., Arneth A., Barnes D. K. A., Burrows M . T., Diamond S. E., Duarte C. M., Kiessling W., Leadley P., Managi S., McElwee P., Midgley G., Ngo H. T., Obura D., Pascual U , Sankaran M . , ShinY. J. & Val A. L. (2023): Overcoming the coupled climate and biodiversity crises and their societal impacts. -Science 380: eabl4881. 91 R E F E R E N C E S QGIS Development Team (2025): QGIS Geographic Information System. QGIS Association. URL: http://www.qgis.org. Reich P. B., Hobbie S. E., Lee T. D., Rich R., Pastore M . A. & Worm K. (2020): Synergistic effects of four climate change drivers on terrestrial carbon cycling. -Nature Geoscience 13: 787-793. Reichstein M., Bahn M., Ciais P., Frank D., Mahecha M.D., Seneviratne S.I., Zscheischler J., Beer C , Buchmann N., Frank D.C., Papale D., Rammig A., Smith P., Thonicke K., van der Velde M., Vicca S., Walz A. & Wattenbach M . (2013): Climate extremes and the carbon cycle. -Nature 500: 287-295. Riahi K., van Vuuren D.P., Kriegler E., Edmonds J., O'Neill B.C., Fujimori S., Bauer N., Calvin K , Dellink R., Fricko O., Lutz W., Popp A., Cuaresma J.C., Kc S., Leimbach M., Jiang L., Kram T., Rao S., Emmerling J., Ebi K , Hasegawa T., Havlik P., Humpenöder F., Da Silva L.A., Smith S., Stehfest E., Bosetti V., Eom J., Gernaat D., Masui T., Rogelj J., Strefler J., Drouet L., Krey V., Luderer G., Harmsen M., Takahashi K , Baumstark L., Doelman J.C., Kainuma M., Klimont Z., Marangoni G., Lotze-Campen H., Obersteiner M., Tabeau A. & Tavoni M . (2017): The Shared Socioeconomic Pathways and their energy, land use, and greenhouse gas emissions implications: An overview. - Global Environmental Change 42: 153-168. Robin X., TurckN., Hainard A., Tiberti N., Lisacek F., Sanchez J. C. & Müller M.(2011): pROC: an open-source package for R and S+ to analyze and compare ROC curves. - BMC Bioinformatics 12: 77. Robinson D., Hayes A. & Couch S. (2025): broom: Convert Statistical Objects into Tidy Tibbies. R package version 1.0.10. URL: . Ruppert J. C , Harmoney K , Henkin Z., Snyman H. A., Sternberg M., Willms W. & Linstädter A. (2015): Quantifying drylands' drought resistance and recovery: the importance of drought intensity, dominant life history and grazing regime. - Global Change Biology 21: 1258-1270. Rustad L., Campbell J., Marion G., Norby R., Mitchell M., Hartley A., Cornelissen J., Gurevitch J. & GCTE-NEWS (2001) A meta-analysis of the response of soil 92 R E F E R E N C E S respiration, net nitrogen mineralization, and aboveground plant growth to experimental ecosystem warming. - Oecologia 126: 543-562. Scherrer D. & Körner C. (2011): Topographically controlled thermal-habitat differentiation buffers alpine plant diversity against climate warming. -Journal of biogeography 3 8: 406-416. Shine K. P., DerwentR. G., Wuebbles D. J., Morcrette J. J. & Apling A. J. (1990): Radiative forcing of climate. - In: Climate change: The IPCC scientific assessment, pp. 41- 68. - Cambridge University Press, New York. South A., Michael S. & Massicotte P. (2024): rnaturalearthdata: World Vector Map Data from Natural Earth Used in 'rnaturalearth'. R package version 1.0.0. . Stampfli A. & Zeiter M. (2004): Plant regeneration directs changes in grassland composition after extreme drought: a 13-year study in southern Switzerland. - Journal of Ecology: 568-576. Staude I. R., Segar J., Temperton V. M., Andrade B. O., de Sä Dechoum M., Weidlich E. W. & Overbeck G. E. (2023): Prioritize grassland restoration to bend the curve of biodiversity loss. - Restoration Ecology 31: el3931. Steinbauer M.J., Grytnes J.-A., Jurasinski G., Kulonen A., Lenoir J., Pauli H., Rixen C , Winkler M., Bardy-Durchhalter M., Barni E., Bjorkman A.D., Breiner F.T., Burg S., Czortek P., Dawes M.A., Delimat A., Dullinger S., Erschbamer B., Felde V.A., Fernändez-Arberas O., Fossheim K.F., Gömez-Garcia D., Georges D., Grindrud E.T., Haider S., Haugum S.V., Henriksen H., Herreros M.J., Jaroszewicz B., Jaroszynska F., Kanka R., Kapfer J., Klanderud K , Kühn I, Lamprecht A., Matteodo M., di Cella U.M., Normand S., Ödland A., Olsen S.L., Palacio S., Petey M., Piscovä V., SedlakovaB., Steinbauer K , Stöckli V., Svenning J.-C, Teppa G., Theurillat J.P., Vittoz P., Woodin S.J., Zimmermann N.E. & Wipf S. (2018): Accelerated increase in plant species richness on mountain summits is linked to warming. Nature 556: 231-234. 93 R E F E R E N C E S Stewart L., Simonsen C. E., Svenning J. C , Schmidt N . M . & Pellissier L. (2018): Forecasted homogenization of high Arctic vegetation communities under climate change. - Journalof Biogeography 45: 2576-2587. Suttie J. M., Reynolds S. G. & Batello C. (eds) (2005): Grasslands of the World. Plant Production and Protection Series, 34. - FAO, Rome. Tang F. & Ishwaran H. (2017): Random forest missing data algorithms. -Statistical Analysis and Data Mining: The ASA Data Science Journal 10: 363-377. Thrasher B., Wang W., Michaelis A. & Nemani R. (2021): NEX-GDDP-CMIP6. NASA Center for Climate Simulation. Thrasher B., Wang W., Michaelis A., Melton F., Lee T. & Nemani R. (2022): NASA global daily downscaled projections, CMIP6. - Scientific data 9: 262. Thuiller W., Lavorel S., Araújo M. B., Sykes M. T. & Prentice I. C. (2005): Climate change threats to plant diversity in Europe. -Proceedings of the National Academy of Sciences 102: 8245-8250. Tórók P., Janišová M., Kuzemko A., Rusina S. & Stevanovič Z. D. (2018): Grasslands, their threats and management in Eastern Europe. - In: Squires V.R., Dengler J., Hua L. & Feng H. (eds): Grasslands of the World: Diversity, Management and Conservation, pp. 64-88. - CRC Press, New York. Valkó O., Venn S., Žmihorski M., Biurrun I., Labadessa R. & Loos J. (2018): The challenge of abandonment for the sustainable management of Palaearctic natural and seminatural grasslands. - Hacquetia 17: 5-16. Van Peer L., Nijs I., Bogaert J., Verelst I. & Reheul D. (2001): Survival, gap formation, and recovery dynamics in grassland ecosystems exposed to heat extremes: the role of species richness. - Ecosystems 4: 797-806. Večeřa M., Axmanová I, Padullés Cubino J., Lososová Z., Divíšek J., Knollová I, Ačič S., Biurrun I, Boch S., Bonari G., Campos J. A., Carni A., Carranza M . L., Casella L., Chiarucci A., Cušterevska R., Delbosc P., Dengler J., Fernández-González F., Gégout J.-C, Jandt U., Jansen F., Jašková A., Jiménez-Alfaro B., Kuzemko A., Lebedeva M., Lenoir J., Lysenko T., Moeslund J. E., Pielech R., Ruprecht E., Šibík J., Sile U., Škvore Z., Swacha G., Tatarenko L, Vassilev K , Wohlgemuth T., 94 R E F E R E N C E S Yamalov S. & Chytrý M . (2021): Mapping species richness of plant families in European vegetation. - Journal of Vegetation Science 32: el3035. Vitasse Y., Baumgarten F., Zohner C. M., Kaewthongrach R., Fu Y. H., Walde M. G. & Moser B. (2021): Impact of microclimatic conditions and resource availability on spring and autumn phenology of temperate tree seedlings. -New Phytologist 232: 537-550. Ward S. E., Schulze M . & Roy B. (2018): A long-term perspective on microclimate and spring plant phenology in the Western Cascades. - Ecosphere 9: e02451. Wesche K., Ambarh D., Kamp J., Török P., Treiber J. & Dengler J. (2016): The Palaearctic steppe biome: a new synthesis. - Biodiversity and conservation 25: 2197-2231. Wesche K., Krause B., Culmsee H. & Leuschner C. (2012): Fifty years of change in Central European grassland vegetation: Large losses in species richness and animalpollinated plants. - Biological Conservation 150: 76-85. Westhoff V. & van der Maarel E. (1978): The Braun-Blanquet approach. - In: Whittaker R.H. (ed.): Classification ofplant communities, pp. 287-399. White R. P., Murray S. & Rohweder M . (2000): Pilot Analysis of Global Ecosystems: Grassland Ecosystems. - World Resources Institute. Wickham H. & Bryan J. (2025): readxl: Read Excel Files. R package version 1.4.5. URL: https ://CRAN.R-proj ect. org/package=readxl. Wickham H., Averick M., Bryan J., Chang W., McGowan L., Francois R., Grolemund G., Hayes A., Henry L., Hester J., Kuhn M., Pedersen T., Miller E., Bache S., Müller K , Ooms J., Robinson D., Seidel D., Spinu V., Takahashi K , Vaughan D., Wilke C , Woo K. & Yutani H. (2019): Welcome to the tidyverse. - Journal of Open Source Software 4: 1686. Wild L, Kopecký M., Macek M., Šanda M., Jankovec J. & Haase T. (2019): Climate at ecologically relevant scales: A new temperature and soil moisture logger for longterm microclimate measurement. - Agricultural and Forest Meteorology 268: 40- 47. Wilson J. B., Peet R. K , Dengler J. & Partei M. (2012): Plant species richness: the world records. - Journal of vegetation Science 23: 796-802. 95 R E F E R E N C E S Wingler A. & Hennessy D. (2016): Limitation of grassland productivity by low temperature and seasonality of growth. - Frontiers in plant science 7: 1130. Wolkovich E.M., Cook B.I., Allen J.M., Crimmins T.M., Betancourt J.L., Travers S.E., Pau S., Regetz J., Davies T.J., Kraft N.J.B., Ault T.R., Bolmgren K , Mazer S.J., McCabe G.J., McGill B.J., Parmesan C , SalaminN., SchwartzM.D. & ClelandEE. (2012): Warming experiments underpredict plant phenological responses to climate change. -Nature 485: 494-497. Wright M. N . & Ziegler A. (2017): ranger: A Fast Implementation of Random Forests for High Dimensional Data in C++ and R. - Journal of Statistical Software 11: 1-17. Wu Q., Ren H , Wang Z., Li Z., Liu Y., Wang Z., Li Y., Zhang R , Zhao M., Chang S.X. & Han G. (2020): Additive negative effects of decadal warming and nitrogen addition on grassland community stability. - Journal of Ecology 108: 1442-1452. Xu C , McDowell N.G., Fisher R A . , Wei L., Sevanto S., Christoffersen B.O., Weng E. & Middleton R.S. (2019): Increasing impacts of extreme droughts on vegetation productivity under climate change. - Nature Climate Change 9: 948-953. Yue K., Vangansbeke P., Myers-Smith I.H., Waller D.M., Verheyen K., BernhardtRömermann M., Baeten L., Staude I.R., Bjorkman A.D., Hédl R., Andrews C , Barni E., Becker T., Becker-Scarpitta A., Benito-Alonso J.-L., Bennie J., Berki I, Blüml V., Brunet J., Bullock J.M., Van Calster H , Carbognani M., Chudomelová M., Closset-Kopp D., Dan Turtureanu P., Daskalova G.N., Decocq G., Dick J., Diekmann M., Dirnböck T., Durak T., Eriksson O., Erschbamer B., Graae B.J., Heinken T., Hermy M., Horchler P., Jandt U., Jaroszewicz B., Kanka R., Kollár J., Kopecký M., Kudernatsch T., Lamprecht A., Lenoir J., Macek M., Malicki M., Máliš F., Michelsen O., Mitchell F., Naaf T., Nagel T.A., Newman M., Newton A.C., Nicklas L., Oddi L., Orczewska A., Orsenigo S., Ortmann-Ajkai A., den Ouden J., Pauli H., Peterken G., Petřík P., Pielech R., Pusxas. M., Randin C , Reczyňska K., Rixen C , Schei F.H., Schmidt W., Sebesta J., Stachurska-Swakon A., Standovár T., Šwierkosz K., Teleki B., Theurillat J.-P., Ursu T.-M., Vanneste T., Vellend M., Vergeer P., Vild O., Villar L., Vittoz P., Winkler M., Wipf S., Wu F., Zhang S. & 96 R E F E R E N C E S De Frenne P. (2026): Contrasting thermophilization among forests, grasslands and alpine summits. - Nature: 1-5. Zeileis A. & Grothendieck G. (2005): zoo: S3 Infrastructure for Regular and Irregular Time Series. - Journal of Statistical Software 14: 1-27. Zellweger F., De Frenne P., Lenoir J., Vangansbeke P., Verheyen K., BernhardtRómermann M . , Baeten L., Hédl R., Berki I, Brunet J., Van Calster H., Chudomelová M., Decocq G., Dirnbóck T., Durak T., Heinken T., Jaroszewicz B., Kopecký M., Máliš F., Macek M., Malicki M., Naaf T., Nagel T.A., Ortmann-Ajkai A., Petřík P., Pielech R., Reczyňska K , Schmidt W., Standovár T., Swierkosz K., Teleki B., Vild O., Wulf M. & Coomes D. (2020): Forest microclimate dynamics drive plant responses to warming. - Science 368: 112-115. Zhou L., Hong Y., L i C , Lu C , He Y., Shao J., Sun X., Wang C , Liu R., Liu H , Zhou G. & Zhou X. (2020): Responses of biomass allocation to multi-factor global change: A global synthesis. - Agriculture, Ecosystems & Environment 304: 107115. Zhu Y., Han G , Pellissier L., Li M . - H , Jiang L., Tang L, Gao C. & Ren H. (2026): Climate warming, rather than nitrogen deposition, reduces plant diversity and increases community homogenization in a desert steppe. - Agricultural and Forest Meteorology 311: 110937. 97 APPENDIX Appendix A Species-specific growth probability model performance Table 1 Extracted coefficient of determination (R2 ) from species-specific growth probability models. For generalized linear mixed-effect models (GLMMs), R2 is reported as marginal Nakagawa R2 . For species where the models showed singular fit, the models were refitted as generalized linear models (GLMs), and the R2 is reported as McKelvey-Zavoina R2 . Species R2 Model type Achillea collina 0.132643428 G L M M (Nakagawa R2) Acinos arvensis 0.147710466 G L M M (Nakagawa R2) Adonis vernalis 0.225366866 G L M M (Nakagawa R2) Allium flavum 0.03552783 G L M M (Nakagawa R2) Allium montanum 0.06560521 G L M M (Nakagawa R2) Alyssum alyssoides 0.095461488 G L M M (Nakagawa R2) Alyssum montanum 0.08295146 G L M M (Nakagawa R2) Anthericum ramosum 0.78396995 G L M (McK-Zav R2) Anthyllis vulneraria 0.179367786 G L M (McK-Zav R2) Arabis auriculata 0.699352161 G L M M (Nakagawa R2) Arabis sagittata 0.121773547 G L M M (Nakagawa R2) Arenaria serpyllifolia 0.053270246 G L M M (Nakagawa R2) Arrhenatherum elatius 0.061436468 G L M M (Nakagawa R2) Asperula cynanchica 0.193951768 G L M M (Nakagawa R2) Aster amellus 0.128854913 G L M M (Nakagawa R2) Astragalus onobrychis 0.073955936 G L M M (Nakagawa R2) Avenulapubescens 0.054099384 G L M M (Nakagawa R2) Bupleurum falcatum 0.051933617 G L M (McK-Zav R2) Camelina microcarpa 0.049719794 G L M M (Nakagawa R2) Campanula persicifolia 0.275731058 G L M M (Nakagawa R2) Carex humilis 0.006033845 G L M M (Nakagawa R2) Carex supina 0.04736295 G L M (McK-Zav R2) Centaurea stoebe 0.028973439 G L M M (Nakagawa R2) Centaurea triumfetti 0.296900327 G L M (McK-Zav R2) Cerastiumpumilum agg. 0.263665444 G L M M (Nakagawa R2) Dianthus lumnitzeri 0.01958128 G L M M (Nakagawa R2) Dianthus pontederae JO. 141950423 G L M M (Nakagawa R2) Dorycnium germanicum 0.077152786 G L M M (Nakagawa R2) Echium vulgare 0.078540647 G L M M (Nakagawa R2) Elymus hispidus 0.126041792 G L M M (Nakagawa R2) Eryngium campestre 0.461737728 G L M (McK-Zav R2) 98 APPENDIX Species R2 Model type Erysimum diffusum 0.137661878 G L M (McK-Zav R2) Erysimum odoratum 0.03185983 G L M (McK-Zav R2) Euphorbia cyparissias 0.053498437 G L M M (Nakagawa R2) Festuca csikhegyensis 0.05319164 G L M M (Nakagawa R2) Festuca rupicola 0.106330106 G L M M (Nakagawa R2) Festuca valesiaca 0.060439328 G L M M (Nakagawa R2) Fragaria viridis 0.044094264 G L M (McK-Zav R2) Fumana procumbens 0.222102602 G L M (McK-Zav R2) Galatella linosyris 0.151765419 G L M M (Nakagawa R2) Galium glaucum 0.067538708 G L M M (Nakagawa R2) Galium verum 0.270526353 G L M M (Nakagawa R2) Genista tinctoria 0.132006958 G L M M (Nakagawa R2) Holosteum umbellatum \ 0.257673552 G L M M (Nakagawa R2) Hypericum perforatum 0.020428727 G L M M (Nakagawa R2) Irispumila 0.038971546 G L M (McK-Zav R2) Jovibarba globifera 0.0087986 G L M M (Nakagawa R2) Koeleria macrantha 0.050403229 G L M M (Nakagawa R2) Lamium amplexicaule 0.034009641 G L M M (Nakagawa R2) Linaria genistifolia 0.206915123 G L M M (Nakagawa R2) Medicago falcata 0.550384159 G L M M (Nakagawa R2) Medicago minima 0.003742876 G L M M (Nakagawa R2) Melica ciliata 0.00499711 G L M M (Nakagawa R2) Microthlaspi perfoliatum 0.146395029 G L M M (Nakagawa R2) Minuartia rubra 0.039846123 G L M M (Nakagawa R2) Minuartia setacea 9.17E-04 G L M M (Nakagawa R2) Muscari tenuiflorum 0.097582255 G L M M (Nakagawa R2) Peucedanum cervaria 0.127473818 G L M M (Nakagawa R2) Phleum phleoides 0.102348779 G L M M (Nakagawa R2) Plantago media 0.373009921 G L M (McK-Zav R2) Poa angustifolia 0.012082474 G L M M (Nakagawa R2) Poa badensis 0.192071415 G L M M (Nakagawa R2) Poa bulbosa 0.105265068 G L M M (Nakagawa R2) Potentilla incana 0.072703824 G L M M (Nakagawa R2) Reseda lutea 0.009567876 G L M M (Nakagawa R2) Salvia pratensis 0.105288706 G L M M (Nakagawa R2) Sanguisorba minor 0.03084963 G L M M (Nakagawa R2) Saxifraga tridactylites 0.491820971 G L M (McK-Zav R2) Scabiosa ochroleuca 0.102619477 G L M M (Nakagawa R2) Scorzonera austriaca 0.136200224 G L M (McK-Zav R2) 99 APPENDIX Species Model type Secnrigera varia 0.206069081 G L M M (Nakagawa R2) Sedum album 0.069769314 G L M (McK-Zav R2) Senecio jacobaea 0.017133375 G L M M (Nakagawa R2) Seseli osseum 0.011173335 G L M (McK-Zav R2) Sesleria caerulea 0.040803312 G L M (McK-Zav R2) Silene otites 0.005161583 G L M M (Nakagawa R2) Stachys recta 0.055631751 G L M M (Nakagawa R2) Stipa capillata 0.059594845 G L M (McK-Zav R2) Stipa pulcherrima 0.056848705 G L M (McK-Zav R2) Teucrium chamaedrys 0.093229691 G L M M (Nakagawa R2) Teucrium montanum 0.053751573 G L M (McK-Zav R2) Thalictrum minus 0.725640493 G L M (McK-Zav R2) Thymus pannonicus 0.23148433 G L M M (Nakagawa R2) Thymus praecox 0.080860957 G L M (McK-Zav R2) Verbascum lychnitis 0.088862354 G L M M (Nakagawa R2) Veronica chamaedrys agg. 0.10343593 G L M (McK-Zav R2) Veronica praecox 0.079046707 G L M M (Nakagawa R2) Veronica prostrata 0.077917635 G L M M (Nakagawa R2) Veronica spicata 0.128618675 G L M M (Nakagawa R2) Vincetoxicum hirundinaria 0.75478879 G L M (McK-Zav R2) Viola tricolor 0.071093794 G L M M (Nakagawa R2) 100 APPENDIX Appendix B Monthly precipitation sums under current and future conditions Date Figure 1 Monthly macroclimatic precipitation sums averaged over the period 2015-2025 for current conditions (weather station Brod nad Dyji) and over 2090-2100 for future climate scenarios. Future climate projections are derived from MPI-ESM-l-2-HRmodel. 101 APPENDIX Appendix C Residuals of observed and predicted microclimate temperatures F1: Svatý Kopeček F4: Devin F7: Kotel -3 0 3 6 -4 0 4 -5 0 5 10 F2: Svatý Kopeček F5: Děvín F8: Stolová hora -3 0 3 6 -5.0 -2.5 0.0 2.5 5.0 -5.0 -2.5 0.0 2.5 5.0 7.5 R e s i d u a l s Figure 2 Density plots of daily temperature residuals, facetted by site and vegetation type. Residuals represent the differences between observed and predicted microclimate temperatures. Individual curves correspond to distributions within each month. 102 APPENDIX Appendix D Macroclimate-microclimate relationships 1 2 3 4 5 6 7 8 9 10 11 12 -5 0 5 10-5 0 5 10 0 5 1015 0 5 101520 8 121620 1620242815 20 25 162024 10152025 5 101520-5 0 5 1015 -5 0 5 10 Macroclimate temperature (°C) Figure 3 Relationship between macroclimate temperature (weather station Brod nad Dyjí) and observed microclimate temperature (°C), measured at each of our study sites, facetted by month. 103 Rocky grasslands SSP2-4.5 SSP3-7.0 SSP5-8.5 O 4^ Stipa capillata Asperula cynanchica Galatella linosyrisAlyssum alyssoidesArabis auriculata Microthlaspi perfoliatum Saxifraga tridactylitesLamium amplexicauleKoeleria macrantha Iris pumila Cerastium pumilum agg. Muscari tenuiflorum Linaria genistifolia Holosteum umbellatum Scorzonera austriaca Thymus praecox Camelina microcarpa Festuca valesiaca Verbascum lychnitis Potentilla incana Medicago falcata Echium vulgare Festuca csikhegyensis Hypericum perforatum Allium senescens subsp. montanum Allium flavum Sedum album Scabiosa ochroleuca Alyssum montanum Arrhenatherurn elatius Poa badensisAnthericum ramosum Stachys recta Sanguisorba minor Fumana procumbensViola tricolorDorycnium germanicum Acinos arvensisEuphorbia cyparissiasVeronica spicata Salvia pratensis Teucrium chamaedrys Teucrium montanum Dianthus pontederae Arenaria serpyllifolia Centaurea stoebe Phleum phleoides Galium glaucum Veronica prostrata- -0.2 Stipa capillata Asperula cynanchica Potentilla incana Galium glaucum Sanguisorba minor Veronica prostrate Centaurea stoebe Koeleria macrantha Euphorbia cyparissias Muscari tenuiflorum Festuca csikhegyensis Medicago falcata Linaria genistifolia Echium vulgare Stachys recta Thymus praecox Iris pumila Scorzonera austriaca Festuca valesiaca Salvia pratensis Anthericum ramosum Poa badensis Scabiosa ochroleuca Camelina microcarpa Allium senescens subsp. montanum Galatella linosyris Verbascum lychnitis Dianthus pontederae Teucrium chamaedrys Dorycnium germanicum Alyssum montanum Fumana procumbens Allium flavum Sedum album Arrhenatherurn elatius Hypericum perforatum Acinos arvensis Lamium amplexicauleCerastium pumilum agg. Holosteum umbellatum Microthlaspi perfoliatum Teucrium montanum Alyssum alyssoides Arabis auriculata Veronica spicata Phleum phleoides Arenaria serpyllifolia Viola tricolor Saxifraga tridactylites Galium glaucum Veronica prostrata Sanguisorba minor Centaurea stoebe Potentilla incana Euphorbia cyparissias Stachys recta Koeleria macrantha Salvia pratensis Linaria genistifolia Muscari tenuiflorum Echium vulgare Festuca csikhegyensis Thymus praecox Teucrium chamaedrys Dianthus pontederae Poa badensis Asperula cynanchica Camelina microcarpa Verbascum lychnitis Allium senescens subsp. montanum Phleum phleoides Sedum album Iris pumila Scabiosa ochroleuca Acinos arvensis Allium flavum Hypericum perforatum Alyssum montanum Festuca valesiaca Medicago falcata Scorzonera austriaca Arrhenatherurn elatius Arenaria serpyllifolia Holosteum umbellatum Cerastium pumilum agg. Stipa capillata Lamium amplexicaule Teucrium montanum Microthlaspi perfoliatum Fumana procumbens Alyssum alyssoides Arabis auriculata Anthericum ramosum Viola tricolor Dorycnium germanicum Veronica spicata Galatella linosyris Saxifraga tridactylites -0 15 -0.10 -0 05 0.00 0.05 Growing index (log ratio of growing days) < CD CQ CD > T 3 T 3 CD =3 Q . X m en TD Q) CD O 13 Index 0.1 0 0 I -0.1 --0.2 -0.2 -0.1 0.0 0.1 O CD C/> r -o o CD CQ C/> J , Q)i—t- o' CD O =3 CL CD X Q) O O c/> c/> > Tl "0 m D X Loess grasslands SSP2-4.5 SSP3-7.0 SSP5-8.5 Arabis sagittata Stipa capillata Carex supina Elymus hispidus Astragalus onobrychis Muscari tenuiflorum Stipa pulcherrima Antheneum ramosum Alyssum alyssoidss Galatella linosyris Arabis auriculata Medicago falcata Arrhenatherum elatius Holosteum umbellatum Festuca valesiaca Peucedanum cervaria Scorzonera austriaca Lamium amplexicaule Galium verum Adonis vernalis Veronica spicata Microthlaspi perfoliatum Echium vulgare Eryngium campestre Iris pumila Koeleria macrantha •orycnium germanicum Acinos arvensis Scabiosa ochroleuca Camelina microcarpa Arenaria serpyllifolia Thymus pannonicus Achillea collina Thalictrum minus Euphorbia cyparissias Cerastium pumilum agg. Allium flavum Potentilla incana Teucrium chamaedrys Phleum phleoides Securigera varia Dianthus pontederae Festuca rupicola Viola tricolor Fragaria viridis Veronica prostrata Aster amellus Galium glaucum Centaurea stoebe Sanguisorba minor Salvia pratensis I Arabis sagittata Euphoi bia uypanssias Veronica prostrata Sanguisorba minorGalium glaucum Centaurea stoebe Stipa capillata Medicago falcata Elymus hispidusAnthericum ramosum Potentilla incanaSalvia pratensis Astragalus onobrychis Galium verum Dorycnium germanicum Muscari tenuiflorumStipa pulcherrimaThalictrum minus • Galatella linosyris • Scorzonera austriaca • Koeleria macranthaFragaria viridis • Arrhenatherum elatiusFestuca valesiaca • Adonis vernalisAcinos arvensis Peucedanum cervariaEchium vulgareIris pumila Thymus pannonicusCarex supinaSecurigera varia Teucrium chamaedrysEryngium campestreScabiosa ochroleucaArenaria serpyllifoliaVeronica spicataDianthus pontederaeCamelina microcarpaAchillea collinaAllium flavumPhleum phleoidesFestuca rupicola Holosteum umbellatumLamium amplexicauleMicrothlaspi perfoliatumCerastium pumilum agg.Alyssum alyssoides Arabis auriculata Viola tricolor Aster amellusSanguisorba minor Veronica prostrata Centaurea stoebe Galium glaucum Salvia pratensis Euphorbia cyparissias Potentilla incana Galium verum Arabis sagittata Koeleria macrantha Astragalus onobrychis Acinos arvensis Muscari tenuiflorum Elymus hispidus Fragaria viridis Echium vulgare Dianthus pontederae Arrhenatherum elatius Adonis vernalis Arenaria serpyllifolia Festuca rupicola Phleum phleoides Peucedanum cervaria Teucrium chamaedrys Scabiosa ochroleuca Achillea collina Iris pumila Eryngium campestre Allium flavum Festuca valesiaca Camelina microcarpa Scorzonera austriaca Securigera varia Thymus pannonicus Stipa pulcherrima Medicago falcata Dorycnium germanicum Anthericum ramosum Holosteum umbellatum Thalictrum minus Carex supina Stipa capillata Cerastium pumilum agg. Lamium amplexicaule Aster amellus Veronica spicata Microthlaspi perfoliatum Galatella linosyris Viola tricolor Arabis auriculata Alyssum alyssoides -0.1 0.0 0.1 G r o w i n g index (log ratio of growing days) Dealpine grasslands SSP2-4.5 o Securigera varia Anthericum ramosumAsperula cynanghicaDorycnium germanicumThalictrum minusTeucrium chamaedrysArrhenatherurn elatiufs Poa badensisFestuca valesiaca Vincetoxicum hirundinaria • Allium senescens subsp. montanumGalium verum • Veronica spicata • Scorzonera austriaca Carex supina Iris pumila Thymus praecox Muscan tenuiflorumErysimum odoratum Bupleurum falcatum Teucrium montanum Elymus hispidusSesleria caerulea Centaurea stoebeGalium glaucum • Lipana genistifolia • Hypericum perforatum Sanguisorba minor Aventila pubescensAlyssum montanum Verbascum Jychnitjs Fragana viridis Veronica prgstrata • Allium flavum Sedum album Scabiosa ochroleucaAchillea collina Veronica chamaedrys agg. Alyssum alyssoidesHolosteum umbellatumCamelina micrpcarpaPotentilla mcana Microthlaspi perforatum Euphorbia cyparissiasSalvia pratensis Lamium amplexicaule Viola tricolorpum (urn agg.• —era - Cerastium Sfachys recl_ Arabis auncula(a Festuca csikhegyensisSaxifraga tridactylitesPhleum phleoidesKoeleria macranthaArenaria serpyllifoljaAcinos arvensis Centaurea triumfetti Echium vulgare Campanula persicifoHa • Plantago med a • Festuca rupicola Arabis sagittata Dianthus pontederaeAster amelluslyllr •Anthyliis vuTnerana SSP3-7.0 SSP5-8.5 > "0 "0 m zg x Arabis sagittata Asperula cynanchica Securigera varia Festuca"valesiaca Anthericum ramosum Galium verum Thaijctrum minus Carex sup.ina Fragaria viridis Dorycnium germanicum Centaurea stoebe Galium glaucum Arrhenatherurn elatius Poa badensis Sanguisorba minor Allium senescens subsp. montanum Avenuia pubescens Thymus praecox Iris pumi|a Vincetoxicum hirundinaria Veronica chamaedrys agg. Veronica spicata Erysimum odpratum Potentilla mcana Veronica prostrata Salvia pratensis Bupleurum falcatum Festuca csikhegyensis scorzonera austriaca Euphorbia cypanssias Teucrium chamaedrys Elymus hispidus Musgari tenuiflorum Teucrium montanum Hypericum perforatum Allium flavum Sesleria caerulea Linaria genistifolia Sedum album Campanula persicifolia Cerastium pumilum agg. Alyssum montanum Microthlaspi perfoliatum Koeleria macrantha Verbascum lychnitis Echium vu1 — Centaurea trig.... Holosteum umbellatum Lamium amplexicaule Arenaria serpyllifolia Alyssum alyssoides Viola tricolor Acinos arvensis Camelina microcarpa Dianthus pontederae Plantago media Festuca rupicola Arabis auriculata Anthyliis vulneraria Staghys recta Saxifraga tndactylites -0.20-0.15-0.10-0.050.00 0.05 - - * - • llllllllllllllllllllllllllllllllllll -0.05 11 0 0.05 0. Aster amellus Erysimun Poter pianthus pontederae Veronica chamaedrys agg Campanula persicifc.._ Thymus praecox Allium flavum Teucrium r Thalictr Ach|l Seslerii Saxifraga tr Festuc euphorbia c Phlegm p hli lilie Arenaria serpyllifolia Teucrium chamaedrys- - • mtmt : : ] -0.10-0.050. K) 0. ISO c Index 0.1 0.0 -0.1 G r o w i n g index (log ratio of growing days) APPENDIX Figure 4 Log-ratio Growth Index (LGI) for dry grassland species across three selected climate scenarios (SSP2-4.5, SSP3-7.0, SSP5-8.5). Each plot represents one of three vegetation types (rocky dry grassland, loess dry grassland, Ses/ma-dominated grassland). The LGI represents the relative change in the number of future growth-days compared to the current baseline. Bars are coloured by the direction of change (blue indicates the loss of growth-days, while red indicates a gain in growthdays). Species are ordered along the y-axis from the most negative to the most positive LGI values. 107 APPENDIX Appendix F Growth intensity across vegetation types R o c k y grassland Dealpine g r a s s l a n d S S P 5 - 8 . 5 - Cfl •2 S S P 3 - 7 . 0 m c o S S P 2 - 4 . 5 00 Current # ^ ^ ^ ^ cf <ř ^