Effect of geographical origin, regional adaptation, genotype, and release year on winter hardiness of wheat and triticale accessions evaluated for six decades in trials

Abstract

The overwintering of accessions of three wheat species (bread, durum, spelt) and triticale was evaluated annually from 1960 to 2020 at the Crop Research Institute in Prague by means of trials in wooden-boxes. The set of tested cereal accessions was regularly changed, so that the winter survival ratings of the accessions represented a highly unbalanced set of values. Out of 15,510 winter survival values, 1,991 accessions were classified using a generalized linear model with the logit link function and transformation of calculated coefficients into a nine-point scale to estimate their genotypic Winter Hardiness Potential (WHP 1 = least hardy; WHP 9 = most hardy). The WHP of the winter wheat accessions depended on their geographical origin: for European countries, the mean ranged from WHP 7 for north-eastern countries to WHP 3 for south-western countries. There was a decrease in WHP for accessions released in the 21st century in the Central European region. A significant correlation was found between the cultivar WHPs and their survival in the field after severe winters, and registration of new, more cold tender cultivars increased after warm winters. Dependence of the overwintering index on climatic changes in the period 1960 to 2020 is discussed.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-89291-8.

Introduction

The winter hardiness (WH) refers to the plants’ ability to withstand the adverse factors of winter. It is a very complex trait, as a number of abiotic and biotic factors can interact. The abiotic factors include: frost, winter flooding, ice-encasement, winter desiccation, vertical movements of topsoil due to alternating freezing and thawing of water in the soil (frost heaving), the action of the snow layer, etc.^1,2. Among the biotic factors, it is mainly the effect of moulds during prolonged periods under the snow, where the darkness, temperatures approaching freezing and high humidity cause respiration damage and attack by moulds^3,4. Frost is a dominant factor not only because it can cause direct damage to the plants but it also contributes to other adverse winter factors that can lead to reduced winter survival⁵.

WH is most often determined by the survival of the plants after winter, either as a percentage of survival between autumn and spring plant counts^6,7 or by visual assessment of the crop stand based on a five- or even a ten-point scale (10 = all plants survived; 1 = all plants killed)^8,9. New methods are also available to quantify damage to accessions in the spring^10,11. Winter survival (WS) of small-grain cereal plants varies depending upon a combination of factors: both environmental (weather, soil properties, agronomic practices), and plant-based (developmental and growth stage; as well as morphological, physiological, and biochemical changes - which are genetically determined)^3,5. Additionally, there are differences in winter hardiness (WH) among cultivars; among cereal species - from the hardiest rye, through the various hardy wheat species, to the least hardy barley^11–13. All of these are reasons why determination the WS of accessions through a winter under field conditions may not always allow for a good differentiation of the WH. After some severe winters, all accessions may be killed; or, conversely, after mild winters, all accessions may survive. Often, a whole WH range of accessions may not be differentiated after one winter due to only a partial or complete winter kill (WK).

The need to overcome the above problems associated with assessing winter hardiness under field conditions usually leads to the testing accessions over several winters and/or at several different locations. The purpose of multi-year assessments is to separate environmental and genotypic differences in WS. However, distinguishing accessions with a small but significant difference in WH may also be difficult due to large experimental errors caused by a number of other sources (e.g., uneven soil profile, non-uniformity of the snow cover, absence of differential WK)¹⁴. These obstacles have led to development of new methods aimed at refinement the estimation of the WH potential. Field trials using several standards (checks) with known WH also have limitations due to the fact that these standards change over time, and may also be subject to environmental error¹⁵. Among other methods, the FSI (Field Survival Index) calculation developed by Fowler and Gusta¹⁴ is still successfully applied¹⁶.

Other approaches to overcome the irregular occurrence of winters that result in differential winter survival of accessions include enhancing winter stresses by growing plants in containers during winter, or exposing plants to selected stress factors under laboratory conditions - most commonly freezing in boxes^3,17. In Kharkiv, Ukraine, Yuriev¹⁸ introduced winter hardiness testing of cereals by growing them in boxes placed in an open vegetative house during the autumn and winter seasons. The plants were devoid of snow, thus exposed to severe harsh ambient temperatures – frosts. This test has been used with various modifications¹⁹. Hoeser²⁰ placed wooden plant boxes at least 50 cm above the ground and exposed them to winter weather. When snow fell, the boxes were covered with movable roofs to prevent the formation of a snow cover on the plants. This semi-controlled test has been successfully used in Germany^8,21,22. In both cases, plants are mainly exposed to winter frosts in their natural environment, and their WS is assessed in spring after the plants have recovered. Segeťa¹⁸ used this test to assess multiple winter factors and placed the wooden boxes at two heights above the ground (5 and 50 cm). In addition to frost, he also monitored the effect of temperature fluctuations at the beginning of spring (frost heaving) or winter drought by using portable shelters over parts of the boxes. Later, he extended the sowing of cereals to three dates to investigate the effect of different stages of plant growth and development on overwintering. In this way, WS of cereal accessions tested in the wooden boxes trials has been regularly obtained in our institute, CRI, since 1960^23,24.

With the development of new methods for the genetic mapping of WH using QTL, and, more recently, GWAS, there is a growing need to assess the genotypic WH of a large set of accessions (in the order of hundreds) obtained from multi-year field WS assessments at different sites. Statistical methods using linear model (LM) analyses of winter survival with effects of accession and trial solved by the best linear unbiased estimator (BLUE) or prediction (BLUP) are now often used to estimate the WH from a series of trials^{7,9,21,25–27}.

The objective of this paper is to provide a comprehensive overview on the 60-year series of WS data for nearly two thousand wheat (and triticale) accessions obtained from the annual wooden-box trials performed at the Crop Research Institute in Prague (CRI-Prague). In addition to the method of processing the WS database, the aim was to express genotypic WH, using a nine-point scale, applicable for international comparisons of accessions with regard to genebank descriptors and wheat breeding usage. After obtaining a long-term series of Winter Hardiness Potential (WHP) data of the accessions, we then focused on evaluating them by geographic origin, time of commercial release, and the impact of climate change.

Results

Classification of accessions and treatments

With the exception of three years (1970 to 1972), we collected winter survival data for accessions tested by the wooden box trials from 1960 to 2020. A total of 188 treatments of the wooden box trials were collected (up to 6 treatments each year, a combination of two placements of wooden-boxes with respect to the level above ground, at up to three sowing dates) (Tab. S1), containing 15,510 winter survival values for 19,991 wheat and triticale accessions. After processing the WS data using a generalized linear model, we obtained the coefficient of accession (C_A) and its standard error (STDE_A) for each accession as well as the coefficient of treatment (C_T) and its STDE_T for each treatment.

To find the optimal number of winter survivals per accession (N_A), their dependence on the size of the standard errors (STDE_A) was determined (Fig. S1). Initially, the value of the standard error, STDE_A decreased rapidly as the number of winter survivals of accession increased from 1 to 5 or 6. Thereafter, the STDE_A only gradually decreased as the number of winter survivals increased. From this point, we concluded that 5 to 6 winter survivals of accession in different treatments of the wooden box trials were sufficient to estimate the WH of the accession. Based on these findings, all accessions with a high variability (i.e., with STDE_A higher than 1) or a low number of repetitions (N_A) (all accessions that were only repeated once or twice) were removed from further evaluations. Only accessions repeated three or four times with low variability were retained as well as accessions that repeated more than four times.

Both coefficients (C_A a C_T) were further transformed with an appropriate sigmoidal function to obtain 9-point scales of WHP (winter hardiness potential) for accessions (Fig. S2), and OWI (overwintering index) for treatments.

OWI and the environment

Of 188 calculated OWIs (number of all treatments), 48, 32, and 20 occurred on the first, second, and third sowing dates for boxes placed on raised benches, respectively; and 39, 30, and 19 occurred for the first, second, and third sowing dates for boxes placed on the ground, respectively (Tab. S1). A comparison of the OWI distributions between these 6 experimental treatments of the wooden box trials is shown in Fig. 1a. Lower values of OWIs (lower winter survival of accessions) occurred when placing the boxes on the raised benches compared to those placed on the ground which was true for all 3 sowing dates. However, the mean value of OWI for a given placement of the wooden boxes was very similar for the first and second sowing, while the third date had a lower value (Fig. 1a), indicating higher winter damage to accessions on the third sowing dates. The plants of the third sowing were mostly in the 2 to 3 leaf stage when entering the winter, in contrast to the plants from the first and second sowing that were at the tillering stage (3 to 1 tillers). Therefore, the enhanced winter hardiness of the accessions in the first and second sowing could be related to the enhanced growth stage and higher mass gain of plants in them.

Fig. 1.

Overwintering index of experimental treatments. (a) Box plot of overwintering indices for 6 basic experimental treatments of the wooden box trials. 1, 2 and 3 = sowing dates, wooden-box placement: R = raised bench, G = on the ground. (b) Relationship between the absolute minimum winter soil temperature due to placement of the wooden-boxes (raised bench and ground) and the overwintering index from the first sowing. (c) and (d) Overwintering index in the placement of wooden-boxes (raised bench or ground) during the years 1960 to 2020; a linear trend (dashed red line). Data for 1970–1972 seasons are not available.

Figure 1c and d show how the OWI varied across treatments for wooden boxes placed on the raised bench and on the ground for the first sowing dates, which had a continuous series of evaluations over the entire 60 years of the wooden box trials. To make the series of evaluations as complete as possible, values of an OWI = 1 were added for those treatments where evaluations were not performed due to the winter kill of all accessions, and an OWI = 9 where all accessions survived the winter (compare with Table S1). While the OWI of the treatment with wooden boxes on the raised bench increased from a value of around 3 in the early 1960s to 6 by the end of the second decade in the 21st century, the increase in OWI of the wooden boxes on the ground was moderate, and around 7 (Fig. 1c and d). In summary, for both box placement treatments, the size of the OWI, and the frequency of complete plant survival gradually increased over the 60 years of evaluations, with the majority of these events occurring in recent years (2000–2020).

OWI and climate

The meteorological data characterizing the winter season (December to February) in the CRI-Prague area showed several facts: a gradual increase in the mean winter temperature and minimum air temperature, a constant trend in total precipitation, and a decreasing number of days with snow cover as well as frost days between 1960 and 2020 (Fig. S3). Correlations between these meteorological variables and the OWI values demonstrated the relationships between increases in the OWI and trends in increasing winter temperatures. These correlations were moderately strong for the wooden boxes on the raised bench. The weaker correlations for wooden boxes on the ground might be due to the gradual decline in the number of days these boxes were covered and protected by snow. In summary, the gradual warming of the winters was the main factor that increased the overall survival of the accessions (i.e., the size of the OWI) through the 60 years of wooden box trials.

In addition to the above air meteorological variables, the OWI values were compared with the absolute minimum winter soil temperatures obtained since 1983 for both wooden box placements (Fig. 1b). The correlation between OWI and minimum soil temperature was strong for boxes placed on the raised bench (rS = 0.6), and moderate for boxes placed on the ground (rS = 0.4); indicating that other winter factors also affected the survival of accessions in this treatment. Furthermore, it can be seen from Fig. 1b that the entire range of the OWI values from the high OWI = 9 to the low OWI = 1 was approximately within the range of the minimum soil temperatures from − 10 °C to -20 °C.

WHP of accessions and their winter field survival

The WHP of the accessions ranged from 0.6 to 9.3, rounded from 1 (least hardy) to 9 (most hardy). First, we compared the calculated WHP of the evaluated accessions in the different years using the wooden box trials with the winter field survival of accessions evaluated at 16 field stations of the Central Institute for Supervising and Testing in Agriculture (CISTA) after the severe winters of 2002/2003 and 2011/2012 (Fig. 2a). This comparison showed a very high agreement between the wooden-box determined WHP and accession winter field survival after both severe winters (with a Spearman correlation coefficient of value 0.89). Thus, it is possible to estimate the susceptibility of accessions to winter stresses in the field using the wooden-box trial ratings.

Fig. 2.

Winter Hardiness Potential in tested accessions. (a) Relationship between WHP of wheat acessions tested in wooden-box trials at CRI-Prague and their winter field survival after two severe winters (2002/03 and 2011/12) at CISTA stations. (b) Box-plot of WHP of wheat species evaluated by wooden-box trials over 60 years (TC = triticale, Ts = Tr. spelta, Td = Tr. durum, Ta_w = Tr. aestivum_winter habit, Ta_s = Tr. aestivum_spring habit), (c) Mean WHP of winter wheat accessions with different geographical origins, values in brackets show number of accessions, bars = STDE_WHP. (d) Map of Europe with countries coloured according to the mean WHP of winter wheat accessions (legend top left). The map chart was created in Excel version 2108 (MS Office LTSC 2021).

WHP of different Triticeae species

Lastly, 1616 accessions including: 47 triticale, 7 spelt, 16 durum wheat, 1507 winter wheat, and 39 spring wheat accessions were included for further analysis (Tab. S2). Looking at the distribution of WHP by species (Fig. 2b) revealed the highest mean value (6.4) for triticale, followed by winter spelt (5.7), and the lowest (2.6) for durum wheat. Bread winter wheat accession WHP values ranged from 1 to 9 with a mean of 5.3; and the spring accessions had a mean of 1. The hardiest triticale reached a WHP of 9, while spelt only reached 7, durum wheat being 4, and with bread spring wheat rarely reached more than 1.5. It should be noted that the number of accessions for some of these species was low; while the number of winter bread wheat accessions was very high, and they came from a greater number of countries.

Accessions with the highest hardiness levels (up to WSP 9) came from the Ukraine (e.g., Kharkovskaya 917, Ferrugineum 1239), Russia (Ulyanovka, Ljutescens 116), and Canada (e.g., CDC Raptor and CDC Buteo). The most hardy triticale plants came from Poland (e.g. Alma, Malno), the most hardy spelt came from Germany and the Czech Republic (e.g. Baulaender spelz, Rubiota), the most hardy durum wheat from Poland, and Slovakia (Komnata, IS Pentadur), etc. (see Tab. S2 for a summary). These accessions with high winter hardiness belonged to both the obsolete and advanced cultivar groups.

WHP and geographical origin

The 1507 winter wheat accessions assessed by the provocation trials came from 41 countries - often with only one or a few representatives. For 24 countries that had multiple accessions, the mean WHP was calculated and ranked by their size (Fig. 2c). Although some accessions were few in number (e.g., Finland) or came largely from a single location in a given country (e.g., CAN; obtained from the University of Saskatchewan), they demonstrated a relationship between the geographic origin of the accessions and their mean WHP. When we color-coded the mean WHP of accessions originating from different European countries on a map of Europe, we observed a decreasing mean WHP from the north-east to the south-west of Europe (Fig. 2d).

A detailed distribution of accessions according to WHP classes showed a higher proportion of more hardy accessions in north-east of European countries, compared to countries in south-west of Europe (Fig. 3). In the case of a high number of accessions for some countries (GER, CZE, FRA), an almost normal distribution of WHP accessions can be discerned. Despite a lower mean WHP value (3.1), there were also accessions with high winter hardiness (WHP = 7 or 8) in the French winter wheat group. These were both obsolete traditional cultivars (e.g., Automne rouge barbu) and a cultivar from the late 1980s, Renan (which, however, has the winter-hardy cultivar Mironovskaja 808 in its pedigree as evidenced by Grin Czech). On the other hand, the entries from the northern and north-eastern countries have the highest number of accessions in the WHP classes 6, 7, and 8, and only occasionally accessions with a degree lower than 5; e.g., the Ukrainian accession Intenzivnaya (WHP = 2.1), which has the spring wheat Kazachstanskaja 126 (Grin Czech) as one parent.

Fig. 3.

Distribution of wheat accession winter hardiness potential (WHP) for different geographical origins classified into nine classes. n = number of accessions of that country, n_c = number of accessions per class.

WHP and year of release

The accessions were divided into three groups based on year of release (registration, authorization, use): the first group containing traditional cultivars and landraces and those released from breeding programs before 1960. The second group contained advanced or improved accessions released from 1961 to 1999; and the third group contained the most recent accessions, bred by expanding international breeding programs (and also in a period of significant climate change) from 2000 to 2020 (Fig. 4). We did not have representation for all accessions of different origins in these three groups, with some countries only having representation in the first two groups. Considering the countries from Western and Southern Europe with a low mean WHP (France, UK, Netherlands); and, conversely, countries from Eastern Europe with a high mean WHP (Russia), distribution of accessions remained similar in all three release periods. For Central European countries (Czechia, Germany, Austria, Hungary), the third group (i.e., released from 2000 to 2020) usually showed a decrease in WHP compared to the two groups of accessions with earlier releases where the distribution of WHP did not change significantly and the mean WHP was higher. Only accessions from Poland showed a higher mean WHP value in the first group compared to the second group from the 20th century (however, the number of accessions in the first group was relatively low).

Fig. 4.

Box-plots of winter hardiness potential (WHP) of winter wheat accessions of different geographical origin by year of release (< 1960 = blue; 1961–1999 = orange; 2000–2020 = grey). The letters above the box-plots indicate significant (p < 0.05) differences between compared groups.

There was only a limited number of accessions available for all three release time periods for most countries except Czechia where almost all cultivars and advanced breeding lines released were available during the period of interest (1980–2020). The distribution of WHP for Czech accessions released from 1981 to 2020, summed over 5-year intervals was studied in detail (Fig. 5). From 1981 to 2000, the accessions of Czech origin had a similar range of WHP (from degree 4 to degree 8), and a similar mean WHP around degree 6. For the accessions released from the beginning of the 21st century forward, there were already accessions with a WHP near 3. The mean WHP of Czech accessions gradually decreased from 6 to 5 during the last 20 years, in line with the decrease in the representation of accessions with a WHP higher than 6; only rarely, cultivars with high WHP have appeared in the second decade of 21st century: (it was Nordika with WHP = 8.6 in 2014, and RGT Premiant with WHP = 7.5 in 2017).

Fig. 5.

Box-plot of WHP of winter wheat accessions of Czech origin released in five-year periods from 1981 to 2020. The letters above the box-plots indicate significant (p < 0.05) differences between groups.

Discussion

Based on comparisons of several experimental treatments, provocation wooden box trials have proven to be a reliable method for annual WS evaluations of wheat and triticale accessions over a period of 60 years (1960–2020). The calculated coefficients allowed us to characterize both each accession (by WHP) as well as each experimental treatment (by OWI) in the wooden box trials.

Environmental factors

OWI values allow for an inter-comparison of all experimental treatments obtained over 60 years of wooden-box trials according to the intensity of the stress factors affecting the survival of the accessions. In our experimental setup, the freezing temperature had a dominant effect, especially in the case of the raised bench treatments. However, the sole value of the lowest temperature reached in the treatments does not explain the survival rate of the accessions. The duration of low temperatures, the occurrence of several freeze-thaw events, or warming followed by a rapid decrease in temperature during winter all have a significant effect on the overall survival of cereal accessions^28–31. We cannot rule out the influence of other unfavourable winter factors. For example, the poorer survival of accessions in the third sowing may be related to their insufficient accumulation of assimilates, which are necessary for cold hardening. Soil heaving, where the underground and above-ground parts of the weaker plants may be broken and then drying out, or are directly exposed to frost, cannot be ruled out. Segeťa¹⁸ found soil heaving of several centimetres within the wooden-boxes, and confirmed the greater sensitivity of cereals in early vegetative stages to this stress. On the other hand, we never observed the occurrence of snow mould, probably because the soil with the plants was frozen at the time of the snowfall.

A number of climate studies have described the gradual warming of European winters^32,33, as we ourselves found by comparing several basic meteorological variables from our weather station over the past 60 years. Moreover, our experimental site, located on the outskirts of the ever-growing city of Prague, is gradually being affected by the Prague urban heat island. Despite our employing boxes emplaced at different heights above the ground, this warming of the winters led to an increase in accession survival; thus, reducing the differences in the WH between these accessions, and also reduced our usage of the wooden box method. Areas where the minimum soil temperatures in the boxes were in the range from − 10 ^oC to -20 ^oC appeared to be the most suitable.

Effects of genotype and regional adaptation

WHP represents genetically-based accession characteristics critical for WS, which in turn affects the final yield. In order to cover the widest possible cultivated wheat WH range by our scale, it was necessary to evaluate a set of accessions with the greatest hardiness differences. In our set of accessions, this range was represented by the cold tender spring cultivars and the very hardy Canadian and Ukrainian cultivars. In years past, but also true recently, it was mainly the Ukrainian cultivars from Crimea and surrounding areas that brought high WH into Canadian cultivars¹⁶. This is similar to the winter wheat cultivars bred in other countries (e.g., Renam in France, and Vlada in the Czech Republic).

Comparison of our WHP values with published results for the WH of cultivars evaluated in field trials showed significant agreement, not only for the Czech Republic but also for trials in Finland³⁴ (rS = 0.76), and Canada¹⁶ (rS = 0.97) (see Figs. S4 and S5). Therefore, WHP represents an estimate of the genotypic WH, as opposed to the sub-annual WS, which are strongly influenced by both the year and environments¹⁴.

Despite the different number of bread, durum, and spelt wheat and triticale accessions, their mean WHP corresponded to the published differences in WH among species^11–13. Mean WHP decreased going from triticale through spelt and bread wheat to durum wheat, which had the lowest WHP. The WHP of spring bread wheat accessions varied from 1 to 2 while winter accessions occupied the whole scale from 1 to 9. This wide range of WHP indicates the variability in winter hardiness of wheat depending on where it is grown and bred.

The strong relationship between the geographical origin of cultivars and their WH has been demonstrated in several studies^21,26,35,36. Our study shows this relationship in a greater detail and on a larger scale. For the regions and countries in Europe, this trend shows a clear difference with location; from the most hardy accessions originating from the north-eastern countries to the least hardy ones originating from the south-western countries. This is a trend that mirrors those from the harshest to the mildest winters, and from the occurrence of the coldest temperatures (winters) with the lowest mean temperatures to the warmest winters (Europe Hardiness Zone Maps are available at: https://www.plantmaps.com). Therefore, the WHP makes it not only possible to compare differences among accessions but also between countries or different regions. The mean WHP can be used as a breeding target or as the optimum level of WHP required for growing winter wheat in different countries or regions.

Current trends in cultivar release

It was interesting to see how the mean WHP of wheat accessions changed during the 20th and early 21st century. As noted earlier, the obsolete cultivars from the first half of the 20th century were often similar in WH to advanced and modern cultivars²⁴. Our large data base of tested accessions has allowed us to divide them into three groups for each country according to the period of origin.We confirmed that the mean WHP of the accessions from the different countries did not change very much in both periods of the 20th century. It was only at the turn of the 20th century that the mean WHP of most Central European accessions decreased. This is in contrast to Western European accessions (with low mean WHP) or Eastern European accessions (with high mean WHP), which did not show a marked mean WHP change in the first two decades of the 21st century in our evaluation. We already have mentioned the potential impact of warming winters in Central Europe. Warm winters do not allow for the selection of highly winter-hardy breeding lines; and low hardiness lines are then accumulated when breeding new cultivars. Leišová-Svobodová et al.³⁷ in 355 bread wheat accessions bred in Central Europe between 1900 and 2015 showed that the genetic diversity (based on 40 microsatellite markers) was mainly influenced by the breeding period.

A detailed analysis was performed on the WHP of Czech accessions released between 1980 and 2020, and a gradual decrease in the mean WHP bred after 2000 was observed. The fact that the warming of winters can affect the WH of newly released cultivars was observed in the registered cultivars. After warm winters, the proportion of less hardy cultivars increased in the Czech Republic while the opposite was true after the severe winters of 2002/2003 and 2011/2012. During both of those winters, severe frost damage and killing of overwintering cereal crops occurred³⁸.

Nováček and Fučíková³⁹ pointed to other factors when they showed that after 1990, changes associated with competition from new cultivars of western provenance gradually appeared on the Czech market. Until the end of the 1980s, the highly WH cultivars originating from eastern countries (Ukrainian or Russian) were more often used as parents in the breeding programs. Since the 1990s, competition from less winter-hardy western cultivars with better quality and yield parameters has increased. In the second decade of the 21st century, both the composition of Czech cultivars as well as the selections of the breeding stocks has changed. In summary, the warming climate, together with the political and economic changes, have caused a gradual decline in the WH of the newly registered wheat cultivars in the Czech Republic during the first two decades of the 21st century.

Developmental differences

In terms of the sowing time and the length of vernalization (the period of exposure of the plants to cold which accelerates the transition of wheat to the reproductive phase), there is a group of so-called facultative (sometimes also called semi-winter) wheat genotypes^7,40–43. Winter types which are sown in autumn have the longest vernalization (several weeks), and the highest WH; spring types which are sown in spring have no vernalization requirement, and show the lowest WH; and facultative types, which can be sown in both autumn and spring, have a shorter vernalization requirement and lower WH than winter wheat. Due to the different severity of winters and growing seasons of wheat in different geographical areas and altitudes, these types are sometimes subsumed under spring or winter types^41,44–46. In southern Europe and North Africa, winter and spring wheat genotypes are sown together or with a delay of 2–3 weeks in October to the end of November in order to catch winter rainfall; and they do not belong only to the facultative group^3,47.

As a facultative wheat, one group of Czech cultivars/landraces was preferably grown in the Czech regions in the 19th and the first half of the 20th century⁴⁸. These are Czech accessions marked w(f) in Table S2, for example, Ceska presivka or Postoloprtska presivka. They are characterized with high WH (WHP from 5 to 8), very short or no vernalization, and very strong sensitivity to photoperiod⁴⁸. Their growth and plant development are more suppressed during short days than in winter types. The genetic analysis of the Ceska presivka (CP) was performed by Travníčkova et al.⁴⁹ who found the dominant allele Vrn-B1a and the recessive alleles vrn-A1 and vrn-D1 in the VRN1 locus; in PPD genes, they identified photoperiod-sensitive alleles ppd-D1b and ppd-A1b. Košner and Panková⁵⁰, by studying the substitution of chromosome 3B of CP into different genetic backgrounds of winter and spring cultivars, found the effect of the chromosome 3B of CP on late maturity as well as an interaction with the photoperiodic response. As studies on winter/spring habit^51,52 showed, relationships between vernalization, photoperiod, and plant development play an important role in the regulation of genes involved in cold tolerance in cereals. Also, genetic studies illustrated the importance of VRN and PPD genes for winter hardiness and cold tolerance in wheat and other winter cereals^{6,7,10,53–55}. The above-mentioned facultative Czech wheats thus represent a genetically interesting group for a more detailed study of the relationships between vernalization, photoperiod, and high winter hardiness.

Conclusion

The need to express the genotypic winter hardiness of small-grain cereals as accurately as possible has its application not only in the genetic and biochemical analyses of the studied set of accessions but also in the characterization of accessions stored in gene banks for use in breeding. The annual determination of WS under field conditions is burdened by high environmental variability. The proposed expression of winter hardiness on a general WHP scale of 1 to 9 goes beyond annual evaluation (beyond year-to-year variation in environmental factors) thus providing a basis for the estimation of genotypically determined winter hardiness of genetic resources.

Materials and methods

Plant materials

Seeds of 1991 wheat (Triticum aestivum L., T. durum Desf., T spelta L.) and triticale (x Triticosecale Wittm.) cultivars and accessions were obtained from the Gene Bank (GB) of the Crop Research Institute in Prague (CZE), the Central Institute for Supervising and Testing of Agriculture in Brno (CISTA), and from other cereal breeding stations and companies/research institutions (Tab S1). The country of origin and the year of release (registration, authorization, use of each of the accessions, or the year of termination of breeding line) were obtained from these same sources: (Search Accessions GRIN-Global (vurv.cz); Searching in Database of Plant Varieties / the Czech National List (eagri.cz). In some cases, these data have been corrected or supplemented from other sources (accessed up to June 30, 2022: ECPGR: EURISCO Catalogue (cgiar.org); Genetic resources information and analytical system (GRIS) for wheat and triticale – CIMMYT; Accessions - GRIN-Global Web v 1.10.6.1 (nordic-baltic-genebanks.org); Search Accessions GRIN-Global (ars-grin.gov); EU Plant variety database (v.3.2.1) (europa.eu); Siregal - URGI (inra.fr); GBIS/I (ipk-gatersleben.de). Wheat and triticale accessions (cultivars and lines) were obtained annually and evaluated for WS in the period from 1960 to 2020. Initially, only tens of accessions per year were evaluated while more than 100 accessions have been evaluated each year since the 1980. Within the revolving tests, one third of the accessions were changed each year, and each accession was tested for at least three years. Some accessions had been tested for a number of years, and had served as standards in the past (e.g., Mironovskaya 808 as high winter-hardy standard and San Pastore as low winter-hardy standard). These were the cultivars that had been grown in the fields at the time, and were used as standards at that time. Most entries were assessed in different years. GB or CISTA provided cultivars grown in different countries. GB also provided accessions from the core collections of other genebanks. The list of accessions also includes those that were multiplied in Institute of Wheat and Sunflower in Bulgaria in the 1970s and 1980s for joint trials within the CMEA (Council for Mutual Economic Assistance, which included the European countries of the Eastern Bloc); these accessions have (CMEA) after their name and the source is BGR in Table S2.

Wooden-box trials

The determination of winter hardiness of the accessions was performed annually via a provocation wooden-box test under natural conditions in an experimental facility of the Crop Research Institute in Prague (Fig. 6)). The experimental design has been described in detail in earlier studies (Prášil and Rogalewicz 1989). The wooden-boxes (40 × 30 × 15 cm) were filled with soil (10 parts Alfisol to 1 part Agro CS RKS I substrate, which contains a mixture of peat and clay with NPK fertilizer, trace elements, and dolomitic limestone). Seeds of three accessions were sown in two rows each containing 16 seeds, in each wooden-box.

Fig. 6.

Experimental area for testing the winter hardiness of cereals in wooden-boxes under natural conditions. (a) autumn, (b) winter, (c) spring.

The wooden-boxes were placed on iron tubes (5 cm diameter) that were laid either on the ground or on brick pillars 50 cm above the ground. In the first case, the wooden-boxes are referred to as the ground experimental treatment, and the second as the raised bench experimental treatment. The treatment on the ground allowed the plants to be covered by snow in the event of snowfall, while the treatment on the raised bench was more affected by ground frosts. In some years, a similar set of accessions was sown on three dates at 12 to 14-day intervals to determine the sensitivity of plants to winter stresses at different stages of development. The first sowing was in the last third of September, the second one in the first third of October, and the third one in the middle third of October, respectively. In total, up to six independent experimental treatments of the wooden-box test (two placements of wooden-boxes with respect to the ground at up to three sowing dates) could be used to assess the hardiness of a similar set of accessions during each year. Each experimental treatment represented an independent evaluation of WS accessions. Each accession was present in two or three wooden-boxes, randomly placed in each experimental treatment (i.e., 4 to 6 rows per accession).

The mean and minimum air temperature, precipitation, and the number of frost days and days with a snow cover in the winter (December - February) were obtained from weather stations located in the experimental area. Soil temperature at a depth of 2.5 cm was measured for each experimental treatment of wooden-box placement with weather station starting in 1983. Where the data from the weather station was incomplete; the absolute minimum soil temperature measured in each winter period was utilized.

Winter survival rates for each accession (WS in %) were calculated from the ratio of the number of living plants in spring and autumn for each experimental treatment in a given year. Each WS value was calculated from a minimum of 60 plants per accession/treatment. If complete winter survival or winter killing of accessions occurred in any of the treatments, these treatments were not evaluated (Table S2). Starting in 1960, four researchers (Vladimír Seget’a - the trial manager of the wooden box test in 1960–1982, Konstantin Papasisis in 1983–1991, Pavla Prášilová in 1992–2013, and Jana Musilová in 2014–2021) were responsible for the implementation of the tests at CRI in Prague.

Classification method

The stresses associated with overwintering accession are a combination of the sensitivity of the accession (C_A) and the winter severity of the treatment (C_T); therefore, the data were evaluated using an incomplete two-factor model of WS on accession and treatment. The model was solved by the GLM (Generalized Linear Model) function in R⁵⁶ as ‘pseudobinomial’, i.e., logit link function is applied so that the predicted survival rate of accession A in experimental treatment T is

For each accession with a given number of survival values (N_A), the coefficient of accession (C_A) and its standard error (STDE_A) were calculated. Similarly, for each experimental treatment of the provocation trials with a given number of accessions in the treatment (N_T), the coefficient of treatment (C_T) and its STDE_T were calculated. The histograms of the coefficients (centered to 0) resembled a unimodal distribution, so the coefficients were transformed by an appropriate sigmoidal (logistic) function to values that are almost uniformly distributed in a 0–9 interval; indicating the winter hardiness potential (WHP) of the accession or overwintering index (OWI) of the treatment.

WHP = 9/(1 + EXP(-CA)) + 0.5

OWI = 9/(1 + EXP(-CT)) + 0.5

The transformation was designed so the WHP and OWI values ranged from 0.5 to 9.5, and, when rounded, could provide a nine-point scale.

The WHP degree characterizes the genetic winter hardiness potential of the accession regardless of the time of evaluation (WHP 1 = least hardy; WHP 9 = most hardy).

The OWI degree characterizes the overwintering of the accessions in the experimental treatment regardless of the accessions (OWI 1 = lowest overwintering; OWI 9 = highest overwintering).

The STDE transformation was performed according to the formula:

STDEWHP = STDEA×9×EXP(-CA)/(1 + EXP(-CA))2

STDEOWI = STDET×9×EXP(-CT)/(1 + EXP(-CT))2

Comparison of WHP accessions with their winter field survival

Routine assessment data of wheat accessions grown at field stations of the Central Institute for Supervising and Testing in Agriculture was used to compare the calculated winter hardiness potential of wheat accessions with their field survival after harsh winters. For these studies, we selected 16 sites that represented climatic regions suitable for wheat agricultural production in the Czech Republic (Czechia) and significant differences in winter field survival of wheat cultivars occurred: Trutnov, Lípa, Čáslav, Jaroměřice, Uherský Ostroh, Branišovice, Chrlice, Pusté Jakartice, Staňkov, Věrovany, Hrubčice, Stupice, Nechanice, Uhřetice, Horažďovice, Kujavy. Winter survival was assessed during the spring by visual inspection using the 1–9 scale where 1 represents 0% plants surviving and 9 represents no apparent damage (i.e., 100% survival). Data on winter survival assessment of wheat in the field were obtained from the CISTA annual reports (Results of VCU Testing | ÚKZÚZ (eagri.cz) for 23 accessions grown after the winter of 2002–2003, and for 30 accessions after the winter of 2011–2012.

Data statistical evaluation

Spearman correlation coefficient (rS), Deming regression line, One-way ANOVA and Kruskal-Wallis test were calculated using XLStat software and Statistica v.14 TIBCO.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

FT

Frost tolerance

OWI

Overwintering index of experimental treatment

WH

Winter hardiness

WHP

Winter hardiness potential of accession

WK

Winter kill

WS

Winter survival

Author contributions

ITP, JM, PP, MC and KK contributed to the conception of the article and preparation of tables and figures. JM, PP and MC collected accession winter survival data from the protocols and basic accession information from the databases listed in the plant materials section. PV, MK, VH and ITP contributed funds and participated in data collection. JH and VP reviewed the data and materials provided by the Czech Gene Bank. JJ performed the statistical analysis. ITP wrote the first draft of the manuscript and all authors commented on previous versions of the manuscript. All authors read and approved the final version of the manuscript.

Funding

This work was supported by the Ministry of Agriculture of the Czech Republic Grant numbers:

QK22010293, QL24010142, RO0423 and MZE-62216/2022–13113/6.4.2.

Data availability

The datasets generated and/or analysed during the current study are not publicly available due to the specific database used, but are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.