Temperate woody perennials survive at low temperatures in winter by entering a dormant stage. This mechanism is characterized by the absence of visible growth in any plant structure containing a meristem and is associated with a slowdown of biological activity, a modification in hormonal balance, and extensive metabolic remodeling (Horvath et al., 2003; Rohde and Bhalerao, 2007).

However, dormancy is not just a survival strategy (Rohde and Bhalerao, 2007). Indeed, the temperature is a conditional factor in woody plants in temperate climates (Cooke et al., 2012; Ríos et al., 2014). Thus, chilling accumulation is required for proper flowering since the lack of winter chill during the winter period can cause physiological disorders such as bud burst delay and erratic and insufficient flowering negatively affecting yield (Viti et al., 2008; Campoy et al., 2011).

Although the need to accumulate chill for flowering and fruiting has been known for over 2 centuries (Knight, 1801), the biological basis behind remains elusive (Lloret et al., 2018; Goeckeritz and Hollender, 2021). However, the importance of chilling requirements (CR) in fruit trees is growing due to the increased release of new cultivars in many fruit tree crops, the expansion to warmer areas, and the lower winter chilling temperatures caused by climate change in many regions of the world (Luedeling et al., 2011).

Lang et al. (1987) proposed three different stages in dormancy: endodormancy, referred to when the regulation of dormancy is triggered by physiological factors; ecodormancy that refers to dormancy regulated by environmental factors; and paradormancy, which is used when growth inhibition arises from another part of the plant (e.g., apical dominance). The key factor to estimate agroclimatic requirements (AR; chilling and heat requirements) of particular cultivars is determining when the tree has overcome the phase of endodormancy and fulfilled its CR. Several approaches have been developed to estimate CR. For this purpose, the date of breaking of endodormancy has been experimentally determined by sequentially collecting shoots from dormant trees during the winter, exposing them to warm temperatures in controlled conditions, and quantifying chilling temperatures up to shoot collecting time when flower buds are considered capable to resume growth (Fadón et al., 2020a). Alternatively, the end of endodormancy has been determined by statistical correlations (Alonso et al., 2005) or partial least squares (PLS) regressions (Luedeling et al., 2013) with data of flowering dates and temperatures from long series of years. Both types of approaches are time-consuming, and in many cases, the results are not directly comparable under different climatic conditions (Fadón and Rodrigo, 2018). In the experimental approach, the use of cut shoots cannot entirely reflect the behavior of adult trees in field conditions, and several factors may affect the establishment of the date of endodormancy release and hence the estimations of CR (Fadón and Rodrigo, 2018). Likewise, statistical correlations require data from long series of years that are only available for a few sets of cultivars that exclude new cultivars, and the results may not fit equally well in different climates and latitudes (Dennis, 2003; Fadón et al., 2020a).

Several models have been used to quantify chilling and heating accumulation through hourly temperature data. The dynamic model has been reported as the most accurate model of chilling quantification (Ruiz et al., 2007; Luedeling and Brown, 2011; Campoy et al., 2012b,2019; Luedeling and Gassner, 2012). However, CR has also been calculated using Utah (Richardson et al., 1974) and Weinberger (1950) models to compare the results with those of the previous studies.

The dynamic model assumes that chilling accumulation is a two-step process, in which previous warm temperatures can counteract the effect of low temperatures during endodormancy. Initially, cold temperatures would promote the accumulation of an intermediate product. However, warm temperatures during this period of accumulation can stimulate the degradation of this precursor and reverse the process. Once the accumulation of this product reaches a critical concentration, it converts to a chill portion (CP), a stable factor, which cannot be destroyed, and promotes the break of the endodormancy (Fishman et al., 1987). The Utah model establishes temperature ranges that contribute differently to chilling accumulation, which is measured as chill units (CU) (Richardson et al., 1974). The chill accumulation takes place within a temperature range of 1.5–12.4°C (Richardson et al., 1974). Finally, the Weinberger model proposes that only temperatures below 7.2°C (45°F) are useful and have an effect in breaking the endodormancy (chilling hours, CH) (Weinberger, 1950).

Nowadays, CR in most cultivars is unknown, mainly due to the problems associated with the existing methods and the lack of a biological marker to determine the chilling fulfillment of a tree in field conditions (Fadón and Rodrigo, 2018). In previous works in apricot, anther differentiation has been related to winter dormancy since it sets a boundary between sporogenous tissue and pollen grain formation stages (Julian et al., 2011). The stamens remain in this anatomically quiescent stage during endodormancy to resume growth and complete the development of pollen grains in ecodormancy (Rodrigo and Herrero, 2002). However, male meiosis takes place once CR has been fulfilled but flower buds do not show external signs of development (Bartolini et al., 2006; Julian et al., 2014).

On the other hand, pollen development is a well-characterized and highly conserved process in angiosperms. Prior to meiosis, microspore mother cells become isolated by a wall with the deposition of a callose layer (McCormick, 2004). Callose is a linear homopolymer widely found in higher plants with important roles in plant growth and development (Chen and Kim, 2009). During pollen development, callose may act as a physical molecular filter involved in stress responses (Dong et al., 2005; Blackmore et al., 2007).

In this work, we evaluate male meiosis as a biological marker of breaking dormancy for estimating AR by histochemical detection of callose deposition around microspore mother cells. For this purpose, chilling and heating requirements of a group of 20 apricot cultivars have been estimated by quantifying chill accumulation until the date of male meiosis and heat accumulation from male meiosis to flowering for over 4 years. To validate the results with those obtained with other existing methods, endodormancy breaking has also been experimentally estimated by studying cut shoots under forcing conditions, by two statistical models that estimate the correlation between blooming dates and seasonal temperatures and the determination of the daily influence of chill accumulation in blooming dates by PLS regression.

Chill accumulation started in October in all the years analyzed and increased exponentially during the winter months. However, significant differences were observed between years, ranging from 89.4 chilling portions (CP) (2011–2012) to 112.9 CP (2003–2004) accumulated at the end of March (Figure 1).

In this work, we have validated male meiosis as a biomarker for the determination of dormancy breaking by comparing the dormancy breaking dates calculated with three other common methodologies in 20 apricot cultivars. The results support our initial hypothesis of male meiosis as a suitable biomarker for endo- to ecodormancy transition.

The AR has been determined in 20 apricot cultivars, for ten of which (“Berdejo,” “Henderson,” “Mitger,” “Muñoz,” “Pandora,” “Pepito del Rubio,” “Stella,” “Sun Glo,” “Tadeo,” and “Veecot”) no previous data were available. This is the first time that a biomarker (male meiosis) and two statistical methods (statistical correlation and PLS regression) have been used to estimate chilling and heat requirements in apricot since all the previous data have been obtained with the experimental method of transferring shoots into a growth chamber sequentially during winter.

The CR reported herein ranged from 30.9 to 84.8 CP (Dynamic model), 470–1,681 CU (Utah model), and 410–1,437 CH (Weinberger model), which are in line with the results previously reported in this species (reviewed in Fadón et al., 2020a). Our estimate of the CR for “Canino” (875.4 CH vs. 750–878 CH), “Luizet” (1,016.3 CH vs. 1,058–1,150 CH), “Moniqui” (983.6 and 1,036.4 CH vs., 799–1,057 CH), and “Paviot” (1,023.7 CH vs. 995–1,296 CH) by using male meiosis as a biomarker of endodormancy breaking was similar to what has been previously determined at the same location (Tabuenca, 1964, 1968, 1979) by using the forcing test (Supplementary Table 4). Three cultivars, “Canino,” “Harcot,” and “Goldrich” were previously studied in different regions, Tuscany (Italy) (Garcia et al., 1999; Viti et al., 2010) and South Africa (Campoy et al., 2012a) compared to Murcia (Spain). Although significant differences were reported between locations, our results under forcing conditions were similar to those reported in Murcia (Spain). In addition, our results showed similar CR values for some cultivars evaluated with the forcing test previously in other regions, such as “Moniqui” in Italy (921.7 and 1,003.1 CU vs. 930–1,140 CU; Guerriero et al., 2006), “Gönci Magyar” in Serbia (1,200.8 CU vs. 1,122–1,310 CU; Ruml et al., 2018), and “Goldrich” in Italy and Serbia (854.0 CU vs. 950–1,030 CU and 834–846 CU, respectively; Guerriero et al., 2006; Ruml et al., 2018; Supplementary Table 1).

The four methods showed the same group of cultivars with higher CR, and similar values were obtained between methods for each cultivar. However, some differences between methods were found for the cultivars with lower CR, “Canino,” “Golrich,” and “Pepito del Rubio” in both experimental methods, and “Harcot,” “Mitger,” “Moniqui 1006,” “Moniqui 2113,” and “Pepito del Rubio” in both statistical methods. Differences between experimental and statistical methods could be due to the different number of years used, 11–18 for statistical methods and 3–8 for experimental methods. Previous studies in apricot using the forcing test have reported differences in CR not only between years, but also between locations, which may be due to a small number of years of experiments (Garcia et al., 1999; Viti et al., 2010; Campoy et al., 2012a).

The classification of the cultivars obtained according to their CR was similar to that obtained by other authors (Ruiz et al., 2007; Campoy et al., 2012a). “Canino” has been considered a medium chill cultivar in previous works (Campoy et al., 2012a), whereas our results showed lower chill portions accumulation in “Canino” and “Veecot” than those in the other cultivars. Thus, we propose an intermediate group that includes cultivars with low-medium CR.

Differences in CR between methods affected heating requirements, showing a significant negative correlation as reported in previous work (Ruiz et al., 2007). Lower values of heat accumulation were found with the male meiosis method whereas higher values were obtained with the forcing test. The heat requirements reported for “Canino” (5,026 GDH) and “Harcot” (4,767 GDH) are in line with those reported in previous studies in Spain [5,418–6,029 GDH (Garcia et al., 1999; Campoy et al., 2012a)]. Our results showed higher heat accumulation for “Goldrich,” “Moniqui,” and “Gönci Magyar” than in other countries. As an example, for “Goldrich,” 4,479 GDH vs. 3,175–3,432 GDH in Serbia (Ruml et al., 2018) and 3,950 or 485–913 GDH in Italy (Guerriero et al., 2006; Viti et al., 2010).

Differences between methods as well as those in previous reports may also be due to the specific limitations of each method. The determination of endodormancy breaking by monitoring shoots under forcing conditions is the only experimental approach currently available and the most widely used to estimate the AR of fruit tree cultivars for more than 60 years (Brown and Kotob, 1957; Fadón et al., 2020a). However, the variability of the factors affecting the experimental design has often led to inconsistencies in the results previously obtained. In apricot, as in other species, the frequency of sampling of shoots during winter may vary between studies, as well as the temperature and photoperiod conditions in the growth chamber (Tabuenca, 1968, 1979; Andrés and Durán, 1999; Valentini et al., 2004; Ruiz et al., 2007; Bartolini et al., 2020). Furthermore, the determination of the endodormancy breaking date through the evaluation of bud growth can be carried out after different periods in the chamber. Several criteria are also used, such as significant increases in fresh (Brown and Kotob, 1957) or dry weight (Tabuenca, 1968), as well as changes in bud phenology (Bennett, 1949). Moreover, most studies evaluate flower buds, but some use vegetative buds (Fernandez et al., 2019). The lack of standardization has made most of the results not applicable to other regions with different climatic conditions (Campoy et al., 2019; Fadón et al., 2020a).

The statistical approaches estimate the date of chilling fulfillment based on a long series of phenological observations. Both correlations of winter temperatures with flowering dates and the PLS analysis have been widely used for the determination of AR in sweet cherry (Luedeling et al., 2013; Fadón et al., 2021), almond (Alonso et al., 2005; Benmoussa et al., 2017a; Díez-Palet et al., 2019), pistachio (Benmoussa et al., 2017b), and apple (Díez-Palet et al., 2019). In our work, results from both statistical methods have shown results similar to those from experimental methods. However, some works indicate the need for about 20 years of flowering date records to obtain consistent results (Luedeling and Gassner, 2012; Luedeling et al., 2013; Fadón et al., 2021), which is one of the main pitfalls of these methods since there are usually not so many years of data for many cultivars, especially for new releases from breeding programs (Fadón et al., 2020a). In addition, these approaches can estimate average chill and heat periods, but cannot provide information about dormancy dynamics in a particular year.

The limitations of the currently available methods, the increasing number of new releases in many fruit tree crops, as well as the decrease in winter chilling due to global warming have led to a number of recent research efforts focused on identifying suitable biomarkers to determine the transition from endo- to ecodormancy (Martínez-Gómez et al., 2017; Lloret et al., 2018). These include starch accumulation within the ovary primordia cells (Fadón et al., 2018a) and hormone regulation (Chmielewski et al., 2017; Vimont et al., 2019; Guillamón et al., 2020), as well as the expression of the DORMANCY-ASSOCIATED MAD-BOX (DAM) and other candidate genes in several Prunus species (Bielenberg et al., 2004; Sánchez-Pérez et al., 2014; Rothkegel et al., 2017; Prudencio et al., 2018, 2019; Balogh et al., 2019; Falavigna et al., 2019; Vimont et al., 2019; Quesada-Traver et al., 2020; Lloret et al., 2021). To deepen insights into the genetics of dormancy, transcriptomic studies have been performed by using flower buds at different dormant stages in apricot (Yu et al., 2020; Canton et al., 2021), sweet cherry (Vimont et al., 2019; Canton et al., 2021), almond (Prudencio et al., 2020), or peach (Yu et al., 2020; Canton et al., 2021). Genes involved in stress response, sugar metabolism, and cell wall assembly contribute to the endodormancy to ecodormancy transition in apricot. Hormone biosynthesis, including ABA catabolism and secondary metabolism, as well as floral and pollen development genes are relevant at ecodormancy (Yu et al., 2020). Sporophytic genes, such as MALE STERILE1 (MS1), ABORTED MICROSPORES (AMS), and MYB103 (Canton et al., 2021), as well as other genes involved in sporopollenin synthesis and deposition found in peach (Ríos et al., 2013), could be useful markers of the ecodormancy stage.

In apricot, the characterization of the phases of pollen development in relation to dormancy established that male meiosis is the first detectable sign of development after endodormancy breaking (Julian et al., 2011), when flower buds are still closed with no observable external phenological changes (Fadón et al., 2018b). This led us to use male meiosis as a biomarker to estimate the AR of cultivars for the first time by using the emergence of callose around the microspore mother cells as a boundary between endo- and ecodormancy. Callose deposition can be considered as a physical filter of molecules that have been reported as a response to several stresses (Chen and Kim, 2009) or degeneration processes, such as ovule abortion (Rodrigo and Herrero, 1998). During pollen development, the accumulation of a layer of callose around the microspore mother cells could provide insulation and possible protection against low winter temperatures.

Since the statistical analyses showed good agreement when male meiosis was compared to the other methods, these results strongly endorse the use of male meiosis as a new and alternative method to determine endodormancy release in apricot. The results obtained by this approach showed slightly higher values than those obtained by the other methods because male meiosis occurs in the early stages of ecodormancy, once the endodormancy breaking had already passed. Despite this limitation, this method has an important advantage since it allowed the determination of endo- and ecodormancy in a short time from samples collected directly from the field, avoiding the variability of results caused by the many factors that may affect the experiments under forced conditions (Dennis, 2003; Fadón and Rodrigo, 2018) and without the need to have a large phenological data set available (Luedeling and Gassner, 2012; Fadón et al., 2020a). Despite the important advantages of this method in estimating CR in apricot, endodormancy breaking and male meiosis are not as closely related processes in other species as sweet cherry, in which several months elapse between both processes (Fadón et al., 2019) and this could lead to overestimated calculations of AR (Fadón et al., 2021). This may be because sweet cherry requires a longer forcing period to flower (Fadón et al., 2020b). However, male meiosis as a biomarker of endodormancy could be an applicable method in those species with a shorter forcing phase in which flowering is more conditioned by the chilling phase, such as almond (Egea et al., 2003) or Japanese plum (Ruiz et al., 2018). Although further studies are needed to determine if this method is valid in other species, our results exhibit male meiosis as a valid approach for apricot, not only for determining AR but also as a basis for physiological and genetic studies of dormancy. In addition, whether a bud is an endo- or ecodormant can be determined in a short time from a sample taken directly from the tree in field conditions. The knowledge of the endo- to ecodormancy transition may provide valuable information to improve and simplify orchard management decisions. The progressive loss of cold hardiness is triggered once the buds have fulfilled their CR (Weiser, 1970a,b) but show no external signs of development (Julian et al., 2007; Hillmann et al., 2021). Determining endodormancy breaking in apricot through the detection of male meiosis could provide insights into whether flower buds are vulnerable to spring frosts, allowing agronomic frost protection decisions to be made. In addition, the date of male meiosis could allow the determination of the optimal application time of agrochemical treatments to advance flowering (Guillamón et al., 2022).

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

SH, JH, JL, and JR conceived the study and designed the experiments. SH collected the shoot and bud samples and performed the microscopic analysis. SH and JA performed the correlation analysis. SH and EF performed the PLS analysis. SH and AH performed the statistical analysis. SH, JL, EF, AH, JA, JH, and JR analyzed the results and wrote the manuscript. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.