De novo super-early progeny in interspecific crosses Pisum sativum L. × P. fulvum Sibth. et Sm

Earliness in crop plants has a crucial role in avoiding the stress of drought and heat, which are the most important challenging stressors in crop production and are predicted to increase in the near future due to global warming. Furthermore, it provides a guarantee of vegetable production in the short growing season of agricultural lands in the northern hemisphere and at high altitudes. The growing human population needs super early plant cultivars for these agricultural lands to meet future global demands. This study examined de novo super-early progeny, referred to as much earlier than that of the earlier parent, which flowered in 13–17 days and pod setting in 18–29 days after germination, discovered in F2 and studied up to F5 derived from interspecific crosses between garden pea (P. sativum L.) and the most distant relative of pea (P. fulvum Sibth. et Sm.). De novo super-early progeny were found to be earlier by about one month than P. sativum and two months than P. fulvum under short day conditions in the F5 population. In respect of days to flowering and pod setting, de novo super-early progeny had a relatively high level of narrow sense heritability (h2 = 82% and 80%, respectively), indicating that the selections for earliness in segregating populations was effective for improvement of extreme early maturing varieties. De novo super-early progeny could be grown under heat stress conditions due to the escape ability. Vegetable types were not only high yielding but also free of any known undesirable traits from the wild species, such as pod dehiscence and non-uniform maturity. It could be considered complementary to “speed breeding”, possibly obtaining more than six generations per year in a suitable climate chamber. Not only de novo super-early progeny but also transgressive segregation for agro-morphological traits can be created via interspecific crosses between P. sativum and P. fulvum, a precious unopened treasure in the second gene pool. Useful progeny obtained from crossing wild species with cultivated species reveal the importance of wild species.


Results
The seed coat color (testa) and surface of P. sativum (ACP 20) were seen as yellowish cream and wrinkled, while these traits were recorded as black and smooth in P. fulvum (AWP 600) ( Table 1). After pollination, the seed coat color and surface were found to be black and smooth in reciprocal interspecific crosses between P. sativum × P. fulvum and P. fulvum × P. sativum. Both F 1 plants derived from reciprocal interspecific crosses had black seed coat and smooth seed coat. The flower color of F 1 plants was orange, while P. sativum and P. fulvum had white and orange flower colors, respectively. The flower color in the plants derived from the F 2 population were separated as three distinct categories of orange, fuchsia and white colors (Fig. 1). Segregations for flower color were fit well to 12 (orange) :3 (fuchsia) :1 (white), dominant epistasis ( Table 2).
Heterosis in F 1 progeny. For days to flowering and days to pod setting traits, the F 1 plants derived from interspecific crosses P. sativum × P. fulvum had negative average heterosis values of − 27% and − 22%, respectively (Fig. 2). Plant height had a heterosis value of 95%, while the first pod height had a heterosis of 195%. Heterosis for the pods per plant, the seeds per pod, pod length and biological yield traits was found to be 49%, − 9%, − 30%, and 27%, respectively. For the seed yield per plant, considerable heterosis was detected at 42%, while heterosis for the harvest index was 38% (Fig. 2).
Transgressive segregation for agro-morphological traits in F 2 and F 3 populations. In the F 2 population, 121 individual plants from four F 1 plants derived from interspecific crosses P. sativum × P. fulvum were grown in the same glasshouse. Plant height for P. sativum and P. fulvum was 137 cm and 41.3 cm, respectively (Table 3). Plant height for F 2 and F 3 populations was determined to be 16-244 cm and 7-252 cm, respec- Table 1. Some agro-morphological and salient traits of parents used in interspecific crosses P. sativum × P. fulvum.  www.nature.com/scientificreports/ While days to flowering of the de novo super early progeny in F 4 were between 24 and 34 days, days to pod setting varied between 28 and 48 days. Days to flowering of de novo super early progeny in F 5 was determined to be between 25 and 40 days and days to pod setting was between 30 and 47 days (Fig. 4). The de novo super early progeny occurred especially in the F 2 and F 3 populations, but not in the parents. De novo super early progeny flowered after 18 days in F 2 population (Fig. 3a), and when segregated for earliness in the later generations, flowered after 25 days in F 5 (Fig. 4). Some of the de novo super early progeny were selected for fresh green pod and seed in the F 5 population (Fig. 1d,e).
Heritability. Narrow-sense heritability ( h 2 ) values were highest for days to flowering (82%) and days to pod setting (80%) ( Table 3). While the narrow-sense heritability of the number of seeds per pod was 50, the plant height and the first pod height were 45. The heritability of biological yield, pod length, number of pods per plant, and seed yield were found to be 37%, 36%, 33% and 28%, respectively. Harvest index had the lowest narrowsense heritability at 16% (Table 3).

Relationships between correlated traits and lines.
Eigenvalues of the principal component analyses (PCAs) in the F 2 population, were found to be greater than 1 for four components. However, the diagram explained 54.5% of the total variance with two components (Fig. 5a). PC1 was closely related to plant height (PH), first pod height (FH), number of pods per plant (PP) and biological yield (BY) at 37.3%. PC2 was related to the number of seeds per pod (SP) and pod length (PL), representing 17.2% of the total variance. The third component was related to days to flowering (DF) and days to pod setting (DP) and explained 16.2% of the total variance. In addition, the progeny in the upper left part of the PCA diagram were those with earlier flowering and pod setting than the others (Fig. 5a). Considering the PCA results of the F 3 population, it was divided into two components. The first component (PC1) was related to almost all traits (DF, DP, PH, FH, PP, BY and SY) with a variance of 42.39%. PC2 represented 13.86% of the total variance with SP and PL traits (Fig. 5b). As in the (DF is days to flowering, DP is days to pod setting, PH is plant height, FH is first pod height, PP is number of pods per plant, SP is number of seeds per plant, PL is pod length, BY is biological yield, SY is seed yield, HI is harvest index). Table 3. Minimum (Min) and maximum (Max) values, means ( X ) ± standard errors ( S X ) and narrow-sense heritability ( h 2 ) for agro-morphological traits in parents and progeny derived from interspecific crosses P. sativum × P. fulvum.
Traits ACP 20 AWP 600    Direct and indirect effects on agro-morphological traits over days to flowering. To examine the direct and indirect effects of agro-morphological traits on days to flowering, path (p) analysis was performed on the F 3 population derived from inter-specific crosses P. sativum × P. fulvum (Table 4). Days to pod setting and first pod height had statistically significant (P < 0.05) direct effects on days to flowering, and the coefficients were p = 0.999* and p = 0.018*, respectively. In addition, plant height with p = − 0.021* had negative direct effect on days to flowering. Plant height (p = − 0.498*), first plant height (p = 0.655*), pods per plant (p = 0.303*) and biological yield (p = -0.288*) had significant indirect effect on days to flowering over days to pod setting (Table 4).

Discussion
Wild pea species are not only useful genetic resources for resistance to different a/biotic stresses [37][38][39][40][41][42][43][44][45][46][47][48][49][50][51][52] , but they also possess nutritional value for food and feed, desirable agronomic traits and advantages in nitrogen fixation 53 . Progeny have successfully been produced by interspecific crosses when P. fulvum was used as a pollen donor 37,41,45,54 . P. fulvum was reported in the second gene pool of the genus Pisum 55 . According to the available literature, reciprocal interspecific crosses were reported by Kosterin et al. 56 as in the present study. www.nature.com/scientificreports/ After hybridization, seed coat color and surface of the seed coat were recorded as black and round inside pods in P. sativum, mother or pollen receiver plant when P. fulvum species was used as a pollinator. A similar color and surface were found in F 1 plants indicating that black seed color and round seed surface were dominant over their counterparts. Gregor Johann Mendel, known as the father of genetics, reported a similar finding on seed coat surface in intraspecific crosses in pea crop 156 years ago 57 . The flower color of both F 1 plants was orange ( Fig. 1) indicating that orange flower color was dominant over white flower color as reported by Kosterin et al. 56 . Kosterin et al. 56 reported that flower color was quantitatively inherited in an F 2 segregation population derived from reciprocal interspecific crosses P. sativum × P. fulvum and P. fulvum × P. sativum. However, flower color in one way interspecific crosses P. sativum × P. fulvum was segregated in the F 2 population as three distinct categories with 12 (orange): 3 (fuchsia): 1(white colors) ( Fig. 1a-c) segregation ratio ( Table 2) inferring as dominant epistasis. Prior to the present study, interspecific crosses P. sativum × P. fulvum were reported in several studies 37,41,45,48,56,[58][59][60] , while hybrids P. fulvum × P. sativum have been introduced in only research study to date 56 .
Not only in the F 2 population but also in the F 3 population, transgressive segregations were investigated for agro-morphological traits. However, differences between maximum and minimum values for some agromorphological traits were declined in the F 3 population indicating that transgressive segregations were reduced from F 2 to F 3 population ( Table 3). The term of transgressive segregation was coined as a phenomenon specific to segregating generations and refers to the fraction of progeny that exceeds the parents in either a negative or positive direction [34][35][36] . This phenomenon is similar to heterosis in first-generation hybrids. Rieseberg et al. 36 defined the creation of transgressive segregations as: (1) mutation frequency in segregating populations; (2) reduced developmental stability; (3) non-additive allelic effects between loci or epistasis; (4) non-additive allelic effects within a locus or overdominance; (5) the unmasking of some recessive alleles that are generally heterozygous in the parents; (6) variation in chromosome number; and (7) the complementary action of additive alleles that are dispersed between the parents. In the present study, except for mutation frequency and variation in chromosome number, five reasons for transgressive segregations in F 2 and F 3 populations could be considered. Flowering time in pea has been considered to be two genetic control systems as a result of an increase in recessive genes for early flowering, and dominant genes for late flowering 63 . Earliness in pea crosses was related with additive and nonadditive genetic effects 64,65 . Transgressive segregations in segregating generations in interspecific crosses in Cicer species have been previously outlined 66,67 . One of the main objectives of most breeding studies is to increase yield, but it is a challenge for breeders to achieve an increase because the yield is affected by genes and environments due to its polygenic nature. Therefore, it is important to benefit from the genetic variations in segregating populations. For example, seed yield is a critical characteristic in terms of yield and this value was recorded as maximum 33.4 g in the best parent (female parent), while it was recorded as 83 g in the F 2 population, more than twice that of the parent (Table 3). Farmers prefer varieties with higher biological yields for forage. In such cases, genotypes www.nature.com/scientificreports/ that produce more vegetative parts rather than seeds are selected. The maximum biological yield in the F 2 and F 3 populations was found to be 237 g and 243 g, respectively, which is almost 3 times higher than the best parent. De novo early progeny was found in segregated populations (Fig. 3, 4 and Table 3). Two progeny flowered in 18 days in F 2 population (Table 2), whereas progeny from these two progeny flowered in 13 days in F 3 (Fig. 3 and Table 3). According to the results of the literature review, no study has reported peas that flowered in 13 days. The use of wild peas in crossbreeding studies has been reported by many researchers to increase genetic diversity 51,68-70 , and a wide variation has been reported in the flowering time of pea crop 12-14 . Watts et al. 71 reported that the flowering time of pea was between 52 and 71 days. Vanhala et al. 15 60 . In the present study, days to flowering was recorded under short day conditions with daily sun hours (day length) between 2-9 h (Fig. 6). Earliness was induced by long light period under controlled conditions and six generations were advanced via speed breeding in pea 10,11 . If de novo super early lines (Fig. 4) can be grown under suitable conditions, it is considered that 8-10 generations per year can be obtained. The earliest progenies in the F 2 population were 3 times earlier than the mean of the  (Table 3 and Fig. 3). The earliest progeny in the F 3 population were 3.9 times earlier than the mean of the female parent and 9.5 times earlier than the mean of the male parent in terms of the number of days to flowering and pod setting (Table 3 and Fig. 3). F 2 and F 3 populations have shown bimodal distribution for flowering time as a segregation ratio of late to early flowering of 9:7 (Fig. 3a,b). Under short-day conditions, this distribution demonstrated that flowering time is controlled by two genes with duplicate recessive epistasis or complementary gene action. Similar findings on 9:7 distribution for flowering time were reported by in F 2 population derived from the intraspecific crosses in pea 27 and chickpea 72 . A bimodal distribution, late and early, was discovered for the flowering node using F 2 population in pea, and the gene responsible for the late flowering was named Sn 73 . Snsn gene pair was responsible for the difference between early and late flowering in intraspecific pea crosses 74 . Approximately 20 loci were pointed out to be involved in pea flowering variation 29 ,  www.nature.com/scientificreports/ with cultivated alleles usually resulting in early flowering and a decline in photoperiod response 75 . Hr, Sn, E, and Lf genes were found to be effective in naturally occurring variation 29 . Hr, on the other hand, has just one naturally existing mutant allele, whereas Sn has both naturally occurring and induced mutant alleles 76 . In short day conditions, hr was reported to be induced early flowering and decreased of the response to photoperiod but not whole loss, while sn was indicated to be caused complete daylength insensitivity 22 . Dominant alleles of E have been stated to cause early flowering and this effect had complex interactions with other loci 29 . Flowering was inhibited on both long and short days conditions by Lf, while accessions having Lf gene was inactivated by the nonsense mutation providing extreme earliness 21,22 . The extreme earliness identified in the present study is considered to be due to variants of Lf gene. There is increasing global concern about the impact of climate change on food production, livelihoods, and food security 77,78 . Global warming is anticipated to be one of the biggest hazards for food and it will have adverse effects on agricultural production. It is estimated that the expected world population will be 8 billion by 2030, which will require an increase of 60% in the current food production 79,80 . The vast majority of the world's population lives in cities and given the reasons for migration from rural areas to the city, it is inevitable that the consumption rate will create even more food deficits 78 . According to the Intergovernmental Panel on Climate Change (IPCC), global warming will exceed 1.5 °C by 2030, leading to permanent loss of the most sensitive ecosystems and crisis for vulnerable people and societies in underdeveloped and developing countries 81 . Drought and heat stresses, which create abiotic stressors that significantly reduce the yield of plants, will be the leading causes of global warming in agriculture 82 . Drought and heat effects are expected to increase with climate change and increasing water shortages 83 . Two types of heat stress in agricultural areas are (1) heat shock, which occurs in daytime and lethal temperatures, and (2) moderate heat, which is higher than optimum temperature in daytime or at night 83 . Since pea cultivation is carried out in rainfed areas in some parts of the world, plants suffer from heat stress. Plants exhibit three different mechanisms for heat tolerance, namely, heat escape, heat tolerance and heat avoidance 84 . Heat escape enables plants to quickly complete their life cycle in a short time and under favorable conditions and in this way, plants tend to escape drought due to early maturity 18,85 . Early flowering genotypes in pea play a crucial role in minimizing bottlenecks such as abiotic and biotic stresses and can lower production costs as a result of low input. De novo super early progeny was seen to have the ability to escape from heat stresses, while late flowering progeny were subjected to heat stresses during flowering and pod setting (Fig. 6). Vegetable types were not only high yielding but also free of any known undesirable traits of the wild species, such as pod dehiscence and non-uniform maturity (Fig. 1).
In addition to problems such as global warming, pandemics that cause the death of thousands of people, such as COVID-19, also adversely affect the world 86 . In vulnerable countries where hunger and malnutrition are already common, pandemics such as COVID-19 also pose a threat to food safety and scarcity 87 . Hunger, malnutrition, and unbalanced nutrition affect more than 820 million people worldwide, and approximately 150 million children are negatively affected by an irregular diet 88 . At the beginning of the COVID-19 outbreak, there was an excessive demand for food in the world 89 . Food deficit is a major problem in such pandemics and thus early varieties play an important role in meeting these food needs. Very early pea progeny, which can be obtained in about one month, will be a unique food source able to meet food needs in a short time.
The high values of narrow sense heritability means that the gain from selection can be achieved by selecting de novo super early progeny. In the present study, days to flowering in de novo super early progeny was fixed with days to flowering of 24-25 days in F 4 and F 5 populations (Fig. 4). Guindon et al. 61 reported that the number of flowering days, number of pods, pod length and number of seeds per pod had high values of heritability, but not plant height. Singh et al. 90 found high broad sense heritability in pea for days to flowering, and days to pod setting. Since earliness in the present study was seen to have high heritability, success can be achieved in selection in early generations.
Based on principal component analyses (PCAs), earliness including days to flowering and pod setting was related in both F 2 and F 3 populations (Fig. 5). Seed yield was correlated with pods per plant, in accordance with the findings of Guindon et al. 61 and Esposito et al. 91 . According to the available literature, this is the first path analysis on days to flowering in interspecific crosses P. sativum × P. fulvum. Days to pod setting, plant height, and first pod height had statistically significant direct effects on days to flowering (Table 4). Path analysis performed in the F 2 population obtained from intraspecific crosses by Singh et al. 90 , showed that plant height and pods per plant had a direct positive effect on seed yield, and a direct negative effect on days to flowering. Similar findings were reported by Singh and Srivastava 92 and Tiwari et al 93 . In the present study, biological yield and pods per plant had a significant effect on seed yield (Table 4) indicating that seed yield selection should be carried out according to biological yield and pods per plant.
In presenting the results obtained in the present study, the following can be suggested for readers; (1) Not only transgressive segregations but also de novo super early progeny were obtained by interspecific crosses between the cultivated pea (P. sativum) and its wild relative P. fulvum. (2) Some de novo super early progeny flowered in 24-25 days in both F 4 and F 5 generations. (3) The heritability was found to be high in the present study since the earliness trait was little affected by the environment. It can be suggested that traits with high heritability will yield successful results in early generation selection studies. (4) De novo super early progeny obtained from interspecific crosses in the breeding study matured without exposure to heat stress and this progeny can escape from heat stress. (5) It is thought that by using de novo super early progeny obtained in breeding studies, peas with a short vegetation period can be grown without the risk of frost exposure in the northern hemisphere. (6) Green pods and seeds of garden pea are important vegetables in Mediterranean cuisine and they have a high price in markets with earliness whenever the crop is purchased by consumers in autumn. (7) It has also been understood that de novo super early progeny can also be cultivated at high altitude without the risk of frost. (8)  www.nature.com/scientificreports/ is estimated that much more than six generations in a year under suitable conditions can be obtained via de novo super early progeny. This shows that these de novo super early progeny can be alternative and complementary materials to the speed breeding approach. (9) This study can be considered an example study for the use of wild peas in breeding studies. (10) Useful progeny obtained from crossing wild species with cultivated species reveal the importance of wild species.

Materials and method
Parents. ACP 20 is the cultivated genotype of P. sativum with mid-early flowering and white flowers, whereas AWP 600 is a wild genotype of P. fulvum with late flowering and orange flower color. According to our records, ACP 20 is a landrace grown as vegetable in Antalya, Turkey, while AWP 600 is originated from Turkey and received from USDA GRIN, USA. The plant materials comply with relevant institutional, national and international guidelines and legislations. As seen in Table 1, ACP 20 has large, cream-colored seeds and weight of 42.8 g per 100 seeds. AWP 600 has small, black-colored seeds and weight of 5.4 g per 100 seeds. It is resistant to seed beetle (Callosobruchus chinensis L.) and powdery mildew caused by Erysiphe pisi DC 37,38 . Both parents, ACP 20 and AWP 600, were grown on the campus of Akdeniz University, Antalya, Turkey (30°38′E, 36°53′N, 33 m above sea level) as a spring-sown crop in 2015.
All progeny derived from interspecific crosses P. sativum × P. fulvum were advanced from F 1 to F 5 (Fig. 1) De novo super-early progeny. Transgressive segregation was defined as the occurrence of progeny with values greater or less than the values of their parents (male and female plants) in segregated generations 35 . These extreme phenotypes are a major mechanism by which extreme or novel adaptations develop [34][35][36] . Different explanations have been provided to take into account the presence of extreme phenotypes in segregating populations 94,95 . However, de novo super-early progeny is here referred to as progeny with flowering or pod setting and maturity much earlier than the early parent (best parent) in segregating generations.
Agro-morphological traits. The following phenological traits were recorded in male and female plants and progeny derived from interspecific crosses P. sativum × P. fulvum. Days to flowering (DF) was recorded as the number of days after germination until the first flowering. Days to pod setting (DP) was recorded as the number of days after germination until the first pod setting. De novo super-early plants were individually harvested and advanced as single plant progeny.
The following agro-morphological traits were recorded on single plant in F 1 and F 2 , while they were obtained from average of five plants from F 3 to F 5 . Plant height (PH) and first pod height (FH) were recorded in cm as the height of a plant from the ground to the top of the plant and as the height from ground to the first pod, respectively. Pods per plant (PP) and seeds per pod (SP) were recorded as total number of pods per plant and seeds per pod, respectively. Pod length (PL) was recorded in cm as the length of a pod. Biological yield (BY) was recorded in grams (g) as the total weight of a plant after harvest, while seed yield (SY) was recorded in g as the weight of seeds per plant after harvest. Harvest index (HI) was calculated as a percentage (%), as the ratio of seed yield per plant to biological yield per plant multiplied by 100. For the pod length, three randomly selected pods on each plant were used and the number of seeds per pod trait of the same pods were recorded. Agro-morphological traits were recorded in F 1 , F 2 and F 3 progeny and the parents. In the F 4 and F 5 populations, days to flowering and days to pod setting were recorded only in early lines and their parents.

Soil analyses.
According to the soil analysis of the experimental field, it was determined that nitrogen and organic matter content were low, pH was found to be as high as 7.62, whereas CaCO 3 was 26.8%. Although iron and zinc contents were considered to be deficient due to the high pH of the soil, other plant nutrition elements were generally considered to be balanced. Agronomic practices. A drip-irrigation system was used and plants were irrigated with well water at threeday intervals to avoid drought stress. Weed control was performed by hand. Fertilization was not applied because the plants supplied 80% of the nitrogen requirement 55 . Heterosis. Hybrid vigor or average heterosis (H A ) estimated for agro-morphological traits in order to test general combining ability between parents was calculated as: