Introduction

Although global soybean (Glycine max (L.) Merill.) production has grown over twofold in the past decade, climate change, shifting weather patterns, and growing global human population are expected to pose major challenges to developing environmentally resilient, high-yielding cultivars that are able to readily withstand the increasing biotic and abiotic stresses (Sneller et al. 2005; Fox et al. 2015; Wang et al. 2017; Kofsky et al. 2018; Gaire et al. 2020; Kilian et al. 2020). Due to the recurrent use of elite cultivars in breeding programs, genetic bottlenecks, and low diversity of founding populations, there is a growing concern regarding the genetic variation found within breeding programs (Gizlice et al. 1993, 1994; Kisha et al. 1998; Cui et al. 2000; Sneller et al. 2005; Hyten et al. 2006; Iquira et al. 2010; Friedrichs et al. 2016). Low genetic diversity can result in diminished yield gains and leave the crop vulnerable to biotic and abiotic stresses (St. Martin and Asiam 1986; Bilyeu and Beuselinck 2005; ). Exotic or under-utilized germplasm has emerged as a genetic reservoir that may hold potentially beneficial genes to help address these issues (Sneller et al. 2005; Fox et al. 2015; Wang et al. 2017; Kofsky et al. 2018; Gaire et al. 2020; Kilian et al. 2020). The following drawbacks are most cited as reasons for avoiding the use of exotic germplasm sources: yield-drag, excessive selection cycles, and time-cost constraints (; Bernardo 2009; Kim et al. 2012). Despite these challenges, there is much literature to support contributions of exotic germplasm (Palomeque et al. 2009a, 2009b; Kim et al. 2011; Rossi et al. 2013; Akpertey et al. 2014; Bellaloui et al. 2017), which highlighted the importance of this under-used germplasm source.

Whereas seed yield, disease resistance, and seed quality traits are often the main breeding objectives, other agronomic traits such as plant height, lodging, and seed weight are also important as they affect yield and can determine the end use of the crop (Zhe et al. 2010; Liang et al. 2016; Teng et al. 2017; Zatybekov et al. 2017; Zhang et al. 2018). The properties of soybean seed such as protein, essential amino acids, oil, metabolizable energy, flavonoids, and other secondary metabolites impart great commercial value to the crop (Willis 2003; Mittal et al. 2004; Patil et al. 2017; Sah 2018). With increasing evidence reported in the literature of the many benefits to human health and nutrition, as well as ongoing need to produce nutritious foods to combat world hunger and food shortage, the demand for soybean and soybean food products is predicted to grow significantly (Friedman and Brandon 2001; McCue and Shetty 2004; Mittal et al. 2004; Lule et al. 2015). Therefore, it becomes imperative to elucidate and thoroughly understand the genetic basis for soybean protein, essential amino acids, oil, metabolizable energy, flavonoids, and other secondary metabolites. Furthermore, the underlying phenotypic correlations among traits should also be investigated as the optimal physical and chemical characteristics of seed required vary for different food processors and manufacturers based on end use (Kinney and Clemente 2005; Whiting et al. 2020).

Soybean seed yield is a complex trait that is often positively correlated with seed weight in that high yield is often associated with late maturity, due to a longer seed-fill period, as well as plant height; however, these relationships have been shown to be environment or population dependent (Posadas et al. 2014; Contreras-Soto et al. 2017). Though simultaneous selection for both seed yield and seed-quality traits is most desirable, the negative correlations between some of these traits, polygenic or quantitative nature of the traits and considerable environmental influence, make the task challenging for breeders to improve the traits concurrently (Zhang et al. 2018).

Previous studies conducted on recombinant inbred soybean line (RIL) populations derived from crosses between exotic elite Chinese and adapted elite Canadian cultivars reported significant variation for seed yield (Palomeque et al. 2009a, 2009b; Rossi et al. 2013). Furthermore, these studies reported that a differential performance was observed among the RILs between Canadian and Chinese mega-environments (Palomeque et al. 2009a, 2009b; Rossi et al. 2013). As these studies were conducted only on RILs, the inclusion of a genomic-diversity panel may help to assess their performance against high-yielding, adapted cultivars to better estimate their agronomic and yield capacity. The objective of this study was to test a diverse panel of soybean cultivars consisting of advanced progeny lines derived from crosses between elite adapted Canadian × elite exotic Chinese cultivars against elite Canadian and Chinese cultivars to determine the potential of improving yield and seed quality traits using alleles from elite exotic germplasm from northeast China. Such an evaluation facilitates direct comparisons between the potential of the exotic-derived lines against high-yielding elite commercial cultivar representatives from both the Canadian and Chinese parental groups, thus facilitating better use of available germplasm sources for crop improvement (Lee et al. 2013; Thivierge et al. 2015; Stasko 2018; Torres et al. 2019). Building further upon the findings of Rossi et al. (2013) and Palomeque et al. (2009a, 2009b), we report results from agronomic evaluations of a diverse panel of 200 soybean lines conducted in 2019 and 2020 at Elora and Woodstock, ON, Canada.

Materials and methods

Results

Seed Yield (kg/ha), seed protein concentration (%, dry basis), and seed oil concentration (%, dry basis) between the genotypic groups

The range of yield observed across all locations was 126–4985 kg/ha (Table 1). Both genotype and genotype × environment interactions were shown to be significant (Table 2). The CD–CH progeny lines outperformed both CD and CH cultivars in terms of seed yield in combined environment analysis (Table 3) with a mean yield and standard error of 2582 ± 75.0 kg/ha, whereas Canadian cultivars yielded 2289 ± 76.8 kg/ha and Chinese cultivars yielded 2263 ± 79.5 kg/ha (Table 3).

Table 1.

Summary of soybean seed yield (kg/ha), seed protein, and seed oil concentration for the four environments: Elora 2019, Elora 2020, Woodstock 2019, and Woodstock 2020.

Table 2.

Type III tests of fixed effects for seed yield (kg/ha), seed protein concentration (%, dry basis), and seed oil concentration (%, dry basis) for combined environment analysis for Elora, ON, and Woodstock, ON, for 2019 and 2020.

Table 3.

Least square means for soybean seed yield (kg/ha), seed protein, and seed oil concentration (dry basis, %) and test of significance of differences for the different genotypic groups of the soybean panel across combined environment analysis for trials conducted at Elora, ON, and Woodstock, ON, for 2019 and 2020.

The range for protein concentration across locations was 34.0%–46.8% (dry basis) (Table 1). For this trait, there was no difference observed between Canadian genotypic group and CD–CH genotypic group (Table 3). However, there was a significant difference between these two groups and the Chinese genotypes, with the latter group containing the lowest protein concentration at 40.6 ± 0.12% (se). Canadian genotypic group had the highest protein concentration at 41.2% to 1 ± 0.11%, with CD–CH group containing 41.1 ± 0.09% (Table 3).

The mean seed oil concentration differed significantly between the three genotypic groups (Table 3). Seed oil concentration across locations varied from 14.9% to 23.1% (dry basis) (Table 1). Canadian genotypic group performed the best of the three for this trait with a seed oil concentration of 20.4% ± 0.07% (dry basis), followed by the CD–CH genotypic group at 19.9% ± 0.06%, with the Chinese genotypic group bearing the lowest oil content at 19.1 ± 0.08% (Table 3).

Correlation analysis between seed yield, seed-quality traits, and agronomic traits

Correlations were calculated to evaluate the relationships between seed yield, seed-quality traits, and select agronomic traits (Table 4). The following traits were evaluated: seed yield, protein concentration, oil concentration, seed weight (g), height (cm), days to maturity, and lodging score (1–5).

Table 4.

Pearson correlation coefficients for soybean seed yield, seed oil concentration, seed protein concentration, and agronomic traits for a panel of 200 soybean lines evaluated from 2019 to 2020 at Elora, ON, and Woodstock, ON.

Yield was positively correlated with height (r = 0.47; p < 0.0001) and lodging score (r = 0.28; p < 0.0001). Yield and oil concentration (r = −0.12; p < 0.0001), as well as yield and protein concentration (r = −0.08; p = 0.0035) showed significant negative correlations. Protein concentration was negatively correlated with yield (r = −0.08; p = 0.0035), oil concentration (r = −0.40; p < 0.0001), height (r = −0.23; p < 0.0001), and days to maturity (r = −0.15; p < 0.0001). A significant positive relationship was observed between protein concentration and seed weight (r = 0.18; p < 0.0001).

Oil concentration showed significant negative relationships with seed yield (r = −0.12; p < 0.0001), protein content (r = −0.40; p < 0.0001), height (r = −0.05; p = 0.0411), seed weight (r = −0.12; p < 0.0001), days to maturity (r = −0.26; p < 0.0001) and lodging score (r = −0.24; p < 0.0001).

Discussion

Soybean seed yield and seed-quality traits continue to be of economic importance and are the primary targets for improvement by breeders. However, due to the recurrent use of elite cultivars in breeding programs, coupled with genetic bottlenecks and low diversity of founding populations, there is growing concern regarding the genetic variation found within breeding programs (Gizlice et al. 1993, 1994; Kisha et al. 1998; Cui et al. 2000; Sneller et al. 2005; Hyten et al. 2006; Iquira et al. 2010; Friedrichs et al. 2016). Low genetic diversity can result in diminished yield gains and leave the crop vulnerable to biotic and abiotic stresses (St. Martin and Asiam 1986; Bilyeu and Beuselinck 2005).

The use of exotic germplasm has been highlighted as a potential solution to this growing concern and has been done with success in other plant species (Ehlers and Foster 1993; Concibido et al. 2003; Sneller et al. 2005; Bernardo 2009; Palomeque et al. 2009a; Diers et al. 2018; Gaire et al. 2020). Exotic germplasm is defined as germplasm that is unadapted to the breeder's target environment and can be a great source of novel genetic variation that can potentially contribute resilience to both biotic and abiotic stresses (Hallauer 2007; Mickelbart et al. 2015; Hoffmann et al. 2018). Despite the challenges associated with the use of exotic germplasm, such as yield drag, deleterious alleles, and the need for excessive backcrossing, there is growing evidence to support the use of exotic germplasm (Concibido et al. 2003; Kabelka et al. 2004; Guzman et al. 2007; Palomeque et al. 2009a; Palomeque et a. 2019b; Rossi et al. 2013; Akpertey et al. 2014; Hegstad et al. 2019).

Since China is the centre of origin of soybean and adapted elite Chinese cultivars may potentially contribute beneficial alleles and help improve seed yield, both Palomeque et al. (2010) and Rossi et al. (2013) evaluated populations derived from Canadian × Chinese crosses to evaluate the behaviour and inheritance of seed yield quantiative trait loci (QTL) in different environments. Building further upon previous works mentioned above, this current study aimed to evaluate a diverse panel of 200 soybean genotypes that consisted of University of Guelph released cultivars and experimental breeding lines, AAFC cultivars, elite Chinese cultivars, and Canadian × Chinese progeny lines derived from different crosses. It was observed that the exotic elite Chinese cultivars had the lowest seed yield compared to the adapted Canadian ones (Table 3). This finding is in line with other studies that reported on evaluations of RIL populations with exotic parentage vs adapted elite cultivars (St. Martin and Asiam 1986; Palomeque et al. 2010; Rossi et al. 2013). Canadian × Chinese RILs, on the other hand, outperformed both the exotic Chinese and adapted elite Canadian cultivars (Table 3). Hegstad et al. (2019) reported on a study that evaluated the performance of exotic germplasm-derived soybean lines against the performance of adapted elite cultivars. They reported that across three field years, the highest yielding line yielded 280 kg/ha more than the adapted elite parent (Hagstad et al. 2019). The average yield difference observed in this study between the Canadian genotypic group and the Canadian × Chinese derived genotypic group was similar to that of Hagstad et al. (2019), with a difference of 293 kg/ha. Kim et al. (2012) reported on a study that included a Chinese cultivar (Tiefeng No. 8), which was used as a donor parent with repeat backcrossing. They reported that a yield increase was observed in at least one of the populations being evaluated (Kim et al. 2012). Diers et al. (2018) reported that in their evaluation of different soybean lines, the exotic founder germplasm lines yielded the lowest compared to adapted parent lines, which is similar to the current observation of exotic elite Chinese cultivars. None of these studies, however, reported on evaluating exotic-derived germplasm against a diverse group of soybean germplasm to assess seed yield, seed quality traits, and agronomic traits. Moreover, Posadas et al. (2014) reported an average yield gain of 567 kg/ha with each cycle of recurrent selection in a population developed from seven unadapted plant introductions. Their study only included the evaluation of a single population derived from three Japanese, three South Korean, and one Chinese plant introduction (Posadas et al. 2014). To the best of our knowledge, the current study is the first to directly evaluate Canadian × Chinese progeny in a diverse pool of soybean lines. The abovementioned studies together with ours provide further support toward the potential for contributing beneficial alleles from exotic germplasm sources for continued crop improvement.

Though soybean seed oil and protein are two major focal points for crop improvement for breeders, the complicated and often quantitative nature of these traits make direct improvement quite challenging. Previous studies have identified similar relationships to those observed in the current study (Table 4) between seed yield, seed quality traits, and agronomic traits (Palomeque et al. 2009a, 2010; Rossi et al. 2013; Contreras-Soto et al. 2017). Contreras-Soto et al. (2017) reported that soybean seed yield, plant height, and days to maturity showed positive correlations, with a general trend of later maturity and taller plants were associated with higher yield. Additionally, Posadas et al. (2014) reported that the magnitude, direction, and significance for plant height, lodging, and seed yield correlations were population dependent.

A significant negative correlation was observed between seed oil and protein concentration (Table 4), which was consistent with previous reports (Hyten et al. 2014; Ficht et al. 2022; Lee et al. 2019). Patil et al. (2018) also reported similar results to those reported herewith regards to the relationships between seed yield and seed-quality traits and concluded that soybean seed composition is significantly affected by the environment. This sentiment is consistent with reports elsewhere in literature (Zhe et al. 2010). It is possible that this was the case in the current study; therefore, expanded environments and further testing will be needed to establish the precise nature of these relationships.

Conclusion

In conclusion, the findings of this study are consistent with previous reports and show the promising potential of using exotic soybean germplasm, specifically from China, in improving yield potential and seed-quality traits in Canada and potentially northern USA. Further studies across multiple years and environments will be needed to verify the genotypic performance of the exotic-derived lines and to rule out environmental or other effects that may have affected seed yield and seed-quality traits during the period of the study. Furthermore, since the first year of field evaluations did not include all 200 cultivars at each environment, inclusion of the full panel across all years and environments would help overcome limitations of an uneven sample size experienced in the current study.