Soybeans are among the leading agricultural crops that are produced on around 6% of the world’s arable land. Soybeans are processed into various products, including high-protein meals, livestock feed, and edible oil [
29]. In Poland, the area under soybeans increased from 7642 to 9210 ha, and soybean production increased from 14,747 to 20,970 tons between 2016 and 2021, and it continues to grow [
2]. This increase resulted from a number of factors, including the development of high-yielding cultivars that are better adapted to the Polish climate. Sowing date is an important determinant of soybean yields [
11,
12,
30,
31,
32]. According to Kumar et al. [
33], climate is the key factor that affects sowing date choices. In Poland, soybeans should be sown when the mean daily soil temperature exceeds 8 °C, i.e., at the turn of April and May [
6,
7]. Early sowing can lead to delayed and uneven seed germination due to low soil temperature, whereas delayed sowing increases the risk of damage caused by spring drought [
22]. In the present study, sowing date influenced the growth, development, and yields of the analyzed soybean cultivars, as well as the length of the growing season. The time from sowing to emergence of soybean seedlings varied across years and cultivars. The most favorable weather conditions for the emergence of soybean seedlings were noted in 2018, which was characterized by the shortest time between sowing and emergence. In north-eastern Poland, frost episodes frequently occur in late April and in the first half of May when soybeans are sown and when seedlings emerge. During the study, the average number of days with frost episodes was determined to be 3.25 for sowing date I, 1.16 for sowing date II, and 0 for sowing date III, which affected the time from sowing to emergence of soybean seedlings. According to Dragańska et al. [
34], late frost episodes at a height of 2 m above ground were noted in north-eastern Poland in the first days of May in the eastern part of the region and in mid-April in the remaining parts of the region between 1981 and 2010. Ground frost at a height of 5 cm above the soil occurred in late May in the east and in mid-May in other parts of the region. The cited authors also reported ground frost episodes in the last ten days of June in the studied period. The average number of days with spring frost episodes ranged from 9 to 16. In the current study, the time from sowing to the emergence of soybean seedlings was influenced by the number of days with unfavorable temperatures (R = 0.76). Other contributing factors were daily minimum temperature (R = −0.44) and, to a lesser extent, daily mean temperature (R = −0.28). Low temperatures considerably delayed seedling emergence, compromised plant health, and delayed seed ripening. Similar observations were made by Uslu and Esendal [
22]. According to Kumagai [
15], early sown soybeans are at greater risk of exposure to extremely low temperatures and late spring frost that can inhibit germination, seedling emergence, and early stand development. The optimal sowing date is determined by the local climate [
35,
36]. In the work of Serafin-Andrzejewska et al. [
7], the growing season was shortened by 14 days when soybeans were sown 20 days past the earliest date, which corresponds with the present findings. Bateman et al. [
37] also found that late-sown plants were unable to harness their growth potential fully. Delayed sowing shortens the growing season and could potentially affect plant height, stand density, and seed yields [
23,
37,
38]. In the current study, the prolonged time from sowing to emergence of soybean seedlings was negatively correlated with seed yields (R = −0.55), which indicates that delayed seedling emergence decreases seed yields. The number of days with frost episodes was correlated with seed yields (R = −0.33), which suggests that frost risk should be considered when selecting sowing dates in a given region. According to Mandić et al. [
39] and Shah et al. [
40], the optimal sowing date and climate-adapted genotypes promote the uptake of soil nutrients and water, thus maximizing seed yields.
In this study, sowing dates exerted varied effects on plant height and yield components across years. In general, late-sown seeds produced the tallest plants. In turn, the mean values of yield components for four years of the experiment were not significantly affected by sowing dates. Soybean cvs. Marlin and Lissabon produced more pods per plant than cv. Aldana and thousand seed weight was highest in cv. Lissabon. In the work of Bateman et al. [
37], plant height increased by 0.3 cm per day when soybeans were sown between 25 March and 2 June, but it decreased by 2.7 cm per day when soybeans were sown late between 2 June and 16 July. Jarecki and Bobrecka-Jamro [
20] found that early sowing increased the number of pods per plant and thousand seed weight relative to the optimal sowing date. Księżak and Bojarszczuk [
41] also reported that yield components in the studied soybean cultivars were influenced by weather conditions during the growing season and sowing date. Between 2017 and 2019, soybeans sown on the optimal date were characterized by the highest seed weight per plant, whereas delayed sowing induced only minor differences in the number of seeds per pod [
41]. In the work of Pedersen and Lauer [
9] and Kumar et al. [
42], the number of pods per plant and the number of seeds per pod were higher in early sown than in late-sown soybeans. In turn, Borowska and Prusiński [
43] reported that the number of pods per plant was the only yield component that was significantly affected by sowing date. In other studies, the number of pods per plant, the number of seeds per pod and seed weight per plant were lower in late-sown soybeans than in early sown soybeans [
44,
45]. Shah et al. [
40] also found that late-sown soybean plants were shorter and produced fewer pods.
Sowing date and weather conditions exerted varied effects on seed yields. The average values in the four-year study indicate that sowing date influenced seed yields, and total seed yields in north-eastern Poland (Region of Warmia and Mazury) were highest when soybeans were sown late (in mid-May). The linear regression analysis revealed a correlation between the length of the growing season and seed yields in late-sown plants (R = 0.47). In a long-term study conducted by Borowska and Prusiński [
43], seed yields peaked when soybeans were sown at the turn of April and May, which corroborates the findings of other authors [
23,
46]. Umburanas et al. [
47] also concluded that optimal sowing dates and seeding rates promote plant growth and increase seed yields. In the cited study, delayed sowing compromised yields by decreasing above-ground biomass per unit area, leaf area index, plant height at harvest, height of the lowest pod, number of pods per unit area, number of seeds per unit area, and seed weight. A higher seeding rate increased seed yields, in particular in late-sown plants, by increasing above-ground biomass per unit area, leaf area index, plant height at harvest, height of the lowest pod, number of pods per unit area, and number of seeds per unit area. In a study by Kumagai and Takahashi [
44], the number of seeds per pod was one of the key determinants of soybean yields. Delayed sowing reduced the number of seeds per pod, mainly due to low temperatures 20 days after the beginning of seed filling. In turn, Mandić et al. [
39] observed a significant reduction in seed yields in all soybean plants that were not sown on the optimal date. Soybeans sown in late April were characterized by a smaller number of pods per plant, lower seed weight per plant, and lower thousand seed weight, which decreased seed yields. These observations were attributed to accelerated plant senescence and the adverse influence of high temperature and low precipitation during seed filling. The cited study was conducted in Serbia, where soybeans are sown at the beginning of April and harvested in September. Therefore, flowering, pod and seed development, and ripening stages take place in July and August when temperatures are high and precipitation is low [
45]. These stressors can decrease soybean yields by up to 74% relative to unstressed plants [
48]. In the present study, seed yields were highest in cv. Merlin (4.00 t ha
−1 on average) and lowest in cv. Aldana (2.67 t ha
−1 on average). Similar results were reported by Borowska and Prusiński [
43], where Merlin was also the highest-yielding cultivar (3.17 t ha
−1), and Aldana was the lowest-yielding (1.91 t ha
−1) cultivar. In the current study, seed protein content was highest in late-sown plants. Seeds of soybean cvs. Lissabon and Merlin were characterized by the highest protein yield. In a four-year study conducted by Borowska and Prusiński [
43], average seed yields were highest in cv. Merlin. In south-eastern Poland, the average seed yield of soybean plants was determined at 4.18 t ha
−1 by Jarecki and Bobrecka-Jamro [
20]. In the cited study, sowing date had no significant influence on seed yields. In 2017, seed yields were significantly higher in late-sown than in early sown plants. Soybeans cv. Aldana were characterized by the lowest seed yields in all years of the study. In the present study, Aldana was also the lowest-yielding cultivar in north-eastern Poland. Seed yields were lowest in 2017 and highest in 2018, which is consistent with the findings of Jarecki and Bobrecka-Jamro [
20] and Księżak and Bojarszczuk [
41]. In the cited studies, seed protein content was significantly higher in late-sown than in early sown plants. In turn, protein and oil yields were not modified by sowing date. In the group of soybean cultivars analyzed by Jarecki et al. [
6], cv. Aldana was characterized by low protein and oil yields. Numerous researchers reported higher protein concentrations in late-sown soybeans [
14,
23,
30,
46]. Mandić et al. [
39] also found that sowing date significantly influenced the protein and oil content of soybeans, especially under water stress in the reproductive stage. Delayed sowing induces a significant decrease in protein content and an increase in the oil content of soybean seeds [
49] because high temperature increases the protein content but has a marginal influence or no effect on oil content [
50]. In a study conducted by Serafin-Andrzejewska et al. [
7] in south-western Poland (Region of Lower Silesia), seed yields were lowest in late-sown soybeans. Therefore, in Lower Silesia, soybeans should be sown in the second or third week of April or at the beginning of May. Soybean cv. Lissabon was characterized by high seed yields [
7]. Delayed sowing also negatively affected seed yields in the work of Bateman et al. [
37] who found that seed yields decreased by more than 26 kg ha
−1 when soybeans were sown past 20 April in the southern USA. Robinson et al. [
14] reported higher seed yields in soybeans sown in April and early May and lower seed yields in soybeans sown in late March and early June. In a study by Kumagai and Takahashi [
44], seed yields were reduced when soybeans were sown around three weeks past the optimal date. In north-eastern China, soybean yields were affected by variations in climatic factors associated with latitude, and in high-altitude regions, yields were positively correlated with temperature but negatively correlated with accumulated sunshine hours. Climate was responsible for −24% to 38% of the variation in seed yields, and temperature was the most significant climatic factor [
51]. A study conducted by Kumagai and Takahashi [
44] in the cool region of northern Japan demonstrated that delayed sowing and, consequently, lower temperature during the reproductive stage decreased seed yields and the values of yield components. Mean daily temperature was negatively and significantly correlated with the fraction of available soil water (FASW), which suggests that excess soil water caused by high precipitation was associated with cold weather. In turn, Borowska and Prusiński [
43] found that total precipitation in June and July was significantly correlated with seed yields in early sown soybeans, whereas total precipitation in August was also significantly correlated with seed yields in soybeans sown on later dates. Seed yields were significantly highest when soybeans were sown at the turn of April and May, whereas seed and protein yields and seed protein content were highest in the medium-early cv. Merlin. Seed yields were also significantly correlated with total precipitation in other studies [
45,
52,
53]. According to Kumagai [
15], water supply plays a particularly important role in soybean production, and soybean yields in northern Japan were influenced by precipitation levels and distribution across years and experimental sites. Thomasz et al. [
54] analyzed the relationship between soil water content and soybean yields in 28 agricultural districts in Argentina. The data provided by local weather stations were used in correlation and regression analyses and to forecast soybean yields. Correlation and regression analyses revealed that, in most cases, soil water content explained at least 50% of the variation in soybean yields.