The experiment was carried out at the Tennessee State University (TSU) Agricultural Research and Education Center (36˚10'43.8"N 86˚49'35.4"W) in Nashville, Tennessee, during the period of May to September 2021. The soil type of the experiment site was silt loam, slightly acidic (approximate pH = 6.1), with soil nitrogen of 0.15%, phosphorus of 188.9 kg/ha, and potassium of 107.0 kg/ha (Table 1). Mean monthly air temperature varied from 22.5˚C to 27.3˚C and monthly precipitation varied from 56.1 to 222.0 mm during the experimental period (Figure 1).

Four mung bean varieties were used in this study: Oklahoma 2000 (OK2000), Oklahoma Berken (Berken), Tennessee State University-1 (TSU-1), and Alabama Agricultural and Mechanical University-1 (AAMU-1). Seeds were obtained from the Oklahoma State University, TSU, and AAMU. The length between the two rows was 45 cm. Three plant spacings of 5 cm, 10 cm, and 15 cm were used, representing 444,444, 222,222, and 148,148 plants /ha, respectively.

The experiment was laid out as a split-plot design with three replications. The density was used as the main plot factor and variety as the subplot factor. In each block, varieties and densities were randomly assigned. Before planting the soil was ploughed and leveled. A pre-plant application of 40 kg N/ha urea (46-0-0) was surface applied. The total number of plots was 36. The plot size was 4 m × 2 m with 10 rows in each plot. Seeds were manually sowed on June 25, and plots were irrigated after sowing to improve seed germination. There were no major insect problems. Weeds were controlled using Roundup before the field was ploughed. During the growing season, weed control was done manually. The mung beans were harvested in late September.

Leaf photosynthesis and transpiration were measured on at least three healthy, fully expanded mature leaves in each plot using the Li-6800 Portable Photosynthesis System (Li-Cor Ins., Lincoln, NE, USA) twice in August. Water use efficiency (WUE) was calculated as the ratio of leaf photosynthesis and transpiration. During measurements, CO₂ concentration was set at 400 ppm and light was set at 1500 quanta/m²/s.

Relative chlorophyll content measurements were taken on three to five leaves in each plot, twice in August and once at the beginning of September, using the SPAD 502 Plus (Konica Minolta Optics, Japan). The height of five randomly selected plants in each plot was measured twice in August. The LAI-2200 Plant Canopy Analyzer (Li-Cor, Lincoln, NE, USA) was used for LAI measurement. We followed the manufacturer’s instructions for row crop measurements. Two light readings were taken above the canopy and 6 were taken at the bottom of the canopy for each plot, in the given sequence of ABBBABBB (A for above canopy and B for below canopy). The LAI was measured in September when the maximum heights were reached.

Fifteen plants were harvested from each plot for yield on October 1. Plants representing the plot were selected sequentially in middle rows and the edge of the plot was avoided. Pods were separated from each plant. Fresh plants were weighed and dried in the oven at 70˚C to constant weight for three days. Six agronomic characteristics were recorded, including the plant dry weight, pod dry weight, the weight of seeds per plant, 100-seed weight, number of pods per plant, and number of seeds per pod. Crop yield was calculated as the weight of seeds per plant × number of plants per hectare and expressed as kg/ha. The harvest index (HI) was calculated as the seed weight divided by the plant dry weight.

The effects of varieties, densities, and their interactions were analyzed using split-plot analysis of variance (ANOVA). Multiple comparisons were conducted using the Least Significant Difference (LSD) method when a significant effect was detected. To test whether there is a significant relationship between yield and its components, we conducted bivariate linear regression. Stepwise multiple regression was further conducted to develop the best regression model of yield with its components. Data analysis was done using SAS software (SAS 9.4, SAS Institute Inc., Cary, NC, USA).

Leaf photosynthesis and transpiration are two important biological processes that directly influence a plant’s growth, productivity, and yield [11] . LAI plays a crucial role in capturing photosynthetically active radiation, as a larger LAI indicates more interception of solar radiation. The height of the plant is an essential aspect of crop growth and is influenced by environmental factors [12] . Dry biomass is a function of LAI and light interception, radiation use efficiency, and the most important determinant of crop yield [8] . In this study, we found no significant difference in leaf photosynthesis, transpiration, water use efficiency (WUE), and LAI among varieties and densities, however, relative chlorophyll content and height significantly varied among the varieties and transpiration varied among the densities (Table 2). Density influenced transpiration and plant dry biomass. No interaction was found between variety and density for all physiological and growth variables.

As no significant difference was found in leaf photosynthesis, transpiration, and the resulting WUE among the varieties, we averaged them over the varieties and found that the mean leaf photosynthetic rate of mung bean was 28.67 µmol CO₂/m²/s and ranged from 27.88 to 29.63 µmol CO₂/m²/s, the mean transpiration rate was 17.07 mmol H₂O/m²/s and ranged from 16.60 to 17.62 mmol H₂O/m²/s, and the mean WUE was 1.69 µmol CO₂/mmol H₂O and ranged from 1.63 to 1.71 µmol CO₂/mmol H₂O. While transpiration and WUE were seldom reported in previous studies, photosynthesis has been widely studied. Our results were comparable to Islam [13] who reported that the photosynthesis among 8 mung bean varieties varied from 23.1 to 27.4 µmol CO₂/m²/s but higher than most other studies. For example, Nazar et al. [14] found that mung bean photosynthesis varies between 14 to 19 µmol CO₂/m²/s for two varieties under the control but can vary from about 8 to 28 µmol CO₂/m²/s under different salt and foliar salicylic acid treatments. Ahmed et al. [15] reported that the photosynthesis of mung bean is about 20 µmol CO₂/m²/s in no water stress conditions. Hossain et al. [16] showed that photosynthesis varied among different developmental stages and during the flowering stage, the photosynthesis of 6 mung bean varieties can reach 24.2 to 43.8 µmol CO₂/m²/s.

Relative chlorophyll content was the highest in TSU-1 (44.18), but it did not differ from OK2000 (Figure 2). Berken had a significantly lower (41.88, 5.2%) chlorophyll content than TSU-1. Plant height was also higher for TSU-1 (63.95 cm) and AAMU-1 (62.34 cm) compared to OK2000 (57.23 cm) and Berken (56.94 cm) (Figure 2). These values were comparable to other previous studies. For example, Khajudparn and Tantasawat [17] reported a mean plant height of 47.6 cm (30.7 - 62.2 cm) among 56 mung bean accessions. Mondal et al. [18] reported that height varied from 40.1 cm to 54.5 cm among three mung bean varieties. Other studies also found that mung bean plants grew to about 50 cm (30 - 60 cm)

*indicates significance at p < 0.1 level; **indicates significance at p < 0.05 level; ***indicates significance at p < 0.01 level.

(e.g., Kabir and Sarkar [3] ; Hasan et al. [19] ; Ahamed et al. [20] ). In this study, we found that LAI and dry biomass were not influenced by variety. The LAI of mung bean on average was 4.86 m²/m² and dry biomass was 2.70 g/plant. The LAI in this study was higher than most other reports although a large variation in LAI of mung bean has been reported in the literature. For example, Khajudparn and Tantasawat [17] evaluated 56 mung bean accessions and found LAI varied from 1.2 to 4, with a mean of 2.4. Mondal et al. [18] found that LAI varied between 2.8 to 5.2. Muchow et al. [2] reported that the maximum LAI can reach up to 6 in mung bean grown in Australia. The plant dry biomass did not differ significantly among the varieties (Table 2). But other studies found that dry biomass may vary among different varieties. For example, Ahamed et al. [19] investigated the effects of cultivars on mung bean in Bangladesh and found that total dry matter varied from 10.72 to 12.32 g/plant.

Density did not influence leaf photosynthesis, but mung beans growing in the highest density plots (5 cm spacing) consumed more water than the other two densities as leaf transpiration was higher (18.16 mmol H₂O/m²/s) and significantly different in the highest density in comparison to the other two densities (Figure 3). There was no change in WUE among the densities. LAI and plant height were not influenced by density, but plant dry biomass differed among different densities (Table 2). More biomass was produced in the highest density plots than in the other density treatments (Figure 3). The impacts of planting density on mung beans have been widely studied and its significant impacts on plant growth have been found in most published studies. Kundu et al. [21] reported that plant height varied from 44.6 cm to 55.6 cm under different planting densities. Mansoor et al. [22] found that plant height varied from 67.53 cm to 72.21 cm among three-row spacings (20, 30, 40 cm). Ahamed et al. [20] found that plant density at 40 cm × 10 cm spacing produced the highest total dry matter of 11.85 g per plant. A few studies also reported no influence by density. For example, Hasan et al. [19] found that plant density (30 cm × 5, 10, 15 cm) did not influence plant height and yield. Overall, mung bean plants growing at high density often grow taller and produce more biomass.

Crop yield is a complex quantitative trait and the final product and vital goal of the farmer [17] . The yield of mung bean is determined by several yield component variables including pod number per plant, seed number per pod, 100-seed weight, and harvest index. The number of pods per plant is a key factor determining crop yield and higher yields are solely related to its multitude of pods bearing branches that produce more pods and number of seeds per pod [12] . The number of seeds per pod is another most important trait of legume crops that gives rise to the highest yields. This is the product harvested and is very important to human nutrition. 100-seed weight is one feature of mung bean varieties and may also be influenced by growing conditions. The harvest index, a measure of the production efficiency of biomass partitioning into yield, determines the capability of the plant to transport photosynthetic material to the economical part [8] .

In this study, results of ANOVA showed that variety had significant impacts on pod dry biomass, number of pods per plant, number of seeds per pod, harvest index, and crop yield, but did not influence 100-seed weight (Table 3). There was no significant influence of planting density on all measured variables. A significant interaction was only found for pod number per plant between variety and density.

Among the three varieties, OK2000 had the highest yield (2724.8 kg/ha), significantly higher (39.6%) than AAMU-1 (1951.8 kg /ha) and (71.8%) higher than TSU-1 (1586.1 kg/ha) but was not significantly different from Berken (2086.2 kg /ha). On average, OK2000 produced a 45.3% higher yield than other varieties. OK2000 also had the highest number of pods per plant (28.70) or 637.8 g/m² among the four varieties which was significantly higher than any other varieties. The maximum number of seeds per pod (11.57) was also obtained from the OK2000 variety. TSU-1 had the lowest number of seeds per pod (9.81) and no significant difference in the number of seeds per pod among the other three varieties (Figure 4). There was also no significant difference in 100-seed weight and the mean value of the four varieties was 4.75 g (4.46 to 4.98 g). In addition, OK2000 had the highest harvest index (20.2%) but was not significantly different from TSU-1 (Figure 4). AAMU-1 had the lowest harvest index (13.3%). On average, OK2000 has a 42.4% higher harvest index than other varieties.

^∆ indicates p < 0.15 level; *indicates significance at p < 0.05 level; ** indicates significance at p < 0.01 level; *** indicates significance at p < 0.001 level.

Yield and its components of mung bean have been investigated in many studies and large variations in yield and its components are reported in the literature (Table 4). Our results were in the range of reported values in previous studies. For example, Ajio et al. [23] reported that the yield of mung bean varied from 919 to 3586 kg/Ha, the number of pods per plant varied from 5.7 to 17.1, and the number of seeds per pod varied from 8.0 to 12.0. Similarly, Khajudparn and Tantasawat [17] found that the mean yield of mung bean is 2038.3 kg/ha (955 - 3176 kg /ha), the number of pods per plant is 23.6 (12.2 - 49.2), the number of seeds per pod is 9.9 (6.1 - 11.6), and 100-seed weight is 5.2 (2.4 - 7.6). One study did not find a significant difference in yield (~734 kg/ ha) among varieties, only pods per plant and seeds per pod were significantly influenced [19] . While Muchow et al. [2] found that mung bean can produce a grain yield of 2500 kg /ha in Australia, the lower yield of mung bean is mostly reported in previous studies. Sarkar et al. [24] conducted a comprehensive field experiment in Bangladesh and found that pods per plant varied from 15.14 to 17.72 among 5 varieties, seeds per pod varied from 9.68 to 10.38, 100-seed weight varied from 2.30 to 3.4 g, harvest index varied from 29.9% to 35.4%, and yield varied from 739.9 to 866.9 kg/ha. Mondal et al. [18] reported that among three varieties, the yield varied from 1496 to 1895 kg/ha, number of pods per plant varied from 21.3 to 33.0, number of seeds per pod varied from 8.6 to 9.3, 100-seed weight varied from 3.2 to 5.5 g, and harvest index varied from 26.4% to 34.6%. Kabir and Sarkar [3]

studied the effect of variety and planting density on the yield of mung bean and found among five varieties, the yield varied from 701.2 to 843.7 kg/ha, number of pods per plant varied from 11.6 to 12.0, number of seeds per pod varied from 9.44 to 10.53, and 100-seed weight varied from 3.02 to 3.60 g. These results and the result from this study suggested that mung bean yield and its components are influenced by many factors including mung bean varieties, environmental conditions in growing areas, and agricultural practices. The number of pods per plant could be influenced by fertilizer rate, as high nitrogen availability can facilitate the production of more branches and canopy development and contribute to high total pod production. The supply of adequate nutrients may facilitate vegetative growth and increase the number of seeds per pod [15] [23] . The 100-seed weight may vary among different varieties which might be attributed to genotypic variation in mung bean [18] [20] [23] [24] . The harvest index for mung bean was found to be around 30% [8] [25] which was slightly higher than the value in this study.

In this study, density did not influence yield and its components, but the significant tests for plant pod biomass (F = 4.03, p = 0.11) and crop yield (F = 2.94, p = 0.15) were close to significant. Both the 10 cm (868.16 g/plot or 108.52 g/m²) and 15 cm (704.18 g/plot or 88.02 g/m²) plant spacing was significantly different from the 5 cm plant spacing (215.3 g/m²) (Figure 3). The high density tended to produce high biomass compared to the medium and low densities. The results were consistent with some previous studies that reported more biomass produced at narrow row spacing than at wider spacing [18] [21] [24] . Kundu et al. [21] found that close row spacing (25 cm) resulted in greater grain (583.0 kg/ha) than wider spaced crop (30 cm) and yield varied from 542.1 to 583.0 kg /ha, and harvest index varied from 21.0% to 20.2%. Rachaputi et al. [26] studied the effects of row spacing and plant density on yield in Australia and found that narrow row spacing resulted in 14% more yield compared to wide rows due to more intercepted radiation. Mansoor et al. [22] also found yield is slightly higher in the high-density treatment (1104 kg/ha). Hasan et al. [19] reported that plant density (30 cm × 5, 10, 15 cm) does not influence plant yield, but high density tends to produce more pods per plant and more seeds per pod. Yield components were also found to be influenced by planting density in several other studies. For example, Ajio et al. [23] reported that among different densities, the number of seeds per pod varied from 12.7 to 16.2, the number of seeds per pod varied from 10.6 to 11.1 and yield varied from 1171 to 2085 kg/ha. Similar findings were reported that as the plant spacing increases so does the number of pods per plant since it allows the plants to access enough nutrients, sunlight, water, and other growth requirements [19] [24] [27] . Sarkar et al. [24] reported that pods per plant varied from 15.21 to 17.07 among 3 plant densities, 100-seeds varied from 2.58 to 2.69 g, harvest index varied from 19.6% to 41.6%, and yield varied from 301.5 to 1183.0 kg/ha, but seeds per pot are not significant (mean is about 10 seeds per pod). Ajio et al. [23] studied two mung bean varieties and four plant densities (10 × 10, 20, 30, 40 cm) in Uganda and found the highest number of pods per plant was recorded at a spacing of 40 × 10 cm, and the highest grain yield at a spacing of 10 × 10 cm. The high dense (10 × 10 cm) spacing resulted in the highest mung bean grain yield in Uganda. Kabir and Sarkar [3] found that planting density influenced all variables except the number of seeds per pod. It is worth noting that these studies used different rows and plant spacings. We used 45 × 5, 10, and 15 cm spacing in this study. The row distance was longer than in most previous studies. The impacts of planting density on yield and its components need to be further explored.

Linear regression analysis showed that yield was significantly correlated with number of pods per plant (Y = 497.39 + 80.295 * X₁, r² = 0.23, p < 0.001, where Y is yield (kg/ha), X₁ is number of pods per plant) and 100-seed weight (Y = −328.92 + 113.19 * X₂, r² = 0.19, p = 0.007, where X₂ is 100-seed weight (g)). The yield of mung beans increased with the number of pods and 100-seed weight. The best regression model of yield and its component was: Y = −304.92 + 6.30 * X₁ + 81.89 * X₂, R² = 0.32, p = 0.002. The number of pods per plant and 100-seed weight contributed almost equally (Path coefficient was 0.37 for number of pods and 0.32 for seed weight). Yield also increased with harvest index (Y= −5.1661 + 11.059 * X, r² = 0.40, p < 0.001). Similar results have been reported in previous studies. For example, Khajudparn and Tantasawat [17] reported that seed yield is significantly and positively corrected with the number of clusters per plant, seeds per pod, dry biomass, and pods per plant of mung bean. Singh et al. [28] evaluated 40 genotypes of mung bean under 4 diverse environments in India and found that yield had a significant positive correlation with the number of pods per plant, number of seeds per pod, 100-seed weight, and harvest index. Other agricultural practices such as nutrient application and irrigation during drought periods that stimulate plant growth and produce high numbers of pods per plant and seeds per pod could also improve crop yield [6] [19] [29] . Summer drought in the southeastern US could reduce leaf photosynthesis and reduced mung bean yield.